diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,82097 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.999827482101268,
+  "eval_steps": 100,
+  "global_step": 11592,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00017251789873199344,
+      "grad_norm": 43.25,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 2.642,
+      "step": 1
+    },
+    {
+      "epoch": 0.0003450357974639869,
+      "grad_norm": 406.0,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 2.9249,
+      "step": 2
+    },
+    {
+      "epoch": 0.0005175536961959803,
+      "grad_norm": 65.0,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 3.0206,
+      "step": 3
+    },
+    {
+      "epoch": 0.0006900715949279737,
+      "grad_norm": 55.75,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 2.7799,
+      "step": 4
+    },
+    {
+      "epoch": 0.0008625894936599673,
+      "grad_norm": 41.5,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 2.7781,
+      "step": 5
+    },
+    {
+      "epoch": 0.0010351073923919607,
+      "grad_norm": 47.0,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 2.8404,
+      "step": 6
+    },
+    {
+      "epoch": 0.001207625291123954,
+      "grad_norm": 49.5,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 2.9029,
+      "step": 7
+    },
+    {
+      "epoch": 0.0013801431898559475,
+      "grad_norm": 47.25,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 2.8079,
+      "step": 8
+    },
+    {
+      "epoch": 0.001552661088587941,
+      "grad_norm": 32.0,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 2.794,
+      "step": 9
+    },
+    {
+      "epoch": 0.0017251789873199345,
+      "grad_norm": 36.5,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 2.7301,
+      "step": 10
+    },
+    {
+      "epoch": 0.001897696886051928,
+      "grad_norm": 38.75,
+      "learning_rate": 4.4e-06,
+      "loss": 2.5871,
+      "step": 11
+    },
+    {
+      "epoch": 0.0020702147847839214,
+      "grad_norm": 44.75,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 2.7294,
+      "step": 12
+    },
+    {
+      "epoch": 0.0022427326835159148,
+      "grad_norm": 41.5,
+      "learning_rate": 5.2e-06,
+      "loss": 2.4985,
+      "step": 13
+    },
+    {
+      "epoch": 0.002415250582247908,
+      "grad_norm": 25.125,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 2.5737,
+      "step": 14
+    },
+    {
+      "epoch": 0.0025877684809799016,
+      "grad_norm": 29.25,
+      "learning_rate": 6e-06,
+      "loss": 2.6294,
+      "step": 15
+    },
+    {
+      "epoch": 0.002760286379711895,
+      "grad_norm": 22.375,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 2.4837,
+      "step": 16
+    },
+    {
+      "epoch": 0.0029328042784438884,
+      "grad_norm": 37.25,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 2.5322,
+      "step": 17
+    },
+    {
+      "epoch": 0.003105322177175882,
+      "grad_norm": 28.125,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 2.3859,
+      "step": 18
+    },
+    {
+      "epoch": 0.0032778400759078757,
+      "grad_norm": 16.25,
+      "learning_rate": 7.600000000000001e-06,
+      "loss": 2.3999,
+      "step": 19
+    },
+    {
+      "epoch": 0.003450357974639869,
+      "grad_norm": 19.25,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 2.418,
+      "step": 20
+    },
+    {
+      "epoch": 0.0036228758733718625,
+      "grad_norm": 15.25,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 2.3585,
+      "step": 21
+    },
+    {
+      "epoch": 0.003795393772103856,
+      "grad_norm": 12.8125,
+      "learning_rate": 8.8e-06,
+      "loss": 2.2626,
+      "step": 22
+    },
+    {
+      "epoch": 0.003967911670835849,
+      "grad_norm": 9.0,
+      "learning_rate": 9.200000000000002e-06,
+      "loss": 2.3482,
+      "step": 23
+    },
+    {
+      "epoch": 0.004140429569567843,
+      "grad_norm": 7.125,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 2.1597,
+      "step": 24
+    },
+    {
+      "epoch": 0.004312947468299836,
+      "grad_norm": 27.25,
+      "learning_rate": 1e-05,
+      "loss": 2.1392,
+      "step": 25
+    },
+    {
+      "epoch": 0.0044854653670318295,
+      "grad_norm": 12.5625,
+      "learning_rate": 1.04e-05,
+      "loss": 2.218,
+      "step": 26
+    },
+    {
+      "epoch": 0.004657983265763823,
+      "grad_norm": 32.75,
+      "learning_rate": 1.0800000000000002e-05,
+      "loss": 2.2254,
+      "step": 27
+    },
+    {
+      "epoch": 0.004830501164495816,
+      "grad_norm": 4.59375,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 2.1076,
+      "step": 28
+    },
+    {
+      "epoch": 0.00500301906322781,
+      "grad_norm": 24.125,
+      "learning_rate": 1.16e-05,
+      "loss": 2.0355,
+      "step": 29
+    },
+    {
+      "epoch": 0.005175536961959803,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.2e-05,
+      "loss": 2.1556,
+      "step": 30
+    },
+    {
+      "epoch": 0.005348054860691797,
+      "grad_norm": 8.75,
+      "learning_rate": 1.2400000000000002e-05,
+      "loss": 2.2392,
+      "step": 31
+    },
+    {
+      "epoch": 0.00552057275942379,
+      "grad_norm": 11.1875,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 2.1823,
+      "step": 32
+    },
+    {
+      "epoch": 0.005693090658155784,
+      "grad_norm": 3.875,
+      "learning_rate": 1.3200000000000002e-05,
+      "loss": 2.0795,
+      "step": 33
+    },
+    {
+      "epoch": 0.005865608556887777,
+      "grad_norm": 20.375,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 2.0121,
+      "step": 34
+    },
+    {
+      "epoch": 0.006038126455619771,
+      "grad_norm": 16.125,
+      "learning_rate": 1.4e-05,
+      "loss": 2.0659,
+      "step": 35
+    },
+    {
+      "epoch": 0.006210644354351764,
+      "grad_norm": 9.625,
+      "learning_rate": 1.4400000000000001e-05,
+      "loss": 2.1046,
+      "step": 36
+    },
+    {
+      "epoch": 0.0063831622530837575,
+      "grad_norm": 3.546875,
+      "learning_rate": 1.48e-05,
+      "loss": 1.968,
+      "step": 37
+    },
+    {
+      "epoch": 0.006555680151815751,
+      "grad_norm": 44.75,
+      "learning_rate": 1.5200000000000002e-05,
+      "loss": 2.0737,
+      "step": 38
+    },
+    {
+      "epoch": 0.006728198050547744,
+      "grad_norm": 5.6875,
+      "learning_rate": 1.5600000000000003e-05,
+      "loss": 2.0571,
+      "step": 39
+    },
+    {
+      "epoch": 0.006900715949279738,
+      "grad_norm": 5.40625,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 2.0511,
+      "step": 40
+    },
+    {
+      "epoch": 0.007073233848011731,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.64e-05,
+      "loss": 1.996,
+      "step": 41
+    },
+    {
+      "epoch": 0.007245751746743725,
+      "grad_norm": 5.21875,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 2.0255,
+      "step": 42
+    },
+    {
+      "epoch": 0.007418269645475718,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.72e-05,
+      "loss": 1.9345,
+      "step": 43
+    },
+    {
+      "epoch": 0.007590787544207712,
+      "grad_norm": 4.25,
+      "learning_rate": 1.76e-05,
+      "loss": 2.069,
+      "step": 44
+    },
+    {
+      "epoch": 0.007763305442939705,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.8e-05,
+      "loss": 2.0591,
+      "step": 45
+    },
+    {
+      "epoch": 0.007935823341671698,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.8400000000000003e-05,
+      "loss": 2.0095,
+      "step": 46
+    },
+    {
+      "epoch": 0.008108341240403692,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.88e-05,
+      "loss": 1.9536,
+      "step": 47
+    },
+    {
+      "epoch": 0.008280859139135685,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 2.0756,
+      "step": 48
+    },
+    {
+      "epoch": 0.00845337703786768,
+      "grad_norm": 2.625,
+      "learning_rate": 1.9600000000000002e-05,
+      "loss": 1.878,
+      "step": 49
+    },
+    {
+      "epoch": 0.008625894936599671,
+      "grad_norm": 6.1875,
+      "learning_rate": 2e-05,
+      "loss": 1.8865,
+      "step": 50
+    },
+    {
+      "epoch": 0.008798412835331665,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.999999962956876e-05,
+      "loss": 1.8895,
+      "step": 51
+    },
+    {
+      "epoch": 0.008970930734063659,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.9999998518275062e-05,
+      "loss": 1.97,
+      "step": 52
+    },
+    {
+      "epoch": 0.009143448632795653,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.9999996666118996e-05,
+      "loss": 1.9397,
+      "step": 53
+    },
+    {
+      "epoch": 0.009315966531527647,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.9999994073100687e-05,
+      "loss": 1.8746,
+      "step": 54
+    },
+    {
+      "epoch": 0.009488484430259639,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9999990739220338e-05,
+      "loss": 1.8565,
+      "step": 55
+    },
+    {
+      "epoch": 0.009661002328991633,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.999998666447819e-05,
+      "loss": 1.8719,
+      "step": 56
+    },
+    {
+      "epoch": 0.009833520227723627,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.999998184887455e-05,
+      "loss": 1.8919,
+      "step": 57
+    },
+    {
+      "epoch": 0.01000603812645562,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.999997629240977e-05,
+      "loss": 1.8949,
+      "step": 58
+    },
+    {
+      "epoch": 0.010178556025187612,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9999969995084264e-05,
+      "loss": 1.8742,
+      "step": 59
+    },
+    {
+      "epoch": 0.010351073923919606,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.99999629568985e-05,
+      "loss": 1.7554,
+      "step": 60
+    },
+    {
+      "epoch": 0.0105235918226516,
+      "grad_norm": 2.875,
+      "learning_rate": 1.9999955177852994e-05,
+      "loss": 1.8714,
+      "step": 61
+    },
+    {
+      "epoch": 0.010696109721383594,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.999994665794833e-05,
+      "loss": 1.776,
+      "step": 62
+    },
+    {
+      "epoch": 0.010868627620115588,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.999993739718513e-05,
+      "loss": 1.8771,
+      "step": 63
+    },
+    {
+      "epoch": 0.01104114551884758,
+      "grad_norm": 1.5,
+      "learning_rate": 1.9999927395564087e-05,
+      "loss": 1.7822,
+      "step": 64
+    },
+    {
+      "epoch": 0.011213663417579574,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.999991665308594e-05,
+      "loss": 1.9068,
+      "step": 65
+    },
+    {
+      "epoch": 0.011386181316311568,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.9999905169751486e-05,
+      "loss": 1.999,
+      "step": 66
+    },
+    {
+      "epoch": 0.011558699215043561,
+      "grad_norm": 1.5,
+      "learning_rate": 1.9999892945561578e-05,
+      "loss": 1.8624,
+      "step": 67
+    },
+    {
+      "epoch": 0.011731217113775554,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.999987998051711e-05,
+      "loss": 1.9439,
+      "step": 68
+    },
+    {
+      "epoch": 0.011903735012507547,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.9999866274619057e-05,
+      "loss": 1.8826,
+      "step": 69
+    },
+    {
+      "epoch": 0.012076252911239541,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.9999851827868428e-05,
+      "loss": 1.7159,
+      "step": 70
+    },
+    {
+      "epoch": 0.012248770809971535,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.9999836640266292e-05,
+      "loss": 1.8327,
+      "step": 71
+    },
+    {
+      "epoch": 0.012421288708703527,
+      "grad_norm": 1.265625,
+      "learning_rate": 1.9999820711813776e-05,
+      "loss": 1.9363,
+      "step": 72
+    },
+    {
+      "epoch": 0.012593806607435521,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.999980404251206e-05,
+      "loss": 1.8603,
+      "step": 73
+    },
+    {
+      "epoch": 0.012766324506167515,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.999978663236238e-05,
+      "loss": 1.8476,
+      "step": 74
+    },
+    {
+      "epoch": 0.012938842404899509,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.9999768481366026e-05,
+      "loss": 1.786,
+      "step": 75
+    },
+    {
+      "epoch": 0.013111360303631503,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.9999749589524338e-05,
+      "loss": 1.8373,
+      "step": 76
+    },
+    {
+      "epoch": 0.013283878202363495,
+      "grad_norm": 1.328125,
+      "learning_rate": 1.999972995683872e-05,
+      "loss": 1.7419,
+      "step": 77
+    },
+    {
+      "epoch": 0.013456396101095489,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9999709583310624e-05,
+      "loss": 1.8733,
+      "step": 78
+    },
+    {
+      "epoch": 0.013628913999827482,
+      "grad_norm": 3.578125,
+      "learning_rate": 1.9999688468941565e-05,
+      "loss": 1.7432,
+      "step": 79
+    },
+    {
+      "epoch": 0.013801431898559476,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.9999666613733102e-05,
+      "loss": 1.8525,
+      "step": 80
+    },
+    {
+      "epoch": 0.013973949797291468,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.9999644017686855e-05,
+      "loss": 1.8241,
+      "step": 81
+    },
+    {
+      "epoch": 0.014146467696023462,
+      "grad_norm": 1.390625,
+      "learning_rate": 1.9999620680804495e-05,
+      "loss": 1.768,
+      "step": 82
+    },
+    {
+      "epoch": 0.014318985594755456,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.999959660308776e-05,
+      "loss": 1.8484,
+      "step": 83
+    },
+    {
+      "epoch": 0.01449150349348745,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.9999571784538428e-05,
+      "loss": 1.8453,
+      "step": 84
+    },
+    {
+      "epoch": 0.014664021392219442,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.9999546225158335e-05,
+      "loss": 1.8202,
+      "step": 85
+    },
+    {
+      "epoch": 0.014836539290951436,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.9999519924949376e-05,
+      "loss": 1.8652,
+      "step": 86
+    },
+    {
+      "epoch": 0.01500905718968343,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.9999492883913506e-05,
+      "loss": 1.7482,
+      "step": 87
+    },
+    {
+      "epoch": 0.015181575088415424,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.999946510205272e-05,
+      "loss": 1.7646,
+      "step": 88
+    },
+    {
+      "epoch": 0.015354092987147417,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.999943657936908e-05,
+      "loss": 1.8786,
+      "step": 89
+    },
+    {
+      "epoch": 0.01552661088587941,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.99994073158647e-05,
+      "loss": 1.7141,
+      "step": 90
+    },
+    {
+      "epoch": 0.015699128784611403,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.9999377311541748e-05,
+      "loss": 1.8988,
+      "step": 91
+    },
+    {
+      "epoch": 0.015871646683343395,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.9999346566402444e-05,
+      "loss": 1.8017,
+      "step": 92
+    },
+    {
+      "epoch": 0.01604416458207539,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.999931508044907e-05,
+      "loss": 1.8711,
+      "step": 93
+    },
+    {
+      "epoch": 0.016216682480807383,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.999928285368395e-05,
+      "loss": 1.7997,
+      "step": 94
+    },
+    {
+      "epoch": 0.01638920037953938,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.9999249886109485e-05,
+      "loss": 1.7473,
+      "step": 95
+    },
+    {
+      "epoch": 0.01656171827827137,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.9999216177728106e-05,
+      "loss": 1.9087,
+      "step": 96
+    },
+    {
+      "epoch": 0.016734236177003363,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.9999181728542316e-05,
+      "loss": 1.8097,
+      "step": 97
+    },
+    {
+      "epoch": 0.01690675407573536,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.9999146538554663e-05,
+      "loss": 1.7901,
+      "step": 98
+    },
+    {
+      "epoch": 0.01707927197446735,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.9999110607767763e-05,
+      "loss": 1.8119,
+      "step": 99
+    },
+    {
+      "epoch": 0.017251789873199343,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.999907393618427e-05,
+      "loss": 1.7627,
+      "step": 100
+    },
+    {
+      "epoch": 0.017251789873199343,
+      "eval_loss": 1.7699980735778809,
+      "eval_runtime": 10.881,
+      "eval_samples_per_second": 94.109,
+      "eval_steps_per_second": 23.527,
+      "step": 100
+    },
+    {
+      "epoch": 0.01742430777193134,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.9999036523806897e-05,
+      "loss": 1.8264,
+      "step": 101
+    },
+    {
+      "epoch": 0.01759682567066333,
+      "grad_norm": 1.0,
+      "learning_rate": 1.9998998370638427e-05,
+      "loss": 1.9073,
+      "step": 102
+    },
+    {
+      "epoch": 0.017769343569395326,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.9998959476681676e-05,
+      "loss": 1.7857,
+      "step": 103
+    },
+    {
+      "epoch": 0.017941861468127318,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.9998919841939536e-05,
+      "loss": 1.8095,
+      "step": 104
+    },
+    {
+      "epoch": 0.01811437936685931,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.9998879466414937e-05,
+      "loss": 1.7011,
+      "step": 105
+    },
+    {
+      "epoch": 0.018286897265591306,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.9998838350110867e-05,
+      "loss": 1.7574,
+      "step": 106
+    },
+    {
+      "epoch": 0.018459415164323298,
+      "grad_norm": 1.375,
+      "learning_rate": 1.9998796493030382e-05,
+      "loss": 1.7882,
+      "step": 107
+    },
+    {
+      "epoch": 0.018631933063055293,
+      "grad_norm": 6.75,
+      "learning_rate": 1.9998753895176576e-05,
+      "loss": 1.7849,
+      "step": 108
+    },
+    {
+      "epoch": 0.018804450961787286,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.9998710556552603e-05,
+      "loss": 1.7914,
+      "step": 109
+    },
+    {
+      "epoch": 0.018976968860519278,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.9998666477161678e-05,
+      "loss": 1.7382,
+      "step": 110
+    },
+    {
+      "epoch": 0.019149486759251273,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.9998621657007068e-05,
+      "loss": 1.7785,
+      "step": 111
+    },
+    {
+      "epoch": 0.019322004657983265,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.9998576096092093e-05,
+      "loss": 1.7812,
+      "step": 112
+    },
+    {
+      "epoch": 0.01949452255671526,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.9998529794420124e-05,
+      "loss": 1.7871,
+      "step": 113
+    },
+    {
+      "epoch": 0.019667040455447253,
+      "grad_norm": 1.125,
+      "learning_rate": 1.9998482751994596e-05,
+      "loss": 1.6526,
+      "step": 114
+    },
+    {
+      "epoch": 0.019839558354179245,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.999843496881899e-05,
+      "loss": 1.7308,
+      "step": 115
+    },
+    {
+      "epoch": 0.02001207625291124,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.999838644489685e-05,
+      "loss": 1.7452,
+      "step": 116
+    },
+    {
+      "epoch": 0.020184594151643233,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.9998337180231768e-05,
+      "loss": 1.7223,
+      "step": 117
+    },
+    {
+      "epoch": 0.020357112050375225,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.9998287174827396e-05,
+      "loss": 1.7875,
+      "step": 118
+    },
+    {
+      "epoch": 0.02052962994910722,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.999823642868744e-05,
+      "loss": 1.7478,
+      "step": 119
+    },
+    {
+      "epoch": 0.020702147847839213,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.9998184941815653e-05,
+      "loss": 1.7594,
+      "step": 120
+    },
+    {
+      "epoch": 0.020874665746571208,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.9998132714215855e-05,
+      "loss": 1.8166,
+      "step": 121
+    },
+    {
+      "epoch": 0.0210471836453032,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.9998079745891918e-05,
+      "loss": 1.7075,
+      "step": 122
+    },
+    {
+      "epoch": 0.021219701544035192,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.999802603684776e-05,
+      "loss": 1.8377,
+      "step": 123
+    },
+    {
+      "epoch": 0.021392219442767188,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.999797158708736e-05,
+      "loss": 1.7289,
+      "step": 124
+    },
+    {
+      "epoch": 0.02156473734149918,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.999791639661476e-05,
+      "loss": 1.8432,
+      "step": 125
+    },
+    {
+      "epoch": 0.021737255240231176,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.999786046543404e-05,
+      "loss": 1.7785,
+      "step": 126
+    },
+    {
+      "epoch": 0.021909773138963168,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.9997803793549347e-05,
+      "loss": 1.7906,
+      "step": 127
+    },
+    {
+      "epoch": 0.02208229103769516,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.999774638096488e-05,
+      "loss": 1.837,
+      "step": 128
+    },
+    {
+      "epoch": 0.022254808936427155,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.9997688227684896e-05,
+      "loss": 1.754,
+      "step": 129
+    },
+    {
+      "epoch": 0.022427326835159148,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.9997629333713697e-05,
+      "loss": 1.8045,
+      "step": 130
+    },
+    {
+      "epoch": 0.02259984473389114,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.999756969905565e-05,
+      "loss": 1.7617,
+      "step": 131
+    },
+    {
+      "epoch": 0.022772362632623135,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.9997509323715166e-05,
+      "loss": 1.8347,
+      "step": 132
+    },
+    {
+      "epoch": 0.022944880531355127,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.999744820769673e-05,
+      "loss": 1.7765,
+      "step": 133
+    },
+    {
+      "epoch": 0.023117398430087123,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.9997386351004864e-05,
+      "loss": 1.8108,
+      "step": 134
+    },
+    {
+      "epoch": 0.023289916328819115,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.9997323753644148e-05,
+      "loss": 1.7178,
+      "step": 135
+    },
+    {
+      "epoch": 0.023462434227551107,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9997260415619223e-05,
+      "loss": 1.7446,
+      "step": 136
+    },
+    {
+      "epoch": 0.023634952126283103,
+      "grad_norm": 1.2421875,
+      "learning_rate": 1.999719633693478e-05,
+      "loss": 1.8163,
+      "step": 137
+    },
+    {
+      "epoch": 0.023807470025015095,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.999713151759557e-05,
+      "loss": 1.7499,
+      "step": 138
+    },
+    {
+      "epoch": 0.02397998792374709,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.999706595760639e-05,
+      "loss": 1.8029,
+      "step": 139
+    },
+    {
+      "epoch": 0.024152505822479083,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.99969996569721e-05,
+      "loss": 1.6801,
+      "step": 140
+    },
+    {
+      "epoch": 0.024325023721211075,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.999693261569761e-05,
+      "loss": 1.7381,
+      "step": 141
+    },
+    {
+      "epoch": 0.02449754161994307,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.999686483378789e-05,
+      "loss": 1.828,
+      "step": 142
+    },
+    {
+      "epoch": 0.024670059518675062,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.9996796311247956e-05,
+      "loss": 1.7018,
+      "step": 143
+    },
+    {
+      "epoch": 0.024842577417407054,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.999672704808289e-05,
+      "loss": 1.8166,
+      "step": 144
+    },
+    {
+      "epoch": 0.02501509531613905,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.9996657044297824e-05,
+      "loss": 1.7758,
+      "step": 145
+    },
+    {
+      "epoch": 0.025187613214871042,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.9996586299897944e-05,
+      "loss": 1.8275,
+      "step": 146
+    },
+    {
+      "epoch": 0.025360131113603038,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.9996514814888483e-05,
+      "loss": 1.7491,
+      "step": 147
+    },
+    {
+      "epoch": 0.02553264901233503,
+      "grad_norm": 1.125,
+      "learning_rate": 1.999644258927475e-05,
+      "loss": 1.8093,
+      "step": 148
+    },
+    {
+      "epoch": 0.025705166911067022,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.9996369623062083e-05,
+      "loss": 1.7722,
+      "step": 149
+    },
+    {
+      "epoch": 0.025877684809799018,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.9996295916255898e-05,
+      "loss": 1.6843,
+      "step": 150
+    },
+    {
+      "epoch": 0.02605020270853101,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.999622146886165e-05,
+      "loss": 1.6833,
+      "step": 151
+    },
+    {
+      "epoch": 0.026222720607263005,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.999614628088486e-05,
+      "loss": 1.779,
+      "step": 152
+    },
+    {
+      "epoch": 0.026395238505994997,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.999607035233109e-05,
+      "loss": 1.7611,
+      "step": 153
+    },
+    {
+      "epoch": 0.02656775640472699,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.999599368320597e-05,
+      "loss": 1.6703,
+      "step": 154
+    },
+    {
+      "epoch": 0.026740274303458985,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9995916273515185e-05,
+      "loss": 1.731,
+      "step": 155
+    },
+    {
+      "epoch": 0.026912792202190977,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.9995838123264463e-05,
+      "loss": 1.746,
+      "step": 156
+    },
+    {
+      "epoch": 0.02708531010092297,
+      "grad_norm": 0.875,
+      "learning_rate": 1.9995759232459594e-05,
+      "loss": 1.6949,
+      "step": 157
+    },
+    {
+      "epoch": 0.027257827999654965,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.9995679601106426e-05,
+      "loss": 1.7415,
+      "step": 158
+    },
+    {
+      "epoch": 0.027430345898386957,
+      "grad_norm": 1.1953125,
+      "learning_rate": 1.9995599229210855e-05,
+      "loss": 1.6621,
+      "step": 159
+    },
+    {
+      "epoch": 0.027602863797118952,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.999551811677884e-05,
+      "loss": 1.6831,
+      "step": 160
+    },
+    {
+      "epoch": 0.027775381695850945,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.9995436263816387e-05,
+      "loss": 1.7927,
+      "step": 161
+    },
+    {
+      "epoch": 0.027947899594582937,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.9995353670329565e-05,
+      "loss": 1.8223,
+      "step": 162
+    },
+    {
+      "epoch": 0.028120417493314932,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.9995270336324487e-05,
+      "loss": 1.7593,
+      "step": 163
+    },
+    {
+      "epoch": 0.028292935392046924,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.9995186261807326e-05,
+      "loss": 1.7412,
+      "step": 164
+    },
+    {
+      "epoch": 0.02846545329077892,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.9995101446784317e-05,
+      "loss": 1.7231,
+      "step": 165
+    },
+    {
+      "epoch": 0.028637971189510912,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.999501589126174e-05,
+      "loss": 1.6749,
+      "step": 166
+    },
+    {
+      "epoch": 0.028810489088242904,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.9994929595245932e-05,
+      "loss": 1.7105,
+      "step": 167
+    },
+    {
+      "epoch": 0.0289830069869749,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.9994842558743293e-05,
+      "loss": 1.7118,
+      "step": 168
+    },
+    {
+      "epoch": 0.029155524885706892,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.9994754781760264e-05,
+      "loss": 1.6403,
+      "step": 169
+    },
+    {
+      "epoch": 0.029328042784438884,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.9994666264303347e-05,
+      "loss": 1.7117,
+      "step": 170
+    },
+    {
+      "epoch": 0.02950056068317088,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.999457700637911e-05,
+      "loss": 1.6726,
+      "step": 171
+    },
+    {
+      "epoch": 0.02967307858190287,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.9994487007994156e-05,
+      "loss": 1.7028,
+      "step": 172
+    },
+    {
+      "epoch": 0.029845596480634867,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.9994396269155153e-05,
+      "loss": 1.7862,
+      "step": 173
+    },
+    {
+      "epoch": 0.03001811437936686,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.9994304789868832e-05,
+      "loss": 1.7097,
+      "step": 174
+    },
+    {
+      "epoch": 0.03019063227809885,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.999421257014196e-05,
+      "loss": 1.7168,
+      "step": 175
+    },
+    {
+      "epoch": 0.030363150176830847,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.999411960998138e-05,
+      "loss": 1.8136,
+      "step": 176
+    },
+    {
+      "epoch": 0.03053566807556284,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.999402590939397e-05,
+      "loss": 1.6356,
+      "step": 177
+    },
+    {
+      "epoch": 0.030708185974294835,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.9993931468386675e-05,
+      "loss": 1.7556,
+      "step": 178
+    },
+    {
+      "epoch": 0.030880703873026827,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.9993836286966492e-05,
+      "loss": 1.7139,
+      "step": 179
+    },
+    {
+      "epoch": 0.03105322177175882,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.999374036514047e-05,
+      "loss": 1.7707,
+      "step": 180
+    },
+    {
+      "epoch": 0.031225739670490815,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.999364370291572e-05,
+      "loss": 1.6573,
+      "step": 181
+    },
+    {
+      "epoch": 0.03139825756922281,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.9993546300299404e-05,
+      "loss": 1.8058,
+      "step": 182
+    },
+    {
+      "epoch": 0.0315707754679548,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9993448157298733e-05,
+      "loss": 1.7508,
+      "step": 183
+    },
+    {
+      "epoch": 0.03174329336668679,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.999334927392098e-05,
+      "loss": 1.7117,
+      "step": 184
+    },
+    {
+      "epoch": 0.031915811265418786,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.9993249650173475e-05,
+      "loss": 1.6337,
+      "step": 185
+    },
+    {
+      "epoch": 0.03208832916415078,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.999314928606359e-05,
+      "loss": 1.7379,
+      "step": 186
+    },
+    {
+      "epoch": 0.03226084706288277,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.999304818159877e-05,
+      "loss": 1.7688,
+      "step": 187
+    },
+    {
+      "epoch": 0.032433364961614766,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.9992946336786502e-05,
+      "loss": 1.6814,
+      "step": 188
+    },
+    {
+      "epoch": 0.03260588286034676,
+      "grad_norm": 0.75,
+      "learning_rate": 1.9992843751634327e-05,
+      "loss": 1.6128,
+      "step": 189
+    },
+    {
+      "epoch": 0.03277840075907876,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9992740426149847e-05,
+      "loss": 1.7347,
+      "step": 190
+    },
+    {
+      "epoch": 0.032950918657810746,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.9992636360340722e-05,
+      "loss": 1.8182,
+      "step": 191
+    },
+    {
+      "epoch": 0.03312343655654274,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.999253155421466e-05,
+      "loss": 1.6953,
+      "step": 192
+    },
+    {
+      "epoch": 0.03329595445527474,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.999242600777942e-05,
+      "loss": 1.7346,
+      "step": 193
+    },
+    {
+      "epoch": 0.033468472354006726,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.999231972104283e-05,
+      "loss": 1.7494,
+      "step": 194
+    },
+    {
+      "epoch": 0.03364099025273872,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.9992212694012757e-05,
+      "loss": 1.7586,
+      "step": 195
+    },
+    {
+      "epoch": 0.03381350815147072,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.9992104926697137e-05,
+      "loss": 1.7297,
+      "step": 196
+    },
+    {
+      "epoch": 0.033986026050202706,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9991996419103947e-05,
+      "loss": 1.6872,
+      "step": 197
+    },
+    {
+      "epoch": 0.0341585439489347,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.999188717124123e-05,
+      "loss": 1.7465,
+      "step": 198
+    },
+    {
+      "epoch": 0.0343310618476667,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.999177718311708e-05,
+      "loss": 1.7722,
+      "step": 199
+    },
+    {
+      "epoch": 0.034503579746398685,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.9991666454739644e-05,
+      "loss": 1.696,
+      "step": 200
+    },
+    {
+      "epoch": 0.034503579746398685,
+      "eval_loss": 1.701870322227478,
+      "eval_runtime": 11.0727,
+      "eval_samples_per_second": 92.48,
+      "eval_steps_per_second": 23.12,
+      "step": 200
+    },
+    {
+      "epoch": 0.03467609764513068,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9991554986117127e-05,
+      "loss": 1.6589,
+      "step": 201
+    },
+    {
+      "epoch": 0.03484861554386268,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.9991442777257786e-05,
+      "loss": 1.7258,
+      "step": 202
+    },
+    {
+      "epoch": 0.03502113344259467,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.9991329828169936e-05,
+      "loss": 1.75,
+      "step": 203
+    },
+    {
+      "epoch": 0.03519365134132666,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.999121613886194e-05,
+      "loss": 1.5888,
+      "step": 204
+    },
+    {
+      "epoch": 0.035366169240058656,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.9991101709342228e-05,
+      "loss": 1.6331,
+      "step": 205
+    },
+    {
+      "epoch": 0.03553868713879065,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.9990986539619274e-05,
+      "loss": 1.7844,
+      "step": 206
+    },
+    {
+      "epoch": 0.03571120503752264,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.999087062970161e-05,
+      "loss": 1.68,
+      "step": 207
+    },
+    {
+      "epoch": 0.035883722936254636,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.999075397959782e-05,
+      "loss": 1.7034,
+      "step": 208
+    },
+    {
+      "epoch": 0.03605624083498663,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.9990636589316556e-05,
+      "loss": 1.7639,
+      "step": 209
+    },
+    {
+      "epoch": 0.03622875873371862,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.9990518458866506e-05,
+      "loss": 1.687,
+      "step": 210
+    },
+    {
+      "epoch": 0.036401276632450616,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.9990399588256425e-05,
+      "loss": 1.6867,
+      "step": 211
+    },
+    {
+      "epoch": 0.03657379453118261,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.999027997749512e-05,
+      "loss": 1.7076,
+      "step": 212
+    },
+    {
+      "epoch": 0.0367463124299146,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.9990159626591455e-05,
+      "loss": 1.6697,
+      "step": 213
+    },
+    {
+      "epoch": 0.036918830328646596,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.999003853555434e-05,
+      "loss": 1.7503,
+      "step": 214
+    },
+    {
+      "epoch": 0.03709134822737859,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9989916704392748e-05,
+      "loss": 1.7105,
+      "step": 215
+    },
+    {
+      "epoch": 0.03726386612611059,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.998979413311571e-05,
+      "loss": 1.7114,
+      "step": 216
+    },
+    {
+      "epoch": 0.037436384024842576,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.99896708217323e-05,
+      "loss": 1.6787,
+      "step": 217
+    },
+    {
+      "epoch": 0.03760890192357457,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.9989546770251662e-05,
+      "loss": 1.714,
+      "step": 218
+    },
+    {
+      "epoch": 0.03778141982230657,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.9989421978682978e-05,
+      "loss": 1.7059,
+      "step": 219
+    },
+    {
+      "epoch": 0.037953937721038555,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.99892964470355e-05,
+      "loss": 1.702,
+      "step": 220
+    },
+    {
+      "epoch": 0.03812645561977055,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9989170175318523e-05,
+      "loss": 1.6866,
+      "step": 221
+    },
+    {
+      "epoch": 0.038298973518502546,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.9989043163541403e-05,
+      "loss": 1.7692,
+      "step": 222
+    },
+    {
+      "epoch": 0.038471491417234535,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9988915411713553e-05,
+      "loss": 1.7088,
+      "step": 223
+    },
+    {
+      "epoch": 0.03864400931596653,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9988786919844437e-05,
+      "loss": 1.6951,
+      "step": 224
+    },
+    {
+      "epoch": 0.038816527214698526,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.998865768794357e-05,
+      "loss": 1.6612,
+      "step": 225
+    },
+    {
+      "epoch": 0.03898904511343052,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.9988527716020532e-05,
+      "loss": 1.6171,
+      "step": 226
+    },
+    {
+      "epoch": 0.03916156301216251,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.998839700408495e-05,
+      "loss": 1.7984,
+      "step": 227
+    },
+    {
+      "epoch": 0.039334080910894506,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9988265552146508e-05,
+      "loss": 1.6983,
+      "step": 228
+    },
+    {
+      "epoch": 0.0395065988096265,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.998813336021494e-05,
+      "loss": 1.6833,
+      "step": 229
+    },
+    {
+      "epoch": 0.03967911670835849,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.9988000428300047e-05,
+      "loss": 1.6599,
+      "step": 230
+    },
+    {
+      "epoch": 0.039851634607090486,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.9987866756411676e-05,
+      "loss": 1.7222,
+      "step": 231
+    },
+    {
+      "epoch": 0.04002415250582248,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.9987732344559724e-05,
+      "loss": 1.7105,
+      "step": 232
+    },
+    {
+      "epoch": 0.04019667040455447,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.9987597192754155e-05,
+      "loss": 1.671,
+      "step": 233
+    },
+    {
+      "epoch": 0.040369188303286466,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.9987461301004984e-05,
+      "loss": 1.6531,
+      "step": 234
+    },
+    {
+      "epoch": 0.04054170620201846,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.9987324669322274e-05,
+      "loss": 1.7921,
+      "step": 235
+    },
+    {
+      "epoch": 0.04071422410075045,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.9987187297716145e-05,
+      "loss": 1.8698,
+      "step": 236
+    },
+    {
+      "epoch": 0.040886741999482445,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.9987049186196782e-05,
+      "loss": 1.7407,
+      "step": 237
+    },
+    {
+      "epoch": 0.04105925989821444,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.9986910334774415e-05,
+      "loss": 1.6458,
+      "step": 238
+    },
+    {
+      "epoch": 0.04123177779694644,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.9986770743459325e-05,
+      "loss": 1.6818,
+      "step": 239
+    },
+    {
+      "epoch": 0.041404295695678425,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.9986630412261857e-05,
+      "loss": 1.6233,
+      "step": 240
+    },
+    {
+      "epoch": 0.04157681359441042,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9986489341192416e-05,
+      "loss": 1.649,
+      "step": 241
+    },
+    {
+      "epoch": 0.041749331493142416,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.998634753026144e-05,
+      "loss": 1.7436,
+      "step": 242
+    },
+    {
+      "epoch": 0.041921849391874405,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.9986204979479443e-05,
+      "loss": 1.6854,
+      "step": 243
+    },
+    {
+      "epoch": 0.0420943672906064,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.9986061688856983e-05,
+      "loss": 1.6537,
+      "step": 244
+    },
+    {
+      "epoch": 0.042266885189338396,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9985917658404677e-05,
+      "loss": 1.767,
+      "step": 245
+    },
+    {
+      "epoch": 0.042439403088070385,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.99857728881332e-05,
+      "loss": 1.68,
+      "step": 246
+    },
+    {
+      "epoch": 0.04261192098680238,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.998562737805327e-05,
+      "loss": 1.6439,
+      "step": 247
+    },
+    {
+      "epoch": 0.042784438885534376,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.9985481128175673e-05,
+      "loss": 1.8946,
+      "step": 248
+    },
+    {
+      "epoch": 0.042956956784266365,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.998533413851124e-05,
+      "loss": 1.6761,
+      "step": 249
+    },
+    {
+      "epoch": 0.04312947468299836,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9985186409070863e-05,
+      "loss": 1.6496,
+      "step": 250
+    },
+    {
+      "epoch": 0.043301992581730356,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.998503793986549e-05,
+      "loss": 1.7303,
+      "step": 251
+    },
+    {
+      "epoch": 0.04347451048046235,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.9984888730906112e-05,
+      "loss": 1.7383,
+      "step": 252
+    },
+    {
+      "epoch": 0.04364702837919434,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.998473878220379e-05,
+      "loss": 1.5984,
+      "step": 253
+    },
+    {
+      "epoch": 0.043819546277926336,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.9984588093769633e-05,
+      "loss": 1.6765,
+      "step": 254
+    },
+    {
+      "epoch": 0.04399206417665833,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.99844366656148e-05,
+      "loss": 1.7208,
+      "step": 255
+    },
+    {
+      "epoch": 0.04416458207539032,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.9984284497750516e-05,
+      "loss": 1.7428,
+      "step": 256
+    },
+    {
+      "epoch": 0.044337099974122315,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.998413159018805e-05,
+      "loss": 1.7136,
+      "step": 257
+    },
+    {
+      "epoch": 0.04450961787285431,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.9983977942938735e-05,
+      "loss": 1.7301,
+      "step": 258
+    },
+    {
+      "epoch": 0.0446821357715863,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.998382355601395e-05,
+      "loss": 1.5718,
+      "step": 259
+    },
+    {
+      "epoch": 0.044854653670318295,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.998366842942513e-05,
+      "loss": 1.6996,
+      "step": 260
+    },
+    {
+      "epoch": 0.04502717156905029,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.998351256318378e-05,
+      "loss": 1.6608,
+      "step": 261
+    },
+    {
+      "epoch": 0.04519968946778228,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.998335595730143e-05,
+      "loss": 1.6553,
+      "step": 262
+    },
+    {
+      "epoch": 0.045372207366514275,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.99831986117897e-05,
+      "loss": 1.6318,
+      "step": 263
+    },
+    {
+      "epoch": 0.04554472526524627,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9983040526660236e-05,
+      "loss": 1.6407,
+      "step": 264
+    },
+    {
+      "epoch": 0.045717243163978266,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.998288170192475e-05,
+      "loss": 1.6473,
+      "step": 265
+    },
+    {
+      "epoch": 0.045889761062710255,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.9982722137595015e-05,
+      "loss": 1.7061,
+      "step": 266
+    },
+    {
+      "epoch": 0.04606227896144225,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.9982561833682848e-05,
+      "loss": 1.6742,
+      "step": 267
+    },
+    {
+      "epoch": 0.046234796860174246,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.9982400790200127e-05,
+      "loss": 1.6009,
+      "step": 268
+    },
+    {
+      "epoch": 0.046407314758906235,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.9982239007158784e-05,
+      "loss": 1.6577,
+      "step": 269
+    },
+    {
+      "epoch": 0.04657983265763823,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.99820764845708e-05,
+      "loss": 1.7364,
+      "step": 270
+    },
+    {
+      "epoch": 0.046752350556370226,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.998191322244822e-05,
+      "loss": 1.6277,
+      "step": 271
+    },
+    {
+      "epoch": 0.046924868455102214,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.9981749220803142e-05,
+      "loss": 1.6939,
+      "step": 272
+    },
+    {
+      "epoch": 0.04709738635383421,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.998158447964771e-05,
+      "loss": 1.7526,
+      "step": 273
+    },
+    {
+      "epoch": 0.047269904252566206,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.998141899899413e-05,
+      "loss": 1.6889,
+      "step": 274
+    },
+    {
+      "epoch": 0.047442422151298194,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.998125277885467e-05,
+      "loss": 1.6421,
+      "step": 275
+    },
+    {
+      "epoch": 0.04761494005003019,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.998108581924163e-05,
+      "loss": 1.6843,
+      "step": 276
+    },
+    {
+      "epoch": 0.047787457948762185,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.9980918120167395e-05,
+      "loss": 1.6682,
+      "step": 277
+    },
+    {
+      "epoch": 0.04795997584749418,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.9980749681644378e-05,
+      "loss": 1.7816,
+      "step": 278
+    },
+    {
+      "epoch": 0.04813249374622617,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.9980580503685064e-05,
+      "loss": 1.6331,
+      "step": 279
+    },
+    {
+      "epoch": 0.048305011644958165,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.9980410586301983e-05,
+      "loss": 1.7004,
+      "step": 280
+    },
+    {
+      "epoch": 0.04847752954369016,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.9980239929507724e-05,
+      "loss": 1.7132,
+      "step": 281
+    },
+    {
+      "epoch": 0.04865004744242215,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.9980068533314937e-05,
+      "loss": 1.6185,
+      "step": 282
+    },
+    {
+      "epoch": 0.048822565341154145,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.9979896397736308e-05,
+      "loss": 1.6983,
+      "step": 283
+    },
+    {
+      "epoch": 0.04899508323988614,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.9979723522784602e-05,
+      "loss": 1.6439,
+      "step": 284
+    },
+    {
+      "epoch": 0.04916760113861813,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.997954990847262e-05,
+      "loss": 1.66,
+      "step": 285
+    },
+    {
+      "epoch": 0.049340119037350125,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9979375554813223e-05,
+      "loss": 1.704,
+      "step": 286
+    },
+    {
+      "epoch": 0.04951263693608212,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.9979200461819334e-05,
+      "loss": 1.7866,
+      "step": 287
+    },
+    {
+      "epoch": 0.04968515483481411,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.997902462950392e-05,
+      "loss": 1.6967,
+      "step": 288
+    },
+    {
+      "epoch": 0.049857672733546105,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9978848057880008e-05,
+      "loss": 1.6726,
+      "step": 289
+    },
+    {
+      "epoch": 0.0500301906322781,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.9978670746960687e-05,
+      "loss": 1.7372,
+      "step": 290
+    },
+    {
+      "epoch": 0.050202708531010096,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.9978492696759084e-05,
+      "loss": 1.714,
+      "step": 291
+    },
+    {
+      "epoch": 0.050375226429742084,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.9978313907288395e-05,
+      "loss": 1.693,
+      "step": 292
+    },
+    {
+      "epoch": 0.05054774432847408,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.997813437856186e-05,
+      "loss": 1.6223,
+      "step": 293
+    },
+    {
+      "epoch": 0.050720262227206075,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.9977954110592787e-05,
+      "loss": 1.7357,
+      "step": 294
+    },
+    {
+      "epoch": 0.050892780125938064,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.9977773103394527e-05,
+      "loss": 1.6369,
+      "step": 295
+    },
+    {
+      "epoch": 0.05106529802467006,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9977591356980493e-05,
+      "loss": 1.7504,
+      "step": 296
+    },
+    {
+      "epoch": 0.051237815923402055,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.997740887136415e-05,
+      "loss": 1.7014,
+      "step": 297
+    },
+    {
+      "epoch": 0.051410333822134044,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.9977225646559013e-05,
+      "loss": 1.6612,
+      "step": 298
+    },
+    {
+      "epoch": 0.05158285172086604,
+      "grad_norm": 0.75,
+      "learning_rate": 1.9977041682578662e-05,
+      "loss": 1.773,
+      "step": 299
+    },
+    {
+      "epoch": 0.051755369619598035,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.997685697943672e-05,
+      "loss": 1.6947,
+      "step": 300
+    },
+    {
+      "epoch": 0.051755369619598035,
+      "eval_loss": 1.6562981605529785,
+      "eval_runtime": 10.8538,
+      "eval_samples_per_second": 94.345,
+      "eval_steps_per_second": 23.586,
+      "step": 300
+    },
+    {
+      "epoch": 0.051927887518330024,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.9976671537146877e-05,
+      "loss": 1.7933,
+      "step": 301
+    },
+    {
+      "epoch": 0.05210040541706202,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.997648535572287e-05,
+      "loss": 1.6316,
+      "step": 302
+    },
+    {
+      "epoch": 0.052272923315794015,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9976298435178493e-05,
+      "loss": 1.6905,
+      "step": 303
+    },
+    {
+      "epoch": 0.05244544121452601,
+      "grad_norm": 1.125,
+      "learning_rate": 1.9976110775527592e-05,
+      "loss": 1.6914,
+      "step": 304
+    },
+    {
+      "epoch": 0.052617959113258,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.997592237678407e-05,
+      "loss": 1.5772,
+      "step": 305
+    },
+    {
+      "epoch": 0.052790477011989995,
+      "grad_norm": 0.75,
+      "learning_rate": 1.9975733238961885e-05,
+      "loss": 1.6322,
+      "step": 306
+    },
+    {
+      "epoch": 0.05296299491072199,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.997554336207505e-05,
+      "loss": 1.6677,
+      "step": 307
+    },
+    {
+      "epoch": 0.05313551280945398,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.9975352746137636e-05,
+      "loss": 1.658,
+      "step": 308
+    },
+    {
+      "epoch": 0.053308030708185974,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.997516139116376e-05,
+      "loss": 1.6323,
+      "step": 309
+    },
+    {
+      "epoch": 0.05348054860691797,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.99749692971676e-05,
+      "loss": 1.7478,
+      "step": 310
+    },
+    {
+      "epoch": 0.05365306650564996,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.9974776464163387e-05,
+      "loss": 1.6917,
+      "step": 311
+    },
+    {
+      "epoch": 0.053825584404381954,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.997458289216541e-05,
+      "loss": 1.7139,
+      "step": 312
+    },
+    {
+      "epoch": 0.05399810230311395,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.9974388581188008e-05,
+      "loss": 1.6154,
+      "step": 313
+    },
+    {
+      "epoch": 0.05417062020184594,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.997419353124558e-05,
+      "loss": 1.5962,
+      "step": 314
+    },
+    {
+      "epoch": 0.054343138100577934,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.997399774235257e-05,
+      "loss": 1.642,
+      "step": 315
+    },
+    {
+      "epoch": 0.05451565599930993,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.997380121452349e-05,
+      "loss": 1.6652,
+      "step": 316
+    },
+    {
+      "epoch": 0.054688173898041925,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.9973603947772893e-05,
+      "loss": 1.6442,
+      "step": 317
+    },
+    {
+      "epoch": 0.054860691796773914,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.99734059421154e-05,
+      "loss": 1.6428,
+      "step": 318
+    },
+    {
+      "epoch": 0.05503320969550591,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.9973207197565678e-05,
+      "loss": 1.6888,
+      "step": 319
+    },
+    {
+      "epoch": 0.055205727594237905,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.9973007714138447e-05,
+      "loss": 1.6007,
+      "step": 320
+    },
+    {
+      "epoch": 0.055378245492969894,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.9972807491848494e-05,
+      "loss": 1.6779,
+      "step": 321
+    },
+    {
+      "epoch": 0.05555076339170189,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.997260653071065e-05,
+      "loss": 1.6187,
+      "step": 322
+    },
+    {
+      "epoch": 0.055723281290433885,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.99724048307398e-05,
+      "loss": 1.6346,
+      "step": 323
+    },
+    {
+      "epoch": 0.05589579918916587,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.9972202391950893e-05,
+      "loss": 1.782,
+      "step": 324
+    },
+    {
+      "epoch": 0.05606831708789787,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.9971999214358918e-05,
+      "loss": 1.6016,
+      "step": 325
+    },
+    {
+      "epoch": 0.056240834986629865,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.9971795297978937e-05,
+      "loss": 1.6269,
+      "step": 326
+    },
+    {
+      "epoch": 0.05641335288536185,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.9971590642826056e-05,
+      "loss": 1.6898,
+      "step": 327
+    },
+    {
+      "epoch": 0.05658587078409385,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.997138524891543e-05,
+      "loss": 1.6505,
+      "step": 328
+    },
+    {
+      "epoch": 0.056758388682825844,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.9971179116262284e-05,
+      "loss": 1.6379,
+      "step": 329
+    },
+    {
+      "epoch": 0.05693090658155784,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.9970972244881886e-05,
+      "loss": 1.6406,
+      "step": 330
+    },
+    {
+      "epoch": 0.05710342448028983,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.997076463478956e-05,
+      "loss": 1.6551,
+      "step": 331
+    },
+    {
+      "epoch": 0.057275942379021824,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9970556286000693e-05,
+      "loss": 1.647,
+      "step": 332
+    },
+    {
+      "epoch": 0.05744846027775382,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.997034719853072e-05,
+      "loss": 1.686,
+      "step": 333
+    },
+    {
+      "epoch": 0.05762097817648581,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.9970137372395123e-05,
+      "loss": 1.6334,
+      "step": 334
+    },
+    {
+      "epoch": 0.057793496075217804,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.9969926807609453e-05,
+      "loss": 1.5861,
+      "step": 335
+    },
+    {
+      "epoch": 0.0579660139739498,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.9969715504189312e-05,
+      "loss": 1.7133,
+      "step": 336
+    },
+    {
+      "epoch": 0.05813853187268179,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.9969503462150352e-05,
+      "loss": 1.6659,
+      "step": 337
+    },
+    {
+      "epoch": 0.058311049771413784,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.9969290681508284e-05,
+      "loss": 1.7634,
+      "step": 338
+    },
+    {
+      "epoch": 0.05848356767014578,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.996907716227887e-05,
+      "loss": 1.7041,
+      "step": 339
+    },
+    {
+      "epoch": 0.05865608556887777,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9968862904477936e-05,
+      "loss": 1.6879,
+      "step": 340
+    },
+    {
+      "epoch": 0.058828603467609764,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.9968647908121342e-05,
+      "loss": 1.6305,
+      "step": 341
+    },
+    {
+      "epoch": 0.05900112136634176,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.996843217322503e-05,
+      "loss": 1.6204,
+      "step": 342
+    },
+    {
+      "epoch": 0.059173639265073755,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.996821569980497e-05,
+      "loss": 1.6207,
+      "step": 343
+    },
+    {
+      "epoch": 0.05934615716380574,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.9967998487877212e-05,
+      "loss": 1.7491,
+      "step": 344
+    },
+    {
+      "epoch": 0.05951867506253774,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.9967780537457842e-05,
+      "loss": 1.6341,
+      "step": 345
+    },
+    {
+      "epoch": 0.059691192961269735,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.996756184856301e-05,
+      "loss": 1.6297,
+      "step": 346
+    },
+    {
+      "epoch": 0.05986371086000172,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.996734242120891e-05,
+      "loss": 1.6929,
+      "step": 347
+    },
+    {
+      "epoch": 0.06003622875873372,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.9967122255411812e-05,
+      "loss": 1.7607,
+      "step": 348
+    },
+    {
+      "epoch": 0.060208746657465714,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9966901351188018e-05,
+      "loss": 1.7494,
+      "step": 349
+    },
+    {
+      "epoch": 0.0603812645561977,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.996667970855389e-05,
+      "loss": 1.6567,
+      "step": 350
+    },
+    {
+      "epoch": 0.0605537824549297,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9966457327525864e-05,
+      "loss": 1.666,
+      "step": 351
+    },
+    {
+      "epoch": 0.060726300353661694,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9966234208120398e-05,
+      "loss": 1.6929,
+      "step": 352
+    },
+    {
+      "epoch": 0.06089881825239368,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.9966010350354032e-05,
+      "loss": 1.7106,
+      "step": 353
+    },
+    {
+      "epoch": 0.06107133615112568,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.996578575424335e-05,
+      "loss": 1.6795,
+      "step": 354
+    },
+    {
+      "epoch": 0.061243854049857674,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.996556041980499e-05,
+      "loss": 1.6676,
+      "step": 355
+    },
+    {
+      "epoch": 0.06141637194858967,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.9965334347055646e-05,
+      "loss": 1.6441,
+      "step": 356
+    },
+    {
+      "epoch": 0.06158888984732166,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.9965107536012067e-05,
+      "loss": 1.7209,
+      "step": 357
+    },
+    {
+      "epoch": 0.061761407746053654,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.996487998669106e-05,
+      "loss": 1.7478,
+      "step": 358
+    },
+    {
+      "epoch": 0.06193392564478565,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.9964651699109476e-05,
+      "loss": 1.6541,
+      "step": 359
+    },
+    {
+      "epoch": 0.06210644354351764,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.996442267328423e-05,
+      "loss": 1.5692,
+      "step": 360
+    },
+    {
+      "epoch": 0.062278961442249633,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.9964192909232297e-05,
+      "loss": 1.6048,
+      "step": 361
+    },
+    {
+      "epoch": 0.06245147934098163,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9963962406970695e-05,
+      "loss": 1.7004,
+      "step": 362
+    },
+    {
+      "epoch": 0.06262399723971362,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.9963731166516494e-05,
+      "loss": 1.61,
+      "step": 363
+    },
+    {
+      "epoch": 0.06279651513844561,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.996349918788684e-05,
+      "loss": 1.641,
+      "step": 364
+    },
+    {
+      "epoch": 0.0629690330371776,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9963266471098905e-05,
+      "loss": 1.6621,
+      "step": 365
+    },
+    {
+      "epoch": 0.0631415509359096,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.996303301616994e-05,
+      "loss": 1.5699,
+      "step": 366
+    },
+    {
+      "epoch": 0.0633140688346416,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.9962798823117232e-05,
+      "loss": 1.7367,
+      "step": 367
+    },
+    {
+      "epoch": 0.06348658673337358,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.996256389195814e-05,
+      "loss": 1.6866,
+      "step": 368
+    },
+    {
+      "epoch": 0.06365910463210558,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.996232822271007e-05,
+      "loss": 1.7105,
+      "step": 369
+    },
+    {
+      "epoch": 0.06383162253083757,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9962091815390475e-05,
+      "loss": 1.5961,
+      "step": 370
+    },
+    {
+      "epoch": 0.06400414042956956,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.9961854670016868e-05,
+      "loss": 1.69,
+      "step": 371
+    },
+    {
+      "epoch": 0.06417665832830156,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.996161678660683e-05,
+      "loss": 1.689,
+      "step": 372
+    },
+    {
+      "epoch": 0.06434917622703355,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.9961378165177972e-05,
+      "loss": 1.6235,
+      "step": 373
+    },
+    {
+      "epoch": 0.06452169412576554,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.9961138805747977e-05,
+      "loss": 1.653,
+      "step": 374
+    },
+    {
+      "epoch": 0.06469421202449754,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.996089870833458e-05,
+      "loss": 1.5524,
+      "step": 375
+    },
+    {
+      "epoch": 0.06486672992322953,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.996065787295557e-05,
+      "loss": 1.6183,
+      "step": 376
+    },
+    {
+      "epoch": 0.06503924782196153,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.9960416299628788e-05,
+      "loss": 1.5917,
+      "step": 377
+    },
+    {
+      "epoch": 0.06521176572069352,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.996017398837213e-05,
+      "loss": 1.7018,
+      "step": 378
+    },
+    {
+      "epoch": 0.06538428361942551,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.995993093920355e-05,
+      "loss": 1.7019,
+      "step": 379
+    },
+    {
+      "epoch": 0.06555680151815751,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9959687152141052e-05,
+      "loss": 1.7515,
+      "step": 380
+    },
+    {
+      "epoch": 0.0657293194168895,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.99594426272027e-05,
+      "loss": 1.6587,
+      "step": 381
+    },
+    {
+      "epoch": 0.06590183731562149,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9959197364406607e-05,
+      "loss": 1.6349,
+      "step": 382
+    },
+    {
+      "epoch": 0.0660743552143535,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.9958951363770946e-05,
+      "loss": 1.6287,
+      "step": 383
+    },
+    {
+      "epoch": 0.06624687311308548,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9958704625313942e-05,
+      "loss": 1.6268,
+      "step": 384
+    },
+    {
+      "epoch": 0.06641939101181747,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9958457149053876e-05,
+      "loss": 1.5422,
+      "step": 385
+    },
+    {
+      "epoch": 0.06659190891054947,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.995820893500908e-05,
+      "loss": 1.5975,
+      "step": 386
+    },
+    {
+      "epoch": 0.06676442680928146,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.9957959983197944e-05,
+      "loss": 1.5995,
+      "step": 387
+    },
+    {
+      "epoch": 0.06693694470801345,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.995771029363891e-05,
+      "loss": 1.6374,
+      "step": 388
+    },
+    {
+      "epoch": 0.06710946260674545,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.995745986635048e-05,
+      "loss": 1.5406,
+      "step": 389
+    },
+    {
+      "epoch": 0.06728198050547744,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.995720870135121e-05,
+      "loss": 1.6726,
+      "step": 390
+    },
+    {
+      "epoch": 0.06745449840420943,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.99569567986597e-05,
+      "loss": 1.6981,
+      "step": 391
+    },
+    {
+      "epoch": 0.06762701630294143,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9956704158294614e-05,
+      "loss": 1.7608,
+      "step": 392
+    },
+    {
+      "epoch": 0.06779953420167342,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9956450780274677e-05,
+      "loss": 1.7732,
+      "step": 393
+    },
+    {
+      "epoch": 0.06797205210040541,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.9956196664618652e-05,
+      "loss": 1.6464,
+      "step": 394
+    },
+    {
+      "epoch": 0.06814456999913741,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.995594181134537e-05,
+      "loss": 1.7028,
+      "step": 395
+    },
+    {
+      "epoch": 0.0683170878978694,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.9955686220473712e-05,
+      "loss": 1.6572,
+      "step": 396
+    },
+    {
+      "epoch": 0.06848960579660139,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9955429892022612e-05,
+      "loss": 1.612,
+      "step": 397
+    },
+    {
+      "epoch": 0.0686621236953334,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.995517282601106e-05,
+      "loss": 1.6071,
+      "step": 398
+    },
+    {
+      "epoch": 0.06883464159406538,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.9954915022458105e-05,
+      "loss": 1.6205,
+      "step": 399
+    },
+    {
+      "epoch": 0.06900715949279737,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.995465648138284e-05,
+      "loss": 1.5928,
+      "step": 400
+    },
+    {
+      "epoch": 0.06900715949279737,
+      "eval_loss": 1.6248373985290527,
+      "eval_runtime": 10.8965,
+      "eval_samples_per_second": 93.975,
+      "eval_steps_per_second": 23.494,
+      "step": 400
+    },
+    {
+      "epoch": 0.06917967739152937,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.9954397202804426e-05,
+      "loss": 1.6632,
+      "step": 401
+    },
+    {
+      "epoch": 0.06935219529026136,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.995413718674207e-05,
+      "loss": 1.663,
+      "step": 402
+    },
+    {
+      "epoch": 0.06952471318899336,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9953876433215035e-05,
+      "loss": 1.6414,
+      "step": 403
+    },
+    {
+      "epoch": 0.06969723108772535,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9953614942242635e-05,
+      "loss": 1.6355,
+      "step": 404
+    },
+    {
+      "epoch": 0.06986974898645734,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.9953352713844253e-05,
+      "loss": 1.517,
+      "step": 405
+    },
+    {
+      "epoch": 0.07004226688518934,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.9953089748039306e-05,
+      "loss": 1.657,
+      "step": 406
+    },
+    {
+      "epoch": 0.07021478478392133,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.9952826044847282e-05,
+      "loss": 1.5441,
+      "step": 407
+    },
+    {
+      "epoch": 0.07038730268265332,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.9952561604287717e-05,
+      "loss": 1.6777,
+      "step": 408
+    },
+    {
+      "epoch": 0.07055982058138532,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.9952296426380198e-05,
+      "loss": 1.6177,
+      "step": 409
+    },
+    {
+      "epoch": 0.07073233848011731,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.9952030511144384e-05,
+      "loss": 1.6362,
+      "step": 410
+    },
+    {
+      "epoch": 0.0709048563788493,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.995176385859996e-05,
+      "loss": 1.5383,
+      "step": 411
+    },
+    {
+      "epoch": 0.0710773742775813,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9951496468766687e-05,
+      "loss": 1.7264,
+      "step": 412
+    },
+    {
+      "epoch": 0.07124989217631329,
+      "grad_norm": 0.75,
+      "learning_rate": 1.9951228341664376e-05,
+      "loss": 1.7352,
+      "step": 413
+    },
+    {
+      "epoch": 0.07142241007504528,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.9950959477312895e-05,
+      "loss": 1.6589,
+      "step": 414
+    },
+    {
+      "epoch": 0.07159492797377728,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9950689875732157e-05,
+      "loss": 1.6992,
+      "step": 415
+    },
+    {
+      "epoch": 0.07176744587250927,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.9950419536942137e-05,
+      "loss": 1.6052,
+      "step": 416
+    },
+    {
+      "epoch": 0.07193996377124126,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.9950148460962867e-05,
+      "loss": 1.561,
+      "step": 417
+    },
+    {
+      "epoch": 0.07211248166997326,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.9949876647814428e-05,
+      "loss": 1.6689,
+      "step": 418
+    },
+    {
+      "epoch": 0.07228499956870525,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.9949604097516956e-05,
+      "loss": 1.5868,
+      "step": 419
+    },
+    {
+      "epoch": 0.07245751746743724,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9949330810090643e-05,
+      "loss": 1.5644,
+      "step": 420
+    },
+    {
+      "epoch": 0.07263003536616924,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.9949056785555738e-05,
+      "loss": 1.5376,
+      "step": 421
+    },
+    {
+      "epoch": 0.07280255326490123,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.9948782023932545e-05,
+      "loss": 1.6105,
+      "step": 422
+    },
+    {
+      "epoch": 0.07297507116363322,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.9948506525241414e-05,
+      "loss": 1.6057,
+      "step": 423
+    },
+    {
+      "epoch": 0.07314758906236522,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.9948230289502758e-05,
+      "loss": 1.7268,
+      "step": 424
+    },
+    {
+      "epoch": 0.07332010696109721,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.9947953316737045e-05,
+      "loss": 1.6582,
+      "step": 425
+    },
+    {
+      "epoch": 0.0734926248598292,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.9947675606964793e-05,
+      "loss": 1.6357,
+      "step": 426
+    },
+    {
+      "epoch": 0.0736651427585612,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.994739716020657e-05,
+      "loss": 1.5585,
+      "step": 427
+    },
+    {
+      "epoch": 0.07383766065729319,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.9947117976483018e-05,
+      "loss": 1.6731,
+      "step": 428
+    },
+    {
+      "epoch": 0.0740101785560252,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.994683805581481e-05,
+      "loss": 1.6636,
+      "step": 429
+    },
+    {
+      "epoch": 0.07418269645475718,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9946557398222686e-05,
+      "loss": 1.7249,
+      "step": 430
+    },
+    {
+      "epoch": 0.07435521435348917,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.9946276003727447e-05,
+      "loss": 1.5798,
+      "step": 431
+    },
+    {
+      "epoch": 0.07452773225222117,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.994599387234993e-05,
+      "loss": 1.6511,
+      "step": 432
+    },
+    {
+      "epoch": 0.07470025015095316,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9945711004111045e-05,
+      "loss": 1.6225,
+      "step": 433
+    },
+    {
+      "epoch": 0.07487276804968515,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.9945427399031744e-05,
+      "loss": 1.5686,
+      "step": 434
+    },
+    {
+      "epoch": 0.07504528594841715,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.9945143057133037e-05,
+      "loss": 1.6841,
+      "step": 435
+    },
+    {
+      "epoch": 0.07521780384714914,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.9944857978435996e-05,
+      "loss": 1.6935,
+      "step": 436
+    },
+    {
+      "epoch": 0.07539032174588113,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.9944572162961735e-05,
+      "loss": 1.5725,
+      "step": 437
+    },
+    {
+      "epoch": 0.07556283964461313,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.9944285610731433e-05,
+      "loss": 1.675,
+      "step": 438
+    },
+    {
+      "epoch": 0.07573535754334512,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.9943998321766318e-05,
+      "loss": 1.7037,
+      "step": 439
+    },
+    {
+      "epoch": 0.07590787544207711,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.9943710296087672e-05,
+      "loss": 1.6607,
+      "step": 440
+    },
+    {
+      "epoch": 0.07608039334080911,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.994342153371684e-05,
+      "loss": 1.6181,
+      "step": 441
+    },
+    {
+      "epoch": 0.0762529112395411,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.9943132034675208e-05,
+      "loss": 1.6794,
+      "step": 442
+    },
+    {
+      "epoch": 0.07642542913827309,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.9942841798984228e-05,
+      "loss": 1.6142,
+      "step": 443
+    },
+    {
+      "epoch": 0.07659794703700509,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.9942550826665404e-05,
+      "loss": 1.7234,
+      "step": 444
+    },
+    {
+      "epoch": 0.07677046493573708,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.994225911774029e-05,
+      "loss": 1.6427,
+      "step": 445
+    },
+    {
+      "epoch": 0.07694298283446907,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.9941966672230494e-05,
+      "loss": 1.7516,
+      "step": 446
+    },
+    {
+      "epoch": 0.07711550073320107,
+      "grad_norm": 1.3828125,
+      "learning_rate": 1.9941673490157694e-05,
+      "loss": 1.6401,
+      "step": 447
+    },
+    {
+      "epoch": 0.07728801863193306,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.9941379571543597e-05,
+      "loss": 1.6463,
+      "step": 448
+    },
+    {
+      "epoch": 0.07746053653066505,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.9941084916409988e-05,
+      "loss": 1.5744,
+      "step": 449
+    },
+    {
+      "epoch": 0.07763305442939705,
+      "grad_norm": 1.3203125,
+      "learning_rate": 1.994078952477869e-05,
+      "loss": 1.6328,
+      "step": 450
+    },
+    {
+      "epoch": 0.07780557232812904,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.9940493396671598e-05,
+      "loss": 1.6568,
+      "step": 451
+    },
+    {
+      "epoch": 0.07797809022686104,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.994019653211064e-05,
+      "loss": 1.5954,
+      "step": 452
+    },
+    {
+      "epoch": 0.07815060812559303,
+      "grad_norm": 1.125,
+      "learning_rate": 1.9939898931117813e-05,
+      "loss": 1.6227,
+      "step": 453
+    },
+    {
+      "epoch": 0.07832312602432502,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.9939600593715166e-05,
+      "loss": 1.5917,
+      "step": 454
+    },
+    {
+      "epoch": 0.07849564392305702,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.99393015199248e-05,
+      "loss": 1.6043,
+      "step": 455
+    },
+    {
+      "epoch": 0.07866816182178901,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.993900170976888e-05,
+      "loss": 1.66,
+      "step": 456
+    },
+    {
+      "epoch": 0.078840679720521,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.9938701163269607e-05,
+      "loss": 1.5998,
+      "step": 457
+    },
+    {
+      "epoch": 0.079013197619253,
+      "grad_norm": 0.875,
+      "learning_rate": 1.9938399880449254e-05,
+      "loss": 1.6249,
+      "step": 458
+    },
+    {
+      "epoch": 0.07918571551798499,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.9938097861330138e-05,
+      "loss": 1.623,
+      "step": 459
+    },
+    {
+      "epoch": 0.07935823341671698,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.9937795105934637e-05,
+      "loss": 1.5886,
+      "step": 460
+    },
+    {
+      "epoch": 0.07953075131544898,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.9937491614285182e-05,
+      "loss": 1.6607,
+      "step": 461
+    },
+    {
+      "epoch": 0.07970326921418097,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.993718738640425e-05,
+      "loss": 1.6488,
+      "step": 462
+    },
+    {
+      "epoch": 0.07987578711291296,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.993688242231439e-05,
+      "loss": 1.5408,
+      "step": 463
+    },
+    {
+      "epoch": 0.08004830501164496,
+      "grad_norm": 0.625,
+      "learning_rate": 1.9936576722038192e-05,
+      "loss": 1.6059,
+      "step": 464
+    },
+    {
+      "epoch": 0.08022082291037695,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.9936270285598306e-05,
+      "loss": 1.5628,
+      "step": 465
+    },
+    {
+      "epoch": 0.08039334080910894,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.993596311301743e-05,
+      "loss": 1.6705,
+      "step": 466
+    },
+    {
+      "epoch": 0.08056585870784094,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.9935655204318323e-05,
+      "loss": 1.6953,
+      "step": 467
+    },
+    {
+      "epoch": 0.08073837660657293,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.99353465595238e-05,
+      "loss": 1.6976,
+      "step": 468
+    },
+    {
+      "epoch": 0.08091089450530492,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.993503717865672e-05,
+      "loss": 1.7231,
+      "step": 469
+    },
+    {
+      "epoch": 0.08108341240403692,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9934727061740013e-05,
+      "loss": 1.6958,
+      "step": 470
+    },
+    {
+      "epoch": 0.08125593030276891,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.993441620879665e-05,
+      "loss": 1.5575,
+      "step": 471
+    },
+    {
+      "epoch": 0.0814284482015009,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.993410461984966e-05,
+      "loss": 1.6419,
+      "step": 472
+    },
+    {
+      "epoch": 0.0816009661002329,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.993379229492213e-05,
+      "loss": 1.5573,
+      "step": 473
+    },
+    {
+      "epoch": 0.08177348399896489,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.99334792340372e-05,
+      "loss": 1.5639,
+      "step": 474
+    },
+    {
+      "epoch": 0.08194600189769688,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.9933165437218057e-05,
+      "loss": 1.6752,
+      "step": 475
+    },
+    {
+      "epoch": 0.08211851979642888,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.993285090448795e-05,
+      "loss": 1.6365,
+      "step": 476
+    },
+    {
+      "epoch": 0.08229103769516087,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.993253563587019e-05,
+      "loss": 1.6038,
+      "step": 477
+    },
+    {
+      "epoch": 0.08246355559389287,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.993221963138813e-05,
+      "loss": 1.6083,
+      "step": 478
+    },
+    {
+      "epoch": 0.08263607349262486,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.993190289106518e-05,
+      "loss": 1.6,
+      "step": 479
+    },
+    {
+      "epoch": 0.08280859139135685,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.99315854149248e-05,
+      "loss": 1.5528,
+      "step": 480
+    },
+    {
+      "epoch": 0.08298110929008885,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.9931267202990524e-05,
+      "loss": 1.6537,
+      "step": 481
+    },
+    {
+      "epoch": 0.08315362718882084,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.9930948255285915e-05,
+      "loss": 1.5673,
+      "step": 482
+    },
+    {
+      "epoch": 0.08332614508755283,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9930628571834608e-05,
+      "loss": 1.6501,
+      "step": 483
+    },
+    {
+      "epoch": 0.08349866298628483,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.993030815266029e-05,
+      "loss": 1.6432,
+      "step": 484
+    },
+    {
+      "epoch": 0.08367118088501682,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.9929986997786698e-05,
+      "loss": 1.5725,
+      "step": 485
+    },
+    {
+      "epoch": 0.08384369878374881,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.992966510723762e-05,
+      "loss": 1.6375,
+      "step": 486
+    },
+    {
+      "epoch": 0.08401621668248081,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.992934248103691e-05,
+      "loss": 1.6855,
+      "step": 487
+    },
+    {
+      "epoch": 0.0841887345812128,
+      "grad_norm": 0.75,
+      "learning_rate": 1.992901911920847e-05,
+      "loss": 1.6522,
+      "step": 488
+    },
+    {
+      "epoch": 0.08436125247994479,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.992869502177625e-05,
+      "loss": 1.6277,
+      "step": 489
+    },
+    {
+      "epoch": 0.08453377037867679,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9928370188764265e-05,
+      "loss": 1.6811,
+      "step": 490
+    },
+    {
+      "epoch": 0.08470628827740878,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.9928044620196582e-05,
+      "loss": 1.7146,
+      "step": 491
+    },
+    {
+      "epoch": 0.08487880617614077,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9927718316097322e-05,
+      "loss": 1.6449,
+      "step": 492
+    },
+    {
+      "epoch": 0.08505132407487277,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9927391276490657e-05,
+      "loss": 1.5758,
+      "step": 493
+    },
+    {
+      "epoch": 0.08522384197360476,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9927063501400817e-05,
+      "loss": 1.5599,
+      "step": 494
+    },
+    {
+      "epoch": 0.08539635987233675,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.9926734990852084e-05,
+      "loss": 1.6281,
+      "step": 495
+    },
+    {
+      "epoch": 0.08556887777106875,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.99264057448688e-05,
+      "loss": 1.7188,
+      "step": 496
+    },
+    {
+      "epoch": 0.08574139566980074,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9926075763475353e-05,
+      "loss": 1.606,
+      "step": 497
+    },
+    {
+      "epoch": 0.08591391356853273,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.9925745046696196e-05,
+      "loss": 1.5543,
+      "step": 498
+    },
+    {
+      "epoch": 0.08608643146726473,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.9925413594555824e-05,
+      "loss": 1.6025,
+      "step": 499
+    },
+    {
+      "epoch": 0.08625894936599672,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.9925081407078798e-05,
+      "loss": 1.69,
+      "step": 500
+    },
+    {
+      "epoch": 0.08625894936599672,
+      "eval_loss": 1.5997848510742188,
+      "eval_runtime": 10.7856,
+      "eval_samples_per_second": 94.941,
+      "eval_steps_per_second": 23.735,
+      "step": 500
+    },
+    {
+      "epoch": 0.08643146726472871,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.9924748484289723e-05,
+      "loss": 1.6629,
+      "step": 501
+    },
+    {
+      "epoch": 0.08660398516346071,
+      "grad_norm": 18.75,
+      "learning_rate": 1.992441482621327e-05,
+      "loss": 1.6557,
+      "step": 502
+    },
+    {
+      "epoch": 0.0867765030621927,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9924080432874158e-05,
+      "loss": 1.5788,
+      "step": 503
+    },
+    {
+      "epoch": 0.0869490209609247,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.992374530429716e-05,
+      "loss": 1.5682,
+      "step": 504
+    },
+    {
+      "epoch": 0.08712153885965669,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.99234094405071e-05,
+      "loss": 1.5251,
+      "step": 505
+    },
+    {
+      "epoch": 0.08729405675838868,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9923072841528862e-05,
+      "loss": 1.5827,
+      "step": 506
+    },
+    {
+      "epoch": 0.08746657465712068,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.9922735507387393e-05,
+      "loss": 1.6767,
+      "step": 507
+    },
+    {
+      "epoch": 0.08763909255585267,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.992239743810767e-05,
+      "loss": 1.4635,
+      "step": 508
+    },
+    {
+      "epoch": 0.08781161045458466,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.9922058633714752e-05,
+      "loss": 1.6189,
+      "step": 509
+    },
+    {
+      "epoch": 0.08798412835331666,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.992171909423373e-05,
+      "loss": 1.5658,
+      "step": 510
+    },
+    {
+      "epoch": 0.08815664625204865,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.9921378819689767e-05,
+      "loss": 1.6472,
+      "step": 511
+    },
+    {
+      "epoch": 0.08832916415078064,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.992103781010807e-05,
+      "loss": 1.6193,
+      "step": 512
+    },
+    {
+      "epoch": 0.08850168204951264,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.9920696065513902e-05,
+      "loss": 1.5667,
+      "step": 513
+    },
+    {
+      "epoch": 0.08867419994824463,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.992035358593258e-05,
+      "loss": 1.5772,
+      "step": 514
+    },
+    {
+      "epoch": 0.08884671784697662,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.992001037138948e-05,
+      "loss": 1.6495,
+      "step": 515
+    },
+    {
+      "epoch": 0.08901923574570862,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.991966642191003e-05,
+      "loss": 1.6067,
+      "step": 516
+    },
+    {
+      "epoch": 0.08919175364444061,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.991932173751971e-05,
+      "loss": 1.5596,
+      "step": 517
+    },
+    {
+      "epoch": 0.0893642715431726,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9918976318244056e-05,
+      "loss": 1.5892,
+      "step": 518
+    },
+    {
+      "epoch": 0.0895367894419046,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.991863016410866e-05,
+      "loss": 1.5691,
+      "step": 519
+    },
+    {
+      "epoch": 0.08970930734063659,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.9918283275139167e-05,
+      "loss": 1.5445,
+      "step": 520
+    },
+    {
+      "epoch": 0.08988182523936858,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.991793565136128e-05,
+      "loss": 1.7354,
+      "step": 521
+    },
+    {
+      "epoch": 0.09005434313810058,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.991758729280074e-05,
+      "loss": 1.5286,
+      "step": 522
+    },
+    {
+      "epoch": 0.09022686103683257,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.9917238199483374e-05,
+      "loss": 1.6222,
+      "step": 523
+    },
+    {
+      "epoch": 0.09039937893556456,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.9916888371435036e-05,
+      "loss": 1.6243,
+      "step": 524
+    },
+    {
+      "epoch": 0.09057189683429656,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.9916537808681643e-05,
+      "loss": 1.5857,
+      "step": 525
+    },
+    {
+      "epoch": 0.09074441473302855,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.9916186511249167e-05,
+      "loss": 1.5818,
+      "step": 526
+    },
+    {
+      "epoch": 0.09091693263176054,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9915834479163634e-05,
+      "loss": 1.5379,
+      "step": 527
+    },
+    {
+      "epoch": 0.09108945053049254,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.991548171245113e-05,
+      "loss": 1.6106,
+      "step": 528
+    },
+    {
+      "epoch": 0.09126196842922453,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.991512821113778e-05,
+      "loss": 1.6018,
+      "step": 529
+    },
+    {
+      "epoch": 0.09143448632795653,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9914773975249782e-05,
+      "loss": 1.5822,
+      "step": 530
+    },
+    {
+      "epoch": 0.09160700422668852,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.991441900481338e-05,
+      "loss": 1.627,
+      "step": 531
+    },
+    {
+      "epoch": 0.09177952212542051,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9914063299854866e-05,
+      "loss": 1.6365,
+      "step": 532
+    },
+    {
+      "epoch": 0.09195204002415251,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.99137068604006e-05,
+      "loss": 1.6972,
+      "step": 533
+    },
+    {
+      "epoch": 0.0921245579228845,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.9913349686476983e-05,
+      "loss": 1.6787,
+      "step": 534
+    },
+    {
+      "epoch": 0.09229707582161649,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9912991778110485e-05,
+      "loss": 1.583,
+      "step": 535
+    },
+    {
+      "epoch": 0.09246959372034849,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.991263313532761e-05,
+      "loss": 1.6112,
+      "step": 536
+    },
+    {
+      "epoch": 0.09264211161908048,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.9912273758154937e-05,
+      "loss": 1.5964,
+      "step": 537
+    },
+    {
+      "epoch": 0.09281462951781247,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.9911913646619094e-05,
+      "loss": 1.6758,
+      "step": 538
+    },
+    {
+      "epoch": 0.09298714741654447,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.991155280074675e-05,
+      "loss": 1.6041,
+      "step": 539
+    },
+    {
+      "epoch": 0.09315966531527646,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.991119122056465e-05,
+      "loss": 1.7256,
+      "step": 540
+    },
+    {
+      "epoch": 0.09333218321400845,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.9910828906099573e-05,
+      "loss": 1.6746,
+      "step": 541
+    },
+    {
+      "epoch": 0.09350470111274045,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.9910465857378367e-05,
+      "loss": 1.5857,
+      "step": 542
+    },
+    {
+      "epoch": 0.09367721901147244,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.991010207442792e-05,
+      "loss": 1.6412,
+      "step": 543
+    },
+    {
+      "epoch": 0.09384973691020443,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.99097375572752e-05,
+      "loss": 1.5818,
+      "step": 544
+    },
+    {
+      "epoch": 0.09402225480893643,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.9909372305947196e-05,
+      "loss": 1.6015,
+      "step": 545
+    },
+    {
+      "epoch": 0.09419477270766842,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.9909006320470977e-05,
+      "loss": 1.7097,
+      "step": 546
+    },
+    {
+      "epoch": 0.09436729060640041,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.9908639600873654e-05,
+      "loss": 1.6574,
+      "step": 547
+    },
+    {
+      "epoch": 0.09453980850513241,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.99082721471824e-05,
+      "loss": 1.6898,
+      "step": 548
+    },
+    {
+      "epoch": 0.0947123264038644,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.9907903959424436e-05,
+      "loss": 1.566,
+      "step": 549
+    },
+    {
+      "epoch": 0.09488484430259639,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9907535037627034e-05,
+      "loss": 1.601,
+      "step": 550
+    },
+    {
+      "epoch": 0.09505736220132839,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.990716538181754e-05,
+      "loss": 1.6122,
+      "step": 551
+    },
+    {
+      "epoch": 0.09522988010006038,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.9906794992023325e-05,
+      "loss": 1.6707,
+      "step": 552
+    },
+    {
+      "epoch": 0.09540239799879237,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.9906423868271837e-05,
+      "loss": 1.6544,
+      "step": 553
+    },
+    {
+      "epoch": 0.09557491589752437,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.990605201059057e-05,
+      "loss": 1.5394,
+      "step": 554
+    },
+    {
+      "epoch": 0.09574743379625636,
+      "grad_norm": 0.875,
+      "learning_rate": 1.9905679419007078e-05,
+      "loss": 1.5378,
+      "step": 555
+    },
+    {
+      "epoch": 0.09591995169498836,
+      "grad_norm": 0.625,
+      "learning_rate": 1.990530609354896e-05,
+      "loss": 1.5238,
+      "step": 556
+    },
+    {
+      "epoch": 0.09609246959372035,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9904932034243872e-05,
+      "loss": 1.6568,
+      "step": 557
+    },
+    {
+      "epoch": 0.09626498749245234,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.9904557241119534e-05,
+      "loss": 1.6512,
+      "step": 558
+    },
+    {
+      "epoch": 0.09643750539118434,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.9904181714203707e-05,
+      "loss": 1.6527,
+      "step": 559
+    },
+    {
+      "epoch": 0.09661002328991633,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.9903805453524213e-05,
+      "loss": 1.6576,
+      "step": 560
+    },
+    {
+      "epoch": 0.09678254118864832,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.990342845910893e-05,
+      "loss": 1.6159,
+      "step": 561
+    },
+    {
+      "epoch": 0.09695505908738032,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.990305073098579e-05,
+      "loss": 1.5827,
+      "step": 562
+    },
+    {
+      "epoch": 0.09712757698611231,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.9902672269182773e-05,
+      "loss": 1.6479,
+      "step": 563
+    },
+    {
+      "epoch": 0.0973000948848443,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.990229307372792e-05,
+      "loss": 1.5307,
+      "step": 564
+    },
+    {
+      "epoch": 0.0974726127835763,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.9901913144649323e-05,
+      "loss": 1.6036,
+      "step": 565
+    },
+    {
+      "epoch": 0.09764513068230829,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.9901532481975133e-05,
+      "loss": 1.5556,
+      "step": 566
+    },
+    {
+      "epoch": 0.09781764858104028,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9901151085733542e-05,
+      "loss": 1.6605,
+      "step": 567
+    },
+    {
+      "epoch": 0.09799016647977228,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.990076895595282e-05,
+      "loss": 1.6237,
+      "step": 568
+    },
+    {
+      "epoch": 0.09816268437850427,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.990038609266127e-05,
+      "loss": 1.5957,
+      "step": 569
+    },
+    {
+      "epoch": 0.09833520227723626,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.9900002495887255e-05,
+      "loss": 1.651,
+      "step": 570
+    },
+    {
+      "epoch": 0.09850772017596826,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.98996181656592e-05,
+      "loss": 1.6785,
+      "step": 571
+    },
+    {
+      "epoch": 0.09868023807470025,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.9899233102005573e-05,
+      "loss": 1.5852,
+      "step": 572
+    },
+    {
+      "epoch": 0.09885275597343224,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.9898847304954906e-05,
+      "loss": 1.5645,
+      "step": 573
+    },
+    {
+      "epoch": 0.09902527387216424,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.989846077453578e-05,
+      "loss": 1.711,
+      "step": 574
+    },
+    {
+      "epoch": 0.09919779177089623,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.989807351077683e-05,
+      "loss": 1.5826,
+      "step": 575
+    },
+    {
+      "epoch": 0.09937030966962822,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.989768551370675e-05,
+      "loss": 1.5717,
+      "step": 576
+    },
+    {
+      "epoch": 0.09954282756836022,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.9897296783354283e-05,
+      "loss": 1.6394,
+      "step": 577
+    },
+    {
+      "epoch": 0.09971534546709221,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.9896907319748227e-05,
+      "loss": 1.7012,
+      "step": 578
+    },
+    {
+      "epoch": 0.0998878633658242,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9896517122917442e-05,
+      "loss": 1.6539,
+      "step": 579
+    },
+    {
+      "epoch": 0.1000603812645562,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.989612619289083e-05,
+      "loss": 1.582,
+      "step": 580
+    },
+    {
+      "epoch": 0.10023289916328819,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.9895734529697356e-05,
+      "loss": 1.5482,
+      "step": 581
+    },
+    {
+      "epoch": 0.10040541706202019,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.9895342133366036e-05,
+      "loss": 1.7008,
+      "step": 582
+    },
+    {
+      "epoch": 0.10057793496075218,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.9894949003925942e-05,
+      "loss": 1.5771,
+      "step": 583
+    },
+    {
+      "epoch": 0.10075045285948417,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.98945551414062e-05,
+      "loss": 1.5932,
+      "step": 584
+    },
+    {
+      "epoch": 0.10092297075821617,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.989416054583599e-05,
+      "loss": 1.6266,
+      "step": 585
+    },
+    {
+      "epoch": 0.10109548865694816,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.9893765217244544e-05,
+      "loss": 1.588,
+      "step": 586
+    },
+    {
+      "epoch": 0.10126800655568015,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.9893369155661152e-05,
+      "loss": 1.6614,
+      "step": 587
+    },
+    {
+      "epoch": 0.10144052445441215,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9892972361115154e-05,
+      "loss": 1.5868,
+      "step": 588
+    },
+    {
+      "epoch": 0.10161304235314414,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.989257483363595e-05,
+      "loss": 1.5864,
+      "step": 589
+    },
+    {
+      "epoch": 0.10178556025187613,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9892176573252993e-05,
+      "loss": 1.5276,
+      "step": 590
+    },
+    {
+      "epoch": 0.10195807815060813,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9891777579995787e-05,
+      "loss": 1.616,
+      "step": 591
+    },
+    {
+      "epoch": 0.10213059604934012,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.9891377853893888e-05,
+      "loss": 1.621,
+      "step": 592
+    },
+    {
+      "epoch": 0.10230311394807211,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.9890977394976915e-05,
+      "loss": 1.5442,
+      "step": 593
+    },
+    {
+      "epoch": 0.10247563184680411,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.9890576203274534e-05,
+      "loss": 1.5709,
+      "step": 594
+    },
+    {
+      "epoch": 0.1026481497455361,
+      "grad_norm": 0.875,
+      "learning_rate": 1.989017427881647e-05,
+      "loss": 1.5972,
+      "step": 595
+    },
+    {
+      "epoch": 0.10282066764426809,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.9889771621632497e-05,
+      "loss": 1.636,
+      "step": 596
+    },
+    {
+      "epoch": 0.10299318554300009,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.9889368231752452e-05,
+      "loss": 1.6101,
+      "step": 597
+    },
+    {
+      "epoch": 0.10316570344173208,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.9888964109206213e-05,
+      "loss": 1.6646,
+      "step": 598
+    },
+    {
+      "epoch": 0.10333822134046407,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.9888559254023723e-05,
+      "loss": 1.5002,
+      "step": 599
+    },
+    {
+      "epoch": 0.10351073923919607,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.988815366623498e-05,
+      "loss": 1.5775,
+      "step": 600
+    },
+    {
+      "epoch": 0.10351073923919607,
+      "eval_loss": 1.5823333263397217,
+      "eval_runtime": 10.7916,
+      "eval_samples_per_second": 94.889,
+      "eval_steps_per_second": 23.722,
+      "step": 600
+    },
+    {
+      "epoch": 0.10368325713792806,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.988774734587003e-05,
+      "loss": 1.6637,
+      "step": 601
+    },
+    {
+      "epoch": 0.10385577503666005,
+      "grad_norm": 0.875,
+      "learning_rate": 1.988734029295897e-05,
+      "loss": 1.556,
+      "step": 602
+    },
+    {
+      "epoch": 0.10402829293539205,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9886932507531966e-05,
+      "loss": 1.5965,
+      "step": 603
+    },
+    {
+      "epoch": 0.10420081083412404,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.9886523989619224e-05,
+      "loss": 1.6596,
+      "step": 604
+    },
+    {
+      "epoch": 0.10437332873285603,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.988611473925101e-05,
+      "loss": 1.6136,
+      "step": 605
+    },
+    {
+      "epoch": 0.10454584663158803,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.988570475645765e-05,
+      "loss": 1.5358,
+      "step": 606
+    },
+    {
+      "epoch": 0.10471836453032002,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.988529404126951e-05,
+      "loss": 1.609,
+      "step": 607
+    },
+    {
+      "epoch": 0.10489088242905202,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.9884882593717018e-05,
+      "loss": 1.6632,
+      "step": 608
+    },
+    {
+      "epoch": 0.10506340032778401,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.9884470413830662e-05,
+      "loss": 1.6104,
+      "step": 609
+    },
+    {
+      "epoch": 0.105235918226516,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.988405750164098e-05,
+      "loss": 1.5754,
+      "step": 610
+    },
+    {
+      "epoch": 0.105408436125248,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.9883643857178554e-05,
+      "loss": 1.614,
+      "step": 611
+    },
+    {
+      "epoch": 0.10558095402397999,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.988322948047404e-05,
+      "loss": 1.6469,
+      "step": 612
+    },
+    {
+      "epoch": 0.10575347192271198,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.988281437155813e-05,
+      "loss": 1.6428,
+      "step": 613
+    },
+    {
+      "epoch": 0.10592598982144398,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.9882398530461582e-05,
+      "loss": 1.6195,
+      "step": 614
+    },
+    {
+      "epoch": 0.10609850772017597,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.98819819572152e-05,
+      "loss": 1.6814,
+      "step": 615
+    },
+    {
+      "epoch": 0.10627102561890796,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.9881564651849855e-05,
+      "loss": 1.5624,
+      "step": 616
+    },
+    {
+      "epoch": 0.10644354351763996,
+      "grad_norm": 0.875,
+      "learning_rate": 1.9881146614396454e-05,
+      "loss": 1.6194,
+      "step": 617
+    },
+    {
+      "epoch": 0.10661606141637195,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.988072784488597e-05,
+      "loss": 1.5799,
+      "step": 618
+    },
+    {
+      "epoch": 0.10678857931510394,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.9880308343349432e-05,
+      "loss": 1.5319,
+      "step": 619
+    },
+    {
+      "epoch": 0.10696109721383594,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.9879888109817916e-05,
+      "loss": 1.5623,
+      "step": 620
+    },
+    {
+      "epoch": 0.10713361511256793,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.9879467144322557e-05,
+      "loss": 1.6315,
+      "step": 621
+    },
+    {
+      "epoch": 0.10730613301129992,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.9879045446894546e-05,
+      "loss": 1.5334,
+      "step": 622
+    },
+    {
+      "epoch": 0.10747865091003192,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.9878623017565115e-05,
+      "loss": 1.5988,
+      "step": 623
+    },
+    {
+      "epoch": 0.10765116880876391,
+      "grad_norm": 0.625,
+      "learning_rate": 1.987819985636557e-05,
+      "loss": 1.537,
+      "step": 624
+    },
+    {
+      "epoch": 0.1078236867074959,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.9877775963327254e-05,
+      "loss": 1.5755,
+      "step": 625
+    },
+    {
+      "epoch": 0.1079962046062279,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.987735133848158e-05,
+      "loss": 1.5418,
+      "step": 626
+    },
+    {
+      "epoch": 0.10816872250495989,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.987692598186e-05,
+      "loss": 1.595,
+      "step": 627
+    },
+    {
+      "epoch": 0.10834124040369188,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.9876499893494028e-05,
+      "loss": 1.4856,
+      "step": 628
+    },
+    {
+      "epoch": 0.10851375830242388,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9876073073415233e-05,
+      "loss": 1.5476,
+      "step": 629
+    },
+    {
+      "epoch": 0.10868627620115587,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.987564552165524e-05,
+      "loss": 1.5228,
+      "step": 630
+    },
+    {
+      "epoch": 0.10885879409988786,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9875217238245715e-05,
+      "loss": 1.5727,
+      "step": 631
+    },
+    {
+      "epoch": 0.10903131199861986,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.9874788223218398e-05,
+      "loss": 1.6439,
+      "step": 632
+    },
+    {
+      "epoch": 0.10920382989735185,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.987435847660507e-05,
+      "loss": 1.4864,
+      "step": 633
+    },
+    {
+      "epoch": 0.10937634779608385,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9873927998437566e-05,
+      "loss": 1.5274,
+      "step": 634
+    },
+    {
+      "epoch": 0.10954886569481584,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.9873496788747778e-05,
+      "loss": 1.6405,
+      "step": 635
+    },
+    {
+      "epoch": 0.10972138359354783,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.987306484756766e-05,
+      "loss": 1.6233,
+      "step": 636
+    },
+    {
+      "epoch": 0.10989390149227983,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9872632174929208e-05,
+      "loss": 1.585,
+      "step": 637
+    },
+    {
+      "epoch": 0.11006641939101182,
+      "grad_norm": 0.75,
+      "learning_rate": 1.9872198770864476e-05,
+      "loss": 1.6029,
+      "step": 638
+    },
+    {
+      "epoch": 0.11023893728974381,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.9871764635405574e-05,
+      "loss": 1.6057,
+      "step": 639
+    },
+    {
+      "epoch": 0.11041145518847581,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.9871329768584666e-05,
+      "loss": 1.5482,
+      "step": 640
+    },
+    {
+      "epoch": 0.1105839730872078,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.987089417043397e-05,
+      "loss": 1.6146,
+      "step": 641
+    },
+    {
+      "epoch": 0.11075649098593979,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.987045784098576e-05,
+      "loss": 1.6118,
+      "step": 642
+    },
+    {
+      "epoch": 0.11092900888467179,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.9870020780272357e-05,
+      "loss": 1.5789,
+      "step": 643
+    },
+    {
+      "epoch": 0.11110152678340378,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.9869582988326145e-05,
+      "loss": 1.5449,
+      "step": 644
+    },
+    {
+      "epoch": 0.11127404468213577,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.9869144465179557e-05,
+      "loss": 1.5871,
+      "step": 645
+    },
+    {
+      "epoch": 0.11144656258086777,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.9868705210865083e-05,
+      "loss": 1.5773,
+      "step": 646
+    },
+    {
+      "epoch": 0.11161908047959976,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.9868265225415263e-05,
+      "loss": 1.6108,
+      "step": 647
+    },
+    {
+      "epoch": 0.11179159837833175,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.9867824508862696e-05,
+      "loss": 1.6162,
+      "step": 648
+    },
+    {
+      "epoch": 0.11196411627706375,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.986738306124003e-05,
+      "loss": 1.5618,
+      "step": 649
+    },
+    {
+      "epoch": 0.11213663417579574,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9866940882579976e-05,
+      "loss": 1.6397,
+      "step": 650
+    },
+    {
+      "epoch": 0.11230915207452773,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.986649797291529e-05,
+      "loss": 1.632,
+      "step": 651
+    },
+    {
+      "epoch": 0.11248166997325973,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.9866054332278784e-05,
+      "loss": 1.6181,
+      "step": 652
+    },
+    {
+      "epoch": 0.11265418787199172,
+      "grad_norm": 0.625,
+      "learning_rate": 1.9865609960703325e-05,
+      "loss": 1.5809,
+      "step": 653
+    },
+    {
+      "epoch": 0.1128267057707237,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9865164858221838e-05,
+      "loss": 1.5046,
+      "step": 654
+    },
+    {
+      "epoch": 0.11299922366945571,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.9864719024867303e-05,
+      "loss": 1.6255,
+      "step": 655
+    },
+    {
+      "epoch": 0.1131717415681877,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.9864272460672738e-05,
+      "loss": 1.613,
+      "step": 656
+    },
+    {
+      "epoch": 0.11334425946691969,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.9863825165671238e-05,
+      "loss": 1.6889,
+      "step": 657
+    },
+    {
+      "epoch": 0.11351677736565169,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.9863377139895935e-05,
+      "loss": 1.6039,
+      "step": 658
+    },
+    {
+      "epoch": 0.11368929526438368,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.986292838338003e-05,
+      "loss": 1.6027,
+      "step": 659
+    },
+    {
+      "epoch": 0.11386181316311568,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.986247889615676e-05,
+      "loss": 1.5537,
+      "step": 660
+    },
+    {
+      "epoch": 0.11403433106184767,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9862028678259427e-05,
+      "loss": 1.5859,
+      "step": 661
+    },
+    {
+      "epoch": 0.11420684896057966,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.986157772972139e-05,
+      "loss": 1.5783,
+      "step": 662
+    },
+    {
+      "epoch": 0.11437936685931166,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.9861126050576063e-05,
+      "loss": 1.5792,
+      "step": 663
+    },
+    {
+      "epoch": 0.11455188475804365,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.9860673640856895e-05,
+      "loss": 1.5931,
+      "step": 664
+    },
+    {
+      "epoch": 0.11472440265677564,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.9860220500597415e-05,
+      "loss": 1.6084,
+      "step": 665
+    },
+    {
+      "epoch": 0.11489692055550764,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9859766629831192e-05,
+      "loss": 1.6543,
+      "step": 666
+    },
+    {
+      "epoch": 0.11506943845423963,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.9859312028591852e-05,
+      "loss": 1.61,
+      "step": 667
+    },
+    {
+      "epoch": 0.11524195635297162,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.9858856696913068e-05,
+      "loss": 1.5262,
+      "step": 668
+    },
+    {
+      "epoch": 0.11541447425170362,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.985840063482858e-05,
+      "loss": 1.5316,
+      "step": 669
+    },
+    {
+      "epoch": 0.11558699215043561,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.9857943842372175e-05,
+      "loss": 1.5745,
+      "step": 670
+    },
+    {
+      "epoch": 0.1157595100491676,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.9857486319577697e-05,
+      "loss": 1.5823,
+      "step": 671
+    },
+    {
+      "epoch": 0.1159320279478996,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.9857028066479042e-05,
+      "loss": 1.5883,
+      "step": 672
+    },
+    {
+      "epoch": 0.11610454584663159,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.9856569083110152e-05,
+      "loss": 1.6125,
+      "step": 673
+    },
+    {
+      "epoch": 0.11627706374536358,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.985610936950504e-05,
+      "loss": 1.5899,
+      "step": 674
+    },
+    {
+      "epoch": 0.11644958164409558,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.9855648925697762e-05,
+      "loss": 1.6004,
+      "step": 675
+    },
+    {
+      "epoch": 0.11662209954282757,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.9855187751722432e-05,
+      "loss": 1.5435,
+      "step": 676
+    },
+    {
+      "epoch": 0.11679461744155956,
+      "grad_norm": 0.625,
+      "learning_rate": 1.9854725847613216e-05,
+      "loss": 1.4848,
+      "step": 677
+    },
+    {
+      "epoch": 0.11696713534029156,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.9854263213404333e-05,
+      "loss": 1.5351,
+      "step": 678
+    },
+    {
+      "epoch": 0.11713965323902355,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.985379984913006e-05,
+      "loss": 1.6054,
+      "step": 679
+    },
+    {
+      "epoch": 0.11731217113775554,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.985333575482472e-05,
+      "loss": 1.5781,
+      "step": 680
+    },
+    {
+      "epoch": 0.11748468903648754,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.9852870930522704e-05,
+      "loss": 1.5546,
+      "step": 681
+    },
+    {
+      "epoch": 0.11765720693521953,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.985240537625845e-05,
+      "loss": 1.7013,
+      "step": 682
+    },
+    {
+      "epoch": 0.11782972483395152,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.985193909206644e-05,
+      "loss": 1.6284,
+      "step": 683
+    },
+    {
+      "epoch": 0.11800224273268352,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.9851472077981226e-05,
+      "loss": 1.6304,
+      "step": 684
+    },
+    {
+      "epoch": 0.1181747606314155,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.98510043340374e-05,
+      "loss": 1.6323,
+      "step": 685
+    },
+    {
+      "epoch": 0.11834727853014751,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.9850535860269628e-05,
+      "loss": 1.6954,
+      "step": 686
+    },
+    {
+      "epoch": 0.1185197964288795,
+      "grad_norm": 0.625,
+      "learning_rate": 1.985006665671261e-05,
+      "loss": 1.5755,
+      "step": 687
+    },
+    {
+      "epoch": 0.11869231432761149,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.984959672340111e-05,
+      "loss": 1.6307,
+      "step": 688
+    },
+    {
+      "epoch": 0.11886483222634349,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9849126060369933e-05,
+      "loss": 1.6133,
+      "step": 689
+    },
+    {
+      "epoch": 0.11903735012507548,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9848654667653964e-05,
+      "loss": 1.649,
+      "step": 690
+    },
+    {
+      "epoch": 0.11920986802380747,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.9848182545288117e-05,
+      "loss": 1.5091,
+      "step": 691
+    },
+    {
+      "epoch": 0.11938238592253947,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.9847709693307375e-05,
+      "loss": 1.6429,
+      "step": 692
+    },
+    {
+      "epoch": 0.11955490382127146,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.9847236111746767e-05,
+      "loss": 1.5849,
+      "step": 693
+    },
+    {
+      "epoch": 0.11972742172000345,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.9846761800641377e-05,
+      "loss": 1.5365,
+      "step": 694
+    },
+    {
+      "epoch": 0.11989993961873545,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.984628676002635e-05,
+      "loss": 1.5412,
+      "step": 695
+    },
+    {
+      "epoch": 0.12007245751746744,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9845810989936882e-05,
+      "loss": 1.4677,
+      "step": 696
+    },
+    {
+      "epoch": 0.12024497541619943,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.984533449040821e-05,
+      "loss": 1.5982,
+      "step": 697
+    },
+    {
+      "epoch": 0.12041749331493143,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.984485726147564e-05,
+      "loss": 1.7411,
+      "step": 698
+    },
+    {
+      "epoch": 0.12059001121366342,
+      "grad_norm": 0.75,
+      "learning_rate": 1.9844379303174537e-05,
+      "loss": 1.5752,
+      "step": 699
+    },
+    {
+      "epoch": 0.1207625291123954,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.9843900615540305e-05,
+      "loss": 1.5738,
+      "step": 700
+    },
+    {
+      "epoch": 0.1207625291123954,
+      "eval_loss": 1.5691492557525635,
+      "eval_runtime": 11.0575,
+      "eval_samples_per_second": 92.607,
+      "eval_steps_per_second": 23.152,
+      "step": 700
+    },
+    {
+      "epoch": 0.12093504701112741,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9843421198608404e-05,
+      "loss": 1.5288,
+      "step": 701
+    },
+    {
+      "epoch": 0.1211075649098594,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.9842941052414356e-05,
+      "loss": 1.5642,
+      "step": 702
+    },
+    {
+      "epoch": 0.12128008280859139,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.9842460176993734e-05,
+      "loss": 1.5566,
+      "step": 703
+    },
+    {
+      "epoch": 0.12145260070732339,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.9841978572382162e-05,
+      "loss": 1.5657,
+      "step": 704
+    },
+    {
+      "epoch": 0.12162511860605538,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9841496238615326e-05,
+      "loss": 1.5543,
+      "step": 705
+    },
+    {
+      "epoch": 0.12179763650478737,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.9841013175728953e-05,
+      "loss": 1.5774,
+      "step": 706
+    },
+    {
+      "epoch": 0.12197015440351937,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9840529383758834e-05,
+      "loss": 1.6671,
+      "step": 707
+    },
+    {
+      "epoch": 0.12214267230225136,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.9840044862740814e-05,
+      "loss": 1.6022,
+      "step": 708
+    },
+    {
+      "epoch": 0.12231519020098335,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.983955961271079e-05,
+      "loss": 1.5639,
+      "step": 709
+    },
+    {
+      "epoch": 0.12248770809971535,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.98390736337047e-05,
+      "loss": 1.6281,
+      "step": 710
+    },
+    {
+      "epoch": 0.12266022599844734,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9838586925758566e-05,
+      "loss": 1.6136,
+      "step": 711
+    },
+    {
+      "epoch": 0.12283274389717934,
+      "grad_norm": 0.625,
+      "learning_rate": 1.9838099488908437e-05,
+      "loss": 1.581,
+      "step": 712
+    },
+    {
+      "epoch": 0.12300526179591133,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.9837611323190423e-05,
+      "loss": 1.6113,
+      "step": 713
+    },
+    {
+      "epoch": 0.12317777969464332,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.9837122428640695e-05,
+      "loss": 1.5996,
+      "step": 714
+    },
+    {
+      "epoch": 0.12335029759337532,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9836632805295477e-05,
+      "loss": 1.7063,
+      "step": 715
+    },
+    {
+      "epoch": 0.12352281549210731,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.9836142453191032e-05,
+      "loss": 1.5825,
+      "step": 716
+    },
+    {
+      "epoch": 0.1236953333908393,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.98356513723637e-05,
+      "loss": 1.5981,
+      "step": 717
+    },
+    {
+      "epoch": 0.1238678512895713,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.9835159562849857e-05,
+      "loss": 1.5686,
+      "step": 718
+    },
+    {
+      "epoch": 0.12404036918830329,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.983466702468594e-05,
+      "loss": 1.5379,
+      "step": 719
+    },
+    {
+      "epoch": 0.12421288708703528,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.9834173757908442e-05,
+      "loss": 1.5893,
+      "step": 720
+    },
+    {
+      "epoch": 0.12438540498576728,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9833679762553905e-05,
+      "loss": 1.6384,
+      "step": 721
+    },
+    {
+      "epoch": 0.12455792288449927,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9833185038658923e-05,
+      "loss": 1.6167,
+      "step": 722
+    },
+    {
+      "epoch": 0.12473044078323126,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.9832689586260157e-05,
+      "loss": 1.5078,
+      "step": 723
+    },
+    {
+      "epoch": 0.12490295868196326,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.9832193405394306e-05,
+      "loss": 1.5687,
+      "step": 724
+    },
+    {
+      "epoch": 0.12507547658069526,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9831696496098135e-05,
+      "loss": 1.561,
+      "step": 725
+    },
+    {
+      "epoch": 0.12524799447942725,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9831198858408456e-05,
+      "loss": 1.5419,
+      "step": 726
+    },
+    {
+      "epoch": 0.12542051237815924,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.983070049236214e-05,
+      "loss": 1.5969,
+      "step": 727
+    },
+    {
+      "epoch": 0.12559303027689123,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.9830201397996104e-05,
+      "loss": 1.4889,
+      "step": 728
+    },
+    {
+      "epoch": 0.12576554817562322,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.9829701575347323e-05,
+      "loss": 1.5763,
+      "step": 729
+    },
+    {
+      "epoch": 0.1259380660743552,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9829201024452834e-05,
+      "loss": 1.5959,
+      "step": 730
+    },
+    {
+      "epoch": 0.12611058397308722,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.982869974534972e-05,
+      "loss": 1.6139,
+      "step": 731
+    },
+    {
+      "epoch": 0.1262831018718192,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.9828197738075114e-05,
+      "loss": 1.5739,
+      "step": 732
+    },
+    {
+      "epoch": 0.1264556197705512,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.982769500266621e-05,
+      "loss": 1.6035,
+      "step": 733
+    },
+    {
+      "epoch": 0.1266281376692832,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.9827191539160253e-05,
+      "loss": 1.5015,
+      "step": 734
+    },
+    {
+      "epoch": 0.12680065556801517,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.9826687347594548e-05,
+      "loss": 1.6663,
+      "step": 735
+    },
+    {
+      "epoch": 0.12697317346674716,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.9826182428006437e-05,
+      "loss": 1.4977,
+      "step": 736
+    },
+    {
+      "epoch": 0.12714569136547918,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.9825676780433342e-05,
+      "loss": 1.5396,
+      "step": 737
+    },
+    {
+      "epoch": 0.12731820926421117,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.9825170404912712e-05,
+      "loss": 1.5833,
+      "step": 738
+    },
+    {
+      "epoch": 0.12749072716294316,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.982466330148207e-05,
+      "loss": 1.6297,
+      "step": 739
+    },
+    {
+      "epoch": 0.12766324506167515,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9824155470178983e-05,
+      "loss": 1.5306,
+      "step": 740
+    },
+    {
+      "epoch": 0.12783576296040713,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.9823646911041076e-05,
+      "loss": 1.529,
+      "step": 741
+    },
+    {
+      "epoch": 0.12800828085913912,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.9823137624106023e-05,
+      "loss": 1.5001,
+      "step": 742
+    },
+    {
+      "epoch": 0.12818079875787114,
+      "grad_norm": 0.625,
+      "learning_rate": 1.982262760941156e-05,
+      "loss": 1.573,
+      "step": 743
+    },
+    {
+      "epoch": 0.12835331665660313,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9822116866995466e-05,
+      "loss": 1.6003,
+      "step": 744
+    },
+    {
+      "epoch": 0.12852583455533512,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.9821605396895584e-05,
+      "loss": 1.6081,
+      "step": 745
+    },
+    {
+      "epoch": 0.1286983524540671,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9821093199149806e-05,
+      "loss": 1.6188,
+      "step": 746
+    },
+    {
+      "epoch": 0.1288708703527991,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.982058027379608e-05,
+      "loss": 1.5679,
+      "step": 747
+    },
+    {
+      "epoch": 0.12904338825153108,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.9820066620872403e-05,
+      "loss": 1.582,
+      "step": 748
+    },
+    {
+      "epoch": 0.1292159061502631,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9819552240416832e-05,
+      "loss": 1.613,
+      "step": 749
+    },
+    {
+      "epoch": 0.1293884240489951,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.9819037132467478e-05,
+      "loss": 1.6215,
+      "step": 750
+    },
+    {
+      "epoch": 0.12956094194772708,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.98185212970625e-05,
+      "loss": 1.5503,
+      "step": 751
+    },
+    {
+      "epoch": 0.12973345984645906,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.9818004734240115e-05,
+      "loss": 1.6983,
+      "step": 752
+    },
+    {
+      "epoch": 0.12990597774519105,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.9817487444038594e-05,
+      "loss": 1.5299,
+      "step": 753
+    },
+    {
+      "epoch": 0.13007849564392307,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.981696942649626e-05,
+      "loss": 1.5083,
+      "step": 754
+    },
+    {
+      "epoch": 0.13025101354265506,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.9816450681651495e-05,
+      "loss": 1.5681,
+      "step": 755
+    },
+    {
+      "epoch": 0.13042353144138705,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.9815931209542723e-05,
+      "loss": 1.5735,
+      "step": 756
+    },
+    {
+      "epoch": 0.13059604934011904,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9815411010208438e-05,
+      "loss": 1.5884,
+      "step": 757
+    },
+    {
+      "epoch": 0.13076856723885102,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9814890083687172e-05,
+      "loss": 1.5198,
+      "step": 758
+    },
+    {
+      "epoch": 0.130941085137583,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9814368430017526e-05,
+      "loss": 1.5571,
+      "step": 759
+    },
+    {
+      "epoch": 0.13111360303631503,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.9813846049238143e-05,
+      "loss": 1.597,
+      "step": 760
+    },
+    {
+      "epoch": 0.13128612093504702,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.981332294138772e-05,
+      "loss": 1.5261,
+      "step": 761
+    },
+    {
+      "epoch": 0.131458638833779,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.981279910650502e-05,
+      "loss": 1.711,
+      "step": 762
+    },
+    {
+      "epoch": 0.131631156732511,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.981227454462885e-05,
+      "loss": 1.6753,
+      "step": 763
+    },
+    {
+      "epoch": 0.13180367463124298,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.9811749255798074e-05,
+      "loss": 1.559,
+      "step": 764
+    },
+    {
+      "epoch": 0.13197619252997497,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.98112232400516e-05,
+      "loss": 1.648,
+      "step": 765
+    },
+    {
+      "epoch": 0.132148710428707,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.9810696497428412e-05,
+      "loss": 1.6339,
+      "step": 766
+    },
+    {
+      "epoch": 0.13232122832743898,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9810169027967524e-05,
+      "loss": 1.5987,
+      "step": 767
+    },
+    {
+      "epoch": 0.13249374622617097,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.9809640831708018e-05,
+      "loss": 1.538,
+      "step": 768
+    },
+    {
+      "epoch": 0.13266626412490296,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.9809111908689028e-05,
+      "loss": 1.4946,
+      "step": 769
+    },
+    {
+      "epoch": 0.13283878202363494,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.9808582258949735e-05,
+      "loss": 1.574,
+      "step": 770
+    },
+    {
+      "epoch": 0.13301129992236693,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.980805188252938e-05,
+      "loss": 1.529,
+      "step": 771
+    },
+    {
+      "epoch": 0.13318381782109895,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.980752077946726e-05,
+      "loss": 1.519,
+      "step": 772
+    },
+    {
+      "epoch": 0.13335633571983094,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.9806988949802722e-05,
+      "loss": 1.6591,
+      "step": 773
+    },
+    {
+      "epoch": 0.13352885361856293,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9806456393575164e-05,
+      "loss": 1.5509,
+      "step": 774
+    },
+    {
+      "epoch": 0.13370137151729491,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.980592311082404e-05,
+      "loss": 1.5898,
+      "step": 775
+    },
+    {
+      "epoch": 0.1338738894160269,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9805389101588868e-05,
+      "loss": 1.5729,
+      "step": 776
+    },
+    {
+      "epoch": 0.13404640731475892,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.9804854365909202e-05,
+      "loss": 1.6551,
+      "step": 777
+    },
+    {
+      "epoch": 0.1342189252134909,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.980431890382466e-05,
+      "loss": 1.5785,
+      "step": 778
+    },
+    {
+      "epoch": 0.1343914431122229,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.9803782715374912e-05,
+      "loss": 1.6861,
+      "step": 779
+    },
+    {
+      "epoch": 0.13456396101095489,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.9803245800599685e-05,
+      "loss": 1.4989,
+      "step": 780
+    },
+    {
+      "epoch": 0.13473647890968687,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.9802708159538755e-05,
+      "loss": 1.6312,
+      "step": 781
+    },
+    {
+      "epoch": 0.13490899680841886,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.9802169792231955e-05,
+      "loss": 1.5757,
+      "step": 782
+    },
+    {
+      "epoch": 0.13508151470715088,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.980163069871917e-05,
+      "loss": 1.649,
+      "step": 783
+    },
+    {
+      "epoch": 0.13525403260588287,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9801090879040342e-05,
+      "loss": 1.6765,
+      "step": 784
+    },
+    {
+      "epoch": 0.13542655050461486,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.9800550333235455e-05,
+      "loss": 1.5596,
+      "step": 785
+    },
+    {
+      "epoch": 0.13559906840334685,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.980000906134457e-05,
+      "loss": 1.5821,
+      "step": 786
+    },
+    {
+      "epoch": 0.13577158630207883,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.9799467063407777e-05,
+      "loss": 1.4838,
+      "step": 787
+    },
+    {
+      "epoch": 0.13594410420081082,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.9798924339465232e-05,
+      "loss": 1.6303,
+      "step": 788
+    },
+    {
+      "epoch": 0.13611662209954284,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.979838088955715e-05,
+      "loss": 1.5389,
+      "step": 789
+    },
+    {
+      "epoch": 0.13628913999827483,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.9797836713723786e-05,
+      "loss": 1.5639,
+      "step": 790
+    },
+    {
+      "epoch": 0.13646165789700682,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9797291812005458e-05,
+      "loss": 1.5259,
+      "step": 791
+    },
+    {
+      "epoch": 0.1366341757957388,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.9796746184442538e-05,
+      "loss": 1.6243,
+      "step": 792
+    },
+    {
+      "epoch": 0.1368066936944708,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.9796199831075445e-05,
+      "loss": 1.5038,
+      "step": 793
+    },
+    {
+      "epoch": 0.13697921159320278,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.9795652751944662e-05,
+      "loss": 1.5741,
+      "step": 794
+    },
+    {
+      "epoch": 0.1371517294919348,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.9795104947090714e-05,
+      "loss": 1.5581,
+      "step": 795
+    },
+    {
+      "epoch": 0.1373242473906668,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9794556416554193e-05,
+      "loss": 1.539,
+      "step": 796
+    },
+    {
+      "epoch": 0.13749676528939878,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.979400716037573e-05,
+      "loss": 1.6923,
+      "step": 797
+    },
+    {
+      "epoch": 0.13766928318813076,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.979345717859602e-05,
+      "loss": 1.5974,
+      "step": 798
+    },
+    {
+      "epoch": 0.13784180108686275,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9792906471255814e-05,
+      "loss": 1.6278,
+      "step": 799
+    },
+    {
+      "epoch": 0.13801431898559474,
+      "grad_norm": 0.75,
+      "learning_rate": 1.9792355038395906e-05,
+      "loss": 1.6722,
+      "step": 800
+    },
+    {
+      "epoch": 0.13801431898559474,
+      "eval_loss": 1.557534098625183,
+      "eval_runtime": 10.7703,
+      "eval_samples_per_second": 95.077,
+      "eval_steps_per_second": 23.769,
+      "step": 800
+    },
+    {
+      "epoch": 0.13818683688432676,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.979180288005715e-05,
+      "loss": 1.5548,
+      "step": 801
+    },
+    {
+      "epoch": 0.13835935478305875,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.9791249996280456e-05,
+      "loss": 1.5351,
+      "step": 802
+    },
+    {
+      "epoch": 0.13853187268179074,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.9790696387106782e-05,
+      "loss": 1.4778,
+      "step": 803
+    },
+    {
+      "epoch": 0.13870439058052272,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9790142052577148e-05,
+      "loss": 1.5723,
+      "step": 804
+    },
+    {
+      "epoch": 0.1388769084792547,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9789586992732615e-05,
+      "loss": 1.642,
+      "step": 805
+    },
+    {
+      "epoch": 0.13904942637798673,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9789031207614312e-05,
+      "loss": 1.5624,
+      "step": 806
+    },
+    {
+      "epoch": 0.13922194427671872,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.978847469726341e-05,
+      "loss": 1.5436,
+      "step": 807
+    },
+    {
+      "epoch": 0.1393944621754507,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.9787917461721143e-05,
+      "loss": 1.5336,
+      "step": 808
+    },
+    {
+      "epoch": 0.1395669800741827,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.9787359501028795e-05,
+      "loss": 1.5797,
+      "step": 809
+    },
+    {
+      "epoch": 0.13973949797291468,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.97868008152277e-05,
+      "loss": 1.4998,
+      "step": 810
+    },
+    {
+      "epoch": 0.13991201587164667,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9786241404359247e-05,
+      "loss": 1.6379,
+      "step": 811
+    },
+    {
+      "epoch": 0.1400845337703787,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.978568126846488e-05,
+      "loss": 1.6053,
+      "step": 812
+    },
+    {
+      "epoch": 0.14025705166911068,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.978512040758611e-05,
+      "loss": 1.6019,
+      "step": 813
+    },
+    {
+      "epoch": 0.14042956956784267,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.9784558821764476e-05,
+      "loss": 1.6052,
+      "step": 814
+    },
+    {
+      "epoch": 0.14060208746657465,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.978399651104159e-05,
+      "loss": 1.6307,
+      "step": 815
+    },
+    {
+      "epoch": 0.14077460536530664,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9783433475459103e-05,
+      "loss": 1.5099,
+      "step": 816
+    },
+    {
+      "epoch": 0.14094712326403863,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.9782869715058738e-05,
+      "loss": 1.5855,
+      "step": 817
+    },
+    {
+      "epoch": 0.14111964116277065,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.978230522988226e-05,
+      "loss": 1.5681,
+      "step": 818
+    },
+    {
+      "epoch": 0.14129215906150264,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.9781740019971485e-05,
+      "loss": 1.6274,
+      "step": 819
+    },
+    {
+      "epoch": 0.14146467696023463,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9781174085368292e-05,
+      "loss": 1.6156,
+      "step": 820
+    },
+    {
+      "epoch": 0.14163719485896661,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.9780607426114606e-05,
+      "loss": 1.6525,
+      "step": 821
+    },
+    {
+      "epoch": 0.1418097127576986,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.9780040042252412e-05,
+      "loss": 1.4553,
+      "step": 822
+    },
+    {
+      "epoch": 0.1419822306564306,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.977947193382374e-05,
+      "loss": 1.5795,
+      "step": 823
+    },
+    {
+      "epoch": 0.1421547485551626,
+      "grad_norm": 0.625,
+      "learning_rate": 1.9778903100870687e-05,
+      "loss": 1.5352,
+      "step": 824
+    },
+    {
+      "epoch": 0.1423272664538946,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9778333543435387e-05,
+      "loss": 1.4478,
+      "step": 825
+    },
+    {
+      "epoch": 0.14249978435262659,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.977776326156004e-05,
+      "loss": 1.609,
+      "step": 826
+    },
+    {
+      "epoch": 0.14267230225135857,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9777192255286897e-05,
+      "loss": 1.6362,
+      "step": 827
+    },
+    {
+      "epoch": 0.14284482015009056,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.977662052465826e-05,
+      "loss": 1.5537,
+      "step": 828
+    },
+    {
+      "epoch": 0.14301733804882258,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.977604806971649e-05,
+      "loss": 1.6069,
+      "step": 829
+    },
+    {
+      "epoch": 0.14318985594755457,
+      "grad_norm": 0.625,
+      "learning_rate": 1.9775474890503996e-05,
+      "loss": 1.5911,
+      "step": 830
+    },
+    {
+      "epoch": 0.14336237384628656,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.9774900987063237e-05,
+      "loss": 1.5657,
+      "step": 831
+    },
+    {
+      "epoch": 0.14353489174501854,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.9774326359436743e-05,
+      "loss": 1.605,
+      "step": 832
+    },
+    {
+      "epoch": 0.14370740964375053,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.9773751007667074e-05,
+      "loss": 1.5102,
+      "step": 833
+    },
+    {
+      "epoch": 0.14387992754248252,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.9773174931796864e-05,
+      "loss": 1.5579,
+      "step": 834
+    },
+    {
+      "epoch": 0.14405244544121454,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.977259813186879e-05,
+      "loss": 1.558,
+      "step": 835
+    },
+    {
+      "epoch": 0.14422496333994653,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.977202060792558e-05,
+      "loss": 1.5766,
+      "step": 836
+    },
+    {
+      "epoch": 0.14439748123867852,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.977144236001003e-05,
+      "loss": 1.5227,
+      "step": 837
+    },
+    {
+      "epoch": 0.1445699991374105,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.977086338816497e-05,
+      "loss": 1.6174,
+      "step": 838
+    },
+    {
+      "epoch": 0.1447425170361425,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9770283692433306e-05,
+      "loss": 1.5583,
+      "step": 839
+    },
+    {
+      "epoch": 0.14491503493487448,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.9769703272857976e-05,
+      "loss": 1.4618,
+      "step": 840
+    },
+    {
+      "epoch": 0.1450875528336065,
+      "grad_norm": 2.84375,
+      "learning_rate": 1.976912212948198e-05,
+      "loss": 1.5849,
+      "step": 841
+    },
+    {
+      "epoch": 0.1452600707323385,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.976854026234838e-05,
+      "loss": 1.5698,
+      "step": 842
+    },
+    {
+      "epoch": 0.14543258863107048,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.9767957671500277e-05,
+      "loss": 1.5552,
+      "step": 843
+    },
+    {
+      "epoch": 0.14560510652980246,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.9767374356980838e-05,
+      "loss": 1.5065,
+      "step": 844
+    },
+    {
+      "epoch": 0.14577762442853445,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.976679031883328e-05,
+      "loss": 1.5852,
+      "step": 845
+    },
+    {
+      "epoch": 0.14595014232726644,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.976620555710087e-05,
+      "loss": 1.6179,
+      "step": 846
+    },
+    {
+      "epoch": 0.14612266022599846,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9765620071826928e-05,
+      "loss": 1.5693,
+      "step": 847
+    },
+    {
+      "epoch": 0.14629517812473045,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.976503386305483e-05,
+      "loss": 1.4705,
+      "step": 848
+    },
+    {
+      "epoch": 0.14646769602346243,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9764446930828015e-05,
+      "loss": 1.5468,
+      "step": 849
+    },
+    {
+      "epoch": 0.14664021392219442,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9763859275189956e-05,
+      "loss": 1.4668,
+      "step": 850
+    },
+    {
+      "epoch": 0.1468127318209264,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.9763270896184195e-05,
+      "loss": 1.5784,
+      "step": 851
+    },
+    {
+      "epoch": 0.1469852497196584,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.9762681793854323e-05,
+      "loss": 1.5303,
+      "step": 852
+    },
+    {
+      "epoch": 0.14715776761839042,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9762091968243982e-05,
+      "loss": 1.5035,
+      "step": 853
+    },
+    {
+      "epoch": 0.1473302855171224,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.9761501419396875e-05,
+      "loss": 1.6431,
+      "step": 854
+    },
+    {
+      "epoch": 0.1475028034158544,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.9760910147356743e-05,
+      "loss": 1.586,
+      "step": 855
+    },
+    {
+      "epoch": 0.14767532131458638,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9760318152167406e-05,
+      "loss": 1.5336,
+      "step": 856
+    },
+    {
+      "epoch": 0.14784783921331837,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.9759725433872713e-05,
+      "loss": 1.5582,
+      "step": 857
+    },
+    {
+      "epoch": 0.1480203571120504,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.9759131992516575e-05,
+      "loss": 1.607,
+      "step": 858
+    },
+    {
+      "epoch": 0.14819287501078238,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.9758537828142966e-05,
+      "loss": 1.5921,
+      "step": 859
+    },
+    {
+      "epoch": 0.14836539290951437,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.9757942940795897e-05,
+      "loss": 1.5577,
+      "step": 860
+    },
+    {
+      "epoch": 0.14853791080824635,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.975734733051945e-05,
+      "loss": 1.6636,
+      "step": 861
+    },
+    {
+      "epoch": 0.14871042870697834,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9756750997357738e-05,
+      "loss": 1.6487,
+      "step": 862
+    },
+    {
+      "epoch": 0.14888294660571033,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.9756153941354955e-05,
+      "loss": 1.5294,
+      "step": 863
+    },
+    {
+      "epoch": 0.14905546450444235,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.9755556162555323e-05,
+      "loss": 1.5988,
+      "step": 864
+    },
+    {
+      "epoch": 0.14922798240317434,
+      "grad_norm": 0.625,
+      "learning_rate": 1.975495766100314e-05,
+      "loss": 1.5239,
+      "step": 865
+    },
+    {
+      "epoch": 0.14940050030190632,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.975435843674274e-05,
+      "loss": 1.6436,
+      "step": 866
+    },
+    {
+      "epoch": 0.1495730182006383,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.975375848981852e-05,
+      "loss": 1.6203,
+      "step": 867
+    },
+    {
+      "epoch": 0.1497455360993703,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.9753157820274924e-05,
+      "loss": 1.6343,
+      "step": 868
+    },
+    {
+      "epoch": 0.1499180539981023,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.975255642815646e-05,
+      "loss": 1.6688,
+      "step": 869
+    },
+    {
+      "epoch": 0.1500905718968343,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.9751954313507674e-05,
+      "loss": 1.6492,
+      "step": 870
+    },
+    {
+      "epoch": 0.1502630897955663,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.9751351476373184e-05,
+      "loss": 1.6285,
+      "step": 871
+    },
+    {
+      "epoch": 0.15043560769429828,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.975074791679765e-05,
+      "loss": 1.6182,
+      "step": 872
+    },
+    {
+      "epoch": 0.15060812559303027,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.9750143634825776e-05,
+      "loss": 1.6117,
+      "step": 873
+    },
+    {
+      "epoch": 0.15078064349176226,
+      "grad_norm": 0.75,
+      "learning_rate": 1.9749538630502346e-05,
+      "loss": 1.5948,
+      "step": 874
+    },
+    {
+      "epoch": 0.15095316139049425,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9748932903872176e-05,
+      "loss": 1.5501,
+      "step": 875
+    },
+    {
+      "epoch": 0.15112567928922627,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.974832645498014e-05,
+      "loss": 1.5538,
+      "step": 876
+    },
+    {
+      "epoch": 0.15129819718795826,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9747719283871172e-05,
+      "loss": 1.534,
+      "step": 877
+    },
+    {
+      "epoch": 0.15147071508669024,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.974711139059025e-05,
+      "loss": 1.5364,
+      "step": 878
+    },
+    {
+      "epoch": 0.15164323298542223,
+      "grad_norm": 0.625,
+      "learning_rate": 1.9746502775182415e-05,
+      "loss": 1.5412,
+      "step": 879
+    },
+    {
+      "epoch": 0.15181575088415422,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.9745893437692757e-05,
+      "loss": 1.4718,
+      "step": 880
+    },
+    {
+      "epoch": 0.15198826878288624,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.9745283378166417e-05,
+      "loss": 1.6271,
+      "step": 881
+    },
+    {
+      "epoch": 0.15216078668161823,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.9744672596648593e-05,
+      "loss": 1.6166,
+      "step": 882
+    },
+    {
+      "epoch": 0.15233330458035022,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.9744061093184537e-05,
+      "loss": 1.5178,
+      "step": 883
+    },
+    {
+      "epoch": 0.1525058224790822,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.974344886781955e-05,
+      "loss": 1.5554,
+      "step": 884
+    },
+    {
+      "epoch": 0.1526783403778142,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9742835920598988e-05,
+      "loss": 1.5596,
+      "step": 885
+    },
+    {
+      "epoch": 0.15285085827654618,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.974222225156827e-05,
+      "loss": 1.6621,
+      "step": 886
+    },
+    {
+      "epoch": 0.1530233761752782,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.974160786077285e-05,
+      "loss": 1.6319,
+      "step": 887
+    },
+    {
+      "epoch": 0.15319589407401019,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9740992748258258e-05,
+      "loss": 1.5838,
+      "step": 888
+    },
+    {
+      "epoch": 0.15336841197274217,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.9740376914070055e-05,
+      "loss": 1.5944,
+      "step": 889
+    },
+    {
+      "epoch": 0.15354092987147416,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9739760358253867e-05,
+      "loss": 1.6164,
+      "step": 890
+    },
+    {
+      "epoch": 0.15371344777020615,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.973914308085538e-05,
+      "loss": 1.6087,
+      "step": 891
+    },
+    {
+      "epoch": 0.15388596566893814,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9738525081920316e-05,
+      "loss": 1.5904,
+      "step": 892
+    },
+    {
+      "epoch": 0.15405848356767016,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.9737906361494467e-05,
+      "loss": 1.5325,
+      "step": 893
+    },
+    {
+      "epoch": 0.15423100146640215,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.973728691962367e-05,
+      "loss": 1.5605,
+      "step": 894
+    },
+    {
+      "epoch": 0.15440351936513413,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.973666675635382e-05,
+      "loss": 1.6023,
+      "step": 895
+    },
+    {
+      "epoch": 0.15457603726386612,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.9736045871730854e-05,
+      "loss": 1.5483,
+      "step": 896
+    },
+    {
+      "epoch": 0.1547485551625981,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.9735424265800775e-05,
+      "loss": 1.587,
+      "step": 897
+    },
+    {
+      "epoch": 0.1549210730613301,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.973480193860964e-05,
+      "loss": 1.5754,
+      "step": 898
+    },
+    {
+      "epoch": 0.15509359096006212,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.973417889020355e-05,
+      "loss": 1.6912,
+      "step": 899
+    },
+    {
+      "epoch": 0.1552661088587941,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.9733555120628666e-05,
+      "loss": 1.4927,
+      "step": 900
+    },
+    {
+      "epoch": 0.1552661088587941,
+      "eval_loss": 1.5473740100860596,
+      "eval_runtime": 10.7618,
+      "eval_samples_per_second": 95.151,
+      "eval_steps_per_second": 23.788,
+      "step": 900
+    },
+    {
+      "epoch": 0.1554386267575261,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.97329306299312e-05,
+      "loss": 1.4499,
+      "step": 901
+    },
+    {
+      "epoch": 0.15561114465625808,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9732305418157423e-05,
+      "loss": 1.5901,
+      "step": 902
+    },
+    {
+      "epoch": 0.15578366255499007,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.9731679485353645e-05,
+      "loss": 1.4815,
+      "step": 903
+    },
+    {
+      "epoch": 0.1559561804537221,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.9731052831566248e-05,
+      "loss": 1.5613,
+      "step": 904
+    },
+    {
+      "epoch": 0.15612869835245408,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.9730425456841652e-05,
+      "loss": 1.5202,
+      "step": 905
+    },
+    {
+      "epoch": 0.15630121625118606,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.9729797361226342e-05,
+      "loss": 1.5046,
+      "step": 906
+    },
+    {
+      "epoch": 0.15647373414991805,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9729168544766848e-05,
+      "loss": 1.6204,
+      "step": 907
+    },
+    {
+      "epoch": 0.15664625204865004,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9728539007509756e-05,
+      "loss": 1.5183,
+      "step": 908
+    },
+    {
+      "epoch": 0.15681876994738203,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.972790874950171e-05,
+      "loss": 1.4783,
+      "step": 909
+    },
+    {
+      "epoch": 0.15699128784611405,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.9727277770789398e-05,
+      "loss": 1.5652,
+      "step": 910
+    },
+    {
+      "epoch": 0.15716380574484604,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.9726646071419573e-05,
+      "loss": 1.4987,
+      "step": 911
+    },
+    {
+      "epoch": 0.15733632364357802,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9726013651439033e-05,
+      "loss": 1.5341,
+      "step": 912
+    },
+    {
+      "epoch": 0.15750884154231,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.972538051089463e-05,
+      "loss": 1.609,
+      "step": 913
+    },
+    {
+      "epoch": 0.157681359441042,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.972474664983327e-05,
+      "loss": 1.5571,
+      "step": 914
+    },
+    {
+      "epoch": 0.157853877339774,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.9724112068301914e-05,
+      "loss": 1.5965,
+      "step": 915
+    },
+    {
+      "epoch": 0.158026395238506,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.972347676634758e-05,
+      "loss": 1.567,
+      "step": 916
+    },
+    {
+      "epoch": 0.158198913137238,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9722840744017332e-05,
+      "loss": 1.5872,
+      "step": 917
+    },
+    {
+      "epoch": 0.15837143103596998,
+      "grad_norm": 0.625,
+      "learning_rate": 1.972220400135829e-05,
+      "loss": 1.5357,
+      "step": 918
+    },
+    {
+      "epoch": 0.15854394893470197,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.9721566538417626e-05,
+      "loss": 1.6097,
+      "step": 919
+    },
+    {
+      "epoch": 0.15871646683343396,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.972092835524257e-05,
+      "loss": 1.5192,
+      "step": 920
+    },
+    {
+      "epoch": 0.15888898473216595,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.9720289451880407e-05,
+      "loss": 1.5389,
+      "step": 921
+    },
+    {
+      "epoch": 0.15906150263089797,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.9719649828378464e-05,
+      "loss": 1.6057,
+      "step": 922
+    },
+    {
+      "epoch": 0.15923402052962995,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9719009484784125e-05,
+      "loss": 1.5346,
+      "step": 923
+    },
+    {
+      "epoch": 0.15940653842836194,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.9718368421144842e-05,
+      "loss": 1.6481,
+      "step": 924
+    },
+    {
+      "epoch": 0.15957905632709393,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.97177266375081e-05,
+      "loss": 1.6184,
+      "step": 925
+    },
+    {
+      "epoch": 0.15975157422582592,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9717084133921456e-05,
+      "loss": 1.6208,
+      "step": 926
+    },
+    {
+      "epoch": 0.1599240921245579,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.97164409104325e-05,
+      "loss": 1.6104,
+      "step": 927
+    },
+    {
+      "epoch": 0.16009661002328993,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.9715796967088888e-05,
+      "loss": 1.6334,
+      "step": 928
+    },
+    {
+      "epoch": 0.16026912792202191,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.971515230393833e-05,
+      "loss": 1.5719,
+      "step": 929
+    },
+    {
+      "epoch": 0.1604416458207539,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.971450692102859e-05,
+      "loss": 1.6079,
+      "step": 930
+    },
+    {
+      "epoch": 0.1606141637194859,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.9713860818407474e-05,
+      "loss": 1.5634,
+      "step": 931
+    },
+    {
+      "epoch": 0.16078668161821788,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.9713213996122857e-05,
+      "loss": 1.5833,
+      "step": 932
+    },
+    {
+      "epoch": 0.1609591995169499,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.9712566454222653e-05,
+      "loss": 1.5331,
+      "step": 933
+    },
+    {
+      "epoch": 0.16113171741568189,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.9711918192754842e-05,
+      "loss": 1.5762,
+      "step": 934
+    },
+    {
+      "epoch": 0.16130423531441387,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.9711269211767446e-05,
+      "loss": 1.63,
+      "step": 935
+    },
+    {
+      "epoch": 0.16147675321314586,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.971061951130855e-05,
+      "loss": 1.6418,
+      "step": 936
+    },
+    {
+      "epoch": 0.16164927111187785,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.9709969091426288e-05,
+      "loss": 1.5717,
+      "step": 937
+    },
+    {
+      "epoch": 0.16182178901060984,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.970931795216884e-05,
+      "loss": 1.524,
+      "step": 938
+    },
+    {
+      "epoch": 0.16199430690934186,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.9708666093584456e-05,
+      "loss": 1.5546,
+      "step": 939
+    },
+    {
+      "epoch": 0.16216682480807385,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.970801351572142e-05,
+      "loss": 1.6206,
+      "step": 940
+    },
+    {
+      "epoch": 0.16233934270680583,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.970736021862809e-05,
+      "loss": 1.5482,
+      "step": 941
+    },
+    {
+      "epoch": 0.16251186060553782,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9706706202352856e-05,
+      "loss": 1.5676,
+      "step": 942
+    },
+    {
+      "epoch": 0.1626843785042698,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.970605146694418e-05,
+      "loss": 1.625,
+      "step": 943
+    },
+    {
+      "epoch": 0.1628568964030018,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.9705396012450563e-05,
+      "loss": 1.5227,
+      "step": 944
+    },
+    {
+      "epoch": 0.16302941430173382,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9704739838920565e-05,
+      "loss": 1.6658,
+      "step": 945
+    },
+    {
+      "epoch": 0.1632019322004658,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.9704082946402805e-05,
+      "loss": 1.5264,
+      "step": 946
+    },
+    {
+      "epoch": 0.1633744500991978,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.9703425334945945e-05,
+      "loss": 1.5827,
+      "step": 947
+    },
+    {
+      "epoch": 0.16354696799792978,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.9702767004598708e-05,
+      "loss": 1.5444,
+      "step": 948
+    },
+    {
+      "epoch": 0.16371948589666177,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.9702107955409862e-05,
+      "loss": 1.5957,
+      "step": 949
+    },
+    {
+      "epoch": 0.16389200379539376,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.9701448187428244e-05,
+      "loss": 1.5006,
+      "step": 950
+    },
+    {
+      "epoch": 0.16406452169412578,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.970078770070272e-05,
+      "loss": 1.5879,
+      "step": 951
+    },
+    {
+      "epoch": 0.16423703959285776,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.9700126495282234e-05,
+      "loss": 1.5882,
+      "step": 952
+    },
+    {
+      "epoch": 0.16440955749158975,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.9699464571215765e-05,
+      "loss": 1.5719,
+      "step": 953
+    },
+    {
+      "epoch": 0.16458207539032174,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9698801928552358e-05,
+      "loss": 1.5701,
+      "step": 954
+    },
+    {
+      "epoch": 0.16475459328905373,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.96981385673411e-05,
+      "loss": 1.6497,
+      "step": 955
+    },
+    {
+      "epoch": 0.16492711118778575,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.9697474487631143e-05,
+      "loss": 1.5415,
+      "step": 956
+    },
+    {
+      "epoch": 0.16509962908651774,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.9696809689471682e-05,
+      "loss": 1.5721,
+      "step": 957
+    },
+    {
+      "epoch": 0.16527214698524972,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.9696144172911974e-05,
+      "loss": 1.5413,
+      "step": 958
+    },
+    {
+      "epoch": 0.1654446648839817,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9695477938001316e-05,
+      "loss": 1.5264,
+      "step": 959
+    },
+    {
+      "epoch": 0.1656171827827137,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.9694810984789074e-05,
+      "loss": 1.5174,
+      "step": 960
+    },
+    {
+      "epoch": 0.1657897006814457,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.969414331332466e-05,
+      "loss": 1.5387,
+      "step": 961
+    },
+    {
+      "epoch": 0.1659622185801777,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9693474923657536e-05,
+      "loss": 1.5397,
+      "step": 962
+    },
+    {
+      "epoch": 0.1661347364789097,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9692805815837224e-05,
+      "loss": 1.5345,
+      "step": 963
+    },
+    {
+      "epoch": 0.16630725437764168,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9692135989913294e-05,
+      "loss": 1.487,
+      "step": 964
+    },
+    {
+      "epoch": 0.16647977227637367,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.969146544593537e-05,
+      "loss": 1.6524,
+      "step": 965
+    },
+    {
+      "epoch": 0.16665229017510566,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.9690794183953128e-05,
+      "loss": 1.667,
+      "step": 966
+    },
+    {
+      "epoch": 0.16682480807383765,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.96901222040163e-05,
+      "loss": 1.5474,
+      "step": 967
+    },
+    {
+      "epoch": 0.16699732597256967,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.968944950617468e-05,
+      "loss": 1.5477,
+      "step": 968
+    },
+    {
+      "epoch": 0.16716984387130165,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.9688776090478096e-05,
+      "loss": 1.5634,
+      "step": 969
+    },
+    {
+      "epoch": 0.16734236177003364,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.9688101956976436e-05,
+      "loss": 1.5918,
+      "step": 970
+    },
+    {
+      "epoch": 0.16751487966876563,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.968742710571965e-05,
+      "loss": 1.6563,
+      "step": 971
+    },
+    {
+      "epoch": 0.16768739756749762,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.968675153675774e-05,
+      "loss": 1.5484,
+      "step": 972
+    },
+    {
+      "epoch": 0.1678599154662296,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.9686075250140745e-05,
+      "loss": 1.5401,
+      "step": 973
+    },
+    {
+      "epoch": 0.16803243336496163,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9685398245918778e-05,
+      "loss": 1.5453,
+      "step": 974
+    },
+    {
+      "epoch": 0.16820495126369361,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.9684720524141988e-05,
+      "loss": 1.5738,
+      "step": 975
+    },
+    {
+      "epoch": 0.1683774691624256,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.9684042084860594e-05,
+      "loss": 1.5588,
+      "step": 976
+    },
+    {
+      "epoch": 0.1685499870611576,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.968336292812485e-05,
+      "loss": 1.5041,
+      "step": 977
+    },
+    {
+      "epoch": 0.16872250495988958,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.9682683053985073e-05,
+      "loss": 1.5881,
+      "step": 978
+    },
+    {
+      "epoch": 0.16889502285862157,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.968200246249164e-05,
+      "loss": 1.5471,
+      "step": 979
+    },
+    {
+      "epoch": 0.16906754075735358,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.9681321153694967e-05,
+      "loss": 1.5248,
+      "step": 980
+    },
+    {
+      "epoch": 0.16924005865608557,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.968063912764553e-05,
+      "loss": 1.5456,
+      "step": 981
+    },
+    {
+      "epoch": 0.16941257655481756,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.967995638439386e-05,
+      "loss": 1.56,
+      "step": 982
+    },
+    {
+      "epoch": 0.16958509445354955,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.967927292399054e-05,
+      "loss": 1.5441,
+      "step": 983
+    },
+    {
+      "epoch": 0.16975761235228154,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9678588746486198e-05,
+      "loss": 1.5252,
+      "step": 984
+    },
+    {
+      "epoch": 0.16993013025101356,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.967790385193153e-05,
+      "loss": 1.54,
+      "step": 985
+    },
+    {
+      "epoch": 0.17010264814974554,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.9677218240377272e-05,
+      "loss": 1.6079,
+      "step": 986
+    },
+    {
+      "epoch": 0.17027516604847753,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.9676531911874223e-05,
+      "loss": 1.5549,
+      "step": 987
+    },
+    {
+      "epoch": 0.17044768394720952,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.967584486647323e-05,
+      "loss": 1.6047,
+      "step": 988
+    },
+    {
+      "epoch": 0.1706202018459415,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.9675157104225188e-05,
+      "loss": 1.6402,
+      "step": 989
+    },
+    {
+      "epoch": 0.1707927197446735,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.9674468625181058e-05,
+      "loss": 1.5565,
+      "step": 990
+    },
+    {
+      "epoch": 0.17096523764340552,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.967377942939184e-05,
+      "loss": 1.4881,
+      "step": 991
+    },
+    {
+      "epoch": 0.1711377555421375,
+      "grad_norm": 0.75,
+      "learning_rate": 1.96730895169086e-05,
+      "loss": 1.6077,
+      "step": 992
+    },
+    {
+      "epoch": 0.1713102734408695,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.9672398887782448e-05,
+      "loss": 1.486,
+      "step": 993
+    },
+    {
+      "epoch": 0.17148279133960148,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.967170754206455e-05,
+      "loss": 1.6017,
+      "step": 994
+    },
+    {
+      "epoch": 0.17165530923833347,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.9671015479806126e-05,
+      "loss": 1.6084,
+      "step": 995
+    },
+    {
+      "epoch": 0.17182782713706546,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.9670322701058447e-05,
+      "loss": 1.5449,
+      "step": 996
+    },
+    {
+      "epoch": 0.17200034503579748,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.966962920587284e-05,
+      "loss": 1.5863,
+      "step": 997
+    },
+    {
+      "epoch": 0.17217286293452946,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.9668934994300684e-05,
+      "loss": 1.5316,
+      "step": 998
+    },
+    {
+      "epoch": 0.17234538083326145,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.9668240066393406e-05,
+      "loss": 1.56,
+      "step": 999
+    },
+    {
+      "epoch": 0.17251789873199344,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.9667544422202497e-05,
+      "loss": 1.5559,
+      "step": 1000
+    },
+    {
+      "epoch": 0.17251789873199344,
+      "eval_loss": 1.538590908050537,
+      "eval_runtime": 11.0756,
+      "eval_samples_per_second": 92.455,
+      "eval_steps_per_second": 23.114,
+      "step": 1000
+    },
+    {
+      "epoch": 0.17269041663072543,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.9666848061779487e-05,
+      "loss": 1.6472,
+      "step": 1001
+    },
+    {
+      "epoch": 0.17286293452945742,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.966615098517598e-05,
+      "loss": 1.4592,
+      "step": 1002
+    },
+    {
+      "epoch": 0.17303545242818943,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.9665453192443603e-05,
+      "loss": 1.5818,
+      "step": 1003
+    },
+    {
+      "epoch": 0.17320797032692142,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.9664754683634064e-05,
+      "loss": 1.6736,
+      "step": 1004
+    },
+    {
+      "epoch": 0.1733804882256534,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.966405545879911e-05,
+      "loss": 1.5858,
+      "step": 1005
+    },
+    {
+      "epoch": 0.1735530061243854,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.9663355517990543e-05,
+      "loss": 1.5809,
+      "step": 1006
+    },
+    {
+      "epoch": 0.1737255240231174,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.966265486126022e-05,
+      "loss": 1.6794,
+      "step": 1007
+    },
+    {
+      "epoch": 0.1738980419218494,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.966195348866005e-05,
+      "loss": 1.6081,
+      "step": 1008
+    },
+    {
+      "epoch": 0.1740705598205814,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9661251400241994e-05,
+      "loss": 1.572,
+      "step": 1009
+    },
+    {
+      "epoch": 0.17424307771931338,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9660548596058068e-05,
+      "loss": 1.4952,
+      "step": 1010
+    },
+    {
+      "epoch": 0.17441559561804537,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.9659845076160345e-05,
+      "loss": 1.6122,
+      "step": 1011
+    },
+    {
+      "epoch": 0.17458811351677736,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.9659140840600934e-05,
+      "loss": 1.4911,
+      "step": 1012
+    },
+    {
+      "epoch": 0.17476063141550935,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.9658435889432022e-05,
+      "loss": 1.4473,
+      "step": 1013
+    },
+    {
+      "epoch": 0.17493314931424137,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.9657730222705828e-05,
+      "loss": 1.5437,
+      "step": 1014
+    },
+    {
+      "epoch": 0.17510566721297335,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.9657023840474637e-05,
+      "loss": 1.5346,
+      "step": 1015
+    },
+    {
+      "epoch": 0.17527818511170534,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.9656316742790778e-05,
+      "loss": 1.6122,
+      "step": 1016
+    },
+    {
+      "epoch": 0.17545070301043733,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.9655608929706636e-05,
+      "loss": 1.635,
+      "step": 1017
+    },
+    {
+      "epoch": 0.17562322090916932,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.965490040127466e-05,
+      "loss": 1.6069,
+      "step": 1018
+    },
+    {
+      "epoch": 0.1757957388079013,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.965419115754733e-05,
+      "loss": 1.5818,
+      "step": 1019
+    },
+    {
+      "epoch": 0.17596825670663332,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.96534811985772e-05,
+      "loss": 1.4853,
+      "step": 1020
+    },
+    {
+      "epoch": 0.1761407746053653,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9652770524416865e-05,
+      "loss": 1.5027,
+      "step": 1021
+    },
+    {
+      "epoch": 0.1763132925040973,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9652059135118976e-05,
+      "loss": 1.5647,
+      "step": 1022
+    },
+    {
+      "epoch": 0.1764858104028293,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.965134703073624e-05,
+      "loss": 1.5769,
+      "step": 1023
+    },
+    {
+      "epoch": 0.17665832830156128,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.9650634211321406e-05,
+      "loss": 1.4441,
+      "step": 1024
+    },
+    {
+      "epoch": 0.17683084620029327,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.9649920676927292e-05,
+      "loss": 1.5214,
+      "step": 1025
+    },
+    {
+      "epoch": 0.17700336409902528,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.964920642760676e-05,
+      "loss": 1.5326,
+      "step": 1026
+    },
+    {
+      "epoch": 0.17717588199775727,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9648491463412724e-05,
+      "loss": 1.5217,
+      "step": 1027
+    },
+    {
+      "epoch": 0.17734839989648926,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9647775784398154e-05,
+      "loss": 1.6123,
+      "step": 1028
+    },
+    {
+      "epoch": 0.17752091779522125,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.9647059390616072e-05,
+      "loss": 1.5351,
+      "step": 1029
+    },
+    {
+      "epoch": 0.17769343569395324,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.964634228211955e-05,
+      "loss": 1.5681,
+      "step": 1030
+    },
+    {
+      "epoch": 0.17786595359268523,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9645624458961722e-05,
+      "loss": 1.5436,
+      "step": 1031
+    },
+    {
+      "epoch": 0.17803847149141724,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.9644905921195763e-05,
+      "loss": 1.6017,
+      "step": 1032
+    },
+    {
+      "epoch": 0.17821098939014923,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9644186668874914e-05,
+      "loss": 1.55,
+      "step": 1033
+    },
+    {
+      "epoch": 0.17838350728888122,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.9643466702052453e-05,
+      "loss": 1.5529,
+      "step": 1034
+    },
+    {
+      "epoch": 0.1785560251876132,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.9642746020781723e-05,
+      "loss": 1.5194,
+      "step": 1035
+    },
+    {
+      "epoch": 0.1787285430863452,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.9642024625116117e-05,
+      "loss": 1.6674,
+      "step": 1036
+    },
+    {
+      "epoch": 0.17890106098507721,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.9641302515109084e-05,
+      "loss": 1.6355,
+      "step": 1037
+    },
+    {
+      "epoch": 0.1790735788838092,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.9640579690814118e-05,
+      "loss": 1.4246,
+      "step": 1038
+    },
+    {
+      "epoch": 0.1792460967825412,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.9639856152284768e-05,
+      "loss": 1.5162,
+      "step": 1039
+    },
+    {
+      "epoch": 0.17941861468127318,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.9639131899574643e-05,
+      "loss": 1.5773,
+      "step": 1040
+    },
+    {
+      "epoch": 0.17959113258000517,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.9638406932737402e-05,
+      "loss": 1.6291,
+      "step": 1041
+    },
+    {
+      "epoch": 0.17976365047873716,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.963768125182675e-05,
+      "loss": 1.554,
+      "step": 1042
+    },
+    {
+      "epoch": 0.17993616837746917,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.9636954856896452e-05,
+      "loss": 1.4847,
+      "step": 1043
+    },
+    {
+      "epoch": 0.18010868627620116,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.9636227748000322e-05,
+      "loss": 1.5396,
+      "step": 1044
+    },
+    {
+      "epoch": 0.18028120417493315,
+      "grad_norm": 0.625,
+      "learning_rate": 1.963549992519223e-05,
+      "loss": 1.5237,
+      "step": 1045
+    },
+    {
+      "epoch": 0.18045372207366514,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.9634771388526103e-05,
+      "loss": 1.5847,
+      "step": 1046
+    },
+    {
+      "epoch": 0.18062623997239713,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9634042138055905e-05,
+      "loss": 1.6293,
+      "step": 1047
+    },
+    {
+      "epoch": 0.18079875787112912,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.9633312173835674e-05,
+      "loss": 1.6017,
+      "step": 1048
+    },
+    {
+      "epoch": 0.18097127576986113,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.963258149591948e-05,
+      "loss": 1.4581,
+      "step": 1049
+    },
+    {
+      "epoch": 0.18114379366859312,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9631850104361467e-05,
+      "loss": 1.615,
+      "step": 1050
+    },
+    {
+      "epoch": 0.1813163115673251,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9631117999215812e-05,
+      "loss": 1.4861,
+      "step": 1051
+    },
+    {
+      "epoch": 0.1814888294660571,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.9630385180536758e-05,
+      "loss": 1.5275,
+      "step": 1052
+    },
+    {
+      "epoch": 0.1816613473647891,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.96296516483786e-05,
+      "loss": 1.5503,
+      "step": 1053
+    },
+    {
+      "epoch": 0.18183386526352108,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.9628917402795677e-05,
+      "loss": 1.5439,
+      "step": 1054
+    },
+    {
+      "epoch": 0.1820063831622531,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9628182443842388e-05,
+      "loss": 1.6039,
+      "step": 1055
+    },
+    {
+      "epoch": 0.18217890106098508,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.962744677157318e-05,
+      "loss": 1.6551,
+      "step": 1056
+    },
+    {
+      "epoch": 0.18235141895971707,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.9626710386042567e-05,
+      "loss": 1.4362,
+      "step": 1057
+    },
+    {
+      "epoch": 0.18252393685844906,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.9625973287305093e-05,
+      "loss": 1.5739,
+      "step": 1058
+    },
+    {
+      "epoch": 0.18269645475718105,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9625235475415377e-05,
+      "loss": 1.5182,
+      "step": 1059
+    },
+    {
+      "epoch": 0.18286897265591306,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.962449695042807e-05,
+      "loss": 1.4969,
+      "step": 1060
+    },
+    {
+      "epoch": 0.18304149055464505,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.9623757712397896e-05,
+      "loss": 1.581,
+      "step": 1061
+    },
+    {
+      "epoch": 0.18321400845337704,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.962301776137962e-05,
+      "loss": 1.6426,
+      "step": 1062
+    },
+    {
+      "epoch": 0.18338652635210903,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.9622277097428058e-05,
+      "loss": 1.5438,
+      "step": 1063
+    },
+    {
+      "epoch": 0.18355904425084102,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.9621535720598085e-05,
+      "loss": 1.5331,
+      "step": 1064
+    },
+    {
+      "epoch": 0.183731562149573,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.9620793630944632e-05,
+      "loss": 1.4752,
+      "step": 1065
+    },
+    {
+      "epoch": 0.18390408004830502,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.962005082852267e-05,
+      "loss": 1.4254,
+      "step": 1066
+    },
+    {
+      "epoch": 0.184076597947037,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.961930731338723e-05,
+      "loss": 1.5386,
+      "step": 1067
+    },
+    {
+      "epoch": 0.184249115845769,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.961856308559341e-05,
+      "loss": 1.5059,
+      "step": 1068
+    },
+    {
+      "epoch": 0.184421633744501,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.961781814519633e-05,
+      "loss": 1.5058,
+      "step": 1069
+    },
+    {
+      "epoch": 0.18459415164323298,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9617072492251187e-05,
+      "loss": 1.5646,
+      "step": 1070
+    },
+    {
+      "epoch": 0.18476666954196497,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.9616326126813224e-05,
+      "loss": 1.5363,
+      "step": 1071
+    },
+    {
+      "epoch": 0.18493918744069698,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.961557904893774e-05,
+      "loss": 1.6104,
+      "step": 1072
+    },
+    {
+      "epoch": 0.18511170533942897,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.9614831258680073e-05,
+      "loss": 1.5752,
+      "step": 1073
+    },
+    {
+      "epoch": 0.18528422323816096,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.9614082756095632e-05,
+      "loss": 1.5502,
+      "step": 1074
+    },
+    {
+      "epoch": 0.18545674113689295,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.961333354123987e-05,
+      "loss": 1.6264,
+      "step": 1075
+    },
+    {
+      "epoch": 0.18562925903562494,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.9612583614168295e-05,
+      "loss": 1.5407,
+      "step": 1076
+    },
+    {
+      "epoch": 0.18580177693435693,
+      "grad_norm": 0.75,
+      "learning_rate": 1.961183297493646e-05,
+      "loss": 1.5649,
+      "step": 1077
+    },
+    {
+      "epoch": 0.18597429483308894,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.961108162359998e-05,
+      "loss": 1.4631,
+      "step": 1078
+    },
+    {
+      "epoch": 0.18614681273182093,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.9610329560214524e-05,
+      "loss": 1.5548,
+      "step": 1079
+    },
+    {
+      "epoch": 0.18631933063055292,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9609576784835803e-05,
+      "loss": 1.52,
+      "step": 1080
+    },
+    {
+      "epoch": 0.1864918485292849,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.960882329751959e-05,
+      "loss": 1.5762,
+      "step": 1081
+    },
+    {
+      "epoch": 0.1866643664280169,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.9608069098321712e-05,
+      "loss": 1.4734,
+      "step": 1082
+    },
+    {
+      "epoch": 0.1868368843267489,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.960731418729804e-05,
+      "loss": 1.4512,
+      "step": 1083
+    },
+    {
+      "epoch": 0.1870094022254809,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.9606558564504503e-05,
+      "loss": 1.584,
+      "step": 1084
+    },
+    {
+      "epoch": 0.1871819201242129,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.9605802229997086e-05,
+      "loss": 1.6183,
+      "step": 1085
+    },
+    {
+      "epoch": 0.18735443802294488,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.9605045183831814e-05,
+      "loss": 1.5524,
+      "step": 1086
+    },
+    {
+      "epoch": 0.18752695592167687,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.960428742606478e-05,
+      "loss": 1.4852,
+      "step": 1087
+    },
+    {
+      "epoch": 0.18769947382040886,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9603528956752126e-05,
+      "loss": 1.5881,
+      "step": 1088
+    },
+    {
+      "epoch": 0.18787199171914087,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.9602769775950044e-05,
+      "loss": 1.5719,
+      "step": 1089
+    },
+    {
+      "epoch": 0.18804450961787286,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.960200988371477e-05,
+      "loss": 1.6392,
+      "step": 1090
+    },
+    {
+      "epoch": 0.18821702751660485,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.9601249280102613e-05,
+      "loss": 1.6041,
+      "step": 1091
+    },
+    {
+      "epoch": 0.18838954541533684,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.9600487965169917e-05,
+      "loss": 1.5457,
+      "step": 1092
+    },
+    {
+      "epoch": 0.18856206331406883,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.9599725938973085e-05,
+      "loss": 1.5748,
+      "step": 1093
+    },
+    {
+      "epoch": 0.18873458121280082,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.959896320156857e-05,
+      "loss": 1.5051,
+      "step": 1094
+    },
+    {
+      "epoch": 0.18890709911153283,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.959819975301289e-05,
+      "loss": 1.5309,
+      "step": 1095
+    },
+    {
+      "epoch": 0.18907961701026482,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9597435593362597e-05,
+      "loss": 1.5352,
+      "step": 1096
+    },
+    {
+      "epoch": 0.1892521349089968,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.9596670722674307e-05,
+      "loss": 1.6632,
+      "step": 1097
+    },
+    {
+      "epoch": 0.1894246528077288,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.9595905141004687e-05,
+      "loss": 1.5162,
+      "step": 1098
+    },
+    {
+      "epoch": 0.1895971707064608,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.959513884841046e-05,
+      "loss": 1.4775,
+      "step": 1099
+    },
+    {
+      "epoch": 0.18976968860519278,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.959437184494839e-05,
+      "loss": 1.4998,
+      "step": 1100
+    },
+    {
+      "epoch": 0.18976968860519278,
+      "eval_loss": 1.5296015739440918,
+      "eval_runtime": 10.937,
+      "eval_samples_per_second": 93.627,
+      "eval_steps_per_second": 23.407,
+      "step": 1100
+    },
+    {
+      "epoch": 0.1899422065039248,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.9593604130675306e-05,
+      "loss": 1.5226,
+      "step": 1101
+    },
+    {
+      "epoch": 0.19011472440265678,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.9592835705648087e-05,
+      "loss": 1.5056,
+      "step": 1102
+    },
+    {
+      "epoch": 0.19028724230138877,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.9592066569923654e-05,
+      "loss": 1.5678,
+      "step": 1103
+    },
+    {
+      "epoch": 0.19045976020012076,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.9591296723559e-05,
+      "loss": 1.5051,
+      "step": 1104
+    },
+    {
+      "epoch": 0.19063227809885275,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.9590526166611153e-05,
+      "loss": 1.534,
+      "step": 1105
+    },
+    {
+      "epoch": 0.19080479599758474,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.9589754899137207e-05,
+      "loss": 1.4893,
+      "step": 1106
+    },
+    {
+      "epoch": 0.19097731389631675,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.9588982921194296e-05,
+      "loss": 1.4873,
+      "step": 1107
+    },
+    {
+      "epoch": 0.19114983179504874,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.9588210232839617e-05,
+      "loss": 1.5466,
+      "step": 1108
+    },
+    {
+      "epoch": 0.19132234969378073,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.9587436834130413e-05,
+      "loss": 1.5609,
+      "step": 1109
+    },
+    {
+      "epoch": 0.19149486759251272,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.9586662725123984e-05,
+      "loss": 1.5532,
+      "step": 1110
+    },
+    {
+      "epoch": 0.1916673854912447,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.958588790587768e-05,
+      "loss": 1.545,
+      "step": 1111
+    },
+    {
+      "epoch": 0.19183990338997672,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9585112376448902e-05,
+      "loss": 1.4745,
+      "step": 1112
+    },
+    {
+      "epoch": 0.1920124212887087,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.9584336136895114e-05,
+      "loss": 1.465,
+      "step": 1113
+    },
+    {
+      "epoch": 0.1921849391874407,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.9583559187273816e-05,
+      "loss": 1.4556,
+      "step": 1114
+    },
+    {
+      "epoch": 0.1923574570861727,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.958278152764257e-05,
+      "loss": 1.5155,
+      "step": 1115
+    },
+    {
+      "epoch": 0.19252997498490468,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.9582003158058996e-05,
+      "loss": 1.5266,
+      "step": 1116
+    },
+    {
+      "epoch": 0.19270249288363667,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9581224078580755e-05,
+      "loss": 1.4815,
+      "step": 1117
+    },
+    {
+      "epoch": 0.19287501078236868,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.9580444289265567e-05,
+      "loss": 1.5429,
+      "step": 1118
+    },
+    {
+      "epoch": 0.19304752868110067,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.957966379017121e-05,
+      "loss": 1.5509,
+      "step": 1119
+    },
+    {
+      "epoch": 0.19322004657983266,
+      "grad_norm": 0.625,
+      "learning_rate": 1.9578882581355497e-05,
+      "loss": 1.5378,
+      "step": 1120
+    },
+    {
+      "epoch": 0.19339256447856465,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.9578100662876314e-05,
+      "loss": 1.5299,
+      "step": 1121
+    },
+    {
+      "epoch": 0.19356508237729664,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.9577318034791586e-05,
+      "loss": 1.5234,
+      "step": 1122
+    },
+    {
+      "epoch": 0.19373760027602863,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.9576534697159298e-05,
+      "loss": 1.4997,
+      "step": 1123
+    },
+    {
+      "epoch": 0.19391011817476064,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.957575065003748e-05,
+      "loss": 1.4906,
+      "step": 1124
+    },
+    {
+      "epoch": 0.19408263607349263,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9574965893484223e-05,
+      "loss": 1.5698,
+      "step": 1125
+    },
+    {
+      "epoch": 0.19425515397222462,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.9574180427557666e-05,
+      "loss": 1.4581,
+      "step": 1126
+    },
+    {
+      "epoch": 0.1944276718709566,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.9573394252316e-05,
+      "loss": 1.5503,
+      "step": 1127
+    },
+    {
+      "epoch": 0.1946001897696886,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.957260736781747e-05,
+      "loss": 1.5962,
+      "step": 1128
+    },
+    {
+      "epoch": 0.19477270766842059,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9571819774120375e-05,
+      "loss": 1.5985,
+      "step": 1129
+    },
+    {
+      "epoch": 0.1949452255671526,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.957103147128306e-05,
+      "loss": 1.5408,
+      "step": 1130
+    },
+    {
+      "epoch": 0.1951177434658846,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.9570242459363937e-05,
+      "loss": 1.6447,
+      "step": 1131
+    },
+    {
+      "epoch": 0.19529026136461658,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.956945273842145e-05,
+      "loss": 1.5605,
+      "step": 1132
+    },
+    {
+      "epoch": 0.19546277926334857,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.9568662308514116e-05,
+      "loss": 1.5723,
+      "step": 1133
+    },
+    {
+      "epoch": 0.19563529716208056,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.9567871169700486e-05,
+      "loss": 1.6163,
+      "step": 1134
+    },
+    {
+      "epoch": 0.19580781506081255,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.956707932203918e-05,
+      "loss": 1.5854,
+      "step": 1135
+    },
+    {
+      "epoch": 0.19598033295954456,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.9566286765588857e-05,
+      "loss": 1.4622,
+      "step": 1136
+    },
+    {
+      "epoch": 0.19615285085827655,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.956549350040824e-05,
+      "loss": 1.5131,
+      "step": 1137
+    },
+    {
+      "epoch": 0.19632536875700854,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9564699526556093e-05,
+      "loss": 1.5931,
+      "step": 1138
+    },
+    {
+      "epoch": 0.19649788665574053,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.9563904844091248e-05,
+      "loss": 1.6076,
+      "step": 1139
+    },
+    {
+      "epoch": 0.19667040455447252,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.956310945307257e-05,
+      "loss": 1.5771,
+      "step": 1140
+    },
+    {
+      "epoch": 0.19684292245320453,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9562313353558992e-05,
+      "loss": 1.5983,
+      "step": 1141
+    },
+    {
+      "epoch": 0.19701544035193652,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.9561516545609493e-05,
+      "loss": 1.6035,
+      "step": 1142
+    },
+    {
+      "epoch": 0.1971879582506685,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.9560719029283104e-05,
+      "loss": 1.6043,
+      "step": 1143
+    },
+    {
+      "epoch": 0.1973604761494005,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.955992080463891e-05,
+      "loss": 1.5247,
+      "step": 1144
+    },
+    {
+      "epoch": 0.1975329940481325,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.9559121871736055e-05,
+      "loss": 1.5868,
+      "step": 1145
+    },
+    {
+      "epoch": 0.19770551194686448,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.9558322230633718e-05,
+      "loss": 1.5682,
+      "step": 1146
+    },
+    {
+      "epoch": 0.1978780298455965,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.955752188139115e-05,
+      "loss": 1.5546,
+      "step": 1147
+    },
+    {
+      "epoch": 0.19805054774432848,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.955672082406764e-05,
+      "loss": 1.5099,
+      "step": 1148
+    },
+    {
+      "epoch": 0.19822306564306047,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.955591905872254e-05,
+      "loss": 1.5612,
+      "step": 1149
+    },
+    {
+      "epoch": 0.19839558354179246,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.9555116585415247e-05,
+      "loss": 1.5709,
+      "step": 1150
+    },
+    {
+      "epoch": 0.19856810144052445,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.9554313404205216e-05,
+      "loss": 1.5448,
+      "step": 1151
+    },
+    {
+      "epoch": 0.19874061933925644,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.955350951515195e-05,
+      "loss": 1.5552,
+      "step": 1152
+    },
+    {
+      "epoch": 0.19891313723798845,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.9552704918315006e-05,
+      "loss": 1.4922,
+      "step": 1153
+    },
+    {
+      "epoch": 0.19908565513672044,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.9551899613753994e-05,
+      "loss": 1.5521,
+      "step": 1154
+    },
+    {
+      "epoch": 0.19925817303545243,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.9551093601528573e-05,
+      "loss": 1.5857,
+      "step": 1155
+    },
+    {
+      "epoch": 0.19943069093418442,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.955028688169846e-05,
+      "loss": 1.586,
+      "step": 1156
+    },
+    {
+      "epoch": 0.1996032088329164,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.9549479454323423e-05,
+      "loss": 1.5905,
+      "step": 1157
+    },
+    {
+      "epoch": 0.1997757267316484,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.954867131946328e-05,
+      "loss": 1.5615,
+      "step": 1158
+    },
+    {
+      "epoch": 0.1999482446303804,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.9547862477177904e-05,
+      "loss": 1.4858,
+      "step": 1159
+    },
+    {
+      "epoch": 0.2001207625291124,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.9547052927527217e-05,
+      "loss": 1.5963,
+      "step": 1160
+    },
+    {
+      "epoch": 0.2002932804278444,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.9546242670571198e-05,
+      "loss": 1.535,
+      "step": 1161
+    },
+    {
+      "epoch": 0.20046579832657638,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.954543170636987e-05,
+      "loss": 1.5349,
+      "step": 1162
+    },
+    {
+      "epoch": 0.20063831622530837,
+      "grad_norm": 0.625,
+      "learning_rate": 1.9544620034983322e-05,
+      "loss": 1.5516,
+      "step": 1163
+    },
+    {
+      "epoch": 0.20081083412404038,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.9543807656471683e-05,
+      "loss": 1.5135,
+      "step": 1164
+    },
+    {
+      "epoch": 0.20098335202277237,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.954299457089514e-05,
+      "loss": 1.6499,
+      "step": 1165
+    },
+    {
+      "epoch": 0.20115586992150436,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9542180778313936e-05,
+      "loss": 1.6134,
+      "step": 1166
+    },
+    {
+      "epoch": 0.20132838782023635,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.954136627878835e-05,
+      "loss": 1.5139,
+      "step": 1167
+    },
+    {
+      "epoch": 0.20150090571896834,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.9540551072378738e-05,
+      "loss": 1.5382,
+      "step": 1168
+    },
+    {
+      "epoch": 0.20167342361770033,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.953973515914549e-05,
+      "loss": 1.4794,
+      "step": 1169
+    },
+    {
+      "epoch": 0.20184594151643234,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.9538918539149054e-05,
+      "loss": 1.4757,
+      "step": 1170
+    },
+    {
+      "epoch": 0.20201845941516433,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.9538101212449932e-05,
+      "loss": 1.5418,
+      "step": 1171
+    },
+    {
+      "epoch": 0.20219097731389632,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.953728317910867e-05,
+      "loss": 1.6522,
+      "step": 1172
+    },
+    {
+      "epoch": 0.2023634952126283,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.953646443918589e-05,
+      "loss": 1.5378,
+      "step": 1173
+    },
+    {
+      "epoch": 0.2025360131113603,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.9535644992742225e-05,
+      "loss": 1.578,
+      "step": 1174
+    },
+    {
+      "epoch": 0.20270853101009229,
+      "grad_norm": 0.625,
+      "learning_rate": 1.9534824839838406e-05,
+      "loss": 1.5171,
+      "step": 1175
+    },
+    {
+      "epoch": 0.2028810489088243,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.953400398053518e-05,
+      "loss": 1.6417,
+      "step": 1176
+    },
+    {
+      "epoch": 0.2030535668075563,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.953318241489337e-05,
+      "loss": 1.4812,
+      "step": 1177
+    },
+    {
+      "epoch": 0.20322608470628828,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9532360142973842e-05,
+      "loss": 1.6705,
+      "step": 1178
+    },
+    {
+      "epoch": 0.20339860260502027,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.9531537164837516e-05,
+      "loss": 1.4646,
+      "step": 1179
+    },
+    {
+      "epoch": 0.20357112050375226,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.9530713480545357e-05,
+      "loss": 1.6521,
+      "step": 1180
+    },
+    {
+      "epoch": 0.20374363840248425,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.9529889090158394e-05,
+      "loss": 1.5087,
+      "step": 1181
+    },
+    {
+      "epoch": 0.20391615630121626,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.9529063993737703e-05,
+      "loss": 1.5173,
+      "step": 1182
+    },
+    {
+      "epoch": 0.20408867419994825,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.952823819134441e-05,
+      "loss": 1.5581,
+      "step": 1183
+    },
+    {
+      "epoch": 0.20426119209868024,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.9527411683039697e-05,
+      "loss": 1.5374,
+      "step": 1184
+    },
+    {
+      "epoch": 0.20443370999741223,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.952658446888479e-05,
+      "loss": 1.523,
+      "step": 1185
+    },
+    {
+      "epoch": 0.20460622789614422,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.952575654894099e-05,
+      "loss": 1.3737,
+      "step": 1186
+    },
+    {
+      "epoch": 0.2047787457948762,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.9524927923269623e-05,
+      "loss": 1.5357,
+      "step": 1187
+    },
+    {
+      "epoch": 0.20495126369360822,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.9524098591932078e-05,
+      "loss": 1.5545,
+      "step": 1188
+    },
+    {
+      "epoch": 0.2051237815923402,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.9523268554989806e-05,
+      "loss": 1.6352,
+      "step": 1189
+    },
+    {
+      "epoch": 0.2052962994910722,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9522437812504285e-05,
+      "loss": 1.5834,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2054688173898042,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.952160636453708e-05,
+      "loss": 1.5966,
+      "step": 1191
+    },
+    {
+      "epoch": 0.20564133528853618,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.9520774211149783e-05,
+      "loss": 1.4937,
+      "step": 1192
+    },
+    {
+      "epoch": 0.2058138531872682,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.951994135240404e-05,
+      "loss": 1.4849,
+      "step": 1193
+    },
+    {
+      "epoch": 0.20598637108600018,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.9519107788361557e-05,
+      "loss": 1.5178,
+      "step": 1194
+    },
+    {
+      "epoch": 0.20615888898473217,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.9518273519084093e-05,
+      "loss": 1.4902,
+      "step": 1195
+    },
+    {
+      "epoch": 0.20633140688346416,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.9517438544633455e-05,
+      "loss": 1.5649,
+      "step": 1196
+    },
+    {
+      "epoch": 0.20650392478219615,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.95166028650715e-05,
+      "loss": 1.5275,
+      "step": 1197
+    },
+    {
+      "epoch": 0.20667644268092814,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.951576648046014e-05,
+      "loss": 1.4538,
+      "step": 1198
+    },
+    {
+      "epoch": 0.20684896057966015,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.951492939086135e-05,
+      "loss": 1.5145,
+      "step": 1199
+    },
+    {
+      "epoch": 0.20702147847839214,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.951409159633713e-05,
+      "loss": 1.3968,
+      "step": 1200
+    },
+    {
+      "epoch": 0.20702147847839214,
+      "eval_loss": 1.5226820707321167,
+      "eval_runtime": 10.8512,
+      "eval_samples_per_second": 94.368,
+      "eval_steps_per_second": 23.592,
+      "step": 1200
+    },
+    {
+      "epoch": 0.20719399637712413,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.951325309694956e-05,
+      "loss": 1.5766,
+      "step": 1201
+    },
+    {
+      "epoch": 0.20736651427585612,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.951241389276076e-05,
+      "loss": 1.4671,
+      "step": 1202
+    },
+    {
+      "epoch": 0.2075390321745881,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9511573983832903e-05,
+      "loss": 1.4517,
+      "step": 1203
+    },
+    {
+      "epoch": 0.2077115500733201,
+      "grad_norm": 4.59375,
+      "learning_rate": 1.9510733370228214e-05,
+      "loss": 1.5248,
+      "step": 1204
+    },
+    {
+      "epoch": 0.2078840679720521,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.950989205200897e-05,
+      "loss": 1.5138,
+      "step": 1205
+    },
+    {
+      "epoch": 0.2080565858707841,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.9509050029237506e-05,
+      "loss": 1.5073,
+      "step": 1206
+    },
+    {
+      "epoch": 0.2082291037695161,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.9508207301976197e-05,
+      "loss": 1.5608,
+      "step": 1207
+    },
+    {
+      "epoch": 0.20840162166824808,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.9507363870287482e-05,
+      "loss": 1.5454,
+      "step": 1208
+    },
+    {
+      "epoch": 0.20857413956698007,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.950651973423385e-05,
+      "loss": 1.4313,
+      "step": 1209
+    },
+    {
+      "epoch": 0.20874665746571205,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.950567489387783e-05,
+      "loss": 1.5429,
+      "step": 1210
+    },
+    {
+      "epoch": 0.20891917536444407,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.950482934928202e-05,
+      "loss": 1.4819,
+      "step": 1211
+    },
+    {
+      "epoch": 0.20909169326317606,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9503983100509067e-05,
+      "loss": 1.5763,
+      "step": 1212
+    },
+    {
+      "epoch": 0.20926421116190805,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.9503136147621662e-05,
+      "loss": 1.4882,
+      "step": 1213
+    },
+    {
+      "epoch": 0.20943672906064004,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9502288490682553e-05,
+      "loss": 1.5377,
+      "step": 1214
+    },
+    {
+      "epoch": 0.20960924695937203,
+      "grad_norm": 0.625,
+      "learning_rate": 1.950144012975454e-05,
+      "loss": 1.5165,
+      "step": 1215
+    },
+    {
+      "epoch": 0.20978176485810404,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.950059106490047e-05,
+      "loss": 1.5467,
+      "step": 1216
+    },
+    {
+      "epoch": 0.20995428275683603,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9499741296183255e-05,
+      "loss": 1.6396,
+      "step": 1217
+    },
+    {
+      "epoch": 0.21012680065556802,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9498890823665846e-05,
+      "loss": 1.6346,
+      "step": 1218
+    },
+    {
+      "epoch": 0.2102993185543,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.9498039647411255e-05,
+      "loss": 1.592,
+      "step": 1219
+    },
+    {
+      "epoch": 0.210471836453032,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.9497187767482538e-05,
+      "loss": 1.5075,
+      "step": 1220
+    },
+    {
+      "epoch": 0.21064435435176398,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.9496335183942814e-05,
+      "loss": 1.5514,
+      "step": 1221
+    },
+    {
+      "epoch": 0.210816872250496,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.949548189685524e-05,
+      "loss": 1.5402,
+      "step": 1222
+    },
+    {
+      "epoch": 0.210989390149228,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.949462790628304e-05,
+      "loss": 1.5735,
+      "step": 1223
+    },
+    {
+      "epoch": 0.21116190804795998,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.9493773212289475e-05,
+      "loss": 1.5145,
+      "step": 1224
+    },
+    {
+      "epoch": 0.21133442594669197,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.9492917814937874e-05,
+      "loss": 1.4504,
+      "step": 1225
+    },
+    {
+      "epoch": 0.21150694384542396,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.9492061714291605e-05,
+      "loss": 1.4625,
+      "step": 1226
+    },
+    {
+      "epoch": 0.21167946174415594,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9491204910414097e-05,
+      "loss": 1.4586,
+      "step": 1227
+    },
+    {
+      "epoch": 0.21185197964288796,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.9490347403368823e-05,
+      "loss": 1.4972,
+      "step": 1228
+    },
+    {
+      "epoch": 0.21202449754161995,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.948948919321932e-05,
+      "loss": 1.5251,
+      "step": 1229
+    },
+    {
+      "epoch": 0.21219701544035194,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.9488630280029154e-05,
+      "loss": 1.5978,
+      "step": 1230
+    },
+    {
+      "epoch": 0.21236953333908393,
+      "grad_norm": 0.625,
+      "learning_rate": 1.948777066386198e-05,
+      "loss": 1.5903,
+      "step": 1231
+    },
+    {
+      "epoch": 0.21254205123781592,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.9486910344781467e-05,
+      "loss": 1.5965,
+      "step": 1232
+    },
+    {
+      "epoch": 0.2127145691365479,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9486049322851358e-05,
+      "loss": 1.5626,
+      "step": 1233
+    },
+    {
+      "epoch": 0.21288708703527992,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.9485187598135445e-05,
+      "loss": 1.5257,
+      "step": 1234
+    },
+    {
+      "epoch": 0.2130596049340119,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9484325170697574e-05,
+      "loss": 1.5239,
+      "step": 1235
+    },
+    {
+      "epoch": 0.2132321228327439,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.9483462040601627e-05,
+      "loss": 1.5665,
+      "step": 1236
+    },
+    {
+      "epoch": 0.2134046407314759,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.948259820791156e-05,
+      "loss": 1.4892,
+      "step": 1237
+    },
+    {
+      "epoch": 0.21357715863020788,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9481733672691365e-05,
+      "loss": 1.5117,
+      "step": 1238
+    },
+    {
+      "epoch": 0.21374967652893986,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.9480868435005098e-05,
+      "loss": 1.6343,
+      "step": 1239
+    },
+    {
+      "epoch": 0.21392219442767188,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9480002494916857e-05,
+      "loss": 1.5402,
+      "step": 1240
+    },
+    {
+      "epoch": 0.21409471232640387,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.9479135852490794e-05,
+      "loss": 1.4328,
+      "step": 1241
+    },
+    {
+      "epoch": 0.21426723022513586,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.9478268507791124e-05,
+      "loss": 1.6286,
+      "step": 1242
+    },
+    {
+      "epoch": 0.21443974812386785,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.9477400460882096e-05,
+      "loss": 1.5382,
+      "step": 1243
+    },
+    {
+      "epoch": 0.21461226602259983,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.9476531711828027e-05,
+      "loss": 1.5822,
+      "step": 1244
+    },
+    {
+      "epoch": 0.21478478392133185,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.9475662260693275e-05,
+      "loss": 1.5945,
+      "step": 1245
+    },
+    {
+      "epoch": 0.21495730182006384,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.9474792107542258e-05,
+      "loss": 1.552,
+      "step": 1246
+    },
+    {
+      "epoch": 0.21512981971879583,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.9473921252439438e-05,
+      "loss": 1.5303,
+      "step": 1247
+    },
+    {
+      "epoch": 0.21530233761752782,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.947304969544934e-05,
+      "loss": 1.4816,
+      "step": 1248
+    },
+    {
+      "epoch": 0.2154748555162598,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.9472177436636523e-05,
+      "loss": 1.5002,
+      "step": 1249
+    },
+    {
+      "epoch": 0.2156473734149918,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.9471304476065624e-05,
+      "loss": 1.5424,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2158198913137238,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.94704308138013e-05,
+      "loss": 1.5329,
+      "step": 1251
+    },
+    {
+      "epoch": 0.2159924092124558,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.94695564499083e-05,
+      "loss": 1.5575,
+      "step": 1252
+    },
+    {
+      "epoch": 0.2161649271111878,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.946868138445138e-05,
+      "loss": 1.5801,
+      "step": 1253
+    },
+    {
+      "epoch": 0.21633744500991978,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9467805617495384e-05,
+      "loss": 1.5072,
+      "step": 1254
+    },
+    {
+      "epoch": 0.21650996290865177,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.9466929149105193e-05,
+      "loss": 1.5689,
+      "step": 1255
+    },
+    {
+      "epoch": 0.21668248080738375,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.9466051979345734e-05,
+      "loss": 1.5702,
+      "step": 1256
+    },
+    {
+      "epoch": 0.21685499870611577,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.9465174108281995e-05,
+      "loss": 1.5755,
+      "step": 1257
+    },
+    {
+      "epoch": 0.21702751660484776,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9464295535979022e-05,
+      "loss": 1.4259,
+      "step": 1258
+    },
+    {
+      "epoch": 0.21720003450357975,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9463416262501898e-05,
+      "loss": 1.4476,
+      "step": 1259
+    },
+    {
+      "epoch": 0.21737255240231174,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.9462536287915773e-05,
+      "loss": 1.5078,
+      "step": 1260
+    },
+    {
+      "epoch": 0.21754507030104372,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9461655612285827e-05,
+      "loss": 1.5395,
+      "step": 1261
+    },
+    {
+      "epoch": 0.2177175881997757,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.946077423567732e-05,
+      "loss": 1.4661,
+      "step": 1262
+    },
+    {
+      "epoch": 0.21789010609850773,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.945989215815554e-05,
+      "loss": 1.55,
+      "step": 1263
+    },
+    {
+      "epoch": 0.21806262399723972,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.9459009379785842e-05,
+      "loss": 1.5413,
+      "step": 1264
+    },
+    {
+      "epoch": 0.2182351418959717,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.9458125900633627e-05,
+      "loss": 1.4655,
+      "step": 1265
+    },
+    {
+      "epoch": 0.2184076597947037,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.945724172076435e-05,
+      "loss": 1.608,
+      "step": 1266
+    },
+    {
+      "epoch": 0.21858017769343568,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.945635684024351e-05,
+      "loss": 1.5109,
+      "step": 1267
+    },
+    {
+      "epoch": 0.2187526955921677,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.945547125913667e-05,
+      "loss": 1.4214,
+      "step": 1268
+    },
+    {
+      "epoch": 0.2189252134908997,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.9454584977509443e-05,
+      "loss": 1.5579,
+      "step": 1269
+    },
+    {
+      "epoch": 0.21909773138963168,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.9453697995427483e-05,
+      "loss": 1.5788,
+      "step": 1270
+    },
+    {
+      "epoch": 0.21927024928836367,
+      "grad_norm": 5.40625,
+      "learning_rate": 1.9452810312956507e-05,
+      "loss": 1.6086,
+      "step": 1271
+    },
+    {
+      "epoch": 0.21944276718709566,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.945192193016228e-05,
+      "loss": 1.5086,
+      "step": 1272
+    },
+    {
+      "epoch": 0.21961528508582764,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.9451032847110615e-05,
+      "loss": 1.6391,
+      "step": 1273
+    },
+    {
+      "epoch": 0.21978780298455966,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.9450143063867385e-05,
+      "loss": 1.5566,
+      "step": 1274
+    },
+    {
+      "epoch": 0.21996032088329165,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.944925258049851e-05,
+      "loss": 1.6582,
+      "step": 1275
+    },
+    {
+      "epoch": 0.22013283878202364,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.9448361397069962e-05,
+      "loss": 1.5291,
+      "step": 1276
+    },
+    {
+      "epoch": 0.22030535668075563,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9447469513647767e-05,
+      "loss": 1.6006,
+      "step": 1277
+    },
+    {
+      "epoch": 0.22047787457948761,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9446576930298e-05,
+      "loss": 1.4931,
+      "step": 1278
+    },
+    {
+      "epoch": 0.2206503924782196,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.9445683647086788e-05,
+      "loss": 1.4638,
+      "step": 1279
+    },
+    {
+      "epoch": 0.22082291037695162,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.944478966408031e-05,
+      "loss": 1.5128,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2209954282756836,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.9443894981344802e-05,
+      "loss": 1.5028,
+      "step": 1281
+    },
+    {
+      "epoch": 0.2211679461744156,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9442999598946545e-05,
+      "loss": 1.5222,
+      "step": 1282
+    },
+    {
+      "epoch": 0.22134046407314759,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.9442103516951878e-05,
+      "loss": 1.5544,
+      "step": 1283
+    },
+    {
+      "epoch": 0.22151298197187957,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.944120673542718e-05,
+      "loss": 1.6135,
+      "step": 1284
+    },
+    {
+      "epoch": 0.22168549987061156,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.94403092544389e-05,
+      "loss": 1.5499,
+      "step": 1285
+    },
+    {
+      "epoch": 0.22185801776934358,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.9439411074053525e-05,
+      "loss": 1.6011,
+      "step": 1286
+    },
+    {
+      "epoch": 0.22203053566807557,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.9438512194337595e-05,
+      "loss": 1.4681,
+      "step": 1287
+    },
+    {
+      "epoch": 0.22220305356680756,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.9437612615357708e-05,
+      "loss": 1.4965,
+      "step": 1288
+    },
+    {
+      "epoch": 0.22237557146553955,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.943671233718051e-05,
+      "loss": 1.5009,
+      "step": 1289
+    },
+    {
+      "epoch": 0.22254808936427153,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9435811359872696e-05,
+      "loss": 1.4423,
+      "step": 1290
+    },
+    {
+      "epoch": 0.22272060726300355,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.9434909683501023e-05,
+      "loss": 1.5931,
+      "step": 1291
+    },
+    {
+      "epoch": 0.22289312516173554,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.9434007308132287e-05,
+      "loss": 1.536,
+      "step": 1292
+    },
+    {
+      "epoch": 0.22306564306046753,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.943310423383334e-05,
+      "loss": 1.4643,
+      "step": 1293
+    },
+    {
+      "epoch": 0.22323816095919952,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.9432200460671096e-05,
+      "loss": 1.5385,
+      "step": 1294
+    },
+    {
+      "epoch": 0.2234106788579315,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.9431295988712504e-05,
+      "loss": 1.6433,
+      "step": 1295
+    },
+    {
+      "epoch": 0.2235831967566635,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.9430390818024575e-05,
+      "loss": 1.5004,
+      "step": 1296
+    },
+    {
+      "epoch": 0.2237557146553955,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.9429484948674374e-05,
+      "loss": 1.5267,
+      "step": 1297
+    },
+    {
+      "epoch": 0.2239282325541275,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9428578380729006e-05,
+      "loss": 1.562,
+      "step": 1298
+    },
+    {
+      "epoch": 0.2241007504528595,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.942767111425564e-05,
+      "loss": 1.5171,
+      "step": 1299
+    },
+    {
+      "epoch": 0.22427326835159148,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.942676314932149e-05,
+      "loss": 1.4994,
+      "step": 1300
+    },
+    {
+      "epoch": 0.22427326835159148,
+      "eval_loss": 1.515897512435913,
+      "eval_runtime": 11.0746,
+      "eval_samples_per_second": 92.464,
+      "eval_steps_per_second": 23.116,
+      "step": 1300
+    },
+    {
+      "epoch": 0.22444578625032346,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.9425854485993828e-05,
+      "loss": 1.6136,
+      "step": 1301
+    },
+    {
+      "epoch": 0.22461830414905545,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.9424945124339965e-05,
+      "loss": 1.5522,
+      "step": 1302
+    },
+    {
+      "epoch": 0.22479082204778747,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.9424035064427286e-05,
+      "loss": 1.4482,
+      "step": 1303
+    },
+    {
+      "epoch": 0.22496333994651946,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.9423124306323197e-05,
+      "loss": 1.6717,
+      "step": 1304
+    },
+    {
+      "epoch": 0.22513585784525145,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.9422212850095183e-05,
+      "loss": 1.4648,
+      "step": 1305
+    },
+    {
+      "epoch": 0.22530837574398344,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.9421300695810773e-05,
+      "loss": 1.5514,
+      "step": 1306
+    },
+    {
+      "epoch": 0.22548089364271542,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9420387843537533e-05,
+      "loss": 1.5525,
+      "step": 1307
+    },
+    {
+      "epoch": 0.2256534115414474,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9419474293343107e-05,
+      "loss": 1.5457,
+      "step": 1308
+    },
+    {
+      "epoch": 0.22582592944017943,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.9418560045295166e-05,
+      "loss": 1.5831,
+      "step": 1309
+    },
+    {
+      "epoch": 0.22599844733891142,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9417645099461446e-05,
+      "loss": 1.407,
+      "step": 1310
+    },
+    {
+      "epoch": 0.2261709652376434,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9416729455909737e-05,
+      "loss": 1.6039,
+      "step": 1311
+    },
+    {
+      "epoch": 0.2263434831363754,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.941581311470787e-05,
+      "loss": 1.5484,
+      "step": 1312
+    },
+    {
+      "epoch": 0.22651600103510738,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9414896075923732e-05,
+      "loss": 1.5051,
+      "step": 1313
+    },
+    {
+      "epoch": 0.22668851893383937,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9413978339625267e-05,
+      "loss": 1.4133,
+      "step": 1314
+    },
+    {
+      "epoch": 0.2268610368325714,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.9413059905880466e-05,
+      "loss": 1.5823,
+      "step": 1315
+    },
+    {
+      "epoch": 0.22703355473130338,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.941214077475737e-05,
+      "loss": 1.5071,
+      "step": 1316
+    },
+    {
+      "epoch": 0.22720607263003537,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.941122094632408e-05,
+      "loss": 1.5494,
+      "step": 1317
+    },
+    {
+      "epoch": 0.22737859052876735,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.9410300420648735e-05,
+      "loss": 1.4567,
+      "step": 1318
+    },
+    {
+      "epoch": 0.22755110842749934,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.9409379197799537e-05,
+      "loss": 1.5043,
+      "step": 1319
+    },
+    {
+      "epoch": 0.22772362632623136,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.9408457277844738e-05,
+      "loss": 1.5131,
+      "step": 1320
+    },
+    {
+      "epoch": 0.22789614422496335,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.9407534660852632e-05,
+      "loss": 1.5591,
+      "step": 1321
+    },
+    {
+      "epoch": 0.22806866212369534,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.940661134689158e-05,
+      "loss": 1.4588,
+      "step": 1322
+    },
+    {
+      "epoch": 0.22824118002242733,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.9405687336029985e-05,
+      "loss": 1.519,
+      "step": 1323
+    },
+    {
+      "epoch": 0.22841369792115931,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.94047626283363e-05,
+      "loss": 1.5361,
+      "step": 1324
+    },
+    {
+      "epoch": 0.2285862158198913,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.9403837223879038e-05,
+      "loss": 1.5026,
+      "step": 1325
+    },
+    {
+      "epoch": 0.22875873371862332,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9402911122726756e-05,
+      "loss": 1.5397,
+      "step": 1326
+    },
+    {
+      "epoch": 0.2289312516173553,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.9401984324948067e-05,
+      "loss": 1.4693,
+      "step": 1327
+    },
+    {
+      "epoch": 0.2291037695160873,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.9401056830611634e-05,
+      "loss": 1.5675,
+      "step": 1328
+    },
+    {
+      "epoch": 0.22927628741481929,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.9400128639786168e-05,
+      "loss": 1.4772,
+      "step": 1329
+    },
+    {
+      "epoch": 0.22944880531355127,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.9399199752540435e-05,
+      "loss": 1.3506,
+      "step": 1330
+    },
+    {
+      "epoch": 0.22962132321228326,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.939827016894326e-05,
+      "loss": 1.4486,
+      "step": 1331
+    },
+    {
+      "epoch": 0.22979384111101528,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.9397339889063506e-05,
+      "loss": 1.5139,
+      "step": 1332
+    },
+    {
+      "epoch": 0.22996635900974727,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.9396408912970096e-05,
+      "loss": 1.5306,
+      "step": 1333
+    },
+    {
+      "epoch": 0.23013887690847926,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.9395477240732003e-05,
+      "loss": 1.4321,
+      "step": 1334
+    },
+    {
+      "epoch": 0.23031139480721124,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.939454487241825e-05,
+      "loss": 1.5289,
+      "step": 1335
+    },
+    {
+      "epoch": 0.23048391270594323,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9393611808097913e-05,
+      "loss": 1.5167,
+      "step": 1336
+    },
+    {
+      "epoch": 0.23065643060467522,
+      "grad_norm": 0.75,
+      "learning_rate": 1.939267804784012e-05,
+      "loss": 1.509,
+      "step": 1337
+    },
+    {
+      "epoch": 0.23082894850340724,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.9391743591714046e-05,
+      "loss": 1.4916,
+      "step": 1338
+    },
+    {
+      "epoch": 0.23100146640213923,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.9390808439788928e-05,
+      "loss": 1.5742,
+      "step": 1339
+    },
+    {
+      "epoch": 0.23117398430087122,
+      "grad_norm": 0.75,
+      "learning_rate": 1.9389872592134044e-05,
+      "loss": 1.5068,
+      "step": 1340
+    },
+    {
+      "epoch": 0.2313465021996032,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.9388936048818725e-05,
+      "loss": 1.6008,
+      "step": 1341
+    },
+    {
+      "epoch": 0.2315190200983352,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.938799880991236e-05,
+      "loss": 1.4813,
+      "step": 1342
+    },
+    {
+      "epoch": 0.2316915379970672,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.9387060875484388e-05,
+      "loss": 1.5196,
+      "step": 1343
+    },
+    {
+      "epoch": 0.2318640558957992,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.9386122245604285e-05,
+      "loss": 1.5514,
+      "step": 1344
+    },
+    {
+      "epoch": 0.2320365737945312,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.938518292034161e-05,
+      "loss": 1.5391,
+      "step": 1345
+    },
+    {
+      "epoch": 0.23220909169326318,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.9384242899765933e-05,
+      "loss": 1.6399,
+      "step": 1346
+    },
+    {
+      "epoch": 0.23238160959199516,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.938330218394691e-05,
+      "loss": 1.6369,
+      "step": 1347
+    },
+    {
+      "epoch": 0.23255412749072715,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.938236077295423e-05,
+      "loss": 1.5085,
+      "step": 1348
+    },
+    {
+      "epoch": 0.23272664538945917,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.9381418666857645e-05,
+      "loss": 1.4524,
+      "step": 1349
+    },
+    {
+      "epoch": 0.23289916328819116,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.938047586572694e-05,
+      "loss": 1.5181,
+      "step": 1350
+    },
+    {
+      "epoch": 0.23307168118692315,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.937953236963198e-05,
+      "loss": 1.5584,
+      "step": 1351
+    },
+    {
+      "epoch": 0.23324419908565514,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.937858817864265e-05,
+      "loss": 1.489,
+      "step": 1352
+    },
+    {
+      "epoch": 0.23341671698438712,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.937764329282891e-05,
+      "loss": 1.595,
+      "step": 1353
+    },
+    {
+      "epoch": 0.2335892348831191,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9376697712260758e-05,
+      "loss": 1.5573,
+      "step": 1354
+    },
+    {
+      "epoch": 0.23376175278185113,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.9375751437008253e-05,
+      "loss": 1.4401,
+      "step": 1355
+    },
+    {
+      "epoch": 0.23393427068058312,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.9374804467141497e-05,
+      "loss": 1.4892,
+      "step": 1356
+    },
+    {
+      "epoch": 0.2341067885793151,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.937385680273065e-05,
+      "loss": 1.5684,
+      "step": 1357
+    },
+    {
+      "epoch": 0.2342793064780471,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.9372908443845926e-05,
+      "loss": 1.6208,
+      "step": 1358
+    },
+    {
+      "epoch": 0.23445182437677908,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.9371959390557573e-05,
+      "loss": 1.5651,
+      "step": 1359
+    },
+    {
+      "epoch": 0.23462434227551107,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.9371009642935913e-05,
+      "loss": 1.4277,
+      "step": 1360
+    },
+    {
+      "epoch": 0.2347968601742431,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.9370059201051304e-05,
+      "loss": 1.5014,
+      "step": 1361
+    },
+    {
+      "epoch": 0.23496937807297508,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.936910806497417e-05,
+      "loss": 1.4315,
+      "step": 1362
+    },
+    {
+      "epoch": 0.23514189597170707,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.936815623477496e-05,
+      "loss": 1.6346,
+      "step": 1363
+    },
+    {
+      "epoch": 0.23531441387043905,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9367203710524204e-05,
+      "loss": 1.5522,
+      "step": 1364
+    },
+    {
+      "epoch": 0.23548693176917104,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.936625049229247e-05,
+      "loss": 1.5137,
+      "step": 1365
+    },
+    {
+      "epoch": 0.23565944966790303,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.9365296580150374e-05,
+      "loss": 1.473,
+      "step": 1366
+    },
+    {
+      "epoch": 0.23583196756663505,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.9364341974168594e-05,
+      "loss": 1.5776,
+      "step": 1367
+    },
+    {
+      "epoch": 0.23600448546536704,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.9363386674417844e-05,
+      "loss": 1.5133,
+      "step": 1368
+    },
+    {
+      "epoch": 0.23617700336409903,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9362430680968907e-05,
+      "loss": 1.5902,
+      "step": 1369
+    },
+    {
+      "epoch": 0.236349521262831,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9361473993892604e-05,
+      "loss": 1.4907,
+      "step": 1370
+    },
+    {
+      "epoch": 0.236522039161563,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.936051661325982e-05,
+      "loss": 1.569,
+      "step": 1371
+    },
+    {
+      "epoch": 0.23669455706029502,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9359558539141474e-05,
+      "loss": 1.5697,
+      "step": 1372
+    },
+    {
+      "epoch": 0.236867074959027,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.935859977160855e-05,
+      "loss": 1.4963,
+      "step": 1373
+    },
+    {
+      "epoch": 0.237039592857759,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.935764031073208e-05,
+      "loss": 1.5429,
+      "step": 1374
+    },
+    {
+      "epoch": 0.23721211075649098,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9356680156583147e-05,
+      "loss": 1.5264,
+      "step": 1375
+    },
+    {
+      "epoch": 0.23738462865522297,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9355719309232885e-05,
+      "loss": 1.5749,
+      "step": 1376
+    },
+    {
+      "epoch": 0.23755714655395496,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.935475776875248e-05,
+      "loss": 1.6136,
+      "step": 1377
+    },
+    {
+      "epoch": 0.23772966445268698,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.9353795535213167e-05,
+      "loss": 1.4908,
+      "step": 1378
+    },
+    {
+      "epoch": 0.23790218235141897,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9352832608686234e-05,
+      "loss": 1.439,
+      "step": 1379
+    },
+    {
+      "epoch": 0.23807470025015096,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.9351868989243026e-05,
+      "loss": 1.5656,
+      "step": 1380
+    },
+    {
+      "epoch": 0.23824721814888294,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.9350904676954927e-05,
+      "loss": 1.4963,
+      "step": 1381
+    },
+    {
+      "epoch": 0.23841973604761493,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9349939671893384e-05,
+      "loss": 1.5902,
+      "step": 1382
+    },
+    {
+      "epoch": 0.23859225394634692,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.934897397412989e-05,
+      "loss": 1.457,
+      "step": 1383
+    },
+    {
+      "epoch": 0.23876477184507894,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9348007583735985e-05,
+      "loss": 1.5349,
+      "step": 1384
+    },
+    {
+      "epoch": 0.23893728974381093,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.9347040500783272e-05,
+      "loss": 1.4851,
+      "step": 1385
+    },
+    {
+      "epoch": 0.23910980764254292,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.9346072725343394e-05,
+      "loss": 1.5019,
+      "step": 1386
+    },
+    {
+      "epoch": 0.2392823255412749,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.934510425748805e-05,
+      "loss": 1.5755,
+      "step": 1387
+    },
+    {
+      "epoch": 0.2394548434400069,
+      "grad_norm": 0.625,
+      "learning_rate": 1.9344135097288997e-05,
+      "loss": 1.5478,
+      "step": 1388
+    },
+    {
+      "epoch": 0.23962736133873888,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.934316524481803e-05,
+      "loss": 1.5787,
+      "step": 1389
+    },
+    {
+      "epoch": 0.2397998792374709,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.9342194700146998e-05,
+      "loss": 1.5794,
+      "step": 1390
+    },
+    {
+      "epoch": 0.2399723971362029,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9341223463347815e-05,
+      "loss": 1.5546,
+      "step": 1391
+    },
+    {
+      "epoch": 0.24014491503493487,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.9340251534492428e-05,
+      "loss": 1.5541,
+      "step": 1392
+    },
+    {
+      "epoch": 0.24031743293366686,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.933927891365285e-05,
+      "loss": 1.644,
+      "step": 1393
+    },
+    {
+      "epoch": 0.24048995083239885,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.9338305600901135e-05,
+      "loss": 1.514,
+      "step": 1394
+    },
+    {
+      "epoch": 0.24066246873113087,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.933733159630939e-05,
+      "loss": 1.5331,
+      "step": 1395
+    },
+    {
+      "epoch": 0.24083498662986286,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.933635689994978e-05,
+      "loss": 1.4759,
+      "step": 1396
+    },
+    {
+      "epoch": 0.24100750452859485,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.933538151189451e-05,
+      "loss": 1.4967,
+      "step": 1397
+    },
+    {
+      "epoch": 0.24118002242732683,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9334405432215857e-05,
+      "loss": 1.5873,
+      "step": 1398
+    },
+    {
+      "epoch": 0.24135254032605882,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.9333428660986118e-05,
+      "loss": 1.537,
+      "step": 1399
+    },
+    {
+      "epoch": 0.2415250582247908,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9332451198277668e-05,
+      "loss": 1.5899,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2415250582247908,
+      "eval_loss": 1.509263277053833,
+      "eval_runtime": 11.0113,
+      "eval_samples_per_second": 92.996,
+      "eval_steps_per_second": 23.249,
+      "step": 1400
+    },
+    {
+      "epoch": 0.24169757612352283,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.933147304416292e-05,
+      "loss": 1.555,
+      "step": 1401
+    },
+    {
+      "epoch": 0.24187009402225482,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9330494198714347e-05,
+      "loss": 1.5731,
+      "step": 1402
+    },
+    {
+      "epoch": 0.2420426119209868,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.932951466200446e-05,
+      "loss": 1.5459,
+      "step": 1403
+    },
+    {
+      "epoch": 0.2422151298197188,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.9328534434105835e-05,
+      "loss": 1.5302,
+      "step": 1404
+    },
+    {
+      "epoch": 0.24238764771845078,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9327553515091092e-05,
+      "loss": 1.452,
+      "step": 1405
+    },
+    {
+      "epoch": 0.24256016561718277,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.93265719050329e-05,
+      "loss": 1.516,
+      "step": 1406
+    },
+    {
+      "epoch": 0.2427326835159148,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.9325589604003992e-05,
+      "loss": 1.5422,
+      "step": 1407
+    },
+    {
+      "epoch": 0.24290520141464678,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.9324606612077135e-05,
+      "loss": 1.5611,
+      "step": 1408
+    },
+    {
+      "epoch": 0.24307771931337877,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.9323622929325155e-05,
+      "loss": 1.5626,
+      "step": 1409
+    },
+    {
+      "epoch": 0.24325023721211075,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.9322638555820934e-05,
+      "loss": 1.5338,
+      "step": 1410
+    },
+    {
+      "epoch": 0.24342275511084274,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.9321653491637397e-05,
+      "loss": 1.6135,
+      "step": 1411
+    },
+    {
+      "epoch": 0.24359527300957473,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.9320667736847526e-05,
+      "loss": 1.5297,
+      "step": 1412
+    },
+    {
+      "epoch": 0.24376779090830675,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.931968129152435e-05,
+      "loss": 1.5739,
+      "step": 1413
+    },
+    {
+      "epoch": 0.24394030880703874,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.931869415574095e-05,
+      "loss": 1.4873,
+      "step": 1414
+    },
+    {
+      "epoch": 0.24411282670577072,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.9317706329570467e-05,
+      "loss": 1.5725,
+      "step": 1415
+    },
+    {
+      "epoch": 0.2442853446045027,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9316717813086073e-05,
+      "loss": 1.4288,
+      "step": 1416
+    },
+    {
+      "epoch": 0.2444578625032347,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9315728606361012e-05,
+      "loss": 1.4411,
+      "step": 1417
+    },
+    {
+      "epoch": 0.2446303804019667,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.931473870946857e-05,
+      "loss": 1.4901,
+      "step": 1418
+    },
+    {
+      "epoch": 0.2448028983006987,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.9313748122482085e-05,
+      "loss": 1.5489,
+      "step": 1419
+    },
+    {
+      "epoch": 0.2449754161994307,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.9312756845474937e-05,
+      "loss": 1.4666,
+      "step": 1420
+    },
+    {
+      "epoch": 0.24514793409816268,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.931176487852058e-05,
+      "loss": 1.5141,
+      "step": 1421
+    },
+    {
+      "epoch": 0.24532045199689467,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.9310772221692495e-05,
+      "loss": 1.5155,
+      "step": 1422
+    },
+    {
+      "epoch": 0.24549296989562666,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.9309778875064228e-05,
+      "loss": 1.5135,
+      "step": 1423
+    },
+    {
+      "epoch": 0.24566548779435868,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.930878483870937e-05,
+      "loss": 1.509,
+      "step": 1424
+    },
+    {
+      "epoch": 0.24583800569309067,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9307790112701573e-05,
+      "loss": 1.5216,
+      "step": 1425
+    },
+    {
+      "epoch": 0.24601052359182266,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.9306794697114522e-05,
+      "loss": 1.5448,
+      "step": 1426
+    },
+    {
+      "epoch": 0.24618304149055464,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.9305798592021968e-05,
+      "loss": 1.459,
+      "step": 1427
+    },
+    {
+      "epoch": 0.24635555938928663,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.930480179749771e-05,
+      "loss": 1.5335,
+      "step": 1428
+    },
+    {
+      "epoch": 0.24652807728801862,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9303804313615596e-05,
+      "loss": 1.5651,
+      "step": 1429
+    },
+    {
+      "epoch": 0.24670059518675064,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.9302806140449525e-05,
+      "loss": 1.5367,
+      "step": 1430
+    },
+    {
+      "epoch": 0.24687311308548263,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9301807278073447e-05,
+      "loss": 1.503,
+      "step": 1431
+    },
+    {
+      "epoch": 0.24704563098421461,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.9300807726561368e-05,
+      "loss": 1.5215,
+      "step": 1432
+    },
+    {
+      "epoch": 0.2472181488829466,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.929980748598734e-05,
+      "loss": 1.5557,
+      "step": 1433
+    },
+    {
+      "epoch": 0.2473906667816786,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.9298806556425462e-05,
+      "loss": 1.495,
+      "step": 1434
+    },
+    {
+      "epoch": 0.24756318468041058,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9297804937949894e-05,
+      "loss": 1.5105,
+      "step": 1435
+    },
+    {
+      "epoch": 0.2477357025791426,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.929680263063484e-05,
+      "loss": 1.5457,
+      "step": 1436
+    },
+    {
+      "epoch": 0.24790822047787459,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.929579963455456e-05,
+      "loss": 1.5825,
+      "step": 1437
+    },
+    {
+      "epoch": 0.24808073837660657,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.929479594978336e-05,
+      "loss": 1.4412,
+      "step": 1438
+    },
+    {
+      "epoch": 0.24825325627533856,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.9293791576395597e-05,
+      "loss": 1.5191,
+      "step": 1439
+    },
+    {
+      "epoch": 0.24842577417407055,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9292786514465685e-05,
+      "loss": 1.4836,
+      "step": 1440
+    },
+    {
+      "epoch": 0.24859829207280254,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.9291780764068085e-05,
+      "loss": 1.4466,
+      "step": 1441
+    },
+    {
+      "epoch": 0.24877080997153456,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.9290774325277305e-05,
+      "loss": 1.6016,
+      "step": 1442
+    },
+    {
+      "epoch": 0.24894332787026655,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.9289767198167918e-05,
+      "loss": 1.5898,
+      "step": 1443
+    },
+    {
+      "epoch": 0.24911584576899853,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.9288759382814526e-05,
+      "loss": 1.5457,
+      "step": 1444
+    },
+    {
+      "epoch": 0.24928836366773052,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9287750879291802e-05,
+      "loss": 1.4422,
+      "step": 1445
+    },
+    {
+      "epoch": 0.2494608815664625,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.928674168767446e-05,
+      "loss": 1.6013,
+      "step": 1446
+    },
+    {
+      "epoch": 0.24963339946519453,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.928573180803727e-05,
+      "loss": 1.5306,
+      "step": 1447
+    },
+    {
+      "epoch": 0.24980591736392652,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9284721240455045e-05,
+      "loss": 1.4907,
+      "step": 1448
+    },
+    {
+      "epoch": 0.2499784352626585,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9283709985002655e-05,
+      "loss": 1.5263,
+      "step": 1449
+    },
+    {
+      "epoch": 0.2501509531613905,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.9282698041755027e-05,
+      "loss": 1.5184,
+      "step": 1450
+    },
+    {
+      "epoch": 0.2503234710601225,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.9281685410787125e-05,
+      "loss": 1.5459,
+      "step": 1451
+    },
+    {
+      "epoch": 0.2504959889588545,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.928067209217397e-05,
+      "loss": 1.6034,
+      "step": 1452
+    },
+    {
+      "epoch": 0.25066850685758646,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9279658085990642e-05,
+      "loss": 1.5055,
+      "step": 1453
+    },
+    {
+      "epoch": 0.2508410247563185,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.927864339231226e-05,
+      "loss": 1.4582,
+      "step": 1454
+    },
+    {
+      "epoch": 0.25101354265505044,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.9277628011214e-05,
+      "loss": 1.5079,
+      "step": 1455
+    },
+    {
+      "epoch": 0.25118606055378245,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.9276611942771083e-05,
+      "loss": 1.4217,
+      "step": 1456
+    },
+    {
+      "epoch": 0.25135857845251447,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.92755951870588e-05,
+      "loss": 1.5337,
+      "step": 1457
+    },
+    {
+      "epoch": 0.25153109635124643,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.927457774415246e-05,
+      "loss": 1.4909,
+      "step": 1458
+    },
+    {
+      "epoch": 0.25170361424997845,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.9273559614127455e-05,
+      "loss": 1.4921,
+      "step": 1459
+    },
+    {
+      "epoch": 0.2518761321487104,
+      "grad_norm": 0.625,
+      "learning_rate": 1.9272540797059208e-05,
+      "loss": 1.4744,
+      "step": 1460
+    },
+    {
+      "epoch": 0.2520486500474424,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9271521293023202e-05,
+      "loss": 1.5649,
+      "step": 1461
+    },
+    {
+      "epoch": 0.25222116794617444,
+      "grad_norm": 0.625,
+      "learning_rate": 1.927050110209497e-05,
+      "loss": 1.5705,
+      "step": 1462
+    },
+    {
+      "epoch": 0.2523936858449064,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9269480224350087e-05,
+      "loss": 1.6142,
+      "step": 1463
+    },
+    {
+      "epoch": 0.2525662037436384,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.9268458659864194e-05,
+      "loss": 1.6079,
+      "step": 1464
+    },
+    {
+      "epoch": 0.2527387216423704,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.926743640871297e-05,
+      "loss": 1.5139,
+      "step": 1465
+    },
+    {
+      "epoch": 0.2529112395411024,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9266413470972153e-05,
+      "loss": 1.5096,
+      "step": 1466
+    },
+    {
+      "epoch": 0.25308375743983436,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.9265389846717522e-05,
+      "loss": 1.5319,
+      "step": 1467
+    },
+    {
+      "epoch": 0.2532562753385664,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.926436553602492e-05,
+      "loss": 1.61,
+      "step": 1468
+    },
+    {
+      "epoch": 0.2534287932372984,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9263340538970237e-05,
+      "loss": 1.5454,
+      "step": 1469
+    },
+    {
+      "epoch": 0.25360131113603035,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9262314855629405e-05,
+      "loss": 1.5231,
+      "step": 1470
+    },
+    {
+      "epoch": 0.25377382903476237,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9261288486078414e-05,
+      "loss": 1.477,
+      "step": 1471
+    },
+    {
+      "epoch": 0.2539463469334943,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.9260261430393306e-05,
+      "loss": 1.4894,
+      "step": 1472
+    },
+    {
+      "epoch": 0.25411886483222634,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.9259233688650167e-05,
+      "loss": 1.4668,
+      "step": 1473
+    },
+    {
+      "epoch": 0.25429138273095836,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.9258205260925143e-05,
+      "loss": 1.4983,
+      "step": 1474
+    },
+    {
+      "epoch": 0.2544639006296903,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.925717614729443e-05,
+      "loss": 1.5638,
+      "step": 1475
+    },
+    {
+      "epoch": 0.25463641852842234,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.9256146347834262e-05,
+      "loss": 1.5986,
+      "step": 1476
+    },
+    {
+      "epoch": 0.2548089364271543,
+      "grad_norm": 1.2578125,
+      "learning_rate": 1.9255115862620938e-05,
+      "loss": 1.4828,
+      "step": 1477
+    },
+    {
+      "epoch": 0.2549814543258863,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.9254084691730804e-05,
+      "loss": 1.6261,
+      "step": 1478
+    },
+    {
+      "epoch": 0.25515397222461833,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.9253052835240252e-05,
+      "loss": 1.5251,
+      "step": 1479
+    },
+    {
+      "epoch": 0.2553264901233503,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.925202029322573e-05,
+      "loss": 1.6185,
+      "step": 1480
+    },
+    {
+      "epoch": 0.2554990080220823,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.9250987065763737e-05,
+      "loss": 1.5007,
+      "step": 1481
+    },
+    {
+      "epoch": 0.25567152592081427,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.9249953152930818e-05,
+      "loss": 1.6077,
+      "step": 1482
+    },
+    {
+      "epoch": 0.2558440438195463,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.9248918554803576e-05,
+      "loss": 1.4667,
+      "step": 1483
+    },
+    {
+      "epoch": 0.25601656171827825,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.9247883271458653e-05,
+      "loss": 1.5333,
+      "step": 1484
+    },
+    {
+      "epoch": 0.25618907961701026,
+      "grad_norm": 1.40625,
+      "learning_rate": 1.9246847302972754e-05,
+      "loss": 1.5543,
+      "step": 1485
+    },
+    {
+      "epoch": 0.2563615975157423,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9245810649422633e-05,
+      "loss": 1.5163,
+      "step": 1486
+    },
+    {
+      "epoch": 0.25653411541447424,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.924477331088509e-05,
+      "loss": 1.5476,
+      "step": 1487
+    },
+    {
+      "epoch": 0.25670663331320626,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.924373528743697e-05,
+      "loss": 1.4741,
+      "step": 1488
+    },
+    {
+      "epoch": 0.2568791512119382,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.924269657915519e-05,
+      "loss": 1.4808,
+      "step": 1489
+    },
+    {
+      "epoch": 0.25705166911067023,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.9241657186116688e-05,
+      "loss": 1.4003,
+      "step": 1490
+    },
+    {
+      "epoch": 0.25722418700940225,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.9240617108398482e-05,
+      "loss": 1.4317,
+      "step": 1491
+    },
+    {
+      "epoch": 0.2573967049081342,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.923957634607762e-05,
+      "loss": 1.4747,
+      "step": 1492
+    },
+    {
+      "epoch": 0.2575692228068662,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.9238534899231216e-05,
+      "loss": 1.5198,
+      "step": 1493
+    },
+    {
+      "epoch": 0.2577417407055982,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.923749276793642e-05,
+      "loss": 1.4781,
+      "step": 1494
+    },
+    {
+      "epoch": 0.2579142586043302,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.9236449952270437e-05,
+      "loss": 1.5068,
+      "step": 1495
+    },
+    {
+      "epoch": 0.25808677650306217,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.923540645231053e-05,
+      "loss": 1.5955,
+      "step": 1496
+    },
+    {
+      "epoch": 0.2582592944017942,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.923436226813401e-05,
+      "loss": 1.5487,
+      "step": 1497
+    },
+    {
+      "epoch": 0.2584318123005262,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.9233317399818237e-05,
+      "loss": 1.6363,
+      "step": 1498
+    },
+    {
+      "epoch": 0.25860433019925816,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.9232271847440614e-05,
+      "loss": 1.588,
+      "step": 1499
+    },
+    {
+      "epoch": 0.2587768480979902,
+      "grad_norm": 1.140625,
+      "learning_rate": 1.923122561107861e-05,
+      "loss": 1.5108,
+      "step": 1500
+    },
+    {
+      "epoch": 0.2587768480979902,
+      "eval_loss": 1.5035676956176758,
+      "eval_runtime": 10.9566,
+      "eval_samples_per_second": 93.46,
+      "eval_steps_per_second": 23.365,
+      "step": 1500
+    },
+    {
+      "epoch": 0.25894936599672214,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.923017869080973e-05,
+      "loss": 1.5098,
+      "step": 1501
+    },
+    {
+      "epoch": 0.25912188389545415,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.9229131086711542e-05,
+      "loss": 1.386,
+      "step": 1502
+    },
+    {
+      "epoch": 0.25929440179418617,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.9228082798861656e-05,
+      "loss": 1.5323,
+      "step": 1503
+    },
+    {
+      "epoch": 0.25946691969291813,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.9227033827337735e-05,
+      "loss": 1.5277,
+      "step": 1504
+    },
+    {
+      "epoch": 0.25963943759165015,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.9225984172217497e-05,
+      "loss": 1.5729,
+      "step": 1505
+    },
+    {
+      "epoch": 0.2598119554903821,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.9224933833578706e-05,
+      "loss": 1.5361,
+      "step": 1506
+    },
+    {
+      "epoch": 0.2599844733891141,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.9223882811499175e-05,
+      "loss": 1.465,
+      "step": 1507
+    },
+    {
+      "epoch": 0.26015699128784614,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.9222831106056768e-05,
+      "loss": 1.4906,
+      "step": 1508
+    },
+    {
+      "epoch": 0.2603295091865781,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.922177871732941e-05,
+      "loss": 1.5643,
+      "step": 1509
+    },
+    {
+      "epoch": 0.2605020270853101,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.9220725645395066e-05,
+      "loss": 1.596,
+      "step": 1510
+    },
+    {
+      "epoch": 0.2606745449840421,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.921967189033175e-05,
+      "loss": 1.482,
+      "step": 1511
+    },
+    {
+      "epoch": 0.2608470628827741,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.9218617452217534e-05,
+      "loss": 1.5303,
+      "step": 1512
+    },
+    {
+      "epoch": 0.26101958078150606,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.9217562331130536e-05,
+      "loss": 1.4882,
+      "step": 1513
+    },
+    {
+      "epoch": 0.26119209868023807,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.9216506527148926e-05,
+      "loss": 1.5004,
+      "step": 1514
+    },
+    {
+      "epoch": 0.2613646165789701,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.9215450040350924e-05,
+      "loss": 1.5241,
+      "step": 1515
+    },
+    {
+      "epoch": 0.26153713447770205,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.9214392870814805e-05,
+      "loss": 1.5178,
+      "step": 1516
+    },
+    {
+      "epoch": 0.26170965237643407,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.921333501861889e-05,
+      "loss": 1.5519,
+      "step": 1517
+    },
+    {
+      "epoch": 0.261882170275166,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.9212276483841544e-05,
+      "loss": 1.5706,
+      "step": 1518
+    },
+    {
+      "epoch": 0.26205468817389804,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.9211217266561197e-05,
+      "loss": 1.504,
+      "step": 1519
+    },
+    {
+      "epoch": 0.26222720607263006,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.921015736685632e-05,
+      "loss": 1.5253,
+      "step": 1520
+    },
+    {
+      "epoch": 0.262399723971362,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.920909678480544e-05,
+      "loss": 1.4976,
+      "step": 1521
+    },
+    {
+      "epoch": 0.26257224187009404,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.9208035520487125e-05,
+      "loss": 1.5356,
+      "step": 1522
+    },
+    {
+      "epoch": 0.262744759768826,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.920697357398001e-05,
+      "loss": 1.4926,
+      "step": 1523
+    },
+    {
+      "epoch": 0.262917277667558,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.920591094536276e-05,
+      "loss": 1.539,
+      "step": 1524
+    },
+    {
+      "epoch": 0.26308979556629003,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.9204847634714105e-05,
+      "loss": 1.5248,
+      "step": 1525
+    },
+    {
+      "epoch": 0.263262313465022,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.9203783642112825e-05,
+      "loss": 1.4683,
+      "step": 1526
+    },
+    {
+      "epoch": 0.263434831363754,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.9202718967637745e-05,
+      "loss": 1.6038,
+      "step": 1527
+    },
+    {
+      "epoch": 0.26360734926248597,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9201653611367742e-05,
+      "loss": 1.5553,
+      "step": 1528
+    },
+    {
+      "epoch": 0.263779867161218,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9200587573381747e-05,
+      "loss": 1.5876,
+      "step": 1529
+    },
+    {
+      "epoch": 0.26395238505994995,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9199520853758734e-05,
+      "loss": 1.6082,
+      "step": 1530
+    },
+    {
+      "epoch": 0.26412490295868196,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.9198453452577738e-05,
+      "loss": 1.5431,
+      "step": 1531
+    },
+    {
+      "epoch": 0.264297420857414,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9197385369917834e-05,
+      "loss": 1.5877,
+      "step": 1532
+    },
+    {
+      "epoch": 0.26446993875614594,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.9196316605858152e-05,
+      "loss": 1.5066,
+      "step": 1533
+    },
+    {
+      "epoch": 0.26464245665487796,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.9195247160477874e-05,
+      "loss": 1.496,
+      "step": 1534
+    },
+    {
+      "epoch": 0.2648149745536099,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9194177033856233e-05,
+      "loss": 1.4877,
+      "step": 1535
+    },
+    {
+      "epoch": 0.26498749245234193,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.919310622607251e-05,
+      "loss": 1.532,
+      "step": 1536
+    },
+    {
+      "epoch": 0.26516001035107395,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.9192034737206037e-05,
+      "loss": 1.5576,
+      "step": 1537
+    },
+    {
+      "epoch": 0.2653325282498059,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.91909625673362e-05,
+      "loss": 1.4709,
+      "step": 1538
+    },
+    {
+      "epoch": 0.2655050461485379,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9189889716542423e-05,
+      "loss": 1.5417,
+      "step": 1539
+    },
+    {
+      "epoch": 0.2656775640472699,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9188816184904194e-05,
+      "loss": 1.5503,
+      "step": 1540
+    },
+    {
+      "epoch": 0.2658500819460019,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.9187741972501052e-05,
+      "loss": 1.5246,
+      "step": 1541
+    },
+    {
+      "epoch": 0.26602259984473386,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.9186667079412575e-05,
+      "loss": 1.4977,
+      "step": 1542
+    },
+    {
+      "epoch": 0.2661951177434659,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.91855915057184e-05,
+      "loss": 1.4235,
+      "step": 1543
+    },
+    {
+      "epoch": 0.2663676356421979,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.918451525149821e-05,
+      "loss": 1.6124,
+      "step": 1544
+    },
+    {
+      "epoch": 0.26654015354092986,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.9183438316831743e-05,
+      "loss": 1.5644,
+      "step": 1545
+    },
+    {
+      "epoch": 0.2667126714396619,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.918236070179879e-05,
+      "loss": 1.4849,
+      "step": 1546
+    },
+    {
+      "epoch": 0.26688518933839384,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.9181282406479175e-05,
+      "loss": 1.4897,
+      "step": 1547
+    },
+    {
+      "epoch": 0.26705770723712585,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9180203430952794e-05,
+      "loss": 1.4912,
+      "step": 1548
+    },
+    {
+      "epoch": 0.26723022513585787,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.9179123775299584e-05,
+      "loss": 1.5156,
+      "step": 1549
+    },
+    {
+      "epoch": 0.26740274303458983,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.9178043439599528e-05,
+      "loss": 1.6099,
+      "step": 1550
+    },
+    {
+      "epoch": 0.26757526093332185,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.9176962423932674e-05,
+      "loss": 1.4946,
+      "step": 1551
+    },
+    {
+      "epoch": 0.2677477788320538,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.9175880728379094e-05,
+      "loss": 1.5194,
+      "step": 1552
+    },
+    {
+      "epoch": 0.2679202967307858,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.917479835301894e-05,
+      "loss": 1.5548,
+      "step": 1553
+    },
+    {
+      "epoch": 0.26809281462951784,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.91737152979324e-05,
+      "loss": 1.4607,
+      "step": 1554
+    },
+    {
+      "epoch": 0.2682653325282498,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.917263156319971e-05,
+      "loss": 1.5371,
+      "step": 1555
+    },
+    {
+      "epoch": 0.2684378504269818,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9171547148901158e-05,
+      "loss": 1.5787,
+      "step": 1556
+    },
+    {
+      "epoch": 0.2686103683257138,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.9170462055117086e-05,
+      "loss": 1.4766,
+      "step": 1557
+    },
+    {
+      "epoch": 0.2687828862244458,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.916937628192789e-05,
+      "loss": 1.4528,
+      "step": 1558
+    },
+    {
+      "epoch": 0.26895540412317775,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9168289829414002e-05,
+      "loss": 1.5461,
+      "step": 1559
+    },
+    {
+      "epoch": 0.26912792202190977,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.916720269765592e-05,
+      "loss": 1.4871,
+      "step": 1560
+    },
+    {
+      "epoch": 0.2693004399206418,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.916611488673418e-05,
+      "loss": 1.3621,
+      "step": 1561
+    },
+    {
+      "epoch": 0.26947295781937375,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.9165026396729377e-05,
+      "loss": 1.5283,
+      "step": 1562
+    },
+    {
+      "epoch": 0.26964547571810576,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.9163937227722154e-05,
+      "loss": 1.4035,
+      "step": 1563
+    },
+    {
+      "epoch": 0.2698179936168377,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.9162847379793203e-05,
+      "loss": 1.5184,
+      "step": 1564
+    },
+    {
+      "epoch": 0.26999051151556974,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.9161756853023266e-05,
+      "loss": 1.4852,
+      "step": 1565
+    },
+    {
+      "epoch": 0.27016302941430176,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.9160665647493136e-05,
+      "loss": 1.542,
+      "step": 1566
+    },
+    {
+      "epoch": 0.2703355473130337,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.9159573763283654e-05,
+      "loss": 1.4912,
+      "step": 1567
+    },
+    {
+      "epoch": 0.27050806521176574,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.9158481200475715e-05,
+      "loss": 1.5102,
+      "step": 1568
+    },
+    {
+      "epoch": 0.2706805831104977,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9157387959150265e-05,
+      "loss": 1.4076,
+      "step": 1569
+    },
+    {
+      "epoch": 0.2708531010092297,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.91562940393883e-05,
+      "loss": 1.4054,
+      "step": 1570
+    },
+    {
+      "epoch": 0.2710256189079617,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.9155199441270863e-05,
+      "loss": 1.5908,
+      "step": 1571
+    },
+    {
+      "epoch": 0.2711981368066937,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.915410416487904e-05,
+      "loss": 1.518,
+      "step": 1572
+    },
+    {
+      "epoch": 0.2713706547054257,
+      "grad_norm": 0.625,
+      "learning_rate": 1.9153008210293988e-05,
+      "loss": 1.4726,
+      "step": 1573
+    },
+    {
+      "epoch": 0.27154317260415767,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.9151911577596895e-05,
+      "loss": 1.5939,
+      "step": 1574
+    },
+    {
+      "epoch": 0.2717156905028897,
+      "grad_norm": 0.625,
+      "learning_rate": 1.915081426686901e-05,
+      "loss": 1.5122,
+      "step": 1575
+    },
+    {
+      "epoch": 0.27188820840162164,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.9149716278191625e-05,
+      "loss": 1.4648,
+      "step": 1576
+    },
+    {
+      "epoch": 0.27206072630035366,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.914861761164609e-05,
+      "loss": 1.4813,
+      "step": 1577
+    },
+    {
+      "epoch": 0.2722332441990857,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.91475182673138e-05,
+      "loss": 1.5682,
+      "step": 1578
+    },
+    {
+      "epoch": 0.27240576209781764,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.91464182452762e-05,
+      "loss": 1.6082,
+      "step": 1579
+    },
+    {
+      "epoch": 0.27257827999654966,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.9145317545614787e-05,
+      "loss": 1.6143,
+      "step": 1580
+    },
+    {
+      "epoch": 0.2727507978952816,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.9144216168411105e-05,
+      "loss": 1.5168,
+      "step": 1581
+    },
+    {
+      "epoch": 0.27292331579401363,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9143114113746755e-05,
+      "loss": 1.4507,
+      "step": 1582
+    },
+    {
+      "epoch": 0.27309583369274565,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9142011381703384e-05,
+      "loss": 1.738,
+      "step": 1583
+    },
+    {
+      "epoch": 0.2732683515914776,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.9140907972362684e-05,
+      "loss": 1.5551,
+      "step": 1584
+    },
+    {
+      "epoch": 0.2734408694902096,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.9139803885806413e-05,
+      "loss": 1.4516,
+      "step": 1585
+    },
+    {
+      "epoch": 0.2736133873889416,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9138699122116355e-05,
+      "loss": 1.5982,
+      "step": 1586
+    },
+    {
+      "epoch": 0.2737859052876736,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.913759368137437e-05,
+      "loss": 1.5159,
+      "step": 1587
+    },
+    {
+      "epoch": 0.27395842318640556,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.913648756366235e-05,
+      "loss": 1.5375,
+      "step": 1588
+    },
+    {
+      "epoch": 0.2741309410851376,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.913538076906224e-05,
+      "loss": 1.5159,
+      "step": 1589
+    },
+    {
+      "epoch": 0.2743034589838696,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.913427329765604e-05,
+      "loss": 1.5594,
+      "step": 1590
+    },
+    {
+      "epoch": 0.27447597688260156,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.913316514952581e-05,
+      "loss": 1.5788,
+      "step": 1591
+    },
+    {
+      "epoch": 0.2746484947813336,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.9132056324753634e-05,
+      "loss": 1.4806,
+      "step": 1592
+    },
+    {
+      "epoch": 0.27482101268006554,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.9130946823421666e-05,
+      "loss": 1.6308,
+      "step": 1593
+    },
+    {
+      "epoch": 0.27499353057879755,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.9129836645612107e-05,
+      "loss": 1.5213,
+      "step": 1594
+    },
+    {
+      "epoch": 0.27516604847752957,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.91287257914072e-05,
+      "loss": 1.3756,
+      "step": 1595
+    },
+    {
+      "epoch": 0.27533856637626153,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.912761426088925e-05,
+      "loss": 1.591,
+      "step": 1596
+    },
+    {
+      "epoch": 0.27551108427499355,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.91265020541406e-05,
+      "loss": 1.4783,
+      "step": 1597
+    },
+    {
+      "epoch": 0.2756836021737255,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.9125389171243656e-05,
+      "loss": 1.4279,
+      "step": 1598
+    },
+    {
+      "epoch": 0.2758561200724575,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.912427561228086e-05,
+      "loss": 1.487,
+      "step": 1599
+    },
+    {
+      "epoch": 0.2760286379711895,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.912316137733472e-05,
+      "loss": 1.4767,
+      "step": 1600
+    },
+    {
+      "epoch": 0.2760286379711895,
+      "eval_loss": 1.498151421546936,
+      "eval_runtime": 10.8117,
+      "eval_samples_per_second": 94.712,
+      "eval_steps_per_second": 23.678,
+      "step": 1600
+    },
+    {
+      "epoch": 0.2762011558699215,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9122046466487776e-05,
+      "loss": 1.4793,
+      "step": 1601
+    },
+    {
+      "epoch": 0.2763736737686535,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.912093087982264e-05,
+      "loss": 1.6061,
+      "step": 1602
+    },
+    {
+      "epoch": 0.2765461916673855,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.911981461742195e-05,
+      "loss": 1.4781,
+      "step": 1603
+    },
+    {
+      "epoch": 0.2767187095661175,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9118697679368412e-05,
+      "loss": 1.5748,
+      "step": 1604
+    },
+    {
+      "epoch": 0.27689122746484945,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.911758006574477e-05,
+      "loss": 1.5885,
+      "step": 1605
+    },
+    {
+      "epoch": 0.27706374536358147,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.911646177663383e-05,
+      "loss": 1.533,
+      "step": 1606
+    },
+    {
+      "epoch": 0.2772362632623135,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.9115342812118437e-05,
+      "loss": 1.4983,
+      "step": 1607
+    },
+    {
+      "epoch": 0.27740878116104545,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9114223172281498e-05,
+      "loss": 1.4582,
+      "step": 1608
+    },
+    {
+      "epoch": 0.27758129905977746,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.9113102857205952e-05,
+      "loss": 1.4498,
+      "step": 1609
+    },
+    {
+      "epoch": 0.2777538169585094,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.911198186697481e-05,
+      "loss": 1.575,
+      "step": 1610
+    },
+    {
+      "epoch": 0.27792633485724144,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.9110860201671112e-05,
+      "loss": 1.5972,
+      "step": 1611
+    },
+    {
+      "epoch": 0.27809885275597346,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9109737861377967e-05,
+      "loss": 1.5344,
+      "step": 1612
+    },
+    {
+      "epoch": 0.2782713706547054,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.910861484617852e-05,
+      "loss": 1.4919,
+      "step": 1613
+    },
+    {
+      "epoch": 0.27844388855343744,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.9107491156155974e-05,
+      "loss": 1.4568,
+      "step": 1614
+    },
+    {
+      "epoch": 0.2786164064521694,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.9106366791393573e-05,
+      "loss": 1.5189,
+      "step": 1615
+    },
+    {
+      "epoch": 0.2787889243509014,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.9105241751974624e-05,
+      "loss": 1.4218,
+      "step": 1616
+    },
+    {
+      "epoch": 0.2789614422496334,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.910411603798247e-05,
+      "loss": 1.5128,
+      "step": 1617
+    },
+    {
+      "epoch": 0.2791339601483654,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.910298964950052e-05,
+      "loss": 1.5849,
+      "step": 1618
+    },
+    {
+      "epoch": 0.2793064780470974,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.9101862586612214e-05,
+      "loss": 1.4772,
+      "step": 1619
+    },
+    {
+      "epoch": 0.27947899594582937,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.9100734849401063e-05,
+      "loss": 1.5158,
+      "step": 1620
+    },
+    {
+      "epoch": 0.2796515138445614,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.9099606437950605e-05,
+      "loss": 1.5205,
+      "step": 1621
+    },
+    {
+      "epoch": 0.27982403174329334,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9098477352344443e-05,
+      "loss": 1.4296,
+      "step": 1622
+    },
+    {
+      "epoch": 0.27999654964202536,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.9097347592666232e-05,
+      "loss": 1.5096,
+      "step": 1623
+    },
+    {
+      "epoch": 0.2801690675407574,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.9096217158999667e-05,
+      "loss": 1.4957,
+      "step": 1624
+    },
+    {
+      "epoch": 0.28034158543948934,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.90950860514285e-05,
+      "loss": 1.4951,
+      "step": 1625
+    },
+    {
+      "epoch": 0.28051410333822135,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.909395427003653e-05,
+      "loss": 1.5029,
+      "step": 1626
+    },
+    {
+      "epoch": 0.2806866212369533,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.909282181490761e-05,
+      "loss": 1.512,
+      "step": 1627
+    },
+    {
+      "epoch": 0.28085913913568533,
+      "grad_norm": 0.625,
+      "learning_rate": 1.9091688686125628e-05,
+      "loss": 1.5529,
+      "step": 1628
+    },
+    {
+      "epoch": 0.28103165703441735,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.9090554883774547e-05,
+      "loss": 1.5084,
+      "step": 1629
+    },
+    {
+      "epoch": 0.2812041749331493,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.9089420407938354e-05,
+      "loss": 1.4911,
+      "step": 1630
+    },
+    {
+      "epoch": 0.2813766928318813,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.9088285258701108e-05,
+      "loss": 1.5517,
+      "step": 1631
+    },
+    {
+      "epoch": 0.2815492107306133,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.90871494361469e-05,
+      "loss": 1.3719,
+      "step": 1632
+    },
+    {
+      "epoch": 0.2817217286293453,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.9086012940359887e-05,
+      "loss": 1.4742,
+      "step": 1633
+    },
+    {
+      "epoch": 0.28189424652807726,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.908487577142426e-05,
+      "loss": 1.4639,
+      "step": 1634
+    },
+    {
+      "epoch": 0.2820667644268093,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9083737929424272e-05,
+      "loss": 1.5469,
+      "step": 1635
+    },
+    {
+      "epoch": 0.2822392823255413,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.9082599414444222e-05,
+      "loss": 1.5729,
+      "step": 1636
+    },
+    {
+      "epoch": 0.28241180022427326,
+      "grad_norm": 0.625,
+      "learning_rate": 1.9081460226568456e-05,
+      "loss": 1.5184,
+      "step": 1637
+    },
+    {
+      "epoch": 0.2825843181230053,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.908032036588137e-05,
+      "loss": 1.5415,
+      "step": 1638
+    },
+    {
+      "epoch": 0.28275683602173723,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.9079179832467417e-05,
+      "loss": 1.5335,
+      "step": 1639
+    },
+    {
+      "epoch": 0.28292935392046925,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.9078038626411093e-05,
+      "loss": 1.5226,
+      "step": 1640
+    },
+    {
+      "epoch": 0.28310187181920127,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.9076896747796945e-05,
+      "loss": 1.4499,
+      "step": 1641
+    },
+    {
+      "epoch": 0.28327438971793323,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.9075754196709574e-05,
+      "loss": 1.5911,
+      "step": 1642
+    },
+    {
+      "epoch": 0.28344690761666524,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.9074610973233622e-05,
+      "loss": 1.5831,
+      "step": 1643
+    },
+    {
+      "epoch": 0.2836194255153972,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.9073467077453783e-05,
+      "loss": 1.4548,
+      "step": 1644
+    },
+    {
+      "epoch": 0.2837919434141292,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.9072322509454814e-05,
+      "loss": 1.5288,
+      "step": 1645
+    },
+    {
+      "epoch": 0.2839644613128612,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9071177269321507e-05,
+      "loss": 1.4839,
+      "step": 1646
+    },
+    {
+      "epoch": 0.2841369792115932,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.907003135713871e-05,
+      "loss": 1.5464,
+      "step": 1647
+    },
+    {
+      "epoch": 0.2843094971103252,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.9068884772991313e-05,
+      "loss": 1.6519,
+      "step": 1648
+    },
+    {
+      "epoch": 0.2844820150090572,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.9067737516964274e-05,
+      "loss": 1.4762,
+      "step": 1649
+    },
+    {
+      "epoch": 0.2846545329077892,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9066589589142577e-05,
+      "loss": 1.5852,
+      "step": 1650
+    },
+    {
+      "epoch": 0.28482705080652115,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.9065440989611274e-05,
+      "loss": 1.6728,
+      "step": 1651
+    },
+    {
+      "epoch": 0.28499956870525317,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.9064291718455455e-05,
+      "loss": 1.4933,
+      "step": 1652
+    },
+    {
+      "epoch": 0.2851720866039852,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9063141775760274e-05,
+      "loss": 1.5006,
+      "step": 1653
+    },
+    {
+      "epoch": 0.28534460450271715,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.906199116161092e-05,
+      "loss": 1.4257,
+      "step": 1654
+    },
+    {
+      "epoch": 0.28551712240144916,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.906083987609264e-05,
+      "loss": 1.5513,
+      "step": 1655
+    },
+    {
+      "epoch": 0.2856896403001811,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.9059687919290727e-05,
+      "loss": 1.5997,
+      "step": 1656
+    },
+    {
+      "epoch": 0.28586215819891314,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.9058535291290524e-05,
+      "loss": 1.6176,
+      "step": 1657
+    },
+    {
+      "epoch": 0.28603467609764516,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9057381992177426e-05,
+      "loss": 1.5312,
+      "step": 1658
+    },
+    {
+      "epoch": 0.2862071939963771,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.905622802203688e-05,
+      "loss": 1.6063,
+      "step": 1659
+    },
+    {
+      "epoch": 0.28637971189510913,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.9055073380954372e-05,
+      "loss": 1.5179,
+      "step": 1660
+    },
+    {
+      "epoch": 0.2865522297938411,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.905391806901545e-05,
+      "loss": 1.4903,
+      "step": 1661
+    },
+    {
+      "epoch": 0.2867247476925731,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.9052762086305706e-05,
+      "loss": 1.583,
+      "step": 1662
+    },
+    {
+      "epoch": 0.2868972655913051,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.9051605432910783e-05,
+      "loss": 1.5453,
+      "step": 1663
+    },
+    {
+      "epoch": 0.2870697834900371,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9050448108916373e-05,
+      "loss": 1.474,
+      "step": 1664
+    },
+    {
+      "epoch": 0.2872423013887691,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.904929011440822e-05,
+      "loss": 1.4971,
+      "step": 1665
+    },
+    {
+      "epoch": 0.28741481928750107,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.9048131449472107e-05,
+      "loss": 1.5126,
+      "step": 1666
+    },
+    {
+      "epoch": 0.2875873371862331,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.9046972114193884e-05,
+      "loss": 1.4895,
+      "step": 1667
+    },
+    {
+      "epoch": 0.28775985508496504,
+      "grad_norm": 1.2890625,
+      "learning_rate": 1.904581210865944e-05,
+      "loss": 1.4972,
+      "step": 1668
+    },
+    {
+      "epoch": 0.28793237298369706,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.904465143295471e-05,
+      "loss": 1.5122,
+      "step": 1669
+    },
+    {
+      "epoch": 0.2881048908824291,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.904349008716569e-05,
+      "loss": 1.4859,
+      "step": 1670
+    },
+    {
+      "epoch": 0.28827740878116104,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.904232807137842e-05,
+      "loss": 1.5311,
+      "step": 1671
+    },
+    {
+      "epoch": 0.28844992667989305,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.9041165385678984e-05,
+      "loss": 1.4518,
+      "step": 1672
+    },
+    {
+      "epoch": 0.288622444578625,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9040002030153532e-05,
+      "loss": 1.5663,
+      "step": 1673
+    },
+    {
+      "epoch": 0.28879496247735703,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.903883800488824e-05,
+      "loss": 1.4741,
+      "step": 1674
+    },
+    {
+      "epoch": 0.288967480376089,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.9037673309969347e-05,
+      "loss": 1.5701,
+      "step": 1675
+    },
+    {
+      "epoch": 0.289139998274821,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.9036507945483152e-05,
+      "loss": 1.5381,
+      "step": 1676
+    },
+    {
+      "epoch": 0.289312516173553,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.9035341911515983e-05,
+      "loss": 1.4083,
+      "step": 1677
+    },
+    {
+      "epoch": 0.289485034072285,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.9034175208154227e-05,
+      "loss": 1.5595,
+      "step": 1678
+    },
+    {
+      "epoch": 0.289657551971017,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.903300783548433e-05,
+      "loss": 1.5761,
+      "step": 1679
+    },
+    {
+      "epoch": 0.28983006986974896,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.9031839793592764e-05,
+      "loss": 1.5564,
+      "step": 1680
+    },
+    {
+      "epoch": 0.290002587768481,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.9030671082566076e-05,
+      "loss": 1.4697,
+      "step": 1681
+    },
+    {
+      "epoch": 0.290175105667213,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.9029501702490848e-05,
+      "loss": 1.4602,
+      "step": 1682
+    },
+    {
+      "epoch": 0.29034762356594496,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.9028331653453715e-05,
+      "loss": 1.5515,
+      "step": 1683
+    },
+    {
+      "epoch": 0.290520141464677,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.9027160935541365e-05,
+      "loss": 1.3977,
+      "step": 1684
+    },
+    {
+      "epoch": 0.29069265936340893,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.902598954884052e-05,
+      "loss": 1.5655,
+      "step": 1685
+    },
+    {
+      "epoch": 0.29086517726214095,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.902481749343798e-05,
+      "loss": 1.5035,
+      "step": 1686
+    },
+    {
+      "epoch": 0.29103769516087297,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.9023644769420567e-05,
+      "loss": 1.5112,
+      "step": 1687
+    },
+    {
+      "epoch": 0.2912102130596049,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.902247137687517e-05,
+      "loss": 1.5383,
+      "step": 1688
+    },
+    {
+      "epoch": 0.29138273095833694,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.9021297315888715e-05,
+      "loss": 1.5061,
+      "step": 1689
+    },
+    {
+      "epoch": 0.2915552488570689,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.902012258654819e-05,
+      "loss": 1.5495,
+      "step": 1690
+    },
+    {
+      "epoch": 0.2917277667558009,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9018947188940623e-05,
+      "loss": 1.5872,
+      "step": 1691
+    },
+    {
+      "epoch": 0.2919002846545329,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.901777112315309e-05,
+      "loss": 1.4907,
+      "step": 1692
+    },
+    {
+      "epoch": 0.2920728025532649,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.9016594389272734e-05,
+      "loss": 1.5275,
+      "step": 1693
+    },
+    {
+      "epoch": 0.2922453204519969,
+      "grad_norm": 0.75,
+      "learning_rate": 1.9015416987386725e-05,
+      "loss": 1.5049,
+      "step": 1694
+    },
+    {
+      "epoch": 0.2924178383507289,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.9014238917582297e-05,
+      "loss": 1.6413,
+      "step": 1695
+    },
+    {
+      "epoch": 0.2925903562494609,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.9013060179946722e-05,
+      "loss": 1.4572,
+      "step": 1696
+    },
+    {
+      "epoch": 0.29276287414819285,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.901188077456733e-05,
+      "loss": 1.4106,
+      "step": 1697
+    },
+    {
+      "epoch": 0.29293539204692487,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.901070070153151e-05,
+      "loss": 1.5794,
+      "step": 1698
+    },
+    {
+      "epoch": 0.2931079099456569,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.900951996092667e-05,
+      "loss": 1.5461,
+      "step": 1699
+    },
+    {
+      "epoch": 0.29328042784438885,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.9008338552840308e-05,
+      "loss": 1.4498,
+      "step": 1700
+    },
+    {
+      "epoch": 0.29328042784438885,
+      "eval_loss": 1.4925167560577393,
+      "eval_runtime": 11.161,
+      "eval_samples_per_second": 91.748,
+      "eval_steps_per_second": 22.937,
+      "step": 1700
+    },
+    {
+      "epoch": 0.29345294574312086,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.9007156477359935e-05,
+      "loss": 1.5066,
+      "step": 1701
+    },
+    {
+      "epoch": 0.2936254636418528,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.900597373457313e-05,
+      "loss": 1.4833,
+      "step": 1702
+    },
+    {
+      "epoch": 0.29379798154058484,
+      "grad_norm": 0.875,
+      "learning_rate": 1.900479032456752e-05,
+      "loss": 1.4667,
+      "step": 1703
+    },
+    {
+      "epoch": 0.2939704994393168,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9003606247430774e-05,
+      "loss": 1.5012,
+      "step": 1704
+    },
+    {
+      "epoch": 0.2941430173380488,
+      "grad_norm": 0.75,
+      "learning_rate": 1.9002421503250626e-05,
+      "loss": 1.5049,
+      "step": 1705
+    },
+    {
+      "epoch": 0.29431553523678083,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.900123609211484e-05,
+      "loss": 1.5594,
+      "step": 1706
+    },
+    {
+      "epoch": 0.2944880531355128,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.9000050014111245e-05,
+      "loss": 1.4642,
+      "step": 1707
+    },
+    {
+      "epoch": 0.2946605710342448,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.8998863269327706e-05,
+      "loss": 1.5367,
+      "step": 1708
+    },
+    {
+      "epoch": 0.2948330889329768,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.8997675857852148e-05,
+      "loss": 1.3788,
+      "step": 1709
+    },
+    {
+      "epoch": 0.2950056068317088,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.899648777977255e-05,
+      "loss": 1.5682,
+      "step": 1710
+    },
+    {
+      "epoch": 0.2951781247304408,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.8995299035176914e-05,
+      "loss": 1.4895,
+      "step": 1711
+    },
+    {
+      "epoch": 0.29535064262917277,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.8994109624153327e-05,
+      "loss": 1.3757,
+      "step": 1712
+    },
+    {
+      "epoch": 0.2955231605279048,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.89929195467899e-05,
+      "loss": 1.4907,
+      "step": 1713
+    },
+    {
+      "epoch": 0.29569567842663674,
+      "grad_norm": 0.875,
+      "learning_rate": 1.8991728803174804e-05,
+      "loss": 1.3716,
+      "step": 1714
+    },
+    {
+      "epoch": 0.29586819632536876,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.899053739339625e-05,
+      "loss": 1.5605,
+      "step": 1715
+    },
+    {
+      "epoch": 0.2960407142241008,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.8989345317542516e-05,
+      "loss": 1.5747,
+      "step": 1716
+    },
+    {
+      "epoch": 0.29621323212283274,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.8988152575701912e-05,
+      "loss": 1.5484,
+      "step": 1717
+    },
+    {
+      "epoch": 0.29638575002156475,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.8986959167962806e-05,
+      "loss": 1.4819,
+      "step": 1718
+    },
+    {
+      "epoch": 0.2965582679202967,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.8985765094413607e-05,
+      "loss": 1.526,
+      "step": 1719
+    },
+    {
+      "epoch": 0.29673078581902873,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.8984570355142787e-05,
+      "loss": 1.5645,
+      "step": 1720
+    },
+    {
+      "epoch": 0.2969033037177607,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.8983374950238854e-05,
+      "loss": 1.5015,
+      "step": 1721
+    },
+    {
+      "epoch": 0.2970758216164927,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.8982178879790377e-05,
+      "loss": 1.5595,
+      "step": 1722
+    },
+    {
+      "epoch": 0.2972483395152247,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.8980982143885965e-05,
+      "loss": 1.5058,
+      "step": 1723
+    },
+    {
+      "epoch": 0.2974208574139567,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.8979784742614283e-05,
+      "loss": 1.5412,
+      "step": 1724
+    },
+    {
+      "epoch": 0.2975933753126887,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.8978586676064036e-05,
+      "loss": 1.4126,
+      "step": 1725
+    },
+    {
+      "epoch": 0.29776589321142066,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.8977387944323985e-05,
+      "loss": 1.5317,
+      "step": 1726
+    },
+    {
+      "epoch": 0.2979384111101527,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.8976188547482944e-05,
+      "loss": 1.5283,
+      "step": 1727
+    },
+    {
+      "epoch": 0.2981109290088847,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.897498848562977e-05,
+      "loss": 1.5747,
+      "step": 1728
+    },
+    {
+      "epoch": 0.29828344690761666,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.8973787758853367e-05,
+      "loss": 1.4914,
+      "step": 1729
+    },
+    {
+      "epoch": 0.2984559648063487,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.89725863672427e-05,
+      "loss": 1.5348,
+      "step": 1730
+    },
+    {
+      "epoch": 0.29862848270508063,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.8971384310886773e-05,
+      "loss": 1.468,
+      "step": 1731
+    },
+    {
+      "epoch": 0.29880100060381265,
+      "grad_norm": 0.625,
+      "learning_rate": 1.8970181589874637e-05,
+      "loss": 1.4735,
+      "step": 1732
+    },
+    {
+      "epoch": 0.29897351850254467,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.89689782042954e-05,
+      "loss": 1.46,
+      "step": 1733
+    },
+    {
+      "epoch": 0.2991460364012766,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.896777415423822e-05,
+      "loss": 1.5731,
+      "step": 1734
+    },
+    {
+      "epoch": 0.29931855430000864,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.8966569439792294e-05,
+      "loss": 1.4503,
+      "step": 1735
+    },
+    {
+      "epoch": 0.2994910721987406,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.896536406104688e-05,
+      "loss": 1.5463,
+      "step": 1736
+    },
+    {
+      "epoch": 0.2996635900974726,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.8964158018091278e-05,
+      "loss": 1.5302,
+      "step": 1737
+    },
+    {
+      "epoch": 0.2998361079962046,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.8962951311014842e-05,
+      "loss": 1.4495,
+      "step": 1738
+    },
+    {
+      "epoch": 0.3000086258949366,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.896174393990697e-05,
+      "loss": 1.5159,
+      "step": 1739
+    },
+    {
+      "epoch": 0.3001811437936686,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.8960535904857103e-05,
+      "loss": 1.4311,
+      "step": 1740
+    },
+    {
+      "epoch": 0.3003536616924006,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.8959327205954757e-05,
+      "loss": 1.5733,
+      "step": 1741
+    },
+    {
+      "epoch": 0.3005261795911326,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.895811784328947e-05,
+      "loss": 1.4511,
+      "step": 1742
+    },
+    {
+      "epoch": 0.30069869748986455,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.8956907816950837e-05,
+      "loss": 1.4835,
+      "step": 1743
+    },
+    {
+      "epoch": 0.30087121538859657,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.895569712702851e-05,
+      "loss": 1.4882,
+      "step": 1744
+    },
+    {
+      "epoch": 0.3010437332873286,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.895448577361218e-05,
+      "loss": 1.5585,
+      "step": 1745
+    },
+    {
+      "epoch": 0.30121625118606055,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.8953273756791595e-05,
+      "loss": 1.4595,
+      "step": 1746
+    },
+    {
+      "epoch": 0.30138876908479256,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.8952061076656547e-05,
+      "loss": 1.56,
+      "step": 1747
+    },
+    {
+      "epoch": 0.3015612869835245,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.8950847733296877e-05,
+      "loss": 1.4842,
+      "step": 1748
+    },
+    {
+      "epoch": 0.30173380488225654,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.8949633726802484e-05,
+      "loss": 1.5915,
+      "step": 1749
+    },
+    {
+      "epoch": 0.3019063227809885,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.89484190572633e-05,
+      "loss": 1.5447,
+      "step": 1750
+    },
+    {
+      "epoch": 0.3020788406797205,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.8947203724769324e-05,
+      "loss": 1.5063,
+      "step": 1751
+    },
+    {
+      "epoch": 0.30225135857845253,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.894598772941059e-05,
+      "loss": 1.4247,
+      "step": 1752
+    },
+    {
+      "epoch": 0.3024238764771845,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.8944771071277188e-05,
+      "loss": 1.5098,
+      "step": 1753
+    },
+    {
+      "epoch": 0.3025963943759165,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.8943553750459256e-05,
+      "loss": 1.6315,
+      "step": 1754
+    },
+    {
+      "epoch": 0.30276891227464847,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.8942335767046978e-05,
+      "loss": 1.5184,
+      "step": 1755
+    },
+    {
+      "epoch": 0.3029414301733805,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.8941117121130594e-05,
+      "loss": 1.429,
+      "step": 1756
+    },
+    {
+      "epoch": 0.3031139480721125,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.8939897812800385e-05,
+      "loss": 1.5151,
+      "step": 1757
+    },
+    {
+      "epoch": 0.30328646597084447,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.893867784214669e-05,
+      "loss": 1.491,
+      "step": 1758
+    },
+    {
+      "epoch": 0.3034589838695765,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.8937457209259888e-05,
+      "loss": 1.4795,
+      "step": 1759
+    },
+    {
+      "epoch": 0.30363150176830844,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.893623591423041e-05,
+      "loss": 1.5306,
+      "step": 1760
+    },
+    {
+      "epoch": 0.30380401966704046,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.893501395714874e-05,
+      "loss": 1.4723,
+      "step": 1761
+    },
+    {
+      "epoch": 0.3039765375657725,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.893379133810541e-05,
+      "loss": 1.4674,
+      "step": 1762
+    },
+    {
+      "epoch": 0.30414905546450444,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.8932568057190995e-05,
+      "loss": 1.4916,
+      "step": 1763
+    },
+    {
+      "epoch": 0.30432157336323645,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.8931344114496127e-05,
+      "loss": 1.4793,
+      "step": 1764
+    },
+    {
+      "epoch": 0.3044940912619684,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.8930119510111476e-05,
+      "loss": 1.5519,
+      "step": 1765
+    },
+    {
+      "epoch": 0.30466660916070043,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.8928894244127782e-05,
+      "loss": 1.5546,
+      "step": 1766
+    },
+    {
+      "epoch": 0.3048391270594324,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.8927668316635804e-05,
+      "loss": 1.6417,
+      "step": 1767
+    },
+    {
+      "epoch": 0.3050116449581644,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.892644172772638e-05,
+      "loss": 1.5674,
+      "step": 1768
+    },
+    {
+      "epoch": 0.3051841628568964,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.8925214477490373e-05,
+      "loss": 1.5066,
+      "step": 1769
+    },
+    {
+      "epoch": 0.3053566807556284,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.8923986566018717e-05,
+      "loss": 1.571,
+      "step": 1770
+    },
+    {
+      "epoch": 0.3055291986543604,
+      "grad_norm": 0.75,
+      "learning_rate": 1.892275799340237e-05,
+      "loss": 1.5209,
+      "step": 1771
+    },
+    {
+      "epoch": 0.30570171655309236,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.8921528759732363e-05,
+      "loss": 1.388,
+      "step": 1772
+    },
+    {
+      "epoch": 0.3058742344518244,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.892029886509976e-05,
+      "loss": 1.5195,
+      "step": 1773
+    },
+    {
+      "epoch": 0.3060467523505564,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.891906830959568e-05,
+      "loss": 1.5106,
+      "step": 1774
+    },
+    {
+      "epoch": 0.30621927024928836,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.891783709331129e-05,
+      "loss": 1.3891,
+      "step": 1775
+    },
+    {
+      "epoch": 0.30639178814802037,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.8916605216337807e-05,
+      "loss": 1.5229,
+      "step": 1776
+    },
+    {
+      "epoch": 0.30656430604675233,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.8915372678766497e-05,
+      "loss": 1.4953,
+      "step": 1777
+    },
+    {
+      "epoch": 0.30673682394548435,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.8914139480688672e-05,
+      "loss": 1.5302,
+      "step": 1778
+    },
+    {
+      "epoch": 0.3069093418442163,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.89129056221957e-05,
+      "loss": 1.4382,
+      "step": 1779
+    },
+    {
+      "epoch": 0.3070818597429483,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.8911671103378983e-05,
+      "loss": 1.4877,
+      "step": 1780
+    },
+    {
+      "epoch": 0.30725437764168034,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.8910435924329993e-05,
+      "loss": 1.4971,
+      "step": 1781
+    },
+    {
+      "epoch": 0.3074268955404123,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.8909200085140233e-05,
+      "loss": 1.6221,
+      "step": 1782
+    },
+    {
+      "epoch": 0.3075994134391443,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.890796358590126e-05,
+      "loss": 1.5684,
+      "step": 1783
+    },
+    {
+      "epoch": 0.3077719313378763,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.890672642670469e-05,
+      "loss": 1.4768,
+      "step": 1784
+    },
+    {
+      "epoch": 0.3079444492366083,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.8905488607642172e-05,
+      "loss": 1.5532,
+      "step": 1785
+    },
+    {
+      "epoch": 0.3081169671353403,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.8904250128805418e-05,
+      "loss": 1.5317,
+      "step": 1786
+    },
+    {
+      "epoch": 0.3082894850340723,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.8903010990286174e-05,
+      "loss": 1.5159,
+      "step": 1787
+    },
+    {
+      "epoch": 0.3084620029328043,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.8901771192176248e-05,
+      "loss": 1.4687,
+      "step": 1788
+    },
+    {
+      "epoch": 0.30863452083153625,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.8900530734567492e-05,
+      "loss": 1.5036,
+      "step": 1789
+    },
+    {
+      "epoch": 0.30880703873026827,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.8899289617551803e-05,
+      "loss": 1.5833,
+      "step": 1790
+    },
+    {
+      "epoch": 0.3089795566290003,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.889804784122114e-05,
+      "loss": 1.442,
+      "step": 1791
+    },
+    {
+      "epoch": 0.30915207452773225,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.889680540566749e-05,
+      "loss": 1.468,
+      "step": 1792
+    },
+    {
+      "epoch": 0.30932459242646426,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.8895562310982907e-05,
+      "loss": 1.4224,
+      "step": 1793
+    },
+    {
+      "epoch": 0.3094971103251962,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.8894318557259485e-05,
+      "loss": 1.4192,
+      "step": 1794
+    },
+    {
+      "epoch": 0.30966962822392824,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.889307414458937e-05,
+      "loss": 1.4496,
+      "step": 1795
+    },
+    {
+      "epoch": 0.3098421461226602,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.8891829073064757e-05,
+      "loss": 1.5036,
+      "step": 1796
+    },
+    {
+      "epoch": 0.3100146640213922,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.889058334277789e-05,
+      "loss": 1.5235,
+      "step": 1797
+    },
+    {
+      "epoch": 0.31018718192012423,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.8889336953821055e-05,
+      "loss": 1.6028,
+      "step": 1798
+    },
+    {
+      "epoch": 0.3103596998188562,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.8888089906286598e-05,
+      "loss": 1.3856,
+      "step": 1799
+    },
+    {
+      "epoch": 0.3105322177175882,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.8886842200266905e-05,
+      "loss": 1.473,
+      "step": 1800
+    },
+    {
+      "epoch": 0.3105322177175882,
+      "eval_loss": 1.4874216318130493,
+      "eval_runtime": 10.8464,
+      "eval_samples_per_second": 94.409,
+      "eval_steps_per_second": 23.602,
+      "step": 1800
+    },
+    {
+      "epoch": 0.31070473561632017,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.888559383585441e-05,
+      "loss": 1.6011,
+      "step": 1801
+    },
+    {
+      "epoch": 0.3108772535150522,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.888434481314161e-05,
+      "loss": 1.4107,
+      "step": 1802
+    },
+    {
+      "epoch": 0.3110497714137842,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.888309513222103e-05,
+      "loss": 1.5343,
+      "step": 1803
+    },
+    {
+      "epoch": 0.31122228931251616,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.8881844793185257e-05,
+      "loss": 1.5289,
+      "step": 1804
+    },
+    {
+      "epoch": 0.3113948072112482,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.8880593796126925e-05,
+      "loss": 1.436,
+      "step": 1805
+    },
+    {
+      "epoch": 0.31156732510998014,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.887934214113872e-05,
+      "loss": 1.5412,
+      "step": 1806
+    },
+    {
+      "epoch": 0.31173984300871216,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.887808982831337e-05,
+      "loss": 1.5912,
+      "step": 1807
+    },
+    {
+      "epoch": 0.3119123609074442,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.887683685774365e-05,
+      "loss": 1.4521,
+      "step": 1808
+    },
+    {
+      "epoch": 0.31208487880617614,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.887558322952239e-05,
+      "loss": 1.5337,
+      "step": 1809
+    },
+    {
+      "epoch": 0.31225739670490815,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.887432894374247e-05,
+      "loss": 1.6269,
+      "step": 1810
+    },
+    {
+      "epoch": 0.3124299146036401,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.8873074000496808e-05,
+      "loss": 1.469,
+      "step": 1811
+    },
+    {
+      "epoch": 0.31260243250237213,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.8871818399878387e-05,
+      "loss": 1.495,
+      "step": 1812
+    },
+    {
+      "epoch": 0.3127749504011041,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.887056214198022e-05,
+      "loss": 1.5321,
+      "step": 1813
+    },
+    {
+      "epoch": 0.3129474682998361,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.8869305226895386e-05,
+      "loss": 1.5165,
+      "step": 1814
+    },
+    {
+      "epoch": 0.3131199861985681,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.8868047654717005e-05,
+      "loss": 1.5412,
+      "step": 1815
+    },
+    {
+      "epoch": 0.3132925040973001,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.886678942553824e-05,
+      "loss": 1.4872,
+      "step": 1816
+    },
+    {
+      "epoch": 0.3134650219960321,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.8865530539452316e-05,
+      "loss": 1.5929,
+      "step": 1817
+    },
+    {
+      "epoch": 0.31363753989476406,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.8864270996552494e-05,
+      "loss": 1.5027,
+      "step": 1818
+    },
+    {
+      "epoch": 0.3138100577934961,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.886301079693209e-05,
+      "loss": 1.5103,
+      "step": 1819
+    },
+    {
+      "epoch": 0.3139825756922281,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.8861749940684464e-05,
+      "loss": 1.5923,
+      "step": 1820
+    },
+    {
+      "epoch": 0.31415509359096006,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.8860488427903038e-05,
+      "loss": 1.4984,
+      "step": 1821
+    },
+    {
+      "epoch": 0.31432761148969207,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.8859226258681262e-05,
+      "loss": 1.5425,
+      "step": 1822
+    },
+    {
+      "epoch": 0.31450012938842403,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.885796343311265e-05,
+      "loss": 1.425,
+      "step": 1823
+    },
+    {
+      "epoch": 0.31467264728715605,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.885669995129076e-05,
+      "loss": 1.4554,
+      "step": 1824
+    },
+    {
+      "epoch": 0.314845165185888,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.8855435813309196e-05,
+      "loss": 1.4866,
+      "step": 1825
+    },
+    {
+      "epoch": 0.31501768308462,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.885417101926162e-05,
+      "loss": 1.5125,
+      "step": 1826
+    },
+    {
+      "epoch": 0.31519020098335204,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.885290556924173e-05,
+      "loss": 1.4906,
+      "step": 1827
+    },
+    {
+      "epoch": 0.315362718882084,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.885163946334328e-05,
+      "loss": 1.5114,
+      "step": 1828
+    },
+    {
+      "epoch": 0.315535236780816,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.8850372701660072e-05,
+      "loss": 1.4366,
+      "step": 1829
+    },
+    {
+      "epoch": 0.315707754679548,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.8849105284285954e-05,
+      "loss": 1.4678,
+      "step": 1830
+    },
+    {
+      "epoch": 0.31588027257828,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.8847837211314822e-05,
+      "loss": 1.5652,
+      "step": 1831
+    },
+    {
+      "epoch": 0.316052790477012,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.8846568482840628e-05,
+      "loss": 1.5195,
+      "step": 1832
+    },
+    {
+      "epoch": 0.316225308375744,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.8845299098957366e-05,
+      "loss": 1.5487,
+      "step": 1833
+    },
+    {
+      "epoch": 0.316397826274476,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.8844029059759076e-05,
+      "loss": 1.4679,
+      "step": 1834
+    },
+    {
+      "epoch": 0.31657034417320795,
+      "grad_norm": 0.75,
+      "learning_rate": 1.8842758365339856e-05,
+      "loss": 1.4689,
+      "step": 1835
+    },
+    {
+      "epoch": 0.31674286207193997,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.884148701579384e-05,
+      "loss": 1.5574,
+      "step": 1836
+    },
+    {
+      "epoch": 0.316915379970672,
+      "grad_norm": 1.375,
+      "learning_rate": 1.884021501121523e-05,
+      "loss": 1.5667,
+      "step": 1837
+    },
+    {
+      "epoch": 0.31708789786940395,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.883894235169825e-05,
+      "loss": 1.4987,
+      "step": 1838
+    },
+    {
+      "epoch": 0.31726041576813596,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.8837669037337188e-05,
+      "loss": 1.5074,
+      "step": 1839
+    },
+    {
+      "epoch": 0.3174329336668679,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.883639506822639e-05,
+      "loss": 1.4288,
+      "step": 1840
+    },
+    {
+      "epoch": 0.31760545156559994,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.883512044446023e-05,
+      "loss": 1.5972,
+      "step": 1841
+    },
+    {
+      "epoch": 0.3177779694643319,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.8833845166133145e-05,
+      "loss": 1.3981,
+      "step": 1842
+    },
+    {
+      "epoch": 0.3179504873630639,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.883256923333961e-05,
+      "loss": 1.5039,
+      "step": 1843
+    },
+    {
+      "epoch": 0.31812300526179593,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.8831292646174163e-05,
+      "loss": 1.4579,
+      "step": 1844
+    },
+    {
+      "epoch": 0.3182955231605279,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.8830015404731375e-05,
+      "loss": 1.5289,
+      "step": 1845
+    },
+    {
+      "epoch": 0.3184680410592599,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.8828737509105873e-05,
+      "loss": 1.5776,
+      "step": 1846
+    },
+    {
+      "epoch": 0.31864055895799187,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.882745895939233e-05,
+      "loss": 1.4838,
+      "step": 1847
+    },
+    {
+      "epoch": 0.3188130768567239,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.882617975568547e-05,
+      "loss": 1.4642,
+      "step": 1848
+    },
+    {
+      "epoch": 0.3189855947554559,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.882489989808007e-05,
+      "loss": 1.4736,
+      "step": 1849
+    },
+    {
+      "epoch": 0.31915811265418786,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.882361938667094e-05,
+      "loss": 1.5478,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3193306305529199,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.8822338221552955e-05,
+      "loss": 1.5369,
+      "step": 1851
+    },
+    {
+      "epoch": 0.31950314845165184,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.882105640282103e-05,
+      "loss": 1.4478,
+      "step": 1852
+    },
+    {
+      "epoch": 0.31967566635038386,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.881977393057013e-05,
+      "loss": 1.5034,
+      "step": 1853
+    },
+    {
+      "epoch": 0.3198481842491158,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.881849080489527e-05,
+      "loss": 1.5675,
+      "step": 1854
+    },
+    {
+      "epoch": 0.32002070214784784,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.881720702589151e-05,
+      "loss": 1.5108,
+      "step": 1855
+    },
+    {
+      "epoch": 0.32019322004657985,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.881592259365396e-05,
+      "loss": 1.5053,
+      "step": 1856
+    },
+    {
+      "epoch": 0.3203657379453118,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.881463750827778e-05,
+      "loss": 1.5251,
+      "step": 1857
+    },
+    {
+      "epoch": 0.32053825584404383,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.881335176985818e-05,
+      "loss": 1.4477,
+      "step": 1858
+    },
+    {
+      "epoch": 0.3207107737427758,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.881206537849041e-05,
+      "loss": 1.5098,
+      "step": 1859
+    },
+    {
+      "epoch": 0.3208832916415078,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.8810778334269778e-05,
+      "loss": 1.4897,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3210558095402398,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.880949063729163e-05,
+      "loss": 1.4835,
+      "step": 1861
+    },
+    {
+      "epoch": 0.3212283274389718,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.8808202287651375e-05,
+      "loss": 1.5093,
+      "step": 1862
+    },
+    {
+      "epoch": 0.3214008453377038,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.880691328544446e-05,
+      "loss": 1.5226,
+      "step": 1863
+    },
+    {
+      "epoch": 0.32157336323643576,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.880562363076638e-05,
+      "loss": 1.498,
+      "step": 1864
+    },
+    {
+      "epoch": 0.3217458811351678,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.880433332371268e-05,
+      "loss": 1.543,
+      "step": 1865
+    },
+    {
+      "epoch": 0.3219183990338998,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.8803042364378955e-05,
+      "loss": 1.5232,
+      "step": 1866
+    },
+    {
+      "epoch": 0.32209091693263175,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.8801750752860847e-05,
+      "loss": 1.574,
+      "step": 1867
+    },
+    {
+      "epoch": 0.32226343483136377,
+      "grad_norm": 0.875,
+      "learning_rate": 1.880045848925405e-05,
+      "loss": 1.5205,
+      "step": 1868
+    },
+    {
+      "epoch": 0.32243595273009573,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.87991655736543e-05,
+      "loss": 1.5055,
+      "step": 1869
+    },
+    {
+      "epoch": 0.32260847062882775,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.8797872006157382e-05,
+      "loss": 1.5527,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3227809885275597,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.8796577786859136e-05,
+      "loss": 1.5301,
+      "step": 1871
+    },
+    {
+      "epoch": 0.3229535064262917,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.8795282915855445e-05,
+      "loss": 1.5342,
+      "step": 1872
+    },
+    {
+      "epoch": 0.32312602432502374,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.8793987393242236e-05,
+      "loss": 1.5328,
+      "step": 1873
+    },
+    {
+      "epoch": 0.3232985422237557,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.8792691219115496e-05,
+      "loss": 1.4506,
+      "step": 1874
+    },
+    {
+      "epoch": 0.3234710601224877,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.8791394393571255e-05,
+      "loss": 1.5116,
+      "step": 1875
+    },
+    {
+      "epoch": 0.3236435780212197,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.879009691670558e-05,
+      "loss": 1.5311,
+      "step": 1876
+    },
+    {
+      "epoch": 0.3238160959199517,
+      "grad_norm": 0.75,
+      "learning_rate": 1.8788798788614606e-05,
+      "loss": 1.4899,
+      "step": 1877
+    },
+    {
+      "epoch": 0.3239886138186837,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.8787500009394503e-05,
+      "loss": 1.5621,
+      "step": 1878
+    },
+    {
+      "epoch": 0.3241611317174157,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.8786200579141488e-05,
+      "loss": 1.5273,
+      "step": 1879
+    },
+    {
+      "epoch": 0.3243336496161477,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.878490049795184e-05,
+      "loss": 1.4859,
+      "step": 1880
+    },
+    {
+      "epoch": 0.32450616751487965,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.878359976592187e-05,
+      "loss": 1.4424,
+      "step": 1881
+    },
+    {
+      "epoch": 0.32467868541361167,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.8782298383147946e-05,
+      "loss": 1.4807,
+      "step": 1882
+    },
+    {
+      "epoch": 0.32485120331234363,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.8780996349726488e-05,
+      "loss": 1.5346,
+      "step": 1883
+    },
+    {
+      "epoch": 0.32502372121107564,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.8779693665753954e-05,
+      "loss": 1.5185,
+      "step": 1884
+    },
+    {
+      "epoch": 0.32519623910980766,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.877839033132685e-05,
+      "loss": 1.5993,
+      "step": 1885
+    },
+    {
+      "epoch": 0.3253687570085396,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.8777086346541743e-05,
+      "loss": 1.5468,
+      "step": 1886
+    },
+    {
+      "epoch": 0.32554127490727164,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.8775781711495237e-05,
+      "loss": 1.5053,
+      "step": 1887
+    },
+    {
+      "epoch": 0.3257137928060036,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.8774476426283993e-05,
+      "loss": 1.4857,
+      "step": 1888
+    },
+    {
+      "epoch": 0.3258863107047356,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.8773170491004704e-05,
+      "loss": 1.52,
+      "step": 1889
+    },
+    {
+      "epoch": 0.32605882860346763,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.877186390575413e-05,
+      "loss": 1.5068,
+      "step": 1890
+    },
+    {
+      "epoch": 0.3262313465021996,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.877055667062907e-05,
+      "loss": 1.4402,
+      "step": 1891
+    },
+    {
+      "epoch": 0.3264038644009316,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.8769248785726367e-05,
+      "loss": 1.4986,
+      "step": 1892
+    },
+    {
+      "epoch": 0.32657638229966357,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.8767940251142924e-05,
+      "loss": 1.4641,
+      "step": 1893
+    },
+    {
+      "epoch": 0.3267489001983956,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.8766631066975684e-05,
+      "loss": 1.5435,
+      "step": 1894
+    },
+    {
+      "epoch": 0.3269214180971276,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.8765321233321634e-05,
+      "loss": 1.5584,
+      "step": 1895
+    },
+    {
+      "epoch": 0.32709393599585956,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.876401075027782e-05,
+      "loss": 1.5136,
+      "step": 1896
+    },
+    {
+      "epoch": 0.3272664538945916,
+      "grad_norm": 1.234375,
+      "learning_rate": 1.8762699617941333e-05,
+      "loss": 1.5428,
+      "step": 1897
+    },
+    {
+      "epoch": 0.32743897179332354,
+      "grad_norm": 1.1875,
+      "learning_rate": 1.87613878364093e-05,
+      "loss": 1.5206,
+      "step": 1898
+    },
+    {
+      "epoch": 0.32761148969205556,
+      "grad_norm": 0.75,
+      "learning_rate": 1.8760075405778918e-05,
+      "loss": 1.5195,
+      "step": 1899
+    },
+    {
+      "epoch": 0.3277840075907875,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.8758762326147414e-05,
+      "loss": 1.4737,
+      "step": 1900
+    },
+    {
+      "epoch": 0.3277840075907875,
+      "eval_loss": 1.4825938940048218,
+      "eval_runtime": 10.8069,
+      "eval_samples_per_second": 94.755,
+      "eval_steps_per_second": 23.689,
+      "step": 1900
+    },
+    {
+      "epoch": 0.32795652548951953,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.875744859761207e-05,
+      "loss": 1.6023,
+      "step": 1901
+    },
+    {
+      "epoch": 0.32812904338825155,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.875613422027021e-05,
+      "loss": 1.5445,
+      "step": 1902
+    },
+    {
+      "epoch": 0.3283015612869835,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.875481919421922e-05,
+      "loss": 1.4026,
+      "step": 1903
+    },
+    {
+      "epoch": 0.32847407918571553,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.875350351955652e-05,
+      "loss": 1.5081,
+      "step": 1904
+    },
+    {
+      "epoch": 0.3286465970844475,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.8752187196379585e-05,
+      "loss": 1.5393,
+      "step": 1905
+    },
+    {
+      "epoch": 0.3288191149831795,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.875087022478594e-05,
+      "loss": 1.5503,
+      "step": 1906
+    },
+    {
+      "epoch": 0.3289916328819115,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.8749552604873146e-05,
+      "loss": 1.4422,
+      "step": 1907
+    },
+    {
+      "epoch": 0.3291641507806435,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.874823433673883e-05,
+      "loss": 1.4446,
+      "step": 1908
+    },
+    {
+      "epoch": 0.3293366686793755,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.8746915420480646e-05,
+      "loss": 1.6119,
+      "step": 1909
+    },
+    {
+      "epoch": 0.32950918657810746,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.8745595856196318e-05,
+      "loss": 1.4752,
+      "step": 1910
+    },
+    {
+      "epoch": 0.3296817044768395,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.8744275643983606e-05,
+      "loss": 1.4901,
+      "step": 1911
+    },
+    {
+      "epoch": 0.3298542223755715,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.8742954783940313e-05,
+      "loss": 1.4867,
+      "step": 1912
+    },
+    {
+      "epoch": 0.33002674027430345,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.8741633276164305e-05,
+      "loss": 1.446,
+      "step": 1913
+    },
+    {
+      "epoch": 0.33019925817303547,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.8740311120753482e-05,
+      "loss": 1.5655,
+      "step": 1914
+    },
+    {
+      "epoch": 0.33037177607176743,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.87389883178058e-05,
+      "loss": 1.4538,
+      "step": 1915
+    },
+    {
+      "epoch": 0.33054429397049945,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.8737664867419262e-05,
+      "loss": 1.4357,
+      "step": 1916
+    },
+    {
+      "epoch": 0.3307168118692314,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.8736340769691912e-05,
+      "loss": 1.5557,
+      "step": 1917
+    },
+    {
+      "epoch": 0.3308893297679634,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.873501602472185e-05,
+      "loss": 1.5924,
+      "step": 1918
+    },
+    {
+      "epoch": 0.33106184766669544,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.8733690632607223e-05,
+      "loss": 1.5446,
+      "step": 1919
+    },
+    {
+      "epoch": 0.3312343655654274,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.8732364593446223e-05,
+      "loss": 1.4492,
+      "step": 1920
+    },
+    {
+      "epoch": 0.3314068834641594,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.8731037907337094e-05,
+      "loss": 1.4965,
+      "step": 1921
+    },
+    {
+      "epoch": 0.3315794013628914,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.872971057437812e-05,
+      "loss": 1.49,
+      "step": 1922
+    },
+    {
+      "epoch": 0.3317519192616234,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.872838259466764e-05,
+      "loss": 1.4051,
+      "step": 1923
+    },
+    {
+      "epoch": 0.3319244371603554,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.8727053968304044e-05,
+      "loss": 1.5263,
+      "step": 1924
+    },
+    {
+      "epoch": 0.3320969550590874,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.8725724695385757e-05,
+      "loss": 1.5625,
+      "step": 1925
+    },
+    {
+      "epoch": 0.3322694729578194,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.8724394776011264e-05,
+      "loss": 1.461,
+      "step": 1926
+    },
+    {
+      "epoch": 0.33244199085655135,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.8723064210279096e-05,
+      "loss": 1.4699,
+      "step": 1927
+    },
+    {
+      "epoch": 0.33261450875528337,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.8721732998287825e-05,
+      "loss": 1.4403,
+      "step": 1928
+    },
+    {
+      "epoch": 0.3327870266540153,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.872040114013608e-05,
+      "loss": 1.5245,
+      "step": 1929
+    },
+    {
+      "epoch": 0.33295954455274734,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.871906863592253e-05,
+      "loss": 1.5167,
+      "step": 1930
+    },
+    {
+      "epoch": 0.33313206245147936,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.8717735485745895e-05,
+      "loss": 1.5321,
+      "step": 1931
+    },
+    {
+      "epoch": 0.3333045803502113,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.8716401689704945e-05,
+      "loss": 1.538,
+      "step": 1932
+    },
+    {
+      "epoch": 0.33347709824894334,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.8715067247898493e-05,
+      "loss": 1.4214,
+      "step": 1933
+    },
+    {
+      "epoch": 0.3336496161476753,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.871373216042541e-05,
+      "loss": 1.491,
+      "step": 1934
+    },
+    {
+      "epoch": 0.3338221340464073,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.8712396427384595e-05,
+      "loss": 1.5567,
+      "step": 1935
+    },
+    {
+      "epoch": 0.33399465194513933,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.8711060048875023e-05,
+      "loss": 1.526,
+      "step": 1936
+    },
+    {
+      "epoch": 0.3341671698438713,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.870972302499569e-05,
+      "loss": 1.4795,
+      "step": 1937
+    },
+    {
+      "epoch": 0.3343396877426033,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.8708385355845654e-05,
+      "loss": 1.3779,
+      "step": 1938
+    },
+    {
+      "epoch": 0.33451220564133527,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.870704704152402e-05,
+      "loss": 1.3989,
+      "step": 1939
+    },
+    {
+      "epoch": 0.3346847235400673,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.8705708082129935e-05,
+      "loss": 1.4858,
+      "step": 1940
+    },
+    {
+      "epoch": 0.3348572414387993,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.87043684777626e-05,
+      "loss": 1.5937,
+      "step": 1941
+    },
+    {
+      "epoch": 0.33502975933753126,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.8703028228521263e-05,
+      "loss": 1.5191,
+      "step": 1942
+    },
+    {
+      "epoch": 0.3352022772362633,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.8701687334505215e-05,
+      "loss": 1.4377,
+      "step": 1943
+    },
+    {
+      "epoch": 0.33537479513499524,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.8700345795813794e-05,
+      "loss": 1.4953,
+      "step": 1944
+    },
+    {
+      "epoch": 0.33554731303372726,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.8699003612546397e-05,
+      "loss": 1.5811,
+      "step": 1945
+    },
+    {
+      "epoch": 0.3357198309324592,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.8697660784802463e-05,
+      "loss": 1.577,
+      "step": 1946
+    },
+    {
+      "epoch": 0.33589234883119123,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.8696317312681466e-05,
+      "loss": 1.3651,
+      "step": 1947
+    },
+    {
+      "epoch": 0.33606486672992325,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.869497319628295e-05,
+      "loss": 1.5111,
+      "step": 1948
+    },
+    {
+      "epoch": 0.3362373846286552,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.8693628435706487e-05,
+      "loss": 1.4774,
+      "step": 1949
+    },
+    {
+      "epoch": 0.33640990252738723,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.8692283031051714e-05,
+      "loss": 1.5871,
+      "step": 1950
+    },
+    {
+      "epoch": 0.3365824204261192,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.86909369824183e-05,
+      "loss": 1.4237,
+      "step": 1951
+    },
+    {
+      "epoch": 0.3367549383248512,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.868959028990597e-05,
+      "loss": 1.4145,
+      "step": 1952
+    },
+    {
+      "epoch": 0.3369274562235832,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.8688242953614496e-05,
+      "loss": 1.4504,
+      "step": 1953
+    },
+    {
+      "epoch": 0.3370999741223152,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.8686894973643698e-05,
+      "loss": 1.5252,
+      "step": 1954
+    },
+    {
+      "epoch": 0.3372724920210472,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.868554635009344e-05,
+      "loss": 1.4336,
+      "step": 1955
+    },
+    {
+      "epoch": 0.33744500991977916,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.868419708306364e-05,
+      "loss": 1.4343,
+      "step": 1956
+    },
+    {
+      "epoch": 0.3376175278185112,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.868284717265426e-05,
+      "loss": 1.4825,
+      "step": 1957
+    },
+    {
+      "epoch": 0.33779004571724314,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.8681496618965308e-05,
+      "loss": 1.5555,
+      "step": 1958
+    },
+    {
+      "epoch": 0.33796256361597515,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.8680145422096844e-05,
+      "loss": 1.5049,
+      "step": 1959
+    },
+    {
+      "epoch": 0.33813508151470717,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.867879358214897e-05,
+      "loss": 1.3421,
+      "step": 1960
+    },
+    {
+      "epoch": 0.33830759941343913,
+      "grad_norm": 0.625,
+      "learning_rate": 1.8677441099221836e-05,
+      "loss": 1.4644,
+      "step": 1961
+    },
+    {
+      "epoch": 0.33848011731217115,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.867608797341565e-05,
+      "loss": 1.6562,
+      "step": 1962
+    },
+    {
+      "epoch": 0.3386526352109031,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.8674734204830655e-05,
+      "loss": 1.4937,
+      "step": 1963
+    },
+    {
+      "epoch": 0.3388251531096351,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.867337979356715e-05,
+      "loss": 1.5014,
+      "step": 1964
+    },
+    {
+      "epoch": 0.33899767100836714,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.8672024739725473e-05,
+      "loss": 1.487,
+      "step": 1965
+    },
+    {
+      "epoch": 0.3391701889070991,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.867066904340602e-05,
+      "loss": 1.5148,
+      "step": 1966
+    },
+    {
+      "epoch": 0.3393427068058311,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.8669312704709224e-05,
+      "loss": 1.4718,
+      "step": 1967
+    },
+    {
+      "epoch": 0.3395152247045631,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.8667955723735577e-05,
+      "loss": 1.5579,
+      "step": 1968
+    },
+    {
+      "epoch": 0.3396877426032951,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.8666598100585613e-05,
+      "loss": 1.5187,
+      "step": 1969
+    },
+    {
+      "epoch": 0.3398602605020271,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.8665239835359904e-05,
+      "loss": 1.4936,
+      "step": 1970
+    },
+    {
+      "epoch": 0.3400327784007591,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.866388092815909e-05,
+      "loss": 1.5325,
+      "step": 1971
+    },
+    {
+      "epoch": 0.3402052962994911,
+      "grad_norm": 0.625,
+      "learning_rate": 1.8662521379083843e-05,
+      "loss": 1.439,
+      "step": 1972
+    },
+    {
+      "epoch": 0.34037781419822305,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.8661161188234882e-05,
+      "loss": 1.4927,
+      "step": 1973
+    },
+    {
+      "epoch": 0.34055033209695507,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.8659800355712984e-05,
+      "loss": 1.556,
+      "step": 1974
+    },
+    {
+      "epoch": 0.340722849995687,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.865843888161897e-05,
+      "loss": 1.4048,
+      "step": 1975
+    },
+    {
+      "epoch": 0.34089536789441904,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.86570767660537e-05,
+      "loss": 1.3183,
+      "step": 1976
+    },
+    {
+      "epoch": 0.34106788579315106,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.865571400911809e-05,
+      "loss": 1.4521,
+      "step": 1977
+    },
+    {
+      "epoch": 0.341240403691883,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.8654350610913106e-05,
+      "loss": 1.5282,
+      "step": 1978
+    },
+    {
+      "epoch": 0.34141292159061504,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.8652986571539754e-05,
+      "loss": 1.467,
+      "step": 1979
+    },
+    {
+      "epoch": 0.341585439489347,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.865162189109909e-05,
+      "loss": 1.4525,
+      "step": 1980
+    },
+    {
+      "epoch": 0.341757957388079,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.8650256569692215e-05,
+      "loss": 1.5296,
+      "step": 1981
+    },
+    {
+      "epoch": 0.34193047528681103,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.864889060742029e-05,
+      "loss": 1.5094,
+      "step": 1982
+    },
+    {
+      "epoch": 0.342102993185543,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.8647524004384503e-05,
+      "loss": 1.5443,
+      "step": 1983
+    },
+    {
+      "epoch": 0.342275511084275,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.8646156760686108e-05,
+      "loss": 1.5589,
+      "step": 1984
+    },
+    {
+      "epoch": 0.34244802898300697,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.8644788876426395e-05,
+      "loss": 1.5559,
+      "step": 1985
+    },
+    {
+      "epoch": 0.342620546881739,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.8643420351706707e-05,
+      "loss": 1.4188,
+      "step": 1986
+    },
+    {
+      "epoch": 0.34279306478047095,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.8642051186628434e-05,
+      "loss": 1.499,
+      "step": 1987
+    },
+    {
+      "epoch": 0.34296558267920296,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.864068138129301e-05,
+      "loss": 1.5922,
+      "step": 1988
+    },
+    {
+      "epoch": 0.343138100577935,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.8639310935801922e-05,
+      "loss": 1.4714,
+      "step": 1989
+    },
+    {
+      "epoch": 0.34331061847666694,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.8637939850256697e-05,
+      "loss": 1.5472,
+      "step": 1990
+    },
+    {
+      "epoch": 0.34348313637539896,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.8636568124758917e-05,
+      "loss": 1.4931,
+      "step": 1991
+    },
+    {
+      "epoch": 0.3436556542741309,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.8635195759410205e-05,
+      "loss": 1.4338,
+      "step": 1992
+    },
+    {
+      "epoch": 0.34382817217286293,
+      "grad_norm": 0.625,
+      "learning_rate": 1.8633822754312233e-05,
+      "loss": 1.4708,
+      "step": 1993
+    },
+    {
+      "epoch": 0.34400069007159495,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.863244910956673e-05,
+      "loss": 1.4769,
+      "step": 1994
+    },
+    {
+      "epoch": 0.3441732079703269,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.8631074825275457e-05,
+      "loss": 1.4068,
+      "step": 1995
+    },
+    {
+      "epoch": 0.3443457258690589,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.8629699901540232e-05,
+      "loss": 1.4729,
+      "step": 1996
+    },
+    {
+      "epoch": 0.3445182437677909,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.8628324338462915e-05,
+      "loss": 1.513,
+      "step": 1997
+    },
+    {
+      "epoch": 0.3446907616665229,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.8626948136145422e-05,
+      "loss": 1.5212,
+      "step": 1998
+    },
+    {
+      "epoch": 0.3448632795652549,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.8625571294689705e-05,
+      "loss": 1.5238,
+      "step": 1999
+    },
+    {
+      "epoch": 0.3450357974639869,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.8624193814197774e-05,
+      "loss": 1.542,
+      "step": 2000
+    },
+    {
+      "epoch": 0.3450357974639869,
+      "eval_loss": 1.4780633449554443,
+      "eval_runtime": 11.8304,
+      "eval_samples_per_second": 86.556,
+      "eval_steps_per_second": 21.639,
+      "step": 2000
+    },
+    {
+      "epoch": 0.3452083153627189,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.8622815694771675e-05,
+      "loss": 1.5014,
+      "step": 2001
+    },
+    {
+      "epoch": 0.34538083326145086,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.8621436936513517e-05,
+      "loss": 1.5214,
+      "step": 2002
+    },
+    {
+      "epoch": 0.3455533511601829,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.862005753952544e-05,
+      "loss": 1.4511,
+      "step": 2003
+    },
+    {
+      "epoch": 0.34572586905891484,
+      "grad_norm": 0.625,
+      "learning_rate": 1.8618677503909637e-05,
+      "loss": 1.5055,
+      "step": 2004
+    },
+    {
+      "epoch": 0.34589838695764685,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.8617296829768354e-05,
+      "loss": 1.4664,
+      "step": 2005
+    },
+    {
+      "epoch": 0.34607090485637887,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.8615915517203878e-05,
+      "loss": 1.5586,
+      "step": 2006
+    },
+    {
+      "epoch": 0.34624342275511083,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.8614533566318544e-05,
+      "loss": 1.5295,
+      "step": 2007
+    },
+    {
+      "epoch": 0.34641594065384285,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.861315097721474e-05,
+      "loss": 1.4864,
+      "step": 2008
+    },
+    {
+      "epoch": 0.3465884585525748,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.8611767749994894e-05,
+      "loss": 1.4829,
+      "step": 2009
+    },
+    {
+      "epoch": 0.3467609764513068,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.8610383884761486e-05,
+      "loss": 1.4033,
+      "step": 2010
+    },
+    {
+      "epoch": 0.34693349435003884,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.8608999381617035e-05,
+      "loss": 1.5237,
+      "step": 2011
+    },
+    {
+      "epoch": 0.3471060122487708,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.860761424066412e-05,
+      "loss": 1.5261,
+      "step": 2012
+    },
+    {
+      "epoch": 0.3472785301475028,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.860622846200536e-05,
+      "loss": 1.5967,
+      "step": 2013
+    },
+    {
+      "epoch": 0.3474510480462348,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.860484204574342e-05,
+      "loss": 1.4249,
+      "step": 2014
+    },
+    {
+      "epoch": 0.3476235659449668,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.860345499198102e-05,
+      "loss": 1.5006,
+      "step": 2015
+    },
+    {
+      "epoch": 0.3477960838436988,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.860206730082091e-05,
+      "loss": 1.4401,
+      "step": 2016
+    },
+    {
+      "epoch": 0.34796860174243077,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.860067897236591e-05,
+      "loss": 1.5685,
+      "step": 2017
+    },
+    {
+      "epoch": 0.3481411196411628,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.8599290006718878e-05,
+      "loss": 1.528,
+      "step": 2018
+    },
+    {
+      "epoch": 0.34831363753989475,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.8597900403982705e-05,
+      "loss": 1.3706,
+      "step": 2019
+    },
+    {
+      "epoch": 0.34848615543862677,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.8596510164260347e-05,
+      "loss": 1.4227,
+      "step": 2020
+    },
+    {
+      "epoch": 0.3486586733373587,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.8595119287654808e-05,
+      "loss": 1.4319,
+      "step": 2021
+    },
+    {
+      "epoch": 0.34883119123609074,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.8593727774269122e-05,
+      "loss": 1.5133,
+      "step": 2022
+    },
+    {
+      "epoch": 0.34900370913482276,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.859233562420639e-05,
+      "loss": 1.553,
+      "step": 2023
+    },
+    {
+      "epoch": 0.3491762270335547,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.8590942837569743e-05,
+      "loss": 1.4991,
+      "step": 2024
+    },
+    {
+      "epoch": 0.34934874493228674,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.8589549414462375e-05,
+      "loss": 1.5245,
+      "step": 2025
+    },
+    {
+      "epoch": 0.3495212628310187,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.8588155354987517e-05,
+      "loss": 1.5654,
+      "step": 2026
+    },
+    {
+      "epoch": 0.3496937807297507,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.8586760659248447e-05,
+      "loss": 1.4404,
+      "step": 2027
+    },
+    {
+      "epoch": 0.34986629862848273,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.8585365327348497e-05,
+      "loss": 1.502,
+      "step": 2028
+    },
+    {
+      "epoch": 0.3500388165272147,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.858396935939104e-05,
+      "loss": 1.4986,
+      "step": 2029
+    },
+    {
+      "epoch": 0.3502113344259467,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.8582572755479494e-05,
+      "loss": 1.49,
+      "step": 2030
+    },
+    {
+      "epoch": 0.35038385232467867,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.8581175515717337e-05,
+      "loss": 1.5376,
+      "step": 2031
+    },
+    {
+      "epoch": 0.3505563702234107,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.8579777640208076e-05,
+      "loss": 1.6789,
+      "step": 2032
+    },
+    {
+      "epoch": 0.35072888812214265,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.857837912905528e-05,
+      "loss": 1.5079,
+      "step": 2033
+    },
+    {
+      "epoch": 0.35090140602087466,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.857697998236256e-05,
+      "loss": 1.4556,
+      "step": 2034
+    },
+    {
+      "epoch": 0.3510739239196067,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.857558020023357e-05,
+      "loss": 1.4302,
+      "step": 2035
+    },
+    {
+      "epoch": 0.35124644181833864,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.8574179782772012e-05,
+      "loss": 1.4337,
+      "step": 2036
+    },
+    {
+      "epoch": 0.35141895971707066,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.8572778730081644e-05,
+      "loss": 1.543,
+      "step": 2037
+    },
+    {
+      "epoch": 0.3515914776158026,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.8571377042266267e-05,
+      "loss": 1.5619,
+      "step": 2038
+    },
+    {
+      "epoch": 0.35176399551453463,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.8569974719429716e-05,
+      "loss": 1.4384,
+      "step": 2039
+    },
+    {
+      "epoch": 0.35193651341326665,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.8568571761675893e-05,
+      "loss": 1.5401,
+      "step": 2040
+    },
+    {
+      "epoch": 0.3521090313119986,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.8567168169108735e-05,
+      "loss": 1.4847,
+      "step": 2041
+    },
+    {
+      "epoch": 0.3522815492107306,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.856576394183223e-05,
+      "loss": 1.4928,
+      "step": 2042
+    },
+    {
+      "epoch": 0.3524540671094626,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.856435907995041e-05,
+      "loss": 1.5223,
+      "step": 2043
+    },
+    {
+      "epoch": 0.3526265850081946,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.8562953583567357e-05,
+      "loss": 1.4796,
+      "step": 2044
+    },
+    {
+      "epoch": 0.3527991029069266,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.85615474527872e-05,
+      "loss": 1.4604,
+      "step": 2045
+    },
+    {
+      "epoch": 0.3529716208056586,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.856014068771411e-05,
+      "loss": 1.5434,
+      "step": 2046
+    },
+    {
+      "epoch": 0.3531441387043906,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.855873328845231e-05,
+      "loss": 1.6396,
+      "step": 2047
+    },
+    {
+      "epoch": 0.35331665660312256,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.8557325255106074e-05,
+      "loss": 1.498,
+      "step": 2048
+    },
+    {
+      "epoch": 0.3534891745018546,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.8555916587779713e-05,
+      "loss": 1.4599,
+      "step": 2049
+    },
+    {
+      "epoch": 0.35366169240058654,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.855450728657759e-05,
+      "loss": 1.5217,
+      "step": 2050
+    },
+    {
+      "epoch": 0.35383421029931855,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.855309735160412e-05,
+      "loss": 1.4141,
+      "step": 2051
+    },
+    {
+      "epoch": 0.35400672819805057,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.8551686782963757e-05,
+      "loss": 1.4475,
+      "step": 2052
+    },
+    {
+      "epoch": 0.35417924609678253,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.8550275580761e-05,
+      "loss": 1.5216,
+      "step": 2053
+    },
+    {
+      "epoch": 0.35435176399551455,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.8548863745100403e-05,
+      "loss": 1.5242,
+      "step": 2054
+    },
+    {
+      "epoch": 0.3545242818942465,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.8547451276086565e-05,
+      "loss": 1.5738,
+      "step": 2055
+    },
+    {
+      "epoch": 0.3546967997929785,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.854603817382413e-05,
+      "loss": 1.5531,
+      "step": 2056
+    },
+    {
+      "epoch": 0.35486931769171054,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.854462443841779e-05,
+      "loss": 1.4921,
+      "step": 2057
+    },
+    {
+      "epoch": 0.3550418355904425,
+      "grad_norm": 0.875,
+      "learning_rate": 1.854321006997228e-05,
+      "loss": 1.4861,
+      "step": 2058
+    },
+    {
+      "epoch": 0.3552143534891745,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.8541795068592388e-05,
+      "loss": 1.5131,
+      "step": 2059
+    },
+    {
+      "epoch": 0.3553868713879065,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.8540379434382946e-05,
+      "loss": 1.4874,
+      "step": 2060
+    },
+    {
+      "epoch": 0.3555593892866385,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.853896316744883e-05,
+      "loss": 1.4067,
+      "step": 2061
+    },
+    {
+      "epoch": 0.35573190718537046,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.8537546267894975e-05,
+      "loss": 1.5423,
+      "step": 2062
+    },
+    {
+      "epoch": 0.35590442508410247,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.8536128735826344e-05,
+      "loss": 1.5336,
+      "step": 2063
+    },
+    {
+      "epoch": 0.3560769429828345,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.853471057134796e-05,
+      "loss": 1.5081,
+      "step": 2064
+    },
+    {
+      "epoch": 0.35624946088156645,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.8533291774564887e-05,
+      "loss": 1.4295,
+      "step": 2065
+    },
+    {
+      "epoch": 0.35642197878029847,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.8531872345582247e-05,
+      "loss": 1.5259,
+      "step": 2066
+    },
+    {
+      "epoch": 0.3565944966790304,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.853045228450519e-05,
+      "loss": 1.4738,
+      "step": 2067
+    },
+    {
+      "epoch": 0.35676701457776244,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.8529031591438926e-05,
+      "loss": 1.484,
+      "step": 2068
+    },
+    {
+      "epoch": 0.35693953247649446,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.8527610266488714e-05,
+      "loss": 1.4609,
+      "step": 2069
+    },
+    {
+      "epoch": 0.3571120503752264,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.8526188309759847e-05,
+      "loss": 1.5005,
+      "step": 2070
+    },
+    {
+      "epoch": 0.35728456827395844,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.8524765721357676e-05,
+      "loss": 1.495,
+      "step": 2071
+    },
+    {
+      "epoch": 0.3574570861726904,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.8523342501387595e-05,
+      "loss": 1.48,
+      "step": 2072
+    },
+    {
+      "epoch": 0.3576296040714224,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.8521918649955047e-05,
+      "loss": 1.3926,
+      "step": 2073
+    },
+    {
+      "epoch": 0.35780212197015443,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.8520494167165517e-05,
+      "loss": 1.5589,
+      "step": 2074
+    },
+    {
+      "epoch": 0.3579746398688864,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.851906905312454e-05,
+      "loss": 1.51,
+      "step": 2075
+    },
+    {
+      "epoch": 0.3581471577676184,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.85176433079377e-05,
+      "loss": 1.6039,
+      "step": 2076
+    },
+    {
+      "epoch": 0.35831967566635037,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.8516216931710622e-05,
+      "loss": 1.4755,
+      "step": 2077
+    },
+    {
+      "epoch": 0.3584921935650824,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.8514789924548982e-05,
+      "loss": 1.46,
+      "step": 2078
+    },
+    {
+      "epoch": 0.35866471146381435,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.85133622865585e-05,
+      "loss": 1.4687,
+      "step": 2079
+    },
+    {
+      "epoch": 0.35883722936254636,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.851193401784495e-05,
+      "loss": 1.4535,
+      "step": 2080
+    },
+    {
+      "epoch": 0.3590097472612784,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.8510505118514138e-05,
+      "loss": 1.5059,
+      "step": 2081
+    },
+    {
+      "epoch": 0.35918226516001034,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.8509075588671934e-05,
+      "loss": 1.5184,
+      "step": 2082
+    },
+    {
+      "epoch": 0.35935478305874236,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.850764542842424e-05,
+      "loss": 1.5078,
+      "step": 2083
+    },
+    {
+      "epoch": 0.3595273009574743,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.8506214637877017e-05,
+      "loss": 1.4467,
+      "step": 2084
+    },
+    {
+      "epoch": 0.35969981885620633,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.8504783217136265e-05,
+      "loss": 1.5681,
+      "step": 2085
+    },
+    {
+      "epoch": 0.35987233675493835,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.8503351166308027e-05,
+      "loss": 1.5403,
+      "step": 2086
+    },
+    {
+      "epoch": 0.3600448546536703,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.850191848549841e-05,
+      "loss": 1.5065,
+      "step": 2087
+    },
+    {
+      "epoch": 0.3602173725524023,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.8500485174813545e-05,
+      "loss": 1.5095,
+      "step": 2088
+    },
+    {
+      "epoch": 0.3603898904511343,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.8499051234359627e-05,
+      "loss": 1.4728,
+      "step": 2089
+    },
+    {
+      "epoch": 0.3605624083498663,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.849761666424289e-05,
+      "loss": 1.4538,
+      "step": 2090
+    },
+    {
+      "epoch": 0.36073492624859826,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.8496181464569608e-05,
+      "loss": 1.5757,
+      "step": 2091
+    },
+    {
+      "epoch": 0.3609074441473303,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.8494745635446124e-05,
+      "loss": 1.4783,
+      "step": 2092
+    },
+    {
+      "epoch": 0.3610799620460623,
+      "grad_norm": 0.625,
+      "learning_rate": 1.8493309176978802e-05,
+      "loss": 1.4694,
+      "step": 2093
+    },
+    {
+      "epoch": 0.36125247994479426,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.849187208927407e-05,
+      "loss": 1.4574,
+      "step": 2094
+    },
+    {
+      "epoch": 0.3614249978435263,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.849043437243839e-05,
+      "loss": 1.5345,
+      "step": 2095
+    },
+    {
+      "epoch": 0.36159751574225824,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.8488996026578286e-05,
+      "loss": 1.4606,
+      "step": 2096
+    },
+    {
+      "epoch": 0.36177003364099025,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.848755705180031e-05,
+      "loss": 1.4965,
+      "step": 2097
+    },
+    {
+      "epoch": 0.36194255153972227,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.8486117448211078e-05,
+      "loss": 1.5199,
+      "step": 2098
+    },
+    {
+      "epoch": 0.36211506943845423,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.8484677215917243e-05,
+      "loss": 1.4694,
+      "step": 2099
+    },
+    {
+      "epoch": 0.36228758733718625,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.84832363550255e-05,
+      "loss": 1.6452,
+      "step": 2100
+    },
+    {
+      "epoch": 0.36228758733718625,
+      "eval_loss": 1.473473072052002,
+      "eval_runtime": 10.9071,
+      "eval_samples_per_second": 93.884,
+      "eval_steps_per_second": 23.471,
+      "step": 2100
+    },
+    {
+      "epoch": 0.3624601052359182,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.8481794865642607e-05,
+      "loss": 1.6222,
+      "step": 2101
+    },
+    {
+      "epoch": 0.3626326231346502,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.848035274787535e-05,
+      "loss": 1.4544,
+      "step": 2102
+    },
+    {
+      "epoch": 0.36280514103338224,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.8478910001830578e-05,
+      "loss": 1.4132,
+      "step": 2103
+    },
+    {
+      "epoch": 0.3629776589321142,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.8477466627615172e-05,
+      "loss": 1.468,
+      "step": 2104
+    },
+    {
+      "epoch": 0.3631501768308462,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.8476022625336067e-05,
+      "loss": 1.3662,
+      "step": 2105
+    },
+    {
+      "epoch": 0.3633226947295782,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.8474577995100247e-05,
+      "loss": 1.5097,
+      "step": 2106
+    },
+    {
+      "epoch": 0.3634952126283102,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.847313273701474e-05,
+      "loss": 1.3676,
+      "step": 2107
+    },
+    {
+      "epoch": 0.36366773052704215,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.847168685118661e-05,
+      "loss": 1.5271,
+      "step": 2108
+    },
+    {
+      "epoch": 0.36384024842577417,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.847024033772299e-05,
+      "loss": 1.4227,
+      "step": 2109
+    },
+    {
+      "epoch": 0.3640127663245062,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.846879319673104e-05,
+      "loss": 1.3847,
+      "step": 2110
+    },
+    {
+      "epoch": 0.36418528422323815,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.8467345428317976e-05,
+      "loss": 1.4954,
+      "step": 2111
+    },
+    {
+      "epoch": 0.36435780212197016,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.8465897032591057e-05,
+      "loss": 1.5108,
+      "step": 2112
+    },
+    {
+      "epoch": 0.3645303200207021,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.8464448009657582e-05,
+      "loss": 1.4386,
+      "step": 2113
+    },
+    {
+      "epoch": 0.36470283791943414,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.8462998359624914e-05,
+      "loss": 1.4996,
+      "step": 2114
+    },
+    {
+      "epoch": 0.36487535581816616,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.846154808260045e-05,
+      "loss": 1.5164,
+      "step": 2115
+    },
+    {
+      "epoch": 0.3650478737168981,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.8460097178691634e-05,
+      "loss": 1.5406,
+      "step": 2116
+    },
+    {
+      "epoch": 0.36522039161563014,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.8458645648005957e-05,
+      "loss": 1.4248,
+      "step": 2117
+    },
+    {
+      "epoch": 0.3653929095143621,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.8457193490650957e-05,
+      "loss": 1.4584,
+      "step": 2118
+    },
+    {
+      "epoch": 0.3655654274130941,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.845574070673422e-05,
+      "loss": 1.4591,
+      "step": 2119
+    },
+    {
+      "epoch": 0.36573794531182613,
+      "grad_norm": 0.875,
+      "learning_rate": 1.8454287296363382e-05,
+      "loss": 1.3891,
+      "step": 2120
+    },
+    {
+      "epoch": 0.3659104632105581,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.8452833259646113e-05,
+      "loss": 1.4364,
+      "step": 2121
+    },
+    {
+      "epoch": 0.3660829811092901,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.8451378596690138e-05,
+      "loss": 1.3792,
+      "step": 2122
+    },
+    {
+      "epoch": 0.36625549900802207,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.8449923307603234e-05,
+      "loss": 1.6282,
+      "step": 2123
+    },
+    {
+      "epoch": 0.3664280169067541,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.844846739249321e-05,
+      "loss": 1.5017,
+      "step": 2124
+    },
+    {
+      "epoch": 0.36660053480548604,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.8447010851467936e-05,
+      "loss": 1.52,
+      "step": 2125
+    },
+    {
+      "epoch": 0.36677305270421806,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.844555368463532e-05,
+      "loss": 1.5371,
+      "step": 2126
+    },
+    {
+      "epoch": 0.3669455706029501,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.8444095892103314e-05,
+      "loss": 1.4878,
+      "step": 2127
+    },
+    {
+      "epoch": 0.36711808850168204,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.844263747397992e-05,
+      "loss": 1.4632,
+      "step": 2128
+    },
+    {
+      "epoch": 0.36729060640041405,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.8441178430373197e-05,
+      "loss": 1.4249,
+      "step": 2129
+    },
+    {
+      "epoch": 0.367463124299146,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.843971876139123e-05,
+      "loss": 1.4135,
+      "step": 2130
+    },
+    {
+      "epoch": 0.36763564219787803,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.843825846714216e-05,
+      "loss": 1.3984,
+      "step": 2131
+    },
+    {
+      "epoch": 0.36780816009661005,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.8436797547734185e-05,
+      "loss": 1.4832,
+      "step": 2132
+    },
+    {
+      "epoch": 0.367980677995342,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.8435336003275525e-05,
+      "loss": 1.476,
+      "step": 2133
+    },
+    {
+      "epoch": 0.368153195894074,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.8433873833874473e-05,
+      "loss": 1.4526,
+      "step": 2134
+    },
+    {
+      "epoch": 0.368325713792806,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.843241103963935e-05,
+      "loss": 1.4606,
+      "step": 2135
+    },
+    {
+      "epoch": 0.368498231691538,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.8430947620678522e-05,
+      "loss": 1.463,
+      "step": 2136
+    },
+    {
+      "epoch": 0.36867074959026996,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.8429483577100424e-05,
+      "loss": 1.5067,
+      "step": 2137
+    },
+    {
+      "epoch": 0.368843267489002,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.842801890901351e-05,
+      "loss": 1.4845,
+      "step": 2138
+    },
+    {
+      "epoch": 0.369015785387734,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.842655361652629e-05,
+      "loss": 1.5341,
+      "step": 2139
+    },
+    {
+      "epoch": 0.36918830328646596,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.842508769974733e-05,
+      "loss": 1.5092,
+      "step": 2140
+    },
+    {
+      "epoch": 0.369360821185198,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.8423621158785232e-05,
+      "loss": 1.4552,
+      "step": 2141
+    },
+    {
+      "epoch": 0.36953333908392993,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.8422153993748645e-05,
+      "loss": 1.5087,
+      "step": 2142
+    },
+    {
+      "epoch": 0.36970585698266195,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.8420686204746264e-05,
+      "loss": 1.6148,
+      "step": 2143
+    },
+    {
+      "epoch": 0.36987837488139397,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.8419217791886838e-05,
+      "loss": 1.3866,
+      "step": 2144
+    },
+    {
+      "epoch": 0.37005089278012593,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.8417748755279146e-05,
+      "loss": 1.5025,
+      "step": 2145
+    },
+    {
+      "epoch": 0.37022341067885794,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.8416279095032036e-05,
+      "loss": 1.4978,
+      "step": 2146
+    },
+    {
+      "epoch": 0.3703959285775899,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.8414808811254378e-05,
+      "loss": 1.4778,
+      "step": 2147
+    },
+    {
+      "epoch": 0.3705684464763219,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.841333790405511e-05,
+      "loss": 1.4975,
+      "step": 2148
+    },
+    {
+      "epoch": 0.37074096437505394,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.8411866373543196e-05,
+      "loss": 1.5895,
+      "step": 2149
+    },
+    {
+      "epoch": 0.3709134822737859,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.8410394219827665e-05,
+      "loss": 1.4017,
+      "step": 2150
+    },
+    {
+      "epoch": 0.3710860001725179,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.840892144301758e-05,
+      "loss": 1.5247,
+      "step": 2151
+    },
+    {
+      "epoch": 0.3712585180712499,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.8407448043222052e-05,
+      "loss": 1.5868,
+      "step": 2152
+    },
+    {
+      "epoch": 0.3714310359699819,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.8405974020550238e-05,
+      "loss": 1.4678,
+      "step": 2153
+    },
+    {
+      "epoch": 0.37160355386871385,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.840449937511135e-05,
+      "loss": 1.4668,
+      "step": 2154
+    },
+    {
+      "epoch": 0.37177607176744587,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.8403024107014633e-05,
+      "loss": 1.4756,
+      "step": 2155
+    },
+    {
+      "epoch": 0.3719485896661779,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.8401548216369387e-05,
+      "loss": 1.4489,
+      "step": 2156
+    },
+    {
+      "epoch": 0.37212110756490985,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.8400071703284952e-05,
+      "loss": 1.5294,
+      "step": 2157
+    },
+    {
+      "epoch": 0.37229362546364186,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.839859456787072e-05,
+      "loss": 1.4748,
+      "step": 2158
+    },
+    {
+      "epoch": 0.3724661433623738,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.8397116810236123e-05,
+      "loss": 1.5687,
+      "step": 2159
+    },
+    {
+      "epoch": 0.37263866126110584,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.839563843049065e-05,
+      "loss": 1.4819,
+      "step": 2160
+    },
+    {
+      "epoch": 0.37281117915983786,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.8394159428743818e-05,
+      "loss": 1.507,
+      "step": 2161
+    },
+    {
+      "epoch": 0.3729836970585698,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.839267980510521e-05,
+      "loss": 1.4811,
+      "step": 2162
+    },
+    {
+      "epoch": 0.37315621495730183,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.839119955968444e-05,
+      "loss": 1.4537,
+      "step": 2163
+    },
+    {
+      "epoch": 0.3733287328560338,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.8389718692591177e-05,
+      "loss": 1.5256,
+      "step": 2164
+    },
+    {
+      "epoch": 0.3735012507547658,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.8388237203935134e-05,
+      "loss": 1.3834,
+      "step": 2165
+    },
+    {
+      "epoch": 0.3736737686534978,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.838675509382606e-05,
+      "loss": 1.5364,
+      "step": 2166
+    },
+    {
+      "epoch": 0.3738462865522298,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.8385272362373775e-05,
+      "loss": 1.5554,
+      "step": 2167
+    },
+    {
+      "epoch": 0.3740188044509618,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.8383789009688117e-05,
+      "loss": 1.4733,
+      "step": 2168
+    },
+    {
+      "epoch": 0.37419132234969377,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.8382305035878983e-05,
+      "loss": 1.4997,
+      "step": 2169
+    },
+    {
+      "epoch": 0.3743638402484258,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.8380820441056317e-05,
+      "loss": 1.4882,
+      "step": 2170
+    },
+    {
+      "epoch": 0.37453635814715774,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.8379335225330108e-05,
+      "loss": 1.5061,
+      "step": 2171
+    },
+    {
+      "epoch": 0.37470887604588976,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.8377849388810386e-05,
+      "loss": 1.5237,
+      "step": 2172
+    },
+    {
+      "epoch": 0.3748813939446218,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.8376362931607237e-05,
+      "loss": 1.485,
+      "step": 2173
+    },
+    {
+      "epoch": 0.37505391184335374,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.8374875853830784e-05,
+      "loss": 1.437,
+      "step": 2174
+    },
+    {
+      "epoch": 0.37522642974208575,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.8373388155591197e-05,
+      "loss": 1.4549,
+      "step": 2175
+    },
+    {
+      "epoch": 0.3753989476408177,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.8371899836998697e-05,
+      "loss": 1.464,
+      "step": 2176
+    },
+    {
+      "epoch": 0.37557146553954973,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.8370410898163553e-05,
+      "loss": 1.508,
+      "step": 2177
+    },
+    {
+      "epoch": 0.37574398343828175,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.836892133919606e-05,
+      "loss": 1.5496,
+      "step": 2178
+    },
+    {
+      "epoch": 0.3759165013370137,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.8367431160206586e-05,
+      "loss": 1.4058,
+      "step": 2179
+    },
+    {
+      "epoch": 0.3760890192357457,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.8365940361305528e-05,
+      "loss": 1.4875,
+      "step": 2180
+    },
+    {
+      "epoch": 0.3762615371344777,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.836444894260334e-05,
+      "loss": 1.5064,
+      "step": 2181
+    },
+    {
+      "epoch": 0.3764340550332097,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.8362956904210507e-05,
+      "loss": 1.4744,
+      "step": 2182
+    },
+    {
+      "epoch": 0.37660657293194166,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.8361464246237575e-05,
+      "loss": 1.483,
+      "step": 2183
+    },
+    {
+      "epoch": 0.3767790908306737,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.835997096879512e-05,
+      "loss": 1.4102,
+      "step": 2184
+    },
+    {
+      "epoch": 0.3769516087294057,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.8358477071993787e-05,
+      "loss": 1.5903,
+      "step": 2185
+    },
+    {
+      "epoch": 0.37712412662813766,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.8356982555944245e-05,
+      "loss": 1.4794,
+      "step": 2186
+    },
+    {
+      "epoch": 0.3772966445268697,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.8355487420757218e-05,
+      "loss": 1.4062,
+      "step": 2187
+    },
+    {
+      "epoch": 0.37746916242560163,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.8353991666543477e-05,
+      "loss": 1.4301,
+      "step": 2188
+    },
+    {
+      "epoch": 0.37764168032433365,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.8352495293413833e-05,
+      "loss": 1.4539,
+      "step": 2189
+    },
+    {
+      "epoch": 0.37781419822306567,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.8350998301479147e-05,
+      "loss": 1.5107,
+      "step": 2190
+    },
+    {
+      "epoch": 0.37798671612179763,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.834950069085033e-05,
+      "loss": 1.4468,
+      "step": 2191
+    },
+    {
+      "epoch": 0.37815923402052964,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.8348002461638333e-05,
+      "loss": 1.4679,
+      "step": 2192
+    },
+    {
+      "epoch": 0.3783317519192616,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.834650361395415e-05,
+      "loss": 1.4242,
+      "step": 2193
+    },
+    {
+      "epoch": 0.3785042698179936,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.8345004147908828e-05,
+      "loss": 1.4978,
+      "step": 2194
+    },
+    {
+      "epoch": 0.37867678771672564,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.834350406361346e-05,
+      "loss": 1.4935,
+      "step": 2195
+    },
+    {
+      "epoch": 0.3788493056154576,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.834200336117918e-05,
+      "loss": 1.5364,
+      "step": 2196
+    },
+    {
+      "epoch": 0.3790218235141896,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.8340502040717162e-05,
+      "loss": 1.5463,
+      "step": 2197
+    },
+    {
+      "epoch": 0.3791943414129216,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.833900010233864e-05,
+      "loss": 1.4713,
+      "step": 2198
+    },
+    {
+      "epoch": 0.3793668593116536,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.8337497546154888e-05,
+      "loss": 1.548,
+      "step": 2199
+    },
+    {
+      "epoch": 0.37953937721038555,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.833599437227722e-05,
+      "loss": 1.4916,
+      "step": 2200
+    },
+    {
+      "epoch": 0.37953937721038555,
+      "eval_loss": 1.4695838689804077,
+      "eval_runtime": 10.8414,
+      "eval_samples_per_second": 94.453,
+      "eval_steps_per_second": 23.613,
+      "step": 2200
+    },
+    {
+      "epoch": 0.37971189510911757,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.8334490580817005e-05,
+      "loss": 1.5291,
+      "step": 2201
+    },
+    {
+      "epoch": 0.3798844130078496,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.8332986171885652e-05,
+      "loss": 1.5046,
+      "step": 2202
+    },
+    {
+      "epoch": 0.38005693090658155,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.8331481145594617e-05,
+      "loss": 1.4797,
+      "step": 2203
+    },
+    {
+      "epoch": 0.38022944880531356,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.83299755020554e-05,
+      "loss": 1.5314,
+      "step": 2204
+    },
+    {
+      "epoch": 0.3804019667040455,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.8328469241379546e-05,
+      "loss": 1.5771,
+      "step": 2205
+    },
+    {
+      "epoch": 0.38057448460277754,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.8326962363678656e-05,
+      "loss": 1.474,
+      "step": 2206
+    },
+    {
+      "epoch": 0.38074700250150956,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.8325454869064366e-05,
+      "loss": 1.4271,
+      "step": 2207
+    },
+    {
+      "epoch": 0.3809195204002415,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.8323946757648357e-05,
+      "loss": 1.5268,
+      "step": 2208
+    },
+    {
+      "epoch": 0.38109203829897353,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.8322438029542364e-05,
+      "loss": 1.5208,
+      "step": 2209
+    },
+    {
+      "epoch": 0.3812645561977055,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.8320928684858162e-05,
+      "loss": 1.4641,
+      "step": 2210
+    },
+    {
+      "epoch": 0.3814370740964375,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.8319418723707568e-05,
+      "loss": 1.4852,
+      "step": 2211
+    },
+    {
+      "epoch": 0.3816095919951695,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.8317908146202455e-05,
+      "loss": 1.4708,
+      "step": 2212
+    },
+    {
+      "epoch": 0.3817821098939015,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.831639695245473e-05,
+      "loss": 1.5509,
+      "step": 2213
+    },
+    {
+      "epoch": 0.3819546277926335,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.831488514257636e-05,
+      "loss": 1.5525,
+      "step": 2214
+    },
+    {
+      "epoch": 0.38212714569136547,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.8313372716679344e-05,
+      "loss": 1.4788,
+      "step": 2215
+    },
+    {
+      "epoch": 0.3822996635900975,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.8311859674875728e-05,
+      "loss": 1.5408,
+      "step": 2216
+    },
+    {
+      "epoch": 0.38247218148882944,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.8310346017277618e-05,
+      "loss": 1.4896,
+      "step": 2217
+    },
+    {
+      "epoch": 0.38264469938756146,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.8308831743997147e-05,
+      "loss": 1.4985,
+      "step": 2218
+    },
+    {
+      "epoch": 0.3828172172862935,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.8307316855146507e-05,
+      "loss": 1.5242,
+      "step": 2219
+    },
+    {
+      "epoch": 0.38298973518502544,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.8305801350837926e-05,
+      "loss": 1.5283,
+      "step": 2220
+    },
+    {
+      "epoch": 0.38316225308375745,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.8304285231183683e-05,
+      "loss": 1.4785,
+      "step": 2221
+    },
+    {
+      "epoch": 0.3833347709824894,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.8302768496296105e-05,
+      "loss": 1.4535,
+      "step": 2222
+    },
+    {
+      "epoch": 0.38350728888122143,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.8301251146287557e-05,
+      "loss": 1.5102,
+      "step": 2223
+    },
+    {
+      "epoch": 0.38367980677995345,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.8299733181270455e-05,
+      "loss": 1.4878,
+      "step": 2224
+    },
+    {
+      "epoch": 0.3838523246786854,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.829821460135726e-05,
+      "loss": 1.5581,
+      "step": 2225
+    },
+    {
+      "epoch": 0.3840248425774174,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.8296695406660477e-05,
+      "loss": 1.4887,
+      "step": 2226
+    },
+    {
+      "epoch": 0.3841973604761494,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.829517559729266e-05,
+      "loss": 1.3807,
+      "step": 2227
+    },
+    {
+      "epoch": 0.3843698783748814,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.8293655173366405e-05,
+      "loss": 1.355,
+      "step": 2228
+    },
+    {
+      "epoch": 0.38454239627361336,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.829213413499435e-05,
+      "loss": 1.516,
+      "step": 2229
+    },
+    {
+      "epoch": 0.3847149141723454,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.829061248228919e-05,
+      "loss": 1.4854,
+      "step": 2230
+    },
+    {
+      "epoch": 0.3848874320710774,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.8289090215363653e-05,
+      "loss": 1.4833,
+      "step": 2231
+    },
+    {
+      "epoch": 0.38505994996980936,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.8287567334330522e-05,
+      "loss": 1.6318,
+      "step": 2232
+    },
+    {
+      "epoch": 0.3852324678685414,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.8286043839302618e-05,
+      "loss": 1.5281,
+      "step": 2233
+    },
+    {
+      "epoch": 0.38540498576727333,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.8284519730392813e-05,
+      "loss": 1.5238,
+      "step": 2234
+    },
+    {
+      "epoch": 0.38557750366600535,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.828299500771402e-05,
+      "loss": 1.4661,
+      "step": 2235
+    },
+    {
+      "epoch": 0.38575002156473737,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.8281469671379208e-05,
+      "loss": 1.4864,
+      "step": 2236
+    },
+    {
+      "epoch": 0.3859225394634693,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.8279943721501376e-05,
+      "loss": 1.5738,
+      "step": 2237
+    },
+    {
+      "epoch": 0.38609505736220134,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.8278417158193575e-05,
+      "loss": 1.5128,
+      "step": 2238
+    },
+    {
+      "epoch": 0.3862675752609333,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.827688998156891e-05,
+      "loss": 1.4928,
+      "step": 2239
+    },
+    {
+      "epoch": 0.3864400931596653,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.8275362191740514e-05,
+      "loss": 1.4992,
+      "step": 2240
+    },
+    {
+      "epoch": 0.3866126110583973,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.827383378882158e-05,
+      "loss": 1.5278,
+      "step": 2241
+    },
+    {
+      "epoch": 0.3867851289571293,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.8272304772925342e-05,
+      "loss": 1.494,
+      "step": 2242
+    },
+    {
+      "epoch": 0.3869576468558613,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.8270775144165082e-05,
+      "loss": 1.3458,
+      "step": 2243
+    },
+    {
+      "epoch": 0.3871301647545933,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.8269244902654116e-05,
+      "loss": 1.4868,
+      "step": 2244
+    },
+    {
+      "epoch": 0.3873026826533253,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.8267714048505822e-05,
+      "loss": 1.5205,
+      "step": 2245
+    },
+    {
+      "epoch": 0.38747520055205725,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.8266182581833615e-05,
+      "loss": 1.4998,
+      "step": 2246
+    },
+    {
+      "epoch": 0.38764771845078927,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.8264650502750946e-05,
+      "loss": 1.5161,
+      "step": 2247
+    },
+    {
+      "epoch": 0.3878202363495213,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.8263117811371333e-05,
+      "loss": 1.4975,
+      "step": 2248
+    },
+    {
+      "epoch": 0.38799275424825325,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.8261584507808318e-05,
+      "loss": 1.4637,
+      "step": 2249
+    },
+    {
+      "epoch": 0.38816527214698526,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.8260050592175507e-05,
+      "loss": 1.4393,
+      "step": 2250
+    },
+    {
+      "epoch": 0.3883377900457172,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.8258516064586532e-05,
+      "loss": 1.4482,
+      "step": 2251
+    },
+    {
+      "epoch": 0.38851030794444924,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.825698092515509e-05,
+      "loss": 1.5549,
+      "step": 2252
+    },
+    {
+      "epoch": 0.38868282584318126,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.8255445173994907e-05,
+      "loss": 1.4593,
+      "step": 2253
+    },
+    {
+      "epoch": 0.3888553437419132,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.8253908811219764e-05,
+      "loss": 1.5047,
+      "step": 2254
+    },
+    {
+      "epoch": 0.38902786164064523,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.8252371836943483e-05,
+      "loss": 1.5405,
+      "step": 2255
+    },
+    {
+      "epoch": 0.3892003795393772,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.825083425127993e-05,
+      "loss": 1.4635,
+      "step": 2256
+    },
+    {
+      "epoch": 0.3893728974381092,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.8249296054343026e-05,
+      "loss": 1.5867,
+      "step": 2257
+    },
+    {
+      "epoch": 0.38954541533684117,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.8247757246246726e-05,
+      "loss": 1.4289,
+      "step": 2258
+    },
+    {
+      "epoch": 0.3897179332355732,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.824621782710503e-05,
+      "loss": 1.4538,
+      "step": 2259
+    },
+    {
+      "epoch": 0.3898904511343052,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.8244677797032e-05,
+      "loss": 1.4613,
+      "step": 2260
+    },
+    {
+      "epoch": 0.39006296903303717,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.824313715614172e-05,
+      "loss": 1.5135,
+      "step": 2261
+    },
+    {
+      "epoch": 0.3902354869317692,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.8241595904548336e-05,
+      "loss": 1.4293,
+      "step": 2262
+    },
+    {
+      "epoch": 0.39040800483050114,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.8240054042366026e-05,
+      "loss": 1.4713,
+      "step": 2263
+    },
+    {
+      "epoch": 0.39058052272923316,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.8238511569709033e-05,
+      "loss": 1.5722,
+      "step": 2264
+    },
+    {
+      "epoch": 0.3907530406279652,
+      "grad_norm": 0.625,
+      "learning_rate": 1.823696848669162e-05,
+      "loss": 1.5584,
+      "step": 2265
+    },
+    {
+      "epoch": 0.39092555852669714,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.823542479342812e-05,
+      "loss": 1.5078,
+      "step": 2266
+    },
+    {
+      "epoch": 0.39109807642542915,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.823388049003289e-05,
+      "loss": 1.5375,
+      "step": 2267
+    },
+    {
+      "epoch": 0.3912705943241611,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.823233557662035e-05,
+      "loss": 1.5285,
+      "step": 2268
+    },
+    {
+      "epoch": 0.39144311222289313,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.823079005330495e-05,
+      "loss": 1.4881,
+      "step": 2269
+    },
+    {
+      "epoch": 0.3916156301216251,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.8229243920201194e-05,
+      "loss": 1.5836,
+      "step": 2270
+    },
+    {
+      "epoch": 0.3917881480203571,
+      "grad_norm": 0.625,
+      "learning_rate": 1.822769717742363e-05,
+      "loss": 1.498,
+      "step": 2271
+    },
+    {
+      "epoch": 0.3919606659190891,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.822614982508685e-05,
+      "loss": 1.5081,
+      "step": 2272
+    },
+    {
+      "epoch": 0.3921331838178211,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.8224601863305495e-05,
+      "loss": 1.4186,
+      "step": 2273
+    },
+    {
+      "epoch": 0.3923057017165531,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.822305329219424e-05,
+      "loss": 1.4982,
+      "step": 2274
+    },
+    {
+      "epoch": 0.39247821961528506,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.8221504111867817e-05,
+      "loss": 1.4796,
+      "step": 2275
+    },
+    {
+      "epoch": 0.3926507375140171,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.8219954322441e-05,
+      "loss": 1.5749,
+      "step": 2276
+    },
+    {
+      "epoch": 0.3928232554127491,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.8218403924028608e-05,
+      "loss": 1.4309,
+      "step": 2277
+    },
+    {
+      "epoch": 0.39299577331148106,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.82168529167455e-05,
+      "loss": 1.5298,
+      "step": 2278
+    },
+    {
+      "epoch": 0.3931682912102131,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.8215301300706584e-05,
+      "loss": 1.4536,
+      "step": 2279
+    },
+    {
+      "epoch": 0.39334080910894503,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.821374907602682e-05,
+      "loss": 1.4692,
+      "step": 2280
+    },
+    {
+      "epoch": 0.39351332700767705,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.8212196242821206e-05,
+      "loss": 1.4213,
+      "step": 2281
+    },
+    {
+      "epoch": 0.39368584490640907,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.8210642801204775e-05,
+      "loss": 1.4134,
+      "step": 2282
+    },
+    {
+      "epoch": 0.393858362805141,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.820908875129263e-05,
+      "loss": 1.4815,
+      "step": 2283
+    },
+    {
+      "epoch": 0.39403088070387304,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.820753409319989e-05,
+      "loss": 1.4424,
+      "step": 2284
+    },
+    {
+      "epoch": 0.394203398602605,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.8205978827041745e-05,
+      "loss": 1.4839,
+      "step": 2285
+    },
+    {
+      "epoch": 0.394375916501337,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.8204422952933416e-05,
+      "loss": 1.4836,
+      "step": 2286
+    },
+    {
+      "epoch": 0.394548434400069,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.8202866470990172e-05,
+      "loss": 1.5224,
+      "step": 2287
+    },
+    {
+      "epoch": 0.394720952298801,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.8201309381327324e-05,
+      "loss": 1.3705,
+      "step": 2288
+    },
+    {
+      "epoch": 0.394893470197533,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.819975168406023e-05,
+      "loss": 1.4925,
+      "step": 2289
+    },
+    {
+      "epoch": 0.395065988096265,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.81981933793043e-05,
+      "loss": 1.5281,
+      "step": 2290
+    },
+    {
+      "epoch": 0.395238505994997,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.8196634467174982e-05,
+      "loss": 1.5847,
+      "step": 2291
+    },
+    {
+      "epoch": 0.39541102389372895,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.8195074947787764e-05,
+      "loss": 1.5413,
+      "step": 2292
+    },
+    {
+      "epoch": 0.39558354179246097,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.8193514821258188e-05,
+      "loss": 1.306,
+      "step": 2293
+    },
+    {
+      "epoch": 0.395756059691193,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.819195408770184e-05,
+      "loss": 1.4656,
+      "step": 2294
+    },
+    {
+      "epoch": 0.39592857758992495,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.819039274723435e-05,
+      "loss": 1.4589,
+      "step": 2295
+    },
+    {
+      "epoch": 0.39610109548865696,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.818883079997139e-05,
+      "loss": 1.5112,
+      "step": 2296
+    },
+    {
+      "epoch": 0.3962736133873889,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.8187268246028672e-05,
+      "loss": 1.4949,
+      "step": 2297
+    },
+    {
+      "epoch": 0.39644613128612094,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.818570508552197e-05,
+      "loss": 1.4868,
+      "step": 2298
+    },
+    {
+      "epoch": 0.39661864918485296,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.818414131856709e-05,
+      "loss": 1.521,
+      "step": 2299
+    },
+    {
+      "epoch": 0.3967911670835849,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.818257694527988e-05,
+      "loss": 1.3609,
+      "step": 2300
+    },
+    {
+      "epoch": 0.3967911670835849,
+      "eval_loss": 1.4656120538711548,
+      "eval_runtime": 10.8637,
+      "eval_samples_per_second": 94.259,
+      "eval_steps_per_second": 23.565,
+      "step": 2300
+    },
+    {
+      "epoch": 0.39696368498231693,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.8181011965776244e-05,
+      "loss": 1.4911,
+      "step": 2301
+    },
+    {
+      "epoch": 0.3971362028810489,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.8179446380172127e-05,
+      "loss": 1.6361,
+      "step": 2302
+    },
+    {
+      "epoch": 0.3973087207797809,
+      "grad_norm": 0.625,
+      "learning_rate": 1.8177880188583513e-05,
+      "loss": 1.4886,
+      "step": 2303
+    },
+    {
+      "epoch": 0.39748123867851287,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.8176313391126438e-05,
+      "loss": 1.3953,
+      "step": 2304
+    },
+    {
+      "epoch": 0.3976537565772449,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.817474598791698e-05,
+      "loss": 1.5849,
+      "step": 2305
+    },
+    {
+      "epoch": 0.3978262744759769,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.8173177979071256e-05,
+      "loss": 1.434,
+      "step": 2306
+    },
+    {
+      "epoch": 0.39799879237470887,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.8171609364705443e-05,
+      "loss": 1.6028,
+      "step": 2307
+    },
+    {
+      "epoch": 0.3981713102734409,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.8170040144935747e-05,
+      "loss": 1.4931,
+      "step": 2308
+    },
+    {
+      "epoch": 0.39834382817217284,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.816847031987843e-05,
+      "loss": 1.4978,
+      "step": 2309
+    },
+    {
+      "epoch": 0.39851634607090486,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.8166899889649795e-05,
+      "loss": 1.4645,
+      "step": 2310
+    },
+    {
+      "epoch": 0.3986888639696369,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.8165328854366183e-05,
+      "loss": 1.5468,
+      "step": 2311
+    },
+    {
+      "epoch": 0.39886138186836884,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.8163757214143993e-05,
+      "loss": 1.5296,
+      "step": 2312
+    },
+    {
+      "epoch": 0.39903389976710085,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.8162184969099658e-05,
+      "loss": 1.4244,
+      "step": 2313
+    },
+    {
+      "epoch": 0.3992064176658328,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.816061211934966e-05,
+      "loss": 1.5313,
+      "step": 2314
+    },
+    {
+      "epoch": 0.39937893556456483,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.8159038665010528e-05,
+      "loss": 1.4613,
+      "step": 2315
+    },
+    {
+      "epoch": 0.3995514534632968,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.8157464606198832e-05,
+      "loss": 1.4243,
+      "step": 2316
+    },
+    {
+      "epoch": 0.3997239713620288,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.8155889943031186e-05,
+      "loss": 1.5109,
+      "step": 2317
+    },
+    {
+      "epoch": 0.3998964892607608,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.815431467562425e-05,
+      "loss": 1.5602,
+      "step": 2318
+    },
+    {
+      "epoch": 0.4000690071594928,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.8152738804094737e-05,
+      "loss": 1.499,
+      "step": 2319
+    },
+    {
+      "epoch": 0.4002415250582248,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.815116232855939e-05,
+      "loss": 1.4352,
+      "step": 2320
+    },
+    {
+      "epoch": 0.40041404295695676,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.8149585249135008e-05,
+      "loss": 1.552,
+      "step": 2321
+    },
+    {
+      "epoch": 0.4005865608556888,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.814800756593843e-05,
+      "loss": 1.5098,
+      "step": 2322
+    },
+    {
+      "epoch": 0.4007590787544208,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.814642927908654e-05,
+      "loss": 1.5874,
+      "step": 2323
+    },
+    {
+      "epoch": 0.40093159665315276,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.814485038869627e-05,
+      "loss": 1.5504,
+      "step": 2324
+    },
+    {
+      "epoch": 0.40110411455188477,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.814327089488459e-05,
+      "loss": 1.5221,
+      "step": 2325
+    },
+    {
+      "epoch": 0.40127663245061673,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.814169079776852e-05,
+      "loss": 1.512,
+      "step": 2326
+    },
+    {
+      "epoch": 0.40144915034934875,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.8140110097465123e-05,
+      "loss": 1.5012,
+      "step": 2327
+    },
+    {
+      "epoch": 0.40162166824808077,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.8138528794091514e-05,
+      "loss": 1.5045,
+      "step": 2328
+    },
+    {
+      "epoch": 0.4017941861468127,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.813694688776483e-05,
+      "loss": 1.4746,
+      "step": 2329
+    },
+    {
+      "epoch": 0.40196670404554474,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.8135364378602288e-05,
+      "loss": 1.5166,
+      "step": 2330
+    },
+    {
+      "epoch": 0.4021392219442767,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.8133781266721114e-05,
+      "loss": 1.4975,
+      "step": 2331
+    },
+    {
+      "epoch": 0.4023117398430087,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.8132197552238608e-05,
+      "loss": 1.5039,
+      "step": 2332
+    },
+    {
+      "epoch": 0.4024842577417407,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.813061323527209e-05,
+      "loss": 1.5281,
+      "step": 2333
+    },
+    {
+      "epoch": 0.4026567756404727,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.8129028315938944e-05,
+      "loss": 1.5463,
+      "step": 2334
+    },
+    {
+      "epoch": 0.4028292935392047,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.8127442794356585e-05,
+      "loss": 1.4572,
+      "step": 2335
+    },
+    {
+      "epoch": 0.4030018114379367,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.8125856670642485e-05,
+      "loss": 1.5668,
+      "step": 2336
+    },
+    {
+      "epoch": 0.4031743293366687,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.8124269944914147e-05,
+      "loss": 1.4871,
+      "step": 2337
+    },
+    {
+      "epoch": 0.40334684723540065,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.812268261728913e-05,
+      "loss": 1.5133,
+      "step": 2338
+    },
+    {
+      "epoch": 0.40351936513413267,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.812109468788503e-05,
+      "loss": 1.3911,
+      "step": 2339
+    },
+    {
+      "epoch": 0.4036918830328647,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.8119506156819495e-05,
+      "loss": 1.5736,
+      "step": 2340
+    },
+    {
+      "epoch": 0.40386440093159665,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.811791702421021e-05,
+      "loss": 1.5105,
+      "step": 2341
+    },
+    {
+      "epoch": 0.40403691883032866,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.811632729017491e-05,
+      "loss": 1.5973,
+      "step": 2342
+    },
+    {
+      "epoch": 0.4042094367290606,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.8114736954831367e-05,
+      "loss": 1.4065,
+      "step": 2343
+    },
+    {
+      "epoch": 0.40438195462779264,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.8113146018297413e-05,
+      "loss": 1.4329,
+      "step": 2344
+    },
+    {
+      "epoch": 0.4045544725265246,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.81115544806909e-05,
+      "loss": 1.55,
+      "step": 2345
+    },
+    {
+      "epoch": 0.4047269904252566,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.8109962342129757e-05,
+      "loss": 1.4663,
+      "step": 2346
+    },
+    {
+      "epoch": 0.40489950832398863,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.8108369602731928e-05,
+      "loss": 1.5616,
+      "step": 2347
+    },
+    {
+      "epoch": 0.4050720262227206,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.810677626261541e-05,
+      "loss": 1.4522,
+      "step": 2348
+    },
+    {
+      "epoch": 0.4052445441214526,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.810518232189826e-05,
+      "loss": 1.5224,
+      "step": 2349
+    },
+    {
+      "epoch": 0.40541706202018457,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.8103587780698556e-05,
+      "loss": 1.428,
+      "step": 2350
+    },
+    {
+      "epoch": 0.4055895799189166,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.8101992639134438e-05,
+      "loss": 1.5132,
+      "step": 2351
+    },
+    {
+      "epoch": 0.4057620978176486,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.810039689732408e-05,
+      "loss": 1.4774,
+      "step": 2352
+    },
+    {
+      "epoch": 0.40593461571638056,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.8098800555385707e-05,
+      "loss": 1.5203,
+      "step": 2353
+    },
+    {
+      "epoch": 0.4061071336151126,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.8097203613437586e-05,
+      "loss": 1.4957,
+      "step": 2354
+    },
+    {
+      "epoch": 0.40627965151384454,
+      "grad_norm": 0.625,
+      "learning_rate": 1.8095606071598028e-05,
+      "loss": 1.4267,
+      "step": 2355
+    },
+    {
+      "epoch": 0.40645216941257656,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.8094007929985387e-05,
+      "loss": 1.5352,
+      "step": 2356
+    },
+    {
+      "epoch": 0.4066246873113086,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.8092409188718064e-05,
+      "loss": 1.5219,
+      "step": 2357
+    },
+    {
+      "epoch": 0.40679720521004054,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.8090809847914506e-05,
+      "loss": 1.4196,
+      "step": 2358
+    },
+    {
+      "epoch": 0.40696972310877255,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.80892099076932e-05,
+      "loss": 1.5713,
+      "step": 2359
+    },
+    {
+      "epoch": 0.4071422410075045,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.808760936817268e-05,
+      "loss": 1.4479,
+      "step": 2360
+    },
+    {
+      "epoch": 0.40731475890623653,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.8086008229471527e-05,
+      "loss": 1.5341,
+      "step": 2361
+    },
+    {
+      "epoch": 0.4074872768049685,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.8084406491708358e-05,
+      "loss": 1.5566,
+      "step": 2362
+    },
+    {
+      "epoch": 0.4076597947037005,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.8082804155001842e-05,
+      "loss": 1.4349,
+      "step": 2363
+    },
+    {
+      "epoch": 0.4078323126024325,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.8081201219470694e-05,
+      "loss": 1.4505,
+      "step": 2364
+    },
+    {
+      "epoch": 0.4080048305011645,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.807959768523366e-05,
+      "loss": 1.4418,
+      "step": 2365
+    },
+    {
+      "epoch": 0.4081773483998965,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.807799355240955e-05,
+      "loss": 1.5607,
+      "step": 2366
+    },
+    {
+      "epoch": 0.40834986629862846,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.8076388821117204e-05,
+      "loss": 1.5266,
+      "step": 2367
+    },
+    {
+      "epoch": 0.4085223841973605,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.807478349147551e-05,
+      "loss": 1.5094,
+      "step": 2368
+    },
+    {
+      "epoch": 0.4086949020960925,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.8073177563603403e-05,
+      "loss": 1.4372,
+      "step": 2369
+    },
+    {
+      "epoch": 0.40886741999482445,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.8071571037619856e-05,
+      "loss": 1.5019,
+      "step": 2370
+    },
+    {
+      "epoch": 0.40903993789355647,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.8069963913643893e-05,
+      "loss": 1.5055,
+      "step": 2371
+    },
+    {
+      "epoch": 0.40921245579228843,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.806835619179458e-05,
+      "loss": 1.4157,
+      "step": 2372
+    },
+    {
+      "epoch": 0.40938497369102045,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.8066747872191028e-05,
+      "loss": 1.4324,
+      "step": 2373
+    },
+    {
+      "epoch": 0.4095574915897524,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.806513895495239e-05,
+      "loss": 1.4795,
+      "step": 2374
+    },
+    {
+      "epoch": 0.4097300094884844,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.8063529440197866e-05,
+      "loss": 1.504,
+      "step": 2375
+    },
+    {
+      "epoch": 0.40990252738721644,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.8061919328046695e-05,
+      "loss": 1.4982,
+      "step": 2376
+    },
+    {
+      "epoch": 0.4100750452859484,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.806030861861817e-05,
+      "loss": 1.4587,
+      "step": 2377
+    },
+    {
+      "epoch": 0.4102475631846804,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.8058697312031615e-05,
+      "loss": 1.4981,
+      "step": 2378
+    },
+    {
+      "epoch": 0.4104200810834124,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.8057085408406415e-05,
+      "loss": 1.4198,
+      "step": 2379
+    },
+    {
+      "epoch": 0.4105925989821444,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.805547290786198e-05,
+      "loss": 1.4076,
+      "step": 2380
+    },
+    {
+      "epoch": 0.4107651168808764,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.8053859810517785e-05,
+      "loss": 1.4019,
+      "step": 2381
+    },
+    {
+      "epoch": 0.4109376347796084,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.805224611649333e-05,
+      "loss": 1.4232,
+      "step": 2382
+    },
+    {
+      "epoch": 0.4111101526783404,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.805063182590817e-05,
+      "loss": 1.4582,
+      "step": 2383
+    },
+    {
+      "epoch": 0.41128267057707235,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.8049016938881897e-05,
+      "loss": 1.4875,
+      "step": 2384
+    },
+    {
+      "epoch": 0.41145518847580437,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.8047401455534162e-05,
+      "loss": 1.4733,
+      "step": 2385
+    },
+    {
+      "epoch": 0.4116277063745364,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.8045785375984642e-05,
+      "loss": 1.5484,
+      "step": 2386
+    },
+    {
+      "epoch": 0.41180022427326834,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.8044168700353073e-05,
+      "loss": 1.4825,
+      "step": 2387
+    },
+    {
+      "epoch": 0.41197274217200036,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.804255142875922e-05,
+      "loss": 1.5638,
+      "step": 2388
+    },
+    {
+      "epoch": 0.4121452600707323,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.8040933561322905e-05,
+      "loss": 1.5131,
+      "step": 2389
+    },
+    {
+      "epoch": 0.41231777796946434,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.8039315098163993e-05,
+      "loss": 1.421,
+      "step": 2390
+    },
+    {
+      "epoch": 0.4124902958681963,
+      "grad_norm": 1.0,
+      "learning_rate": 1.8037696039402385e-05,
+      "loss": 1.4433,
+      "step": 2391
+    },
+    {
+      "epoch": 0.4126628137669283,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.8036076385158034e-05,
+      "loss": 1.4687,
+      "step": 2392
+    },
+    {
+      "epoch": 0.41283533166566033,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.803445613555093e-05,
+      "loss": 1.5166,
+      "step": 2393
+    },
+    {
+      "epoch": 0.4130078495643923,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.8032835290701115e-05,
+      "loss": 1.4204,
+      "step": 2394
+    },
+    {
+      "epoch": 0.4131803674631243,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.803121385072867e-05,
+      "loss": 1.4271,
+      "step": 2395
+    },
+    {
+      "epoch": 0.41335288536185627,
+      "grad_norm": 1.359375,
+      "learning_rate": 1.802959181575372e-05,
+      "loss": 1.4582,
+      "step": 2396
+    },
+    {
+      "epoch": 0.4135254032605883,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.802796918589644e-05,
+      "loss": 1.5298,
+      "step": 2397
+    },
+    {
+      "epoch": 0.4136979211593203,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.802634596127704e-05,
+      "loss": 1.459,
+      "step": 2398
+    },
+    {
+      "epoch": 0.41387043905805226,
+      "grad_norm": 0.625,
+      "learning_rate": 1.8024722142015784e-05,
+      "loss": 1.4324,
+      "step": 2399
+    },
+    {
+      "epoch": 0.4140429569567843,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.8023097728232967e-05,
+      "loss": 1.437,
+      "step": 2400
+    },
+    {
+      "epoch": 0.4140429569567843,
+      "eval_loss": 1.4613255262374878,
+      "eval_runtime": 10.8914,
+      "eval_samples_per_second": 94.019,
+      "eval_steps_per_second": 23.505,
+      "step": 2400
+    },
+    {
+      "epoch": 0.41421547485551624,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.802147272004894e-05,
+      "loss": 1.4324,
+      "step": 2401
+    },
+    {
+      "epoch": 0.41438799275424826,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.8019847117584092e-05,
+      "loss": 1.441,
+      "step": 2402
+    },
+    {
+      "epoch": 0.4145605106529803,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.8018220920958864e-05,
+      "loss": 1.4981,
+      "step": 2403
+    },
+    {
+      "epoch": 0.41473302855171224,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.8016594130293725e-05,
+      "loss": 1.5251,
+      "step": 2404
+    },
+    {
+      "epoch": 0.41490554645044425,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.8014966745709202e-05,
+      "loss": 1.5364,
+      "step": 2405
+    },
+    {
+      "epoch": 0.4150780643491762,
+      "grad_norm": 0.75,
+      "learning_rate": 1.8013338767325866e-05,
+      "loss": 1.4724,
+      "step": 2406
+    },
+    {
+      "epoch": 0.41525058224790823,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.8011710195264323e-05,
+      "loss": 1.5819,
+      "step": 2407
+    },
+    {
+      "epoch": 0.4154231001466402,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.8010081029645232e-05,
+      "loss": 1.4644,
+      "step": 2408
+    },
+    {
+      "epoch": 0.4155956180453722,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.8008451270589288e-05,
+      "loss": 1.519,
+      "step": 2409
+    },
+    {
+      "epoch": 0.4157681359441042,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.8006820918217233e-05,
+      "loss": 1.4889,
+      "step": 2410
+    },
+    {
+      "epoch": 0.4159406538428362,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.8005189972649856e-05,
+      "loss": 1.4611,
+      "step": 2411
+    },
+    {
+      "epoch": 0.4161131717415682,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.800355843400799e-05,
+      "loss": 1.5453,
+      "step": 2412
+    },
+    {
+      "epoch": 0.41628568964030016,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.8001926302412503e-05,
+      "loss": 1.4353,
+      "step": 2413
+    },
+    {
+      "epoch": 0.4164582075390322,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.8000293577984318e-05,
+      "loss": 1.4559,
+      "step": 2414
+    },
+    {
+      "epoch": 0.4166307254377642,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.7998660260844397e-05,
+      "loss": 1.395,
+      "step": 2415
+    },
+    {
+      "epoch": 0.41680324333649615,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.7997026351113746e-05,
+      "loss": 1.3569,
+      "step": 2416
+    },
+    {
+      "epoch": 0.41697576123522817,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.7995391848913414e-05,
+      "loss": 1.5016,
+      "step": 2417
+    },
+    {
+      "epoch": 0.41714827913396013,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.7993756754364497e-05,
+      "loss": 1.5362,
+      "step": 2418
+    },
+    {
+      "epoch": 0.41732079703269215,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.7992121067588134e-05,
+      "loss": 1.4501,
+      "step": 2419
+    },
+    {
+      "epoch": 0.4174933149314241,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.7990484788705506e-05,
+      "loss": 1.4937,
+      "step": 2420
+    },
+    {
+      "epoch": 0.4176658328301561,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.7988847917837832e-05,
+      "loss": 1.4663,
+      "step": 2421
+    },
+    {
+      "epoch": 0.41783835072888814,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.7987210455106395e-05,
+      "loss": 1.5524,
+      "step": 2422
+    },
+    {
+      "epoch": 0.4180108686276201,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.7985572400632496e-05,
+      "loss": 1.5553,
+      "step": 2423
+    },
+    {
+      "epoch": 0.4181833865263521,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.7983933754537498e-05,
+      "loss": 1.4366,
+      "step": 2424
+    },
+    {
+      "epoch": 0.4183559044250841,
+      "grad_norm": 24.875,
+      "learning_rate": 1.7982294516942804e-05,
+      "loss": 1.5607,
+      "step": 2425
+    },
+    {
+      "epoch": 0.4185284223238161,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.7980654687969853e-05,
+      "loss": 1.4021,
+      "step": 2426
+    },
+    {
+      "epoch": 0.4187009402225481,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.797901426774014e-05,
+      "loss": 1.4621,
+      "step": 2427
+    },
+    {
+      "epoch": 0.4188734581212801,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.7977373256375194e-05,
+      "loss": 1.4756,
+      "step": 2428
+    },
+    {
+      "epoch": 0.4190459760200121,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.7975731653996594e-05,
+      "loss": 1.4211,
+      "step": 2429
+    },
+    {
+      "epoch": 0.41921849391874405,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.7974089460725958e-05,
+      "loss": 1.4056,
+      "step": 2430
+    },
+    {
+      "epoch": 0.41939101181747607,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.797244667668495e-05,
+      "loss": 1.4528,
+      "step": 2431
+    },
+    {
+      "epoch": 0.4195635297162081,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.7970803301995278e-05,
+      "loss": 1.4564,
+      "step": 2432
+    },
+    {
+      "epoch": 0.41973604761494004,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.796915933677869e-05,
+      "loss": 1.3605,
+      "step": 2433
+    },
+    {
+      "epoch": 0.41990856551367206,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.7967514781156988e-05,
+      "loss": 1.4233,
+      "step": 2434
+    },
+    {
+      "epoch": 0.420081083412404,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.7965869635252005e-05,
+      "loss": 1.3589,
+      "step": 2435
+    },
+    {
+      "epoch": 0.42025360131113604,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.7964223899185633e-05,
+      "loss": 1.564,
+      "step": 2436
+    },
+    {
+      "epoch": 0.420426119209868,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.7962577573079785e-05,
+      "loss": 1.4458,
+      "step": 2437
+    },
+    {
+      "epoch": 0.4205986371086,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.796093065705644e-05,
+      "loss": 1.4488,
+      "step": 2438
+    },
+    {
+      "epoch": 0.42077115500733203,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.795928315123761e-05,
+      "loss": 1.6364,
+      "step": 2439
+    },
+    {
+      "epoch": 0.420943672906064,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.7957635055745354e-05,
+      "loss": 1.5654,
+      "step": 2440
+    },
+    {
+      "epoch": 0.421116190804796,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.7955986370701768e-05,
+      "loss": 1.5302,
+      "step": 2441
+    },
+    {
+      "epoch": 0.42128870870352797,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.7954337096229e-05,
+      "loss": 1.4639,
+      "step": 2442
+    },
+    {
+      "epoch": 0.42146122660226,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.7952687232449242e-05,
+      "loss": 1.422,
+      "step": 2443
+    },
+    {
+      "epoch": 0.421633744500992,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.7951036779484723e-05,
+      "loss": 1.4756,
+      "step": 2444
+    },
+    {
+      "epoch": 0.42180626239972396,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.7949385737457716e-05,
+      "loss": 1.4052,
+      "step": 2445
+    },
+    {
+      "epoch": 0.421978780298456,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.7947734106490545e-05,
+      "loss": 1.4661,
+      "step": 2446
+    },
+    {
+      "epoch": 0.42215129819718794,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.7946081886705574e-05,
+      "loss": 1.5564,
+      "step": 2447
+    },
+    {
+      "epoch": 0.42232381609591996,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.7944429078225204e-05,
+      "loss": 1.4868,
+      "step": 2448
+    },
+    {
+      "epoch": 0.4224963339946519,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.794277568117189e-05,
+      "loss": 1.4422,
+      "step": 2449
+    },
+    {
+      "epoch": 0.42266885189338393,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.794112169566813e-05,
+      "loss": 1.4882,
+      "step": 2450
+    },
+    {
+      "epoch": 0.42284136979211595,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.793946712183645e-05,
+      "loss": 1.6392,
+      "step": 2451
+    },
+    {
+      "epoch": 0.4230138876908479,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.793781195979944e-05,
+      "loss": 1.517,
+      "step": 2452
+    },
+    {
+      "epoch": 0.42318640558957993,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.793615620967972e-05,
+      "loss": 1.4967,
+      "step": 2453
+    },
+    {
+      "epoch": 0.4233589234883119,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.7934499871599962e-05,
+      "loss": 1.5063,
+      "step": 2454
+    },
+    {
+      "epoch": 0.4235314413870439,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.7932842945682877e-05,
+      "loss": 1.4665,
+      "step": 2455
+    },
+    {
+      "epoch": 0.4237039592857759,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.7931185432051216e-05,
+      "loss": 1.4857,
+      "step": 2456
+    },
+    {
+      "epoch": 0.4238764771845079,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.7929527330827788e-05,
+      "loss": 1.4581,
+      "step": 2457
+    },
+    {
+      "epoch": 0.4240489950832399,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.7927868642135427e-05,
+      "loss": 1.5367,
+      "step": 2458
+    },
+    {
+      "epoch": 0.42422151298197186,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.792620936609702e-05,
+      "loss": 1.5317,
+      "step": 2459
+    },
+    {
+      "epoch": 0.4243940308807039,
+      "grad_norm": 0.625,
+      "learning_rate": 1.79245495028355e-05,
+      "loss": 1.4564,
+      "step": 2460
+    },
+    {
+      "epoch": 0.4245665487794359,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.7922889052473834e-05,
+      "loss": 1.4537,
+      "step": 2461
+    },
+    {
+      "epoch": 0.42473906667816785,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.792122801513505e-05,
+      "loss": 1.4462,
+      "step": 2462
+    },
+    {
+      "epoch": 0.42491158457689987,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.7919566390942193e-05,
+      "loss": 1.4502,
+      "step": 2463
+    },
+    {
+      "epoch": 0.42508410247563183,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.791790418001838e-05,
+      "loss": 1.5052,
+      "step": 2464
+    },
+    {
+      "epoch": 0.42525662037436385,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.7916241382486745e-05,
+      "loss": 1.4147,
+      "step": 2465
+    },
+    {
+      "epoch": 0.4254291382730958,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.791457799847049e-05,
+      "loss": 1.4445,
+      "step": 2466
+    },
+    {
+      "epoch": 0.4256016561718278,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.7912914028092845e-05,
+      "loss": 1.5408,
+      "step": 2467
+    },
+    {
+      "epoch": 0.42577417407055984,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.7911249471477085e-05,
+      "loss": 1.4918,
+      "step": 2468
+    },
+    {
+      "epoch": 0.4259466919692918,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.7909584328746533e-05,
+      "loss": 1.3957,
+      "step": 2469
+    },
+    {
+      "epoch": 0.4261192098680238,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.790791860002455e-05,
+      "loss": 1.4524,
+      "step": 2470
+    },
+    {
+      "epoch": 0.4262917277667558,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.790625228543455e-05,
+      "loss": 1.5006,
+      "step": 2471
+    },
+    {
+      "epoch": 0.4264642456654878,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.790458538509998e-05,
+      "loss": 1.6025,
+      "step": 2472
+    },
+    {
+      "epoch": 0.4266367635642198,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.790291789914433e-05,
+      "loss": 1.5626,
+      "step": 2473
+    },
+    {
+      "epoch": 0.4268092814629518,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.7901249827691148e-05,
+      "loss": 1.5277,
+      "step": 2474
+    },
+    {
+      "epoch": 0.4269817993616838,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.7899581170864003e-05,
+      "loss": 1.4732,
+      "step": 2475
+    },
+    {
+      "epoch": 0.42715431726041575,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.7897911928786535e-05,
+      "loss": 1.5614,
+      "step": 2476
+    },
+    {
+      "epoch": 0.42732683515914777,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.7896242101582393e-05,
+      "loss": 1.5511,
+      "step": 2477
+    },
+    {
+      "epoch": 0.4274993530578797,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.7894571689375305e-05,
+      "loss": 1.4692,
+      "step": 2478
+    },
+    {
+      "epoch": 0.42767187095661174,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.7892900692289016e-05,
+      "loss": 1.4351,
+      "step": 2479
+    },
+    {
+      "epoch": 0.42784438885534376,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.7891229110447325e-05,
+      "loss": 1.4667,
+      "step": 2480
+    },
+    {
+      "epoch": 0.4280169067540757,
+      "grad_norm": 0.625,
+      "learning_rate": 1.7889556943974078e-05,
+      "loss": 1.4757,
+      "step": 2481
+    },
+    {
+      "epoch": 0.42818942465280774,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.7887884192993155e-05,
+      "loss": 1.5014,
+      "step": 2482
+    },
+    {
+      "epoch": 0.4283619425515397,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.7886210857628485e-05,
+      "loss": 1.5169,
+      "step": 2483
+    },
+    {
+      "epoch": 0.4285344604502717,
+      "grad_norm": 0.75,
+      "learning_rate": 1.788453693800404e-05,
+      "loss": 1.4848,
+      "step": 2484
+    },
+    {
+      "epoch": 0.42870697834900373,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.7882862434243835e-05,
+      "loss": 1.4559,
+      "step": 2485
+    },
+    {
+      "epoch": 0.4288794962477357,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.7881187346471924e-05,
+      "loss": 1.5918,
+      "step": 2486
+    },
+    {
+      "epoch": 0.4290520141464677,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.7879511674812413e-05,
+      "loss": 1.4414,
+      "step": 2487
+    },
+    {
+      "epoch": 0.42922453204519967,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.7877835419389444e-05,
+      "loss": 1.4403,
+      "step": 2488
+    },
+    {
+      "epoch": 0.4293970499439317,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.78761585803272e-05,
+      "loss": 1.57,
+      "step": 2489
+    },
+    {
+      "epoch": 0.4295695678426637,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.787448115774992e-05,
+      "loss": 1.4331,
+      "step": 2490
+    },
+    {
+      "epoch": 0.42974208574139566,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.787280315178187e-05,
+      "loss": 1.4587,
+      "step": 2491
+    },
+    {
+      "epoch": 0.4299146036401277,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.7871124562547372e-05,
+      "loss": 1.537,
+      "step": 2492
+    },
+    {
+      "epoch": 0.43008712153885964,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.7869445390170787e-05,
+      "loss": 1.4619,
+      "step": 2493
+    },
+    {
+      "epoch": 0.43025963943759166,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.7867765634776516e-05,
+      "loss": 1.4177,
+      "step": 2494
+    },
+    {
+      "epoch": 0.4304321573363236,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.786608529648901e-05,
+      "loss": 1.4781,
+      "step": 2495
+    },
+    {
+      "epoch": 0.43060467523505563,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.786440437543275e-05,
+      "loss": 1.5143,
+      "step": 2496
+    },
+    {
+      "epoch": 0.43077719313378765,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.786272287173228e-05,
+      "loss": 1.383,
+      "step": 2497
+    },
+    {
+      "epoch": 0.4309497110325196,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.786104078551217e-05,
+      "loss": 1.551,
+      "step": 2498
+    },
+    {
+      "epoch": 0.4311222289312516,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.7859358116897034e-05,
+      "loss": 1.4651,
+      "step": 2499
+    },
+    {
+      "epoch": 0.4312947468299836,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.7857674866011546e-05,
+      "loss": 1.5387,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4312947468299836,
+      "eval_loss": 1.4578332901000977,
+      "eval_runtime": 10.8462,
+      "eval_samples_per_second": 94.411,
+      "eval_steps_per_second": 23.603,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4314672647287156,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.785599103298041e-05,
+      "loss": 1.3842,
+      "step": 2501
+    },
+    {
+      "epoch": 0.4316397826274476,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.7854306617928366e-05,
+      "loss": 1.5601,
+      "step": 2502
+    },
+    {
+      "epoch": 0.4318123005261796,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.7852621620980216e-05,
+      "loss": 1.4351,
+      "step": 2503
+    },
+    {
+      "epoch": 0.4319848184249116,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.785093604226079e-05,
+      "loss": 1.5191,
+      "step": 2504
+    },
+    {
+      "epoch": 0.43215733632364356,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.784924988189497e-05,
+      "loss": 1.4955,
+      "step": 2505
+    },
+    {
+      "epoch": 0.4323298542223756,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.7847563140007665e-05,
+      "loss": 1.435,
+      "step": 2506
+    },
+    {
+      "epoch": 0.4325023721211076,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.7845875816723855e-05,
+      "loss": 1.4883,
+      "step": 2507
+    },
+    {
+      "epoch": 0.43267489001983955,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.7844187912168543e-05,
+      "loss": 1.4455,
+      "step": 2508
+    },
+    {
+      "epoch": 0.43284740791857157,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.784249942646678e-05,
+      "loss": 1.4211,
+      "step": 2509
+    },
+    {
+      "epoch": 0.43301992581730353,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.784081035974365e-05,
+      "loss": 1.4285,
+      "step": 2510
+    },
+    {
+      "epoch": 0.43319244371603555,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.7839120712124297e-05,
+      "loss": 1.3663,
+      "step": 2511
+    },
+    {
+      "epoch": 0.4333649616147675,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.7837430483733908e-05,
+      "loss": 1.5058,
+      "step": 2512
+    },
+    {
+      "epoch": 0.4335374795134995,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.783573967469769e-05,
+      "loss": 1.555,
+      "step": 2513
+    },
+    {
+      "epoch": 0.43370999741223154,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.7834048285140923e-05,
+      "loss": 1.5479,
+      "step": 2514
+    },
+    {
+      "epoch": 0.4338825153109635,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.7832356315188907e-05,
+      "loss": 1.5398,
+      "step": 2515
+    },
+    {
+      "epoch": 0.4340550332096955,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.7830663764966995e-05,
+      "loss": 1.4866,
+      "step": 2516
+    },
+    {
+      "epoch": 0.4342275511084275,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.7828970634600584e-05,
+      "loss": 1.4726,
+      "step": 2517
+    },
+    {
+      "epoch": 0.4344000690071595,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.7827276924215113e-05,
+      "loss": 1.43,
+      "step": 2518
+    },
+    {
+      "epoch": 0.4345725869058915,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.7825582633936058e-05,
+      "loss": 1.4749,
+      "step": 2519
+    },
+    {
+      "epoch": 0.4347451048046235,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.7823887763888944e-05,
+      "loss": 1.5154,
+      "step": 2520
+    },
+    {
+      "epoch": 0.4349176227033555,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.782219231419934e-05,
+      "loss": 1.4378,
+      "step": 2521
+    },
+    {
+      "epoch": 0.43509014060208745,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.7820496284992853e-05,
+      "loss": 1.5858,
+      "step": 2522
+    },
+    {
+      "epoch": 0.43526265850081947,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.781879967639514e-05,
+      "loss": 1.5369,
+      "step": 2523
+    },
+    {
+      "epoch": 0.4354351763995514,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.781710248853189e-05,
+      "loss": 1.4153,
+      "step": 2524
+    },
+    {
+      "epoch": 0.43560769429828344,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.7815404721528848e-05,
+      "loss": 1.4206,
+      "step": 2525
+    },
+    {
+      "epoch": 0.43578021219701546,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.7813706375511784e-05,
+      "loss": 1.5314,
+      "step": 2526
+    },
+    {
+      "epoch": 0.4359527300957474,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.7812007450606536e-05,
+      "loss": 1.4995,
+      "step": 2527
+    },
+    {
+      "epoch": 0.43612524799447944,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.781030794693896e-05,
+      "loss": 1.4864,
+      "step": 2528
+    },
+    {
+      "epoch": 0.4362977658932114,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.7808607864634976e-05,
+      "loss": 1.475,
+      "step": 2529
+    },
+    {
+      "epoch": 0.4364702837919434,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.7806907203820525e-05,
+      "loss": 1.4901,
+      "step": 2530
+    },
+    {
+      "epoch": 0.43664280169067543,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.780520596462161e-05,
+      "loss": 1.4808,
+      "step": 2531
+    },
+    {
+      "epoch": 0.4368153195894074,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.780350414716427e-05,
+      "loss": 1.5526,
+      "step": 2532
+    },
+    {
+      "epoch": 0.4369878374881394,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.7801801751574583e-05,
+      "loss": 1.4987,
+      "step": 2533
+    },
+    {
+      "epoch": 0.43716035538687137,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.780009877797867e-05,
+      "loss": 1.455,
+      "step": 2534
+    },
+    {
+      "epoch": 0.4373328732856034,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.779839522650271e-05,
+      "loss": 1.5205,
+      "step": 2535
+    },
+    {
+      "epoch": 0.4375053911843354,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.7796691097272902e-05,
+      "loss": 1.5363,
+      "step": 2536
+    },
+    {
+      "epoch": 0.43767790908306736,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.77949863904155e-05,
+      "loss": 1.4779,
+      "step": 2537
+    },
+    {
+      "epoch": 0.4378504269817994,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.77932811060568e-05,
+      "loss": 1.4271,
+      "step": 2538
+    },
+    {
+      "epoch": 0.43802294488053134,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.7791575244323143e-05,
+      "loss": 1.5411,
+      "step": 2539
+    },
+    {
+      "epoch": 0.43819546277926336,
+      "grad_norm": 0.625,
+      "learning_rate": 1.7789868805340908e-05,
+      "loss": 1.551,
+      "step": 2540
+    },
+    {
+      "epoch": 0.4383679806779953,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.7788161789236518e-05,
+      "loss": 1.4683,
+      "step": 2541
+    },
+    {
+      "epoch": 0.43854049857672733,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.778645419613644e-05,
+      "loss": 1.4342,
+      "step": 2542
+    },
+    {
+      "epoch": 0.43871301647545935,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.7784746026167184e-05,
+      "loss": 1.4436,
+      "step": 2543
+    },
+    {
+      "epoch": 0.4388855343741913,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.77830372794553e-05,
+      "loss": 1.4579,
+      "step": 2544
+    },
+    {
+      "epoch": 0.4390580522729233,
+      "grad_norm": 0.625,
+      "learning_rate": 1.7781327956127385e-05,
+      "loss": 1.5987,
+      "step": 2545
+    },
+    {
+      "epoch": 0.4392305701716553,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.7779618056310074e-05,
+      "loss": 1.5104,
+      "step": 2546
+    },
+    {
+      "epoch": 0.4394030880703873,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.777790758013005e-05,
+      "loss": 1.5887,
+      "step": 2547
+    },
+    {
+      "epoch": 0.4395756059691193,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.777619652771403e-05,
+      "loss": 1.5033,
+      "step": 2548
+    },
+    {
+      "epoch": 0.4397481238678513,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.7774484899188788e-05,
+      "loss": 1.5848,
+      "step": 2549
+    },
+    {
+      "epoch": 0.4399206417665833,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.7772772694681123e-05,
+      "loss": 1.5275,
+      "step": 2550
+    },
+    {
+      "epoch": 0.44009315966531526,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.777105991431789e-05,
+      "loss": 1.4234,
+      "step": 2551
+    },
+    {
+      "epoch": 0.4402656775640473,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.7769346558225987e-05,
+      "loss": 1.5312,
+      "step": 2552
+    },
+    {
+      "epoch": 0.44043819546277924,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.7767632626532344e-05,
+      "loss": 1.5402,
+      "step": 2553
+    },
+    {
+      "epoch": 0.44061071336151125,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.7765918119363942e-05,
+      "loss": 1.4823,
+      "step": 2554
+    },
+    {
+      "epoch": 0.44078323126024327,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.77642030368478e-05,
+      "loss": 1.5659,
+      "step": 2555
+    },
+    {
+      "epoch": 0.44095574915897523,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.7762487379110984e-05,
+      "loss": 1.518,
+      "step": 2556
+    },
+    {
+      "epoch": 0.44112826705770725,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.7760771146280603e-05,
+      "loss": 1.4378,
+      "step": 2557
+    },
+    {
+      "epoch": 0.4413007849564392,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.7759054338483803e-05,
+      "loss": 1.4941,
+      "step": 2558
+    },
+    {
+      "epoch": 0.4414733028551712,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.7757336955847775e-05,
+      "loss": 1.5745,
+      "step": 2559
+    },
+    {
+      "epoch": 0.44164582075390324,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.7755618998499757e-05,
+      "loss": 1.5342,
+      "step": 2560
+    },
+    {
+      "epoch": 0.4418183386526352,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.7753900466567024e-05,
+      "loss": 1.5415,
+      "step": 2561
+    },
+    {
+      "epoch": 0.4419908565513672,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.7752181360176895e-05,
+      "loss": 1.4617,
+      "step": 2562
+    },
+    {
+      "epoch": 0.4421633744500992,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.7750461679456737e-05,
+      "loss": 1.5255,
+      "step": 2563
+    },
+    {
+      "epoch": 0.4423358923488312,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.7748741424533947e-05,
+      "loss": 1.454,
+      "step": 2564
+    },
+    {
+      "epoch": 0.4425084102475632,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.7747020595535976e-05,
+      "loss": 1.5193,
+      "step": 2565
+    },
+    {
+      "epoch": 0.44268092814629517,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.7745299192590317e-05,
+      "loss": 1.4076,
+      "step": 2566
+    },
+    {
+      "epoch": 0.4428534460450272,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.7743577215824494e-05,
+      "loss": 1.4981,
+      "step": 2567
+    },
+    {
+      "epoch": 0.44302596394375915,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.774185466536609e-05,
+      "loss": 1.5218,
+      "step": 2568
+    },
+    {
+      "epoch": 0.44319848184249117,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.774013154134272e-05,
+      "loss": 1.5258,
+      "step": 2569
+    },
+    {
+      "epoch": 0.4433709997412231,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.7738407843882037e-05,
+      "loss": 1.5437,
+      "step": 2570
+    },
+    {
+      "epoch": 0.44354351763995514,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.773668357311175e-05,
+      "loss": 1.45,
+      "step": 2571
+    },
+    {
+      "epoch": 0.44371603553868716,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.773495872915961e-05,
+      "loss": 1.4216,
+      "step": 2572
+    },
+    {
+      "epoch": 0.4438885534374191,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.773323331215339e-05,
+      "loss": 1.342,
+      "step": 2573
+    },
+    {
+      "epoch": 0.44406107133615114,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.7731507322220932e-05,
+      "loss": 1.5443,
+      "step": 2574
+    },
+    {
+      "epoch": 0.4442335892348831,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.77297807594901e-05,
+      "loss": 1.5211,
+      "step": 2575
+    },
+    {
+      "epoch": 0.4444061071336151,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.7728053624088812e-05,
+      "loss": 1.4566,
+      "step": 2576
+    },
+    {
+      "epoch": 0.44457862503234713,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.7726325916145027e-05,
+      "loss": 1.4819,
+      "step": 2577
+    },
+    {
+      "epoch": 0.4447511429310791,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.7724597635786737e-05,
+      "loss": 1.4937,
+      "step": 2578
+    },
+    {
+      "epoch": 0.4449236608298111,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.7722868783141992e-05,
+      "loss": 1.5356,
+      "step": 2579
+    },
+    {
+      "epoch": 0.44509617872854307,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.772113935833887e-05,
+      "loss": 1.4269,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4452686966272751,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.7719409361505503e-05,
+      "loss": 1.5314,
+      "step": 2581
+    },
+    {
+      "epoch": 0.4454412145260071,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.7717678792770056e-05,
+      "loss": 1.4201,
+      "step": 2582
+    },
+    {
+      "epoch": 0.44561373242473906,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.771594765226074e-05,
+      "loss": 1.4507,
+      "step": 2583
+    },
+    {
+      "epoch": 0.4457862503234711,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.7714215940105813e-05,
+      "loss": 1.4426,
+      "step": 2584
+    },
+    {
+      "epoch": 0.44595876822220304,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.771248365643357e-05,
+      "loss": 1.4883,
+      "step": 2585
+    },
+    {
+      "epoch": 0.44613128612093506,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.7710750801372345e-05,
+      "loss": 1.4473,
+      "step": 2586
+    },
+    {
+      "epoch": 0.446303804019667,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.7709017375050525e-05,
+      "loss": 1.4357,
+      "step": 2587
+    },
+    {
+      "epoch": 0.44647632191839903,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.7707283377596526e-05,
+      "loss": 1.372,
+      "step": 2588
+    },
+    {
+      "epoch": 0.44664883981713105,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.770554880913882e-05,
+      "loss": 1.5774,
+      "step": 2589
+    },
+    {
+      "epoch": 0.446821357715863,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.770381366980591e-05,
+      "loss": 1.4317,
+      "step": 2590
+    },
+    {
+      "epoch": 0.446993875614595,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.7702077959726346e-05,
+      "loss": 1.5231,
+      "step": 2591
+    },
+    {
+      "epoch": 0.447166393513327,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.7700341679028725e-05,
+      "loss": 1.4977,
+      "step": 2592
+    },
+    {
+      "epoch": 0.447338911412059,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.769860482784168e-05,
+      "loss": 1.5746,
+      "step": 2593
+    },
+    {
+      "epoch": 0.447511429310791,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.769686740629388e-05,
+      "loss": 1.474,
+      "step": 2594
+    },
+    {
+      "epoch": 0.447683947209523,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.7695129414514057e-05,
+      "loss": 1.4457,
+      "step": 2595
+    },
+    {
+      "epoch": 0.447856465108255,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.769339085263096e-05,
+      "loss": 1.5435,
+      "step": 2596
+    },
+    {
+      "epoch": 0.44802898300698696,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.76916517207734e-05,
+      "loss": 1.4355,
+      "step": 2597
+    },
+    {
+      "epoch": 0.448201500905719,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.7689912019070223e-05,
+      "loss": 1.4525,
+      "step": 2598
+    },
+    {
+      "epoch": 0.44837401880445094,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.7688171747650313e-05,
+      "loss": 1.4333,
+      "step": 2599
+    },
+    {
+      "epoch": 0.44854653670318295,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.7686430906642602e-05,
+      "loss": 1.5239,
+      "step": 2600
+    },
+    {
+      "epoch": 0.44854653670318295,
+      "eval_loss": 1.4545358419418335,
+      "eval_runtime": 11.0064,
+      "eval_samples_per_second": 93.037,
+      "eval_steps_per_second": 23.259,
+      "step": 2600
+    },
+    {
+      "epoch": 0.44871905460191497,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.7684689496176065e-05,
+      "loss": 1.4694,
+      "step": 2601
+    },
+    {
+      "epoch": 0.44889157250064693,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.7682947516379706e-05,
+      "loss": 1.4667,
+      "step": 2602
+    },
+    {
+      "epoch": 0.44906409039937895,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.7681204967382597e-05,
+      "loss": 1.5195,
+      "step": 2603
+    },
+    {
+      "epoch": 0.4492366082981109,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.767946184931383e-05,
+      "loss": 1.4439,
+      "step": 2604
+    },
+    {
+      "epoch": 0.4494091261968429,
+      "grad_norm": 0.625,
+      "learning_rate": 1.7677718162302546e-05,
+      "loss": 1.5034,
+      "step": 2605
+    },
+    {
+      "epoch": 0.44958164409557494,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.7675973906477924e-05,
+      "loss": 1.4178,
+      "step": 2606
+    },
+    {
+      "epoch": 0.4497541619943069,
+      "grad_norm": 0.625,
+      "learning_rate": 1.7674229081969195e-05,
+      "loss": 1.3571,
+      "step": 2607
+    },
+    {
+      "epoch": 0.4499266798930389,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.7672483688905622e-05,
+      "loss": 1.3652,
+      "step": 2608
+    },
+    {
+      "epoch": 0.4500991977917709,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.767073772741652e-05,
+      "loss": 1.4889,
+      "step": 2609
+    },
+    {
+      "epoch": 0.4502717156905029,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.766899119763124e-05,
+      "loss": 1.5149,
+      "step": 2610
+    },
+    {
+      "epoch": 0.4504442335892349,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.7667244099679172e-05,
+      "loss": 1.5902,
+      "step": 2611
+    },
+    {
+      "epoch": 0.45061675148796687,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.7665496433689754e-05,
+      "loss": 1.458,
+      "step": 2612
+    },
+    {
+      "epoch": 0.4507892693866989,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.7663748199792463e-05,
+      "loss": 1.4436,
+      "step": 2613
+    },
+    {
+      "epoch": 0.45096178728543085,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.7661999398116824e-05,
+      "loss": 1.3657,
+      "step": 2614
+    },
+    {
+      "epoch": 0.45113430518416286,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.7660250028792392e-05,
+      "loss": 1.5338,
+      "step": 2615
+    },
+    {
+      "epoch": 0.4513068230828948,
+      "grad_norm": 0.625,
+      "learning_rate": 1.7658500091948774e-05,
+      "loss": 1.4237,
+      "step": 2616
+    },
+    {
+      "epoch": 0.45147934098162684,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.7656749587715617e-05,
+      "loss": 1.4443,
+      "step": 2617
+    },
+    {
+      "epoch": 0.45165185888035886,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.765499851622261e-05,
+      "loss": 1.5702,
+      "step": 2618
+    },
+    {
+      "epoch": 0.4518243767790908,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.765324687759948e-05,
+      "loss": 1.4471,
+      "step": 2619
+    },
+    {
+      "epoch": 0.45199689467782284,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.7651494671976003e-05,
+      "loss": 1.5225,
+      "step": 2620
+    },
+    {
+      "epoch": 0.4521694125765548,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.7649741899481997e-05,
+      "loss": 1.4638,
+      "step": 2621
+    },
+    {
+      "epoch": 0.4523419304752868,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.7647988560247305e-05,
+      "loss": 1.4911,
+      "step": 2622
+    },
+    {
+      "epoch": 0.45251444837401883,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.764623465440184e-05,
+      "loss": 1.396,
+      "step": 2623
+    },
+    {
+      "epoch": 0.4526869662727508,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.764448018207553e-05,
+      "loss": 1.4932,
+      "step": 2624
+    },
+    {
+      "epoch": 0.4528594841714828,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.764272514339837e-05,
+      "loss": 1.4553,
+      "step": 2625
+    },
+    {
+      "epoch": 0.45303200207021477,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.764096953850037e-05,
+      "loss": 1.5165,
+      "step": 2626
+    },
+    {
+      "epoch": 0.4532045199689468,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.7639213367511608e-05,
+      "loss": 1.5094,
+      "step": 2627
+    },
+    {
+      "epoch": 0.45337703786767874,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.763745663056219e-05,
+      "loss": 1.4608,
+      "step": 2628
+    },
+    {
+      "epoch": 0.45354955576641076,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.7635699327782257e-05,
+      "loss": 1.4027,
+      "step": 2629
+    },
+    {
+      "epoch": 0.4537220736651428,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.7633941459302013e-05,
+      "loss": 1.4297,
+      "step": 2630
+    },
+    {
+      "epoch": 0.45389459156387474,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.763218302525169e-05,
+      "loss": 1.3769,
+      "step": 2631
+    },
+    {
+      "epoch": 0.45406710946260675,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.7630424025761554e-05,
+      "loss": 1.585,
+      "step": 2632
+    },
+    {
+      "epoch": 0.4542396273613387,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.7628664460961928e-05,
+      "loss": 1.4986,
+      "step": 2633
+    },
+    {
+      "epoch": 0.45441214526007073,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.7626904330983176e-05,
+      "loss": 1.5802,
+      "step": 2634
+    },
+    {
+      "epoch": 0.45458466315880275,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.7625143635955697e-05,
+      "loss": 1.3242,
+      "step": 2635
+    },
+    {
+      "epoch": 0.4547571810575347,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.7623382376009928e-05,
+      "loss": 1.5192,
+      "step": 2636
+    },
+    {
+      "epoch": 0.4549296989562667,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.7621620551276366e-05,
+      "loss": 1.4385,
+      "step": 2637
+    },
+    {
+      "epoch": 0.4551022168549987,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.761985816188553e-05,
+      "loss": 1.4715,
+      "step": 2638
+    },
+    {
+      "epoch": 0.4552747347537307,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.7618095207967988e-05,
+      "loss": 1.5298,
+      "step": 2639
+    },
+    {
+      "epoch": 0.4554472526524627,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.7616331689654352e-05,
+      "loss": 1.4869,
+      "step": 2640
+    },
+    {
+      "epoch": 0.4556197705511947,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.7614567607075278e-05,
+      "loss": 1.5208,
+      "step": 2641
+    },
+    {
+      "epoch": 0.4557922884499267,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.761280296036146e-05,
+      "loss": 1.4899,
+      "step": 2642
+    },
+    {
+      "epoch": 0.45596480634865866,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.761103774964363e-05,
+      "loss": 1.3997,
+      "step": 2643
+    },
+    {
+      "epoch": 0.4561373242473907,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.7609271975052563e-05,
+      "loss": 1.4327,
+      "step": 2644
+    },
+    {
+      "epoch": 0.45630984214612264,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.7607505636719085e-05,
+      "loss": 1.4269,
+      "step": 2645
+    },
+    {
+      "epoch": 0.45648236004485465,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.7605738734774062e-05,
+      "loss": 1.4966,
+      "step": 2646
+    },
+    {
+      "epoch": 0.45665487794358667,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.7603971269348383e-05,
+      "loss": 1.4565,
+      "step": 2647
+    },
+    {
+      "epoch": 0.45682739584231863,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.7602203240573004e-05,
+      "loss": 1.3994,
+      "step": 2648
+    },
+    {
+      "epoch": 0.45699991374105065,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.7600434648578906e-05,
+      "loss": 1.4751,
+      "step": 2649
+    },
+    {
+      "epoch": 0.4571724316397826,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.7598665493497122e-05,
+      "loss": 1.5806,
+      "step": 2650
+    },
+    {
+      "epoch": 0.4573449495385146,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.759689577545872e-05,
+      "loss": 1.5674,
+      "step": 2651
+    },
+    {
+      "epoch": 0.45751746743724664,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.759512549459481e-05,
+      "loss": 1.478,
+      "step": 2652
+    },
+    {
+      "epoch": 0.4576899853359786,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.7593354651036544e-05,
+      "loss": 1.4421,
+      "step": 2653
+    },
+    {
+      "epoch": 0.4578625032347106,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.759158324491512e-05,
+      "loss": 1.5771,
+      "step": 2654
+    },
+    {
+      "epoch": 0.4580350211334426,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.758981127636178e-05,
+      "loss": 1.4393,
+      "step": 2655
+    },
+    {
+      "epoch": 0.4582075390321746,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.7588038745507797e-05,
+      "loss": 1.5219,
+      "step": 2656
+    },
+    {
+      "epoch": 0.45838005693090655,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.7586265652484488e-05,
+      "loss": 1.4692,
+      "step": 2657
+    },
+    {
+      "epoch": 0.45855257482963857,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.758449199742322e-05,
+      "loss": 1.4827,
+      "step": 2658
+    },
+    {
+      "epoch": 0.4587250927283706,
+      "grad_norm": 1.2734375,
+      "learning_rate": 1.7582717780455395e-05,
+      "loss": 1.509,
+      "step": 2659
+    },
+    {
+      "epoch": 0.45889761062710255,
+      "grad_norm": 1.3359375,
+      "learning_rate": 1.7580943001712457e-05,
+      "loss": 1.61,
+      "step": 2660
+    },
+    {
+      "epoch": 0.45907012852583456,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.7579167661325892e-05,
+      "loss": 1.5574,
+      "step": 2661
+    },
+    {
+      "epoch": 0.4592426464245665,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.757739175942723e-05,
+      "loss": 1.4575,
+      "step": 2662
+    },
+    {
+      "epoch": 0.45941516432329854,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.7575615296148044e-05,
+      "loss": 1.429,
+      "step": 2663
+    },
+    {
+      "epoch": 0.45958768222203056,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.757383827161994e-05,
+      "loss": 1.3583,
+      "step": 2664
+    },
+    {
+      "epoch": 0.4597602001207625,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.7572060685974577e-05,
+      "loss": 1.4718,
+      "step": 2665
+    },
+    {
+      "epoch": 0.45993271801949454,
+      "grad_norm": 0.75,
+      "learning_rate": 1.7570282539343643e-05,
+      "loss": 1.3904,
+      "step": 2666
+    },
+    {
+      "epoch": 0.4601052359182265,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.7568503831858875e-05,
+      "loss": 1.5472,
+      "step": 2667
+    },
+    {
+      "epoch": 0.4602777538169585,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.7566724563652052e-05,
+      "loss": 1.4417,
+      "step": 2668
+    },
+    {
+      "epoch": 0.46045027171569053,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.7564944734855e-05,
+      "loss": 1.5233,
+      "step": 2669
+    },
+    {
+      "epoch": 0.4606227896144225,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.756316434559957e-05,
+      "loss": 1.4207,
+      "step": 2670
+    },
+    {
+      "epoch": 0.4607953075131545,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.7561383396017672e-05,
+      "loss": 1.5302,
+      "step": 2671
+    },
+    {
+      "epoch": 0.46096782541188647,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.7559601886241245e-05,
+      "loss": 1.4491,
+      "step": 2672
+    },
+    {
+      "epoch": 0.4611403433106185,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.7557819816402273e-05,
+      "loss": 1.5596,
+      "step": 2673
+    },
+    {
+      "epoch": 0.46131286120935044,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.7556037186632787e-05,
+      "loss": 1.4611,
+      "step": 2674
+    },
+    {
+      "epoch": 0.46148537910808246,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.7554253997064854e-05,
+      "loss": 1.4785,
+      "step": 2675
+    },
+    {
+      "epoch": 0.4616578970068145,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.755247024783058e-05,
+      "loss": 1.4997,
+      "step": 2676
+    },
+    {
+      "epoch": 0.46183041490554644,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.7550685939062125e-05,
+      "loss": 1.4593,
+      "step": 2677
+    },
+    {
+      "epoch": 0.46200293280427845,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.754890107089168e-05,
+      "loss": 1.5093,
+      "step": 2678
+    },
+    {
+      "epoch": 0.4621754507030104,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.754711564345147e-05,
+      "loss": 1.436,
+      "step": 2679
+    },
+    {
+      "epoch": 0.46234796860174243,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.754532965687378e-05,
+      "loss": 1.4643,
+      "step": 2680
+    },
+    {
+      "epoch": 0.46252048650047445,
+      "grad_norm": 0.625,
+      "learning_rate": 1.754354311129092e-05,
+      "loss": 1.4679,
+      "step": 2681
+    },
+    {
+      "epoch": 0.4626930043992064,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.7541756006835253e-05,
+      "loss": 1.5252,
+      "step": 2682
+    },
+    {
+      "epoch": 0.4628655222979384,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.753996834363918e-05,
+      "loss": 1.5314,
+      "step": 2683
+    },
+    {
+      "epoch": 0.4630380401966704,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.753818012183514e-05,
+      "loss": 1.4043,
+      "step": 2684
+    },
+    {
+      "epoch": 0.4632105580954024,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.7536391341555613e-05,
+      "loss": 1.5028,
+      "step": 2685
+    },
+    {
+      "epoch": 0.4633830759941344,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.7534602002933128e-05,
+      "loss": 1.4713,
+      "step": 2686
+    },
+    {
+      "epoch": 0.4635555938928664,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.7532812106100247e-05,
+      "loss": 1.4458,
+      "step": 2687
+    },
+    {
+      "epoch": 0.4637281117915984,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.7531021651189578e-05,
+      "loss": 1.4422,
+      "step": 2688
+    },
+    {
+      "epoch": 0.46390062969033036,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.752923063833377e-05,
+      "loss": 1.4305,
+      "step": 2689
+    },
+    {
+      "epoch": 0.4640731475890624,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.7527439067665516e-05,
+      "loss": 1.4937,
+      "step": 2690
+    },
+    {
+      "epoch": 0.46424566548779433,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.7525646939317535e-05,
+      "loss": 1.4594,
+      "step": 2691
+    },
+    {
+      "epoch": 0.46441818338652635,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.752385425342261e-05,
+      "loss": 1.4699,
+      "step": 2692
+    },
+    {
+      "epoch": 0.46459070128525837,
+      "grad_norm": 0.625,
+      "learning_rate": 1.752206101011355e-05,
+      "loss": 1.5202,
+      "step": 2693
+    },
+    {
+      "epoch": 0.46476321918399033,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.752026720952321e-05,
+      "loss": 1.5505,
+      "step": 2694
+    },
+    {
+      "epoch": 0.46493573708272234,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.7518472851784485e-05,
+      "loss": 1.4083,
+      "step": 2695
+    },
+    {
+      "epoch": 0.4651082549814543,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.7516677937030318e-05,
+      "loss": 1.5286,
+      "step": 2696
+    },
+    {
+      "epoch": 0.4652807728801863,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.751488246539368e-05,
+      "loss": 1.5077,
+      "step": 2697
+    },
+    {
+      "epoch": 0.46545329077891834,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.7513086437007593e-05,
+      "loss": 1.5907,
+      "step": 2698
+    },
+    {
+      "epoch": 0.4656258086776503,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.751128985200512e-05,
+      "loss": 1.498,
+      "step": 2699
+    },
+    {
+      "epoch": 0.4657983265763823,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.750949271051936e-05,
+      "loss": 1.5613,
+      "step": 2700
+    },
+    {
+      "epoch": 0.4657983265763823,
+      "eval_loss": 1.451514720916748,
+      "eval_runtime": 11.0249,
+      "eval_samples_per_second": 92.88,
+      "eval_steps_per_second": 23.22,
+      "step": 2700
+    },
+    {
+      "epoch": 0.4659708444751143,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.7507695012683463e-05,
+      "loss": 1.4832,
+      "step": 2701
+    },
+    {
+      "epoch": 0.4661433623738463,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.7505896758630606e-05,
+      "loss": 1.508,
+      "step": 2702
+    },
+    {
+      "epoch": 0.46631588027257825,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.750409794849402e-05,
+      "loss": 1.3877,
+      "step": 2703
+    },
+    {
+      "epoch": 0.46648839817131027,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.7502298582406967e-05,
+      "loss": 1.4419,
+      "step": 2704
+    },
+    {
+      "epoch": 0.4666609160700423,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.7500498660502757e-05,
+      "loss": 1.4838,
+      "step": 2705
+    },
+    {
+      "epoch": 0.46683343396877425,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.7498698182914746e-05,
+      "loss": 1.5293,
+      "step": 2706
+    },
+    {
+      "epoch": 0.46700595186750626,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.749689714977632e-05,
+      "loss": 1.5346,
+      "step": 2707
+    },
+    {
+      "epoch": 0.4671784697662382,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.7495095561220908e-05,
+      "loss": 1.4503,
+      "step": 2708
+    },
+    {
+      "epoch": 0.46735098766497024,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.7493293417381985e-05,
+      "loss": 1.4146,
+      "step": 2709
+    },
+    {
+      "epoch": 0.46752350556370226,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.7491490718393067e-05,
+      "loss": 1.554,
+      "step": 2710
+    },
+    {
+      "epoch": 0.4676960234624342,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.7489687464387705e-05,
+      "loss": 1.5255,
+      "step": 2711
+    },
+    {
+      "epoch": 0.46786854136116623,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.74878836554995e-05,
+      "loss": 1.4677,
+      "step": 2712
+    },
+    {
+      "epoch": 0.4680410592598982,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.748607929186209e-05,
+      "loss": 1.5083,
+      "step": 2713
+    },
+    {
+      "epoch": 0.4682135771586302,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.7484274373609143e-05,
+      "loss": 1.5612,
+      "step": 2714
+    },
+    {
+      "epoch": 0.46838609505736223,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.7482468900874393e-05,
+      "loss": 1.4519,
+      "step": 2715
+    },
+    {
+      "epoch": 0.4685586129560942,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.7480662873791592e-05,
+      "loss": 1.4596,
+      "step": 2716
+    },
+    {
+      "epoch": 0.4687311308548262,
+      "grad_norm": 1.0,
+      "learning_rate": 1.7478856292494543e-05,
+      "loss": 1.5029,
+      "step": 2717
+    },
+    {
+      "epoch": 0.46890364875355817,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.7477049157117093e-05,
+      "loss": 1.4348,
+      "step": 2718
+    },
+    {
+      "epoch": 0.4690761666522902,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.747524146779312e-05,
+      "loss": 1.531,
+      "step": 2719
+    },
+    {
+      "epoch": 0.46924868455102214,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.7473433224656554e-05,
+      "loss": 1.4822,
+      "step": 2720
+    },
+    {
+      "epoch": 0.46942120244975416,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.7471624427841356e-05,
+      "loss": 1.3717,
+      "step": 2721
+    },
+    {
+      "epoch": 0.4695937203484862,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.7469815077481537e-05,
+      "loss": 1.462,
+      "step": 2722
+    },
+    {
+      "epoch": 0.46976623824721814,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.746800517371114e-05,
+      "loss": 1.4464,
+      "step": 2723
+    },
+    {
+      "epoch": 0.46993875614595015,
+      "grad_norm": 0.625,
+      "learning_rate": 1.7466194716664262e-05,
+      "loss": 1.4288,
+      "step": 2724
+    },
+    {
+      "epoch": 0.4701112740446821,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.7464383706475028e-05,
+      "loss": 1.4812,
+      "step": 2725
+    },
+    {
+      "epoch": 0.47028379194341413,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.7462572143277606e-05,
+      "loss": 1.388,
+      "step": 2726
+    },
+    {
+      "epoch": 0.47045630984214615,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.7460760027206215e-05,
+      "loss": 1.4307,
+      "step": 2727
+    },
+    {
+      "epoch": 0.4706288277408781,
+      "grad_norm": 0.75,
+      "learning_rate": 1.7458947358395102e-05,
+      "loss": 1.521,
+      "step": 2728
+    },
+    {
+      "epoch": 0.4708013456396101,
+      "grad_norm": 0.75,
+      "learning_rate": 1.7457134136978566e-05,
+      "loss": 1.525,
+      "step": 2729
+    },
+    {
+      "epoch": 0.4709738635383421,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.7455320363090936e-05,
+      "loss": 1.4898,
+      "step": 2730
+    },
+    {
+      "epoch": 0.4711463814370741,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.7453506036866592e-05,
+      "loss": 1.416,
+      "step": 2731
+    },
+    {
+      "epoch": 0.47131889933580606,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.745169115843995e-05,
+      "loss": 1.4478,
+      "step": 2732
+    },
+    {
+      "epoch": 0.4714914172345381,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.7449875727945463e-05,
+      "loss": 1.4715,
+      "step": 2733
+    },
+    {
+      "epoch": 0.4716639351332701,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.7448059745517635e-05,
+      "loss": 1.5544,
+      "step": 2734
+    },
+    {
+      "epoch": 0.47183645303200206,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.7446243211291003e-05,
+      "loss": 1.4665,
+      "step": 2735
+    },
+    {
+      "epoch": 0.4720089709307341,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.7444426125400148e-05,
+      "loss": 1.4048,
+      "step": 2736
+    },
+    {
+      "epoch": 0.47218148882946603,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.7442608487979692e-05,
+      "loss": 1.4598,
+      "step": 2737
+    },
+    {
+      "epoch": 0.47235400672819805,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.7440790299164295e-05,
+      "loss": 1.5017,
+      "step": 2738
+    },
+    {
+      "epoch": 0.47252652462693007,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.7438971559088658e-05,
+      "loss": 1.4438,
+      "step": 2739
+    },
+    {
+      "epoch": 0.472699042525662,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.743715226788753e-05,
+      "loss": 1.4198,
+      "step": 2740
+    },
+    {
+      "epoch": 0.47287156042439404,
+      "grad_norm": 0.625,
+      "learning_rate": 1.743533242569569e-05,
+      "loss": 1.4941,
+      "step": 2741
+    },
+    {
+      "epoch": 0.473044078323126,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.7433512032647968e-05,
+      "loss": 1.4281,
+      "step": 2742
+    },
+    {
+      "epoch": 0.473216596221858,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.7431691088879228e-05,
+      "loss": 1.4557,
+      "step": 2743
+    },
+    {
+      "epoch": 0.47338911412059004,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.7429869594524375e-05,
+      "loss": 1.4984,
+      "step": 2744
+    },
+    {
+      "epoch": 0.473561632019322,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.742804754971836e-05,
+      "loss": 1.537,
+      "step": 2745
+    },
+    {
+      "epoch": 0.473734149918054,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.742622495459617e-05,
+      "loss": 1.4307,
+      "step": 2746
+    },
+    {
+      "epoch": 0.473906667816786,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.7424401809292833e-05,
+      "loss": 1.5323,
+      "step": 2747
+    },
+    {
+      "epoch": 0.474079185715518,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.742257811394342e-05,
+      "loss": 1.5329,
+      "step": 2748
+    },
+    {
+      "epoch": 0.47425170361424995,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.7420753868683044e-05,
+      "loss": 1.533,
+      "step": 2749
+    },
+    {
+      "epoch": 0.47442422151298197,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.7418929073646855e-05,
+      "loss": 1.4591,
+      "step": 2750
+    },
+    {
+      "epoch": 0.474596739411714,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.741710372897004e-05,
+      "loss": 1.4273,
+      "step": 2751
+    },
+    {
+      "epoch": 0.47476925731044595,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.741527783478784e-05,
+      "loss": 1.4346,
+      "step": 2752
+    },
+    {
+      "epoch": 0.47494177520917796,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.7413451391235524e-05,
+      "loss": 1.4941,
+      "step": 2753
+    },
+    {
+      "epoch": 0.4751142931079099,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.7411624398448408e-05,
+      "loss": 1.4746,
+      "step": 2754
+    },
+    {
+      "epoch": 0.47528681100664194,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.7409796856561847e-05,
+      "loss": 1.4099,
+      "step": 2755
+    },
+    {
+      "epoch": 0.47545932890537396,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.7407968765711234e-05,
+      "loss": 1.5676,
+      "step": 2756
+    },
+    {
+      "epoch": 0.4756318468041059,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.740614012603201e-05,
+      "loss": 1.4248,
+      "step": 2757
+    },
+    {
+      "epoch": 0.47580436470283793,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.740431093765965e-05,
+      "loss": 1.5019,
+      "step": 2758
+    },
+    {
+      "epoch": 0.4759768826015699,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.740248120072967e-05,
+      "loss": 1.4743,
+      "step": 2759
+    },
+    {
+      "epoch": 0.4761494005003019,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.740065091537763e-05,
+      "loss": 1.3471,
+      "step": 2760
+    },
+    {
+      "epoch": 0.4763219183990339,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.7398820081739128e-05,
+      "loss": 1.5532,
+      "step": 2761
+    },
+    {
+      "epoch": 0.4764944362977659,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.739698869994981e-05,
+      "loss": 1.4123,
+      "step": 2762
+    },
+    {
+      "epoch": 0.4766669541964979,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.7395156770145343e-05,
+      "loss": 1.5422,
+      "step": 2763
+    },
+    {
+      "epoch": 0.47683947209522987,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.739332429246146e-05,
+      "loss": 1.5009,
+      "step": 2764
+    },
+    {
+      "epoch": 0.4770119899939619,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.7391491267033916e-05,
+      "loss": 1.3644,
+      "step": 2765
+    },
+    {
+      "epoch": 0.47718450789269384,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.7389657693998515e-05,
+      "loss": 1.4097,
+      "step": 2766
+    },
+    {
+      "epoch": 0.47735702579142586,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.73878235734911e-05,
+      "loss": 1.4486,
+      "step": 2767
+    },
+    {
+      "epoch": 0.4775295436901579,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.7385988905647556e-05,
+      "loss": 1.3957,
+      "step": 2768
+    },
+    {
+      "epoch": 0.47770206158888984,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.73841536906038e-05,
+      "loss": 1.3941,
+      "step": 2769
+    },
+    {
+      "epoch": 0.47787457948762185,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.7382317928495803e-05,
+      "loss": 1.5116,
+      "step": 2770
+    },
+    {
+      "epoch": 0.4780470973863538,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.7380481619459564e-05,
+      "loss": 1.4201,
+      "step": 2771
+    },
+    {
+      "epoch": 0.47821961528508583,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.7378644763631133e-05,
+      "loss": 1.4355,
+      "step": 2772
+    },
+    {
+      "epoch": 0.47839213318381785,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.7376807361146594e-05,
+      "loss": 1.6016,
+      "step": 2773
+    },
+    {
+      "epoch": 0.4785646510825498,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.7374969412142072e-05,
+      "loss": 1.433,
+      "step": 2774
+    },
+    {
+      "epoch": 0.4787371689812818,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.7373130916753737e-05,
+      "loss": 1.4282,
+      "step": 2775
+    },
+    {
+      "epoch": 0.4789096868800138,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.737129187511779e-05,
+      "loss": 1.4499,
+      "step": 2776
+    },
+    {
+      "epoch": 0.4790822047787458,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.736945228737049e-05,
+      "loss": 1.5561,
+      "step": 2777
+    },
+    {
+      "epoch": 0.47925472267747776,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.7367612153648113e-05,
+      "loss": 1.3707,
+      "step": 2778
+    },
+    {
+      "epoch": 0.4794272405762098,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.7365771474086993e-05,
+      "loss": 1.5786,
+      "step": 2779
+    },
+    {
+      "epoch": 0.4795997584749418,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.73639302488235e-05,
+      "loss": 1.5003,
+      "step": 2780
+    },
+    {
+      "epoch": 0.47977227637367376,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.7362088477994043e-05,
+      "loss": 1.5189,
+      "step": 2781
+    },
+    {
+      "epoch": 0.4799447942724058,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.736024616173507e-05,
+      "loss": 1.4467,
+      "step": 2782
+    },
+    {
+      "epoch": 0.48011731217113773,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.735840330018307e-05,
+      "loss": 1.5065,
+      "step": 2783
+    },
+    {
+      "epoch": 0.48028983006986975,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.7356559893474578e-05,
+      "loss": 1.5442,
+      "step": 2784
+    },
+    {
+      "epoch": 0.48046234796860177,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.735471594174616e-05,
+      "loss": 1.41,
+      "step": 2785
+    },
+    {
+      "epoch": 0.4806348658673337,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.735287144513444e-05,
+      "loss": 1.4113,
+      "step": 2786
+    },
+    {
+      "epoch": 0.48080738376606574,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.7351026403776054e-05,
+      "loss": 1.4412,
+      "step": 2787
+    },
+    {
+      "epoch": 0.4809799016647977,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.73491808178077e-05,
+      "loss": 1.4297,
+      "step": 2788
+    },
+    {
+      "epoch": 0.4811524195635297,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.7347334687366114e-05,
+      "loss": 1.6249,
+      "step": 2789
+    },
+    {
+      "epoch": 0.48132493746226174,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.7345488012588064e-05,
+      "loss": 1.3788,
+      "step": 2790
+    },
+    {
+      "epoch": 0.4814974553609937,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.7343640793610366e-05,
+      "loss": 1.476,
+      "step": 2791
+    },
+    {
+      "epoch": 0.4816699732597257,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.7341793030569874e-05,
+      "loss": 1.4452,
+      "step": 2792
+    },
+    {
+      "epoch": 0.4818424911584577,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.733994472360348e-05,
+      "loss": 1.5272,
+      "step": 2793
+    },
+    {
+      "epoch": 0.4820150090571897,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.733809587284812e-05,
+      "loss": 1.416,
+      "step": 2794
+    },
+    {
+      "epoch": 0.48218752695592165,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.733624647844076e-05,
+      "loss": 1.4474,
+      "step": 2795
+    },
+    {
+      "epoch": 0.48236004485465367,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.733439654051843e-05,
+      "loss": 1.5665,
+      "step": 2796
+    },
+    {
+      "epoch": 0.4825325627533857,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.7332546059218174e-05,
+      "loss": 1.4993,
+      "step": 2797
+    },
+    {
+      "epoch": 0.48270508065211765,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.733069503467709e-05,
+      "loss": 1.3398,
+      "step": 2798
+    },
+    {
+      "epoch": 0.48287759855084966,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.7328843467032314e-05,
+      "loss": 1.502,
+      "step": 2799
+    },
+    {
+      "epoch": 0.4830501164495816,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.7326991356421023e-05,
+      "loss": 1.5295,
+      "step": 2800
+    },
+    {
+      "epoch": 0.4830501164495816,
+      "eval_loss": 1.4480316638946533,
+      "eval_runtime": 10.8875,
+      "eval_samples_per_second": 94.053,
+      "eval_steps_per_second": 23.513,
+      "step": 2800
+    },
+    {
+      "epoch": 0.48322263434831364,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.7325138702980427e-05,
+      "loss": 1.4666,
+      "step": 2801
+    },
+    {
+      "epoch": 0.48339515224704566,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.7323285506847788e-05,
+      "loss": 1.4056,
+      "step": 2802
+    },
+    {
+      "epoch": 0.4835676701457776,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.7321431768160402e-05,
+      "loss": 1.4883,
+      "step": 2803
+    },
+    {
+      "epoch": 0.48374018804450963,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.7319577487055603e-05,
+      "loss": 1.4309,
+      "step": 2804
+    },
+    {
+      "epoch": 0.4839127059432416,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.7317722663670767e-05,
+      "loss": 1.5414,
+      "step": 2805
+    },
+    {
+      "epoch": 0.4840852238419736,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.731586729814332e-05,
+      "loss": 1.3881,
+      "step": 2806
+    },
+    {
+      "epoch": 0.48425774174070557,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.7314011390610705e-05,
+      "loss": 1.5173,
+      "step": 2807
+    },
+    {
+      "epoch": 0.4844302596394376,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.731215494121043e-05,
+      "loss": 1.3449,
+      "step": 2808
+    },
+    {
+      "epoch": 0.4846027775381696,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.7310297950080022e-05,
+      "loss": 1.4315,
+      "step": 2809
+    },
+    {
+      "epoch": 0.48477529543690157,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.730844041735707e-05,
+      "loss": 1.4923,
+      "step": 2810
+    },
+    {
+      "epoch": 0.4849478133356336,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.730658234317919e-05,
+      "loss": 1.5253,
+      "step": 2811
+    },
+    {
+      "epoch": 0.48512033123436554,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.7304723727684033e-05,
+      "loss": 1.5575,
+      "step": 2812
+    },
+    {
+      "epoch": 0.48529284913309756,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.7302864571009296e-05,
+      "loss": 1.4681,
+      "step": 2813
+    },
+    {
+      "epoch": 0.4854653670318296,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.7301004873292727e-05,
+      "loss": 1.5213,
+      "step": 2814
+    },
+    {
+      "epoch": 0.48563788493056154,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.7299144634672096e-05,
+      "loss": 1.4503,
+      "step": 2815
+    },
+    {
+      "epoch": 0.48581040282929355,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.729728385528522e-05,
+      "loss": 1.4475,
+      "step": 2816
+    },
+    {
+      "epoch": 0.4859829207280255,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.729542253526997e-05,
+      "loss": 1.507,
+      "step": 2817
+    },
+    {
+      "epoch": 0.48615543862675753,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.7293560674764224e-05,
+      "loss": 1.4533,
+      "step": 2818
+    },
+    {
+      "epoch": 0.48632795652548955,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.7291698273905936e-05,
+      "loss": 1.4052,
+      "step": 2819
+    },
+    {
+      "epoch": 0.4865004744242215,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.7289835332833083e-05,
+      "loss": 1.3862,
+      "step": 2820
+    },
+    {
+      "epoch": 0.4866729923229535,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.7287971851683675e-05,
+      "loss": 1.4409,
+      "step": 2821
+    },
+    {
+      "epoch": 0.4868455102216855,
+      "grad_norm": 0.625,
+      "learning_rate": 1.7286107830595774e-05,
+      "loss": 1.436,
+      "step": 2822
+    },
+    {
+      "epoch": 0.4870180281204175,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.7284243269707478e-05,
+      "loss": 1.4044,
+      "step": 2823
+    },
+    {
+      "epoch": 0.48719054601914946,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.728237816915693e-05,
+      "loss": 1.5035,
+      "step": 2824
+    },
+    {
+      "epoch": 0.4873630639178815,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.7280512529082306e-05,
+      "loss": 1.4622,
+      "step": 2825
+    },
+    {
+      "epoch": 0.4875355818166135,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.727864634962182e-05,
+      "loss": 1.5554,
+      "step": 2826
+    },
+    {
+      "epoch": 0.48770809971534546,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.7276779630913734e-05,
+      "loss": 1.5669,
+      "step": 2827
+    },
+    {
+      "epoch": 0.48788061761407747,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.7274912373096345e-05,
+      "loss": 1.3718,
+      "step": 2828
+    },
+    {
+      "epoch": 0.48805313551280943,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.7273044576307993e-05,
+      "loss": 1.3165,
+      "step": 2829
+    },
+    {
+      "epoch": 0.48822565341154145,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.7271176240687054e-05,
+      "loss": 1.455,
+      "step": 2830
+    },
+    {
+      "epoch": 0.48839817131027347,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.7269307366371948e-05,
+      "loss": 1.5077,
+      "step": 2831
+    },
+    {
+      "epoch": 0.4885706892090054,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.726743795350113e-05,
+      "loss": 1.4871,
+      "step": 2832
+    },
+    {
+      "epoch": 0.48874320710773744,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.7265568002213103e-05,
+      "loss": 1.5284,
+      "step": 2833
+    },
+    {
+      "epoch": 0.4889157250064694,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.7263697512646397e-05,
+      "loss": 1.5449,
+      "step": 2834
+    },
+    {
+      "epoch": 0.4890882429052014,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.7261826484939592e-05,
+      "loss": 1.4365,
+      "step": 2835
+    },
+    {
+      "epoch": 0.4892607608039334,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.725995491923131e-05,
+      "loss": 1.5388,
+      "step": 2836
+    },
+    {
+      "epoch": 0.4894332787026654,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.7258082815660203e-05,
+      "loss": 1.5065,
+      "step": 2837
+    },
+    {
+      "epoch": 0.4896057966013974,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.7256210174364975e-05,
+      "loss": 1.5679,
+      "step": 2838
+    },
+    {
+      "epoch": 0.4897783145001294,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.7254336995484355e-05,
+      "loss": 1.4383,
+      "step": 2839
+    },
+    {
+      "epoch": 0.4899508323988614,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.7252463279157123e-05,
+      "loss": 1.4321,
+      "step": 2840
+    },
+    {
+      "epoch": 0.49012335029759335,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.72505890255221e-05,
+      "loss": 1.47,
+      "step": 2841
+    },
+    {
+      "epoch": 0.49029586819632537,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.724871423471813e-05,
+      "loss": 1.4975,
+      "step": 2842
+    },
+    {
+      "epoch": 0.4904683860950574,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.7246838906884125e-05,
+      "loss": 1.4325,
+      "step": 2843
+    },
+    {
+      "epoch": 0.49064090399378935,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.724496304215901e-05,
+      "loss": 1.5435,
+      "step": 2844
+    },
+    {
+      "epoch": 0.49081342189252136,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.724308664068176e-05,
+      "loss": 1.4481,
+      "step": 2845
+    },
+    {
+      "epoch": 0.4909859397912533,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.7241209702591405e-05,
+      "loss": 1.4997,
+      "step": 2846
+    },
+    {
+      "epoch": 0.49115845768998534,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.7239332228026982e-05,
+      "loss": 1.3803,
+      "step": 2847
+    },
+    {
+      "epoch": 0.49133097558871736,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.72374542171276e-05,
+      "loss": 1.4905,
+      "step": 2848
+    },
+    {
+      "epoch": 0.4915034934874493,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.7235575670032382e-05,
+      "loss": 1.4539,
+      "step": 2849
+    },
+    {
+      "epoch": 0.49167601138618133,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.7233696586880513e-05,
+      "loss": 1.4779,
+      "step": 2850
+    },
+    {
+      "epoch": 0.4918485292849133,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.72318169678112e-05,
+      "loss": 1.4413,
+      "step": 2851
+    },
+    {
+      "epoch": 0.4920210471836453,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.7229936812963697e-05,
+      "loss": 1.5197,
+      "step": 2852
+    },
+    {
+      "epoch": 0.49219356508237727,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.7228056122477307e-05,
+      "loss": 1.5067,
+      "step": 2853
+    },
+    {
+      "epoch": 0.4923660829811093,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.7226174896491354e-05,
+      "loss": 1.5387,
+      "step": 2854
+    },
+    {
+      "epoch": 0.4925386008798413,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.7224293135145213e-05,
+      "loss": 1.5383,
+      "step": 2855
+    },
+    {
+      "epoch": 0.49271111877857326,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.722241083857829e-05,
+      "loss": 1.4639,
+      "step": 2856
+    },
+    {
+      "epoch": 0.4928836366773053,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.7220528006930056e-05,
+      "loss": 1.441,
+      "step": 2857
+    },
+    {
+      "epoch": 0.49305615457603724,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.7218644640339986e-05,
+      "loss": 1.5281,
+      "step": 2858
+    },
+    {
+      "epoch": 0.49322867247476926,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.7216760738947614e-05,
+      "loss": 1.4893,
+      "step": 2859
+    },
+    {
+      "epoch": 0.4934011903735013,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.721487630289252e-05,
+      "loss": 1.4929,
+      "step": 2860
+    },
+    {
+      "epoch": 0.49357370827223324,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.7212991332314303e-05,
+      "loss": 1.4151,
+      "step": 2861
+    },
+    {
+      "epoch": 0.49374622617096525,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.7211105827352624e-05,
+      "loss": 1.5097,
+      "step": 2862
+    },
+    {
+      "epoch": 0.4939187440696972,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.7209219788147166e-05,
+      "loss": 1.559,
+      "step": 2863
+    },
+    {
+      "epoch": 0.49409126196842923,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.720733321483766e-05,
+      "loss": 1.399,
+      "step": 2864
+    },
+    {
+      "epoch": 0.4942637798671612,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.7205446107563876e-05,
+      "loss": 1.4557,
+      "step": 2865
+    },
+    {
+      "epoch": 0.4944362977658932,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.7203558466465626e-05,
+      "loss": 1.5592,
+      "step": 2866
+    },
+    {
+      "epoch": 0.4946088156646252,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.7201670291682754e-05,
+      "loss": 1.4551,
+      "step": 2867
+    },
+    {
+      "epoch": 0.4947813335633572,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.7199781583355144e-05,
+      "loss": 1.5005,
+      "step": 2868
+    },
+    {
+      "epoch": 0.4949538514620892,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.719789234162273e-05,
+      "loss": 1.3738,
+      "step": 2869
+    },
+    {
+      "epoch": 0.49512636936082116,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.719600256662548e-05,
+      "loss": 1.5196,
+      "step": 2870
+    },
+    {
+      "epoch": 0.4952988872595532,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.7194112258503395e-05,
+      "loss": 1.4637,
+      "step": 2871
+    },
+    {
+      "epoch": 0.4954714051582852,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.7192221417396524e-05,
+      "loss": 1.5419,
+      "step": 2872
+    },
+    {
+      "epoch": 0.49564392305701716,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.7190330043444953e-05,
+      "loss": 1.5005,
+      "step": 2873
+    },
+    {
+      "epoch": 0.49581644095574917,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.71884381367888e-05,
+      "loss": 1.5498,
+      "step": 2874
+    },
+    {
+      "epoch": 0.49598895885448113,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.7186545697568236e-05,
+      "loss": 1.4867,
+      "step": 2875
+    },
+    {
+      "epoch": 0.49616147675321315,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.7184652725923465e-05,
+      "loss": 1.5072,
+      "step": 2876
+    },
+    {
+      "epoch": 0.49633399465194517,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.7182759221994727e-05,
+      "loss": 1.4238,
+      "step": 2877
+    },
+    {
+      "epoch": 0.4965065125506771,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.7180865185922307e-05,
+      "loss": 1.4501,
+      "step": 2878
+    },
+    {
+      "epoch": 0.49667903044940914,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.7178970617846524e-05,
+      "loss": 1.4374,
+      "step": 2879
+    },
+    {
+      "epoch": 0.4968515483481411,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.7177075517907745e-05,
+      "loss": 1.4367,
+      "step": 2880
+    },
+    {
+      "epoch": 0.4970240662468731,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.7175179886246365e-05,
+      "loss": 1.4153,
+      "step": 2881
+    },
+    {
+      "epoch": 0.4971965841456051,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.7173283723002825e-05,
+      "loss": 1.5074,
+      "step": 2882
+    },
+    {
+      "epoch": 0.4973691020443371,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.7171387028317606e-05,
+      "loss": 1.4974,
+      "step": 2883
+    },
+    {
+      "epoch": 0.4975416199430691,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.716948980233123e-05,
+      "loss": 1.4671,
+      "step": 2884
+    },
+    {
+      "epoch": 0.4977141378418011,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.716759204518425e-05,
+      "loss": 1.4103,
+      "step": 2885
+    },
+    {
+      "epoch": 0.4978866557405331,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.7165693757017267e-05,
+      "loss": 1.5011,
+      "step": 2886
+    },
+    {
+      "epoch": 0.49805917363926505,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.7163794937970916e-05,
+      "loss": 1.4511,
+      "step": 2887
+    },
+    {
+      "epoch": 0.49823169153799707,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.7161895588185878e-05,
+      "loss": 1.5568,
+      "step": 2888
+    },
+    {
+      "epoch": 0.4984042094367291,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.7159995707802863e-05,
+      "loss": 1.5091,
+      "step": 2889
+    },
+    {
+      "epoch": 0.49857672733546105,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.7158095296962627e-05,
+      "loss": 1.4798,
+      "step": 2890
+    },
+    {
+      "epoch": 0.49874924523419306,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.7156194355805968e-05,
+      "loss": 1.4575,
+      "step": 2891
+    },
+    {
+      "epoch": 0.498921763132925,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.7154292884473712e-05,
+      "loss": 1.5086,
+      "step": 2892
+    },
+    {
+      "epoch": 0.49909428103165704,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.7152390883106743e-05,
+      "loss": 1.4749,
+      "step": 2893
+    },
+    {
+      "epoch": 0.49926679893038906,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.7150488351845965e-05,
+      "loss": 1.5043,
+      "step": 2894
+    },
+    {
+      "epoch": 0.499439316829121,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.7148585290832333e-05,
+      "loss": 1.539,
+      "step": 2895
+    },
+    {
+      "epoch": 0.49961183472785303,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.7146681700206834e-05,
+      "loss": 1.434,
+      "step": 2896
+    },
+    {
+      "epoch": 0.499784352626585,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.71447775801105e-05,
+      "loss": 1.4453,
+      "step": 2897
+    },
+    {
+      "epoch": 0.499956870525317,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.71428729306844e-05,
+      "loss": 1.456,
+      "step": 2898
+    },
+    {
+      "epoch": 0.500129388424049,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.7140967752069645e-05,
+      "loss": 1.4775,
+      "step": 2899
+    },
+    {
+      "epoch": 0.500301906322781,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.713906204440738e-05,
+      "loss": 1.3852,
+      "step": 2900
+    },
+    {
+      "epoch": 0.500301906322781,
+      "eval_loss": 1.445255160331726,
+      "eval_runtime": 10.763,
+      "eval_samples_per_second": 95.141,
+      "eval_steps_per_second": 23.785,
+      "step": 2900
+    },
+    {
+      "epoch": 0.500474424221513,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.713715580783879e-05,
+      "loss": 1.5041,
+      "step": 2901
+    },
+    {
+      "epoch": 0.500646942120245,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.71352490425051e-05,
+      "loss": 1.4627,
+      "step": 2902
+    },
+    {
+      "epoch": 0.5008194600189769,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.7133341748547586e-05,
+      "loss": 1.503,
+      "step": 2903
+    },
+    {
+      "epoch": 0.500991977917709,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.7131433926107536e-05,
+      "loss": 1.4275,
+      "step": 2904
+    },
+    {
+      "epoch": 0.501164495816441,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.7129525575326307e-05,
+      "loss": 1.49,
+      "step": 2905
+    },
+    {
+      "epoch": 0.5013370137151729,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.7127616696345273e-05,
+      "loss": 1.5291,
+      "step": 2906
+    },
+    {
+      "epoch": 0.501509531613905,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.7125707289305862e-05,
+      "loss": 1.4875,
+      "step": 2907
+    },
+    {
+      "epoch": 0.501682049512637,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.7123797354349524e-05,
+      "loss": 1.4227,
+      "step": 2908
+    },
+    {
+      "epoch": 0.5018545674113689,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.7121886891617774e-05,
+      "loss": 1.4348,
+      "step": 2909
+    },
+    {
+      "epoch": 0.5020270853101009,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.711997590125214e-05,
+      "loss": 1.4808,
+      "step": 2910
+    },
+    {
+      "epoch": 0.502199603208833,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.7118064383394206e-05,
+      "loss": 1.4416,
+      "step": 2911
+    },
+    {
+      "epoch": 0.5023721211075649,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.7116152338185584e-05,
+      "loss": 1.4522,
+      "step": 2912
+    },
+    {
+      "epoch": 0.5025446390062969,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.711423976576794e-05,
+      "loss": 1.4111,
+      "step": 2913
+    },
+    {
+      "epoch": 0.5027171569050289,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.7112326666282953e-05,
+      "loss": 1.4947,
+      "step": 2914
+    },
+    {
+      "epoch": 0.5028896748037609,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.7110413039872372e-05,
+      "loss": 1.4555,
+      "step": 2915
+    },
+    {
+      "epoch": 0.5030621927024929,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.710849888667796e-05,
+      "loss": 1.5209,
+      "step": 2916
+    },
+    {
+      "epoch": 0.5032347106012248,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.710658420684154e-05,
+      "loss": 1.4655,
+      "step": 2917
+    },
+    {
+      "epoch": 0.5034072284999569,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.7104669000504955e-05,
+      "loss": 1.4579,
+      "step": 2918
+    },
+    {
+      "epoch": 0.5035797463986889,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.71027532678101e-05,
+      "loss": 1.5267,
+      "step": 2919
+    },
+    {
+      "epoch": 0.5037522642974208,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.7100837008898903e-05,
+      "loss": 1.484,
+      "step": 2920
+    },
+    {
+      "epoch": 0.5039247821961529,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.709892022391333e-05,
+      "loss": 1.4201,
+      "step": 2921
+    },
+    {
+      "epoch": 0.5040973000948848,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.7097002912995392e-05,
+      "loss": 1.4935,
+      "step": 2922
+    },
+    {
+      "epoch": 0.5042698179936168,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.7095085076287135e-05,
+      "loss": 1.5663,
+      "step": 2923
+    },
+    {
+      "epoch": 0.5044423358923489,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.7093166713930642e-05,
+      "loss": 1.4487,
+      "step": 2924
+    },
+    {
+      "epoch": 0.5046148537910808,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.709124782606804e-05,
+      "loss": 1.3968,
+      "step": 2925
+    },
+    {
+      "epoch": 0.5047873716898128,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.708932841284149e-05,
+      "loss": 1.3832,
+      "step": 2926
+    },
+    {
+      "epoch": 0.5049598895885448,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.7087408474393193e-05,
+      "loss": 1.4526,
+      "step": 2927
+    },
+    {
+      "epoch": 0.5051324074872768,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.7085488010865398e-05,
+      "loss": 1.4902,
+      "step": 2928
+    },
+    {
+      "epoch": 0.5053049253860088,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.7083567022400376e-05,
+      "loss": 1.4707,
+      "step": 2929
+    },
+    {
+      "epoch": 0.5054774432847408,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.7081645509140448e-05,
+      "loss": 1.491,
+      "step": 2930
+    },
+    {
+      "epoch": 0.5056499611834728,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.707972347122797e-05,
+      "loss": 1.4669,
+      "step": 2931
+    },
+    {
+      "epoch": 0.5058224790822048,
+      "grad_norm": 0.625,
+      "learning_rate": 1.7077800908805343e-05,
+      "loss": 1.4441,
+      "step": 2932
+    },
+    {
+      "epoch": 0.5059949969809368,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.7075877822015003e-05,
+      "loss": 1.4583,
+      "step": 2933
+    },
+    {
+      "epoch": 0.5061675148796687,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.707395421099942e-05,
+      "loss": 1.5226,
+      "step": 2934
+    },
+    {
+      "epoch": 0.5063400327784008,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.707203007590111e-05,
+      "loss": 1.49,
+      "step": 2935
+    },
+    {
+      "epoch": 0.5065125506771327,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.707010541686262e-05,
+      "loss": 1.479,
+      "step": 2936
+    },
+    {
+      "epoch": 0.5066850685758647,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.706818023402655e-05,
+      "loss": 1.4068,
+      "step": 2937
+    },
+    {
+      "epoch": 0.5068575864745968,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.7066254527535518e-05,
+      "loss": 1.4789,
+      "step": 2938
+    },
+    {
+      "epoch": 0.5070301043733287,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.7064328297532203e-05,
+      "loss": 1.5497,
+      "step": 2939
+    },
+    {
+      "epoch": 0.5072026222720607,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.706240154415931e-05,
+      "loss": 1.4872,
+      "step": 2940
+    },
+    {
+      "epoch": 0.5073751401707928,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.706047426755958e-05,
+      "loss": 1.4383,
+      "step": 2941
+    },
+    {
+      "epoch": 0.5075476580695247,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.70585464678758e-05,
+      "loss": 1.4348,
+      "step": 2942
+    },
+    {
+      "epoch": 0.5077201759682567,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.7056618145250797e-05,
+      "loss": 1.4141,
+      "step": 2943
+    },
+    {
+      "epoch": 0.5078926938669887,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.7054689299827425e-05,
+      "loss": 1.5521,
+      "step": 2944
+    },
+    {
+      "epoch": 0.5080652117657207,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.7052759931748593e-05,
+      "loss": 1.4734,
+      "step": 2945
+    },
+    {
+      "epoch": 0.5082377296644527,
+      "grad_norm": 0.625,
+      "learning_rate": 1.7050830041157234e-05,
+      "loss": 1.4657,
+      "step": 2946
+    },
+    {
+      "epoch": 0.5084102475631846,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.7048899628196337e-05,
+      "loss": 1.4567,
+      "step": 2947
+    },
+    {
+      "epoch": 0.5085827654619167,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.7046968693008903e-05,
+      "loss": 1.4203,
+      "step": 2948
+    },
+    {
+      "epoch": 0.5087552833606487,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.7045037235738005e-05,
+      "loss": 1.4839,
+      "step": 2949
+    },
+    {
+      "epoch": 0.5089278012593806,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.7043105256526723e-05,
+      "loss": 1.5014,
+      "step": 2950
+    },
+    {
+      "epoch": 0.5091003191581126,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.70411727555182e-05,
+      "loss": 1.4932,
+      "step": 2951
+    },
+    {
+      "epoch": 0.5092728370568447,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.70392397328556e-05,
+      "loss": 1.5016,
+      "step": 2952
+    },
+    {
+      "epoch": 0.5094453549555766,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.7037306188682142e-05,
+      "loss": 1.4502,
+      "step": 2953
+    },
+    {
+      "epoch": 0.5096178728543086,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.7035372123141067e-05,
+      "loss": 1.4817,
+      "step": 2954
+    },
+    {
+      "epoch": 0.5097903907530407,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.703343753637567e-05,
+      "loss": 1.4951,
+      "step": 2955
+    },
+    {
+      "epoch": 0.5099629086517726,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.703150242852927e-05,
+      "loss": 1.5164,
+      "step": 2956
+    },
+    {
+      "epoch": 0.5101354265505046,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.7029566799745233e-05,
+      "loss": 1.4479,
+      "step": 2957
+    },
+    {
+      "epoch": 0.5103079444492367,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.702763065016697e-05,
+      "loss": 1.4953,
+      "step": 2958
+    },
+    {
+      "epoch": 0.5104804623479686,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.7025693979937915e-05,
+      "loss": 1.5578,
+      "step": 2959
+    },
+    {
+      "epoch": 0.5106529802467006,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.7023756789201553e-05,
+      "loss": 1.4928,
+      "step": 2960
+    },
+    {
+      "epoch": 0.5108254981454325,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.70218190781014e-05,
+      "loss": 1.4897,
+      "step": 2961
+    },
+    {
+      "epoch": 0.5109980160441646,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.7019880846781017e-05,
+      "loss": 1.5874,
+      "step": 2962
+    },
+    {
+      "epoch": 0.5111705339428966,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.7017942095383997e-05,
+      "loss": 1.5494,
+      "step": 2963
+    },
+    {
+      "epoch": 0.5113430518416285,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.701600282405398e-05,
+      "loss": 1.3314,
+      "step": 2964
+    },
+    {
+      "epoch": 0.5115155697403606,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.7014063032934632e-05,
+      "loss": 1.3961,
+      "step": 2965
+    },
+    {
+      "epoch": 0.5116880876390926,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.701212272216967e-05,
+      "loss": 1.5077,
+      "step": 2966
+    },
+    {
+      "epoch": 0.5118606055378245,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.701018189190284e-05,
+      "loss": 1.4418,
+      "step": 2967
+    },
+    {
+      "epoch": 0.5120331234365565,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.7008240542277933e-05,
+      "loss": 1.3953,
+      "step": 2968
+    },
+    {
+      "epoch": 0.5122056413352886,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.7006298673438784e-05,
+      "loss": 1.4377,
+      "step": 2969
+    },
+    {
+      "epoch": 0.5123781592340205,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.7004356285529246e-05,
+      "loss": 1.4127,
+      "step": 2970
+    },
+    {
+      "epoch": 0.5125506771327525,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.7002413378693236e-05,
+      "loss": 1.5438,
+      "step": 2971
+    },
+    {
+      "epoch": 0.5127231950314846,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.7000469953074682e-05,
+      "loss": 1.4592,
+      "step": 2972
+    },
+    {
+      "epoch": 0.5128957129302165,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.6998526008817577e-05,
+      "loss": 1.4799,
+      "step": 2973
+    },
+    {
+      "epoch": 0.5130682308289485,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.6996581546065933e-05,
+      "loss": 1.5223,
+      "step": 2974
+    },
+    {
+      "epoch": 0.5132407487276806,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.6994636564963817e-05,
+      "loss": 1.4836,
+      "step": 2975
+    },
+    {
+      "epoch": 0.5134132666264125,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.6992691065655316e-05,
+      "loss": 1.3625,
+      "step": 2976
+    },
+    {
+      "epoch": 0.5135857845251445,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.699074504828457e-05,
+      "loss": 1.4768,
+      "step": 2977
+    },
+    {
+      "epoch": 0.5137583024238764,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.6988798512995747e-05,
+      "loss": 1.5272,
+      "step": 2978
+    },
+    {
+      "epoch": 0.5139308203226085,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.6986851459933067e-05,
+      "loss": 1.4577,
+      "step": 2979
+    },
+    {
+      "epoch": 0.5141033382213405,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.698490388924077e-05,
+      "loss": 1.5431,
+      "step": 2980
+    },
+    {
+      "epoch": 0.5142758561200724,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.6982955801063155e-05,
+      "loss": 1.469,
+      "step": 2981
+    },
+    {
+      "epoch": 0.5144483740188045,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.698100719554454e-05,
+      "loss": 1.4332,
+      "step": 2982
+    },
+    {
+      "epoch": 0.5146208919175365,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.6979058072829292e-05,
+      "loss": 1.552,
+      "step": 2983
+    },
+    {
+      "epoch": 0.5147934098162684,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.6977108433061812e-05,
+      "loss": 1.4409,
+      "step": 2984
+    },
+    {
+      "epoch": 0.5149659277150004,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.697515827638655e-05,
+      "loss": 1.4671,
+      "step": 2985
+    },
+    {
+      "epoch": 0.5151384456137325,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.6973207602947977e-05,
+      "loss": 1.5082,
+      "step": 2986
+    },
+    {
+      "epoch": 0.5153109635124644,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.697125641289062e-05,
+      "loss": 1.4577,
+      "step": 2987
+    },
+    {
+      "epoch": 0.5154834814111964,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.696930470635902e-05,
+      "loss": 1.4511,
+      "step": 2988
+    },
+    {
+      "epoch": 0.5156559993099284,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.6967352483497785e-05,
+      "loss": 1.4251,
+      "step": 2989
+    },
+    {
+      "epoch": 0.5158285172086604,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.6965399744451544e-05,
+      "loss": 1.3204,
+      "step": 2990
+    },
+    {
+      "epoch": 0.5160010351073924,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.696344648936497e-05,
+      "loss": 1.5144,
+      "step": 2991
+    },
+    {
+      "epoch": 0.5161735530061243,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.696149271838277e-05,
+      "loss": 1.4833,
+      "step": 2992
+    },
+    {
+      "epoch": 0.5163460709048564,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.695953843164969e-05,
+      "loss": 1.5172,
+      "step": 2993
+    },
+    {
+      "epoch": 0.5165185888035884,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.695758362931052e-05,
+      "loss": 1.6615,
+      "step": 2994
+    },
+    {
+      "epoch": 0.5166911067023203,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.695562831151008e-05,
+      "loss": 1.4564,
+      "step": 2995
+    },
+    {
+      "epoch": 0.5168636246010524,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.6953672478393235e-05,
+      "loss": 1.3986,
+      "step": 2996
+    },
+    {
+      "epoch": 0.5170361424997844,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.6951716130104884e-05,
+      "loss": 1.4119,
+      "step": 2997
+    },
+    {
+      "epoch": 0.5172086603985163,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.6949759266789963e-05,
+      "loss": 1.4871,
+      "step": 2998
+    },
+    {
+      "epoch": 0.5173811782972484,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.6947801888593452e-05,
+      "loss": 1.4817,
+      "step": 2999
+    },
+    {
+      "epoch": 0.5175536961959804,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.694584399566037e-05,
+      "loss": 1.4331,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5175536961959804,
+      "eval_loss": 1.4426935911178589,
+      "eval_runtime": 11.0486,
+      "eval_samples_per_second": 92.681,
+      "eval_steps_per_second": 23.17,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5177262140947123,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.6943885588135763e-05,
+      "loss": 1.4386,
+      "step": 3001
+    },
+    {
+      "epoch": 0.5178987319934443,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.694192666616472e-05,
+      "loss": 1.4496,
+      "step": 3002
+    },
+    {
+      "epoch": 0.5180712498921763,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.693996722989238e-05,
+      "loss": 1.4264,
+      "step": 3003
+    },
+    {
+      "epoch": 0.5182437677909083,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.6938007279463903e-05,
+      "loss": 1.5147,
+      "step": 3004
+    },
+    {
+      "epoch": 0.5184162856896403,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.6936046815024495e-05,
+      "loss": 1.5204,
+      "step": 3005
+    },
+    {
+      "epoch": 0.5185888035883723,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.69340858367194e-05,
+      "loss": 1.4578,
+      "step": 3006
+    },
+    {
+      "epoch": 0.5187613214871043,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.6932124344693904e-05,
+      "loss": 1.4739,
+      "step": 3007
+    },
+    {
+      "epoch": 0.5189338393858363,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.693016233909332e-05,
+      "loss": 1.4174,
+      "step": 3008
+    },
+    {
+      "epoch": 0.5191063572845682,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.692819982006301e-05,
+      "loss": 1.4694,
+      "step": 3009
+    },
+    {
+      "epoch": 0.5192788751833003,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.692623678774836e-05,
+      "loss": 1.4741,
+      "step": 3010
+    },
+    {
+      "epoch": 0.5194513930820323,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.692427324229482e-05,
+      "loss": 1.4905,
+      "step": 3011
+    },
+    {
+      "epoch": 0.5196239109807642,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.692230918384785e-05,
+      "loss": 1.4186,
+      "step": 3012
+    },
+    {
+      "epoch": 0.5197964288794963,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.6920344612552965e-05,
+      "loss": 1.4845,
+      "step": 3013
+    },
+    {
+      "epoch": 0.5199689467782282,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.691837952855571e-05,
+      "loss": 1.5879,
+      "step": 3014
+    },
+    {
+      "epoch": 0.5201414646769602,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.6916413932001673e-05,
+      "loss": 1.4893,
+      "step": 3015
+    },
+    {
+      "epoch": 0.5203139825756923,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.6914447823036474e-05,
+      "loss": 1.5236,
+      "step": 3016
+    },
+    {
+      "epoch": 0.5204865004744242,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.691248120180578e-05,
+      "loss": 1.5532,
+      "step": 3017
+    },
+    {
+      "epoch": 0.5206590183731562,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.691051406845528e-05,
+      "loss": 1.4801,
+      "step": 3018
+    },
+    {
+      "epoch": 0.5208315362718882,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.6908546423130726e-05,
+      "loss": 1.4341,
+      "step": 3019
+    },
+    {
+      "epoch": 0.5210040541706202,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.6906578265977885e-05,
+      "loss": 1.5411,
+      "step": 3020
+    },
+    {
+      "epoch": 0.5211765720693522,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.690460959714257e-05,
+      "loss": 1.4389,
+      "step": 3021
+    },
+    {
+      "epoch": 0.5213490899680842,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.6902640416770635e-05,
+      "loss": 1.5066,
+      "step": 3022
+    },
+    {
+      "epoch": 0.5215216078668162,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.6900670725007968e-05,
+      "loss": 1.4621,
+      "step": 3023
+    },
+    {
+      "epoch": 0.5216941257655482,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.6898700522000498e-05,
+      "loss": 1.4797,
+      "step": 3024
+    },
+    {
+      "epoch": 0.5218666436642801,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.6896729807894187e-05,
+      "loss": 1.5472,
+      "step": 3025
+    },
+    {
+      "epoch": 0.5220391615630121,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.6894758582835038e-05,
+      "loss": 1.4273,
+      "step": 3026
+    },
+    {
+      "epoch": 0.5222116794617442,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.6892786846969095e-05,
+      "loss": 1.4918,
+      "step": 3027
+    },
+    {
+      "epoch": 0.5223841973604761,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.689081460044243e-05,
+      "loss": 1.4427,
+      "step": 3028
+    },
+    {
+      "epoch": 0.5225567152592081,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.6888841843401164e-05,
+      "loss": 1.3589,
+      "step": 3029
+    },
+    {
+      "epoch": 0.5227292331579402,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.688686857599146e-05,
+      "loss": 1.3723,
+      "step": 3030
+    },
+    {
+      "epoch": 0.5229017510566721,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.688489479835949e-05,
+      "loss": 1.5148,
+      "step": 3031
+    },
+    {
+      "epoch": 0.5230742689554041,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.6882920510651497e-05,
+      "loss": 1.4609,
+      "step": 3032
+    },
+    {
+      "epoch": 0.5232467868541362,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.6880945713013747e-05,
+      "loss": 1.3951,
+      "step": 3033
+    },
+    {
+      "epoch": 0.5234193047528681,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.6878970405592545e-05,
+      "loss": 1.4334,
+      "step": 3034
+    },
+    {
+      "epoch": 0.5235918226516001,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.6876994588534234e-05,
+      "loss": 1.4421,
+      "step": 3035
+    },
+    {
+      "epoch": 0.523764340550332,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.6875018261985193e-05,
+      "loss": 1.4525,
+      "step": 3036
+    },
+    {
+      "epoch": 0.5239368584490641,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.6873041426091845e-05,
+      "loss": 1.42,
+      "step": 3037
+    },
+    {
+      "epoch": 0.5241093763477961,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.6871064081000643e-05,
+      "loss": 1.5279,
+      "step": 3038
+    },
+    {
+      "epoch": 0.524281894246528,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.686908622685808e-05,
+      "loss": 1.4593,
+      "step": 3039
+    },
+    {
+      "epoch": 0.5244544121452601,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.6867107863810687e-05,
+      "loss": 1.5149,
+      "step": 3040
+    },
+    {
+      "epoch": 0.5246269300439921,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.6865128992005037e-05,
+      "loss": 1.3865,
+      "step": 3041
+    },
+    {
+      "epoch": 0.524799447942724,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.6863149611587738e-05,
+      "loss": 1.4426,
+      "step": 3042
+    },
+    {
+      "epoch": 0.524971965841456,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.686116972270543e-05,
+      "loss": 1.4312,
+      "step": 3043
+    },
+    {
+      "epoch": 0.5251444837401881,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.68591893255048e-05,
+      "loss": 1.5589,
+      "step": 3044
+    },
+    {
+      "epoch": 0.52531700163892,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.685720842013257e-05,
+      "loss": 1.441,
+      "step": 3045
+    },
+    {
+      "epoch": 0.525489519537652,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.6855227006735492e-05,
+      "loss": 1.3209,
+      "step": 3046
+    },
+    {
+      "epoch": 0.5256620374363841,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.6853245085460362e-05,
+      "loss": 1.4729,
+      "step": 3047
+    },
+    {
+      "epoch": 0.525834555335116,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.6851262656454016e-05,
+      "loss": 1.5194,
+      "step": 3048
+    },
+    {
+      "epoch": 0.526007073233848,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.684927971986332e-05,
+      "loss": 1.5129,
+      "step": 3049
+    },
+    {
+      "epoch": 0.5261795911325801,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.684729627583519e-05,
+      "loss": 1.4712,
+      "step": 3050
+    },
+    {
+      "epoch": 0.526352109031312,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.684531232451657e-05,
+      "loss": 1.473,
+      "step": 3051
+    },
+    {
+      "epoch": 0.526524626930044,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.684332786605444e-05,
+      "loss": 1.4668,
+      "step": 3052
+    },
+    {
+      "epoch": 0.5266971448287759,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.684134290059582e-05,
+      "loss": 1.3984,
+      "step": 3053
+    },
+    {
+      "epoch": 0.526869662727508,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.6839357428287775e-05,
+      "loss": 1.447,
+      "step": 3054
+    },
+    {
+      "epoch": 0.52704218062624,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.6837371449277395e-05,
+      "loss": 1.4159,
+      "step": 3055
+    },
+    {
+      "epoch": 0.5272146985249719,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.6835384963711815e-05,
+      "loss": 1.5567,
+      "step": 3056
+    },
+    {
+      "epoch": 0.527387216423704,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.683339797173821e-05,
+      "loss": 1.5057,
+      "step": 3057
+    },
+    {
+      "epoch": 0.527559734322436,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.6831410473503788e-05,
+      "loss": 1.5606,
+      "step": 3058
+    },
+    {
+      "epoch": 0.5277322522211679,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.682942246915579e-05,
+      "loss": 1.4107,
+      "step": 3059
+    },
+    {
+      "epoch": 0.5279047701198999,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.6827433958841504e-05,
+      "loss": 1.3407,
+      "step": 3060
+    },
+    {
+      "epoch": 0.528077288018632,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.682544494270825e-05,
+      "loss": 1.5163,
+      "step": 3061
+    },
+    {
+      "epoch": 0.5282498059173639,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.682345542090339e-05,
+      "loss": 1.4255,
+      "step": 3062
+    },
+    {
+      "epoch": 0.5284223238160959,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.6821465393574315e-05,
+      "loss": 1.3823,
+      "step": 3063
+    },
+    {
+      "epoch": 0.528594841714828,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.681947486086846e-05,
+      "loss": 1.489,
+      "step": 3064
+    },
+    {
+      "epoch": 0.5287673596135599,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.6817483822933298e-05,
+      "loss": 1.3998,
+      "step": 3065
+    },
+    {
+      "epoch": 0.5289398775122919,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.681549227991634e-05,
+      "loss": 1.4426,
+      "step": 3066
+    },
+    {
+      "epoch": 0.5291123954110238,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.6813500231965127e-05,
+      "loss": 1.486,
+      "step": 3067
+    },
+    {
+      "epoch": 0.5292849133097559,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.6811507679227242e-05,
+      "loss": 1.4526,
+      "step": 3068
+    },
+    {
+      "epoch": 0.5294574312084879,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.6809514621850313e-05,
+      "loss": 1.4536,
+      "step": 3069
+    },
+    {
+      "epoch": 0.5296299491072198,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.680752105998199e-05,
+      "loss": 1.4325,
+      "step": 3070
+    },
+    {
+      "epoch": 0.5298024670059519,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.680552699376997e-05,
+      "loss": 1.3994,
+      "step": 3071
+    },
+    {
+      "epoch": 0.5299749849046839,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.6803532423361985e-05,
+      "loss": 1.516,
+      "step": 3072
+    },
+    {
+      "epoch": 0.5301475028034158,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.6801537348905813e-05,
+      "loss": 1.4652,
+      "step": 3073
+    },
+    {
+      "epoch": 0.5303200207021479,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.6799541770549256e-05,
+      "loss": 1.5399,
+      "step": 3074
+    },
+    {
+      "epoch": 0.5304925386008799,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.6797545688440156e-05,
+      "loss": 1.4818,
+      "step": 3075
+    },
+    {
+      "epoch": 0.5306650564996118,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.67955491027264e-05,
+      "loss": 1.4993,
+      "step": 3076
+    },
+    {
+      "epoch": 0.5308375743983438,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.6793552013555905e-05,
+      "loss": 1.3808,
+      "step": 3077
+    },
+    {
+      "epoch": 0.5310100922970759,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.679155442107663e-05,
+      "loss": 1.4844,
+      "step": 3078
+    },
+    {
+      "epoch": 0.5311826101958078,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.6789556325436566e-05,
+      "loss": 1.4248,
+      "step": 3079
+    },
+    {
+      "epoch": 0.5313551280945398,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.678755772678375e-05,
+      "loss": 1.4678,
+      "step": 3080
+    },
+    {
+      "epoch": 0.5315276459932718,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.6785558625266243e-05,
+      "loss": 1.4931,
+      "step": 3081
+    },
+    {
+      "epoch": 0.5317001638920038,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.6783559021032158e-05,
+      "loss": 1.3671,
+      "step": 3082
+    },
+    {
+      "epoch": 0.5318726817907358,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.678155891422963e-05,
+      "loss": 1.5186,
+      "step": 3083
+    },
+    {
+      "epoch": 0.5320451996894677,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.677955830500685e-05,
+      "loss": 1.3654,
+      "step": 3084
+    },
+    {
+      "epoch": 0.5322177175881998,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.6777557193512025e-05,
+      "loss": 1.4802,
+      "step": 3085
+    },
+    {
+      "epoch": 0.5323902354869318,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.6775555579893422e-05,
+      "loss": 1.5506,
+      "step": 3086
+    },
+    {
+      "epoch": 0.5325627533856637,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.677355346429932e-05,
+      "loss": 1.392,
+      "step": 3087
+    },
+    {
+      "epoch": 0.5327352712843958,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.6771550846878056e-05,
+      "loss": 1.5255,
+      "step": 3088
+    },
+    {
+      "epoch": 0.5329077891831278,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.6769547727777997e-05,
+      "loss": 1.4502,
+      "step": 3089
+    },
+    {
+      "epoch": 0.5330803070818597,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.6767544107147542e-05,
+      "loss": 1.4886,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5332528249805918,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.6765539985135134e-05,
+      "loss": 1.4446,
+      "step": 3091
+    },
+    {
+      "epoch": 0.5334253428793237,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.6763535361889252e-05,
+      "loss": 1.4988,
+      "step": 3092
+    },
+    {
+      "epoch": 0.5335978607780557,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.676153023755841e-05,
+      "loss": 1.4266,
+      "step": 3093
+    },
+    {
+      "epoch": 0.5337703786767877,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.675952461229116e-05,
+      "loss": 1.5449,
+      "step": 3094
+    },
+    {
+      "epoch": 0.5339428965755197,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.6757518486236088e-05,
+      "loss": 1.469,
+      "step": 3095
+    },
+    {
+      "epoch": 0.5341154144742517,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.6755511859541827e-05,
+      "loss": 1.5697,
+      "step": 3096
+    },
+    {
+      "epoch": 0.5342879323729837,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.6753504732357035e-05,
+      "loss": 1.4817,
+      "step": 3097
+    },
+    {
+      "epoch": 0.5344604502717157,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.6751497104830416e-05,
+      "loss": 1.5511,
+      "step": 3098
+    },
+    {
+      "epoch": 0.5346329681704477,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.6749488977110706e-05,
+      "loss": 1.4915,
+      "step": 3099
+    },
+    {
+      "epoch": 0.5348054860691797,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.6747480349346678e-05,
+      "loss": 1.4205,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5348054860691797,
+      "eval_loss": 1.4400311708450317,
+      "eval_runtime": 10.8978,
+      "eval_samples_per_second": 93.964,
+      "eval_steps_per_second": 23.491,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5349780039679116,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.674547122168715e-05,
+      "loss": 1.4605,
+      "step": 3101
+    },
+    {
+      "epoch": 0.5351505218666437,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.6743461594280962e-05,
+      "loss": 1.4807,
+      "step": 3102
+    },
+    {
+      "epoch": 0.5353230397653757,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.6741451467277006e-05,
+      "loss": 1.4194,
+      "step": 3103
+    },
+    {
+      "epoch": 0.5354955576641076,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.6739440840824203e-05,
+      "loss": 1.4102,
+      "step": 3104
+    },
+    {
+      "epoch": 0.5356680755628397,
+      "grad_norm": 0.625,
+      "learning_rate": 1.6737429715071512e-05,
+      "loss": 1.4654,
+      "step": 3105
+    },
+    {
+      "epoch": 0.5358405934615716,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.673541809016793e-05,
+      "loss": 1.3424,
+      "step": 3106
+    },
+    {
+      "epoch": 0.5360131113603036,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.673340596626249e-05,
+      "loss": 1.4412,
+      "step": 3107
+    },
+    {
+      "epoch": 0.5361856292590357,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.6731393343504266e-05,
+      "loss": 1.4353,
+      "step": 3108
+    },
+    {
+      "epoch": 0.5363581471577676,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.6729380222042363e-05,
+      "loss": 1.4199,
+      "step": 3109
+    },
+    {
+      "epoch": 0.5365306650564996,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.672736660202592e-05,
+      "loss": 1.4368,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5367031829552316,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.6725352483604133e-05,
+      "loss": 1.4803,
+      "step": 3111
+    },
+    {
+      "epoch": 0.5368757008539636,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.672333786692621e-05,
+      "loss": 1.5334,
+      "step": 3112
+    },
+    {
+      "epoch": 0.5370482187526956,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.6721322752141404e-05,
+      "loss": 1.412,
+      "step": 3113
+    },
+    {
+      "epoch": 0.5372207366514276,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.6719307139399018e-05,
+      "loss": 1.5448,
+      "step": 3114
+    },
+    {
+      "epoch": 0.5373932545501596,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.671729102884837e-05,
+      "loss": 1.3743,
+      "step": 3115
+    },
+    {
+      "epoch": 0.5375657724488916,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.6715274420638833e-05,
+      "loss": 1.3251,
+      "step": 3116
+    },
+    {
+      "epoch": 0.5377382903476235,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.671325731491981e-05,
+      "loss": 1.4829,
+      "step": 3117
+    },
+    {
+      "epoch": 0.5379108082463555,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.6711239711840736e-05,
+      "loss": 1.4443,
+      "step": 3118
+    },
+    {
+      "epoch": 0.5380833261450876,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.670922161155109e-05,
+      "loss": 1.5348,
+      "step": 3119
+    },
+    {
+      "epoch": 0.5382558440438195,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.670720301420039e-05,
+      "loss": 1.4214,
+      "step": 3120
+    },
+    {
+      "epoch": 0.5384283619425515,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.670518391993818e-05,
+      "loss": 1.5121,
+      "step": 3121
+    },
+    {
+      "epoch": 0.5386008798412836,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.6703164328914048e-05,
+      "loss": 1.4731,
+      "step": 3122
+    },
+    {
+      "epoch": 0.5387733977400155,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.670114424127762e-05,
+      "loss": 1.4324,
+      "step": 3123
+    },
+    {
+      "epoch": 0.5389459156387475,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.6699123657178553e-05,
+      "loss": 1.4381,
+      "step": 3124
+    },
+    {
+      "epoch": 0.5391184335374796,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.6697102576766552e-05,
+      "loss": 1.45,
+      "step": 3125
+    },
+    {
+      "epoch": 0.5392909514362115,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.6695081000191345e-05,
+      "loss": 1.414,
+      "step": 3126
+    },
+    {
+      "epoch": 0.5394634693349435,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.6693058927602704e-05,
+      "loss": 1.5434,
+      "step": 3127
+    },
+    {
+      "epoch": 0.5396359872336755,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.6691036359150435e-05,
+      "loss": 1.4602,
+      "step": 3128
+    },
+    {
+      "epoch": 0.5398085051324075,
+      "grad_norm": 0.625,
+      "learning_rate": 1.668901329498439e-05,
+      "loss": 1.5231,
+      "step": 3129
+    },
+    {
+      "epoch": 0.5399810230311395,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.6686989735254442e-05,
+      "loss": 1.3867,
+      "step": 3130
+    },
+    {
+      "epoch": 0.5401535409298714,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.6684965680110514e-05,
+      "loss": 1.4377,
+      "step": 3131
+    },
+    {
+      "epoch": 0.5403260588286035,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.668294112970256e-05,
+      "loss": 1.523,
+      "step": 3132
+    },
+    {
+      "epoch": 0.5404985767273355,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.6680916084180566e-05,
+      "loss": 1.5506,
+      "step": 3133
+    },
+    {
+      "epoch": 0.5406710946260674,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.667889054369457e-05,
+      "loss": 1.5098,
+      "step": 3134
+    },
+    {
+      "epoch": 0.5408436125247994,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.6676864508394624e-05,
+      "loss": 1.4797,
+      "step": 3135
+    },
+    {
+      "epoch": 0.5410161304235315,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.667483797843084e-05,
+      "loss": 1.5096,
+      "step": 3136
+    },
+    {
+      "epoch": 0.5411886483222634,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.6672810953953352e-05,
+      "loss": 1.5928,
+      "step": 3137
+    },
+    {
+      "epoch": 0.5413611662209954,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.6670783435112334e-05,
+      "loss": 1.4941,
+      "step": 3138
+    },
+    {
+      "epoch": 0.5415336841197275,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.6668755422058e-05,
+      "loss": 1.4992,
+      "step": 3139
+    },
+    {
+      "epoch": 0.5417062020184594,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.6666726914940594e-05,
+      "loss": 1.4193,
+      "step": 3140
+    },
+    {
+      "epoch": 0.5418787199171914,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.6664697913910405e-05,
+      "loss": 1.4341,
+      "step": 3141
+    },
+    {
+      "epoch": 0.5420512378159233,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.6662668419117748e-05,
+      "loss": 1.4427,
+      "step": 3142
+    },
+    {
+      "epoch": 0.5422237557146554,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.666063843071299e-05,
+      "loss": 1.4435,
+      "step": 3143
+    },
+    {
+      "epoch": 0.5423962736133874,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.6658607948846513e-05,
+      "loss": 1.4828,
+      "step": 3144
+    },
+    {
+      "epoch": 0.5425687915121193,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.6656576973668754e-05,
+      "loss": 1.5457,
+      "step": 3145
+    },
+    {
+      "epoch": 0.5427413094108514,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.6654545505330184e-05,
+      "loss": 1.374,
+      "step": 3146
+    },
+    {
+      "epoch": 0.5429138273095834,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.6652513543981302e-05,
+      "loss": 1.496,
+      "step": 3147
+    },
+    {
+      "epoch": 0.5430863452083153,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.6650481089772652e-05,
+      "loss": 1.4306,
+      "step": 3148
+    },
+    {
+      "epoch": 0.5432588631070474,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.664844814285481e-05,
+      "loss": 1.3942,
+      "step": 3149
+    },
+    {
+      "epoch": 0.5434313810057794,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.6646414703378387e-05,
+      "loss": 1.4703,
+      "step": 3150
+    },
+    {
+      "epoch": 0.5436038989045113,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.6644380771494032e-05,
+      "loss": 1.6057,
+      "step": 3151
+    },
+    {
+      "epoch": 0.5437764168032433,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.664234634735243e-05,
+      "loss": 1.3779,
+      "step": 3152
+    },
+    {
+      "epoch": 0.5439489347019754,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.6640311431104314e-05,
+      "loss": 1.4757,
+      "step": 3153
+    },
+    {
+      "epoch": 0.5441214526007073,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.6638276022900434e-05,
+      "loss": 1.4549,
+      "step": 3154
+    },
+    {
+      "epoch": 0.5442939704994393,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.6636240122891587e-05,
+      "loss": 1.4703,
+      "step": 3155
+    },
+    {
+      "epoch": 0.5444664883981714,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.663420373122861e-05,
+      "loss": 1.528,
+      "step": 3156
+    },
+    {
+      "epoch": 0.5446390062969033,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.6632166848062367e-05,
+      "loss": 1.4214,
+      "step": 3157
+    },
+    {
+      "epoch": 0.5448115241956353,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.6630129473543762e-05,
+      "loss": 1.5069,
+      "step": 3158
+    },
+    {
+      "epoch": 0.5449840420943672,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.662809160782374e-05,
+      "loss": 1.4371,
+      "step": 3159
+    },
+    {
+      "epoch": 0.5451565599930993,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.662605325105328e-05,
+      "loss": 1.5044,
+      "step": 3160
+    },
+    {
+      "epoch": 0.5453290778918313,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.662401440338339e-05,
+      "loss": 1.4541,
+      "step": 3161
+    },
+    {
+      "epoch": 0.5455015957905632,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.6621975064965122e-05,
+      "loss": 1.5276,
+      "step": 3162
+    },
+    {
+      "epoch": 0.5456741136892953,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.661993523594957e-05,
+      "loss": 1.5695,
+      "step": 3163
+    },
+    {
+      "epoch": 0.5458466315880273,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.6617894916487852e-05,
+      "loss": 1.5393,
+      "step": 3164
+    },
+    {
+      "epoch": 0.5460191494867592,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.6615854106731127e-05,
+      "loss": 1.4206,
+      "step": 3165
+    },
+    {
+      "epoch": 0.5461916673854913,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.661381280683059e-05,
+      "loss": 1.5095,
+      "step": 3166
+    },
+    {
+      "epoch": 0.5463641852842233,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.661177101693748e-05,
+      "loss": 1.5578,
+      "step": 3167
+    },
+    {
+      "epoch": 0.5465367031829552,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.6609728737203058e-05,
+      "loss": 1.5203,
+      "step": 3168
+    },
+    {
+      "epoch": 0.5467092210816872,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.660768596777863e-05,
+      "loss": 1.5172,
+      "step": 3169
+    },
+    {
+      "epoch": 0.5468817389804193,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.660564270881554e-05,
+      "loss": 1.4111,
+      "step": 3170
+    },
+    {
+      "epoch": 0.5470542568791512,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.6603598960465166e-05,
+      "loss": 1.4433,
+      "step": 3171
+    },
+    {
+      "epoch": 0.5472267747778832,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.660155472287892e-05,
+      "loss": 1.3998,
+      "step": 3172
+    },
+    {
+      "epoch": 0.5473992926766152,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.659950999620825e-05,
+      "loss": 1.4611,
+      "step": 3173
+    },
+    {
+      "epoch": 0.5475718105753472,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.6597464780604646e-05,
+      "loss": 1.4887,
+      "step": 3174
+    },
+    {
+      "epoch": 0.5477443284740792,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.6595419076219625e-05,
+      "loss": 1.5055,
+      "step": 3175
+    },
+    {
+      "epoch": 0.5479168463728111,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.659337288320475e-05,
+      "loss": 1.4551,
+      "step": 3176
+    },
+    {
+      "epoch": 0.5480893642715432,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.6591326201711612e-05,
+      "loss": 1.5612,
+      "step": 3177
+    },
+    {
+      "epoch": 0.5482618821702752,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.6589279031891847e-05,
+      "loss": 1.4727,
+      "step": 3178
+    },
+    {
+      "epoch": 0.5484344000690071,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.658723137389712e-05,
+      "loss": 1.5481,
+      "step": 3179
+    },
+    {
+      "epoch": 0.5486069179677392,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.6585183227879132e-05,
+      "loss": 1.4579,
+      "step": 3180
+    },
+    {
+      "epoch": 0.5487794358664712,
+      "grad_norm": 0.75,
+      "learning_rate": 1.6583134593989627e-05,
+      "loss": 1.466,
+      "step": 3181
+    },
+    {
+      "epoch": 0.5489519537652031,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.658108547238038e-05,
+      "loss": 1.4213,
+      "step": 3182
+    },
+    {
+      "epoch": 0.5491244716639352,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.6579035863203197e-05,
+      "loss": 1.457,
+      "step": 3183
+    },
+    {
+      "epoch": 0.5492969895626671,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.6576985766609926e-05,
+      "loss": 1.4649,
+      "step": 3184
+    },
+    {
+      "epoch": 0.5494695074613991,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.657493518275246e-05,
+      "loss": 1.426,
+      "step": 3185
+    },
+    {
+      "epoch": 0.5496420253601311,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.6572884111782716e-05,
+      "loss": 1.3793,
+      "step": 3186
+    },
+    {
+      "epoch": 0.5498145432588631,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.6570832553852643e-05,
+      "loss": 1.3756,
+      "step": 3187
+    },
+    {
+      "epoch": 0.5499870611575951,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.6568780509114237e-05,
+      "loss": 1.458,
+      "step": 3188
+    },
+    {
+      "epoch": 0.5501595790563271,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.656672797771953e-05,
+      "loss": 1.3458,
+      "step": 3189
+    },
+    {
+      "epoch": 0.5503320969550591,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.6564674959820585e-05,
+      "loss": 1.4256,
+      "step": 3190
+    },
+    {
+      "epoch": 0.5505046148537911,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.6562621455569495e-05,
+      "loss": 1.4807,
+      "step": 3191
+    },
+    {
+      "epoch": 0.5506771327525231,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.6560567465118407e-05,
+      "loss": 1.5037,
+      "step": 3192
+    },
+    {
+      "epoch": 0.550849650651255,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.655851298861949e-05,
+      "loss": 1.606,
+      "step": 3193
+    },
+    {
+      "epoch": 0.5510221685499871,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.6556458026224948e-05,
+      "loss": 1.4176,
+      "step": 3194
+    },
+    {
+      "epoch": 0.551194686448719,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.655440257808703e-05,
+      "loss": 1.3671,
+      "step": 3195
+    },
+    {
+      "epoch": 0.551367204347451,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.6552346644358014e-05,
+      "loss": 1.4281,
+      "step": 3196
+    },
+    {
+      "epoch": 0.5515397222461831,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.6550290225190217e-05,
+      "loss": 1.3226,
+      "step": 3197
+    },
+    {
+      "epoch": 0.551712240144915,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.6548233320735997e-05,
+      "loss": 1.2947,
+      "step": 3198
+    },
+    {
+      "epoch": 0.551884758043647,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.6546175931147733e-05,
+      "loss": 1.4838,
+      "step": 3199
+    },
+    {
+      "epoch": 0.552057275942379,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.6544118056577856e-05,
+      "loss": 1.3835,
+      "step": 3200
+    },
+    {
+      "epoch": 0.552057275942379,
+      "eval_loss": 1.4378929138183594,
+      "eval_runtime": 11.3818,
+      "eval_samples_per_second": 89.968,
+      "eval_steps_per_second": 22.492,
+      "step": 3200
+    },
+    {
+      "epoch": 0.552229793841111,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.6542059697178822e-05,
+      "loss": 1.5091,
+      "step": 3201
+    },
+    {
+      "epoch": 0.552402311739843,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.6540000853103132e-05,
+      "loss": 1.4104,
+      "step": 3202
+    },
+    {
+      "epoch": 0.552574829638575,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.653794152450331e-05,
+      "loss": 1.3491,
+      "step": 3203
+    },
+    {
+      "epoch": 0.552747347537307,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.653588171153193e-05,
+      "loss": 1.4691,
+      "step": 3204
+    },
+    {
+      "epoch": 0.552919865436039,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.6533821414341597e-05,
+      "loss": 1.473,
+      "step": 3205
+    },
+    {
+      "epoch": 0.553092383334771,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.6531760633084948e-05,
+      "loss": 1.5411,
+      "step": 3206
+    },
+    {
+      "epoch": 0.553264901233503,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.652969936791466e-05,
+      "loss": 1.4346,
+      "step": 3207
+    },
+    {
+      "epoch": 0.553437419132235,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.6527637618983443e-05,
+      "loss": 1.4768,
+      "step": 3208
+    },
+    {
+      "epoch": 0.553609937030967,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.6525575386444042e-05,
+      "loss": 1.4556,
+      "step": 3209
+    },
+    {
+      "epoch": 0.5537824549296989,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.6523512670449246e-05,
+      "loss": 1.4614,
+      "step": 3210
+    },
+    {
+      "epoch": 0.553954972828431,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.6521449471151867e-05,
+      "loss": 1.3232,
+      "step": 3211
+    },
+    {
+      "epoch": 0.5541274907271629,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.6519385788704766e-05,
+      "loss": 1.2933,
+      "step": 3212
+    },
+    {
+      "epoch": 0.5543000086258949,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.6517321623260828e-05,
+      "loss": 1.4747,
+      "step": 3213
+    },
+    {
+      "epoch": 0.554472526524627,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.6515256974972985e-05,
+      "loss": 1.4267,
+      "step": 3214
+    },
+    {
+      "epoch": 0.5546450444233589,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.6513191843994195e-05,
+      "loss": 1.4805,
+      "step": 3215
+    },
+    {
+      "epoch": 0.5548175623220909,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.6511126230477458e-05,
+      "loss": 1.374,
+      "step": 3216
+    },
+    {
+      "epoch": 0.5549900802208229,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.6509060134575807e-05,
+      "loss": 1.4438,
+      "step": 3217
+    },
+    {
+      "epoch": 0.5551625981195549,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.6506993556442307e-05,
+      "loss": 1.3694,
+      "step": 3218
+    },
+    {
+      "epoch": 0.5553351160182869,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.650492649623007e-05,
+      "loss": 1.4282,
+      "step": 3219
+    },
+    {
+      "epoch": 0.5555076339170189,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.6502858954092233e-05,
+      "loss": 1.4742,
+      "step": 3220
+    },
+    {
+      "epoch": 0.5556801518157509,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.6500790930181973e-05,
+      "loss": 1.4559,
+      "step": 3221
+    },
+    {
+      "epoch": 0.5558526697144829,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.64987224246525e-05,
+      "loss": 1.474,
+      "step": 3222
+    },
+    {
+      "epoch": 0.5560251876132148,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.6496653437657068e-05,
+      "loss": 1.5381,
+      "step": 3223
+    },
+    {
+      "epoch": 0.5561977055119469,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.6494583969348953e-05,
+      "loss": 1.4106,
+      "step": 3224
+    },
+    {
+      "epoch": 0.5563702234106789,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.6492514019881476e-05,
+      "loss": 1.5662,
+      "step": 3225
+    },
+    {
+      "epoch": 0.5565427413094108,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.6490443589407996e-05,
+      "loss": 1.4847,
+      "step": 3226
+    },
+    {
+      "epoch": 0.5567152592081428,
+      "grad_norm": 0.625,
+      "learning_rate": 1.6488372678081902e-05,
+      "loss": 1.4211,
+      "step": 3227
+    },
+    {
+      "epoch": 0.5568877771068749,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.6486301286056617e-05,
+      "loss": 1.4936,
+      "step": 3228
+    },
+    {
+      "epoch": 0.5570602950056068,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.64842294134856e-05,
+      "loss": 1.3648,
+      "step": 3229
+    },
+    {
+      "epoch": 0.5572328129043388,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.6482157060522363e-05,
+      "loss": 1.5151,
+      "step": 3230
+    },
+    {
+      "epoch": 0.5574053308030709,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.6480084227320422e-05,
+      "loss": 1.5417,
+      "step": 3231
+    },
+    {
+      "epoch": 0.5575778487018028,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.647801091403335e-05,
+      "loss": 1.4683,
+      "step": 3232
+    },
+    {
+      "epoch": 0.5577503666005348,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.647593712081476e-05,
+      "loss": 1.5107,
+      "step": 3233
+    },
+    {
+      "epoch": 0.5579228844992667,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.647386284781828e-05,
+      "loss": 1.398,
+      "step": 3234
+    },
+    {
+      "epoch": 0.5580954023979988,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.6471788095197587e-05,
+      "loss": 1.4902,
+      "step": 3235
+    },
+    {
+      "epoch": 0.5582679202967308,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.64697128631064e-05,
+      "loss": 1.457,
+      "step": 3236
+    },
+    {
+      "epoch": 0.5584404381954627,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.6467637151698457e-05,
+      "loss": 1.6169,
+      "step": 3237
+    },
+    {
+      "epoch": 0.5586129560941948,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.646556096112754e-05,
+      "loss": 1.5738,
+      "step": 3238
+    },
+    {
+      "epoch": 0.5587854739929268,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.6463484291547472e-05,
+      "loss": 1.3692,
+      "step": 3239
+    },
+    {
+      "epoch": 0.5589579918916587,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.64614071431121e-05,
+      "loss": 1.4634,
+      "step": 3240
+    },
+    {
+      "epoch": 0.5591305097903908,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.6459329515975313e-05,
+      "loss": 1.4855,
+      "step": 3241
+    },
+    {
+      "epoch": 0.5593030276891228,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.645725141029104e-05,
+      "loss": 1.4872,
+      "step": 3242
+    },
+    {
+      "epoch": 0.5594755455878547,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.6455172826213228e-05,
+      "loss": 1.5038,
+      "step": 3243
+    },
+    {
+      "epoch": 0.5596480634865867,
+      "grad_norm": 0.75,
+      "learning_rate": 1.6453093763895885e-05,
+      "loss": 1.4403,
+      "step": 3244
+    },
+    {
+      "epoch": 0.5598205813853188,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.6451014223493035e-05,
+      "loss": 1.3938,
+      "step": 3245
+    },
+    {
+      "epoch": 0.5599930992840507,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.6448934205158743e-05,
+      "loss": 1.4209,
+      "step": 3246
+    },
+    {
+      "epoch": 0.5601656171827827,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.644685370904711e-05,
+      "loss": 1.4142,
+      "step": 3247
+    },
+    {
+      "epoch": 0.5603381350815148,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.6444772735312272e-05,
+      "loss": 1.508,
+      "step": 3248
+    },
+    {
+      "epoch": 0.5605106529802467,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.6442691284108403e-05,
+      "loss": 1.4892,
+      "step": 3249
+    },
+    {
+      "epoch": 0.5606831708789787,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.6440609355589704e-05,
+      "loss": 1.4895,
+      "step": 3250
+    },
+    {
+      "epoch": 0.5608556887777106,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.6438526949910425e-05,
+      "loss": 1.4742,
+      "step": 3251
+    },
+    {
+      "epoch": 0.5610282066764427,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.643644406722484e-05,
+      "loss": 1.5226,
+      "step": 3252
+    },
+    {
+      "epoch": 0.5612007245751747,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.643436070768726e-05,
+      "loss": 1.4427,
+      "step": 3253
+    },
+    {
+      "epoch": 0.5613732424739066,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.6432276871452036e-05,
+      "loss": 1.4517,
+      "step": 3254
+    },
+    {
+      "epoch": 0.5615457603726387,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.643019255867355e-05,
+      "loss": 1.4601,
+      "step": 3255
+    },
+    {
+      "epoch": 0.5617182782713707,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.642810776950622e-05,
+      "loss": 1.5211,
+      "step": 3256
+    },
+    {
+      "epoch": 0.5618907961701026,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.6426022504104505e-05,
+      "loss": 1.5343,
+      "step": 3257
+    },
+    {
+      "epoch": 0.5620633140688347,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.6423936762622887e-05,
+      "loss": 1.4592,
+      "step": 3258
+    },
+    {
+      "epoch": 0.5622358319675667,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.6421850545215897e-05,
+      "loss": 1.4003,
+      "step": 3259
+    },
+    {
+      "epoch": 0.5624083498662986,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.6419763852038092e-05,
+      "loss": 1.4569,
+      "step": 3260
+    },
+    {
+      "epoch": 0.5625808677650306,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.641767668324407e-05,
+      "loss": 1.4049,
+      "step": 3261
+    },
+    {
+      "epoch": 0.5627533856637627,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.641558903898846e-05,
+      "loss": 1.4446,
+      "step": 3262
+    },
+    {
+      "epoch": 0.5629259035624946,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.6413500919425927e-05,
+      "loss": 1.481,
+      "step": 3263
+    },
+    {
+      "epoch": 0.5630984214612266,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.6411412324711174e-05,
+      "loss": 1.5068,
+      "step": 3264
+    },
+    {
+      "epoch": 0.5632709393599586,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.6409323254998932e-05,
+      "loss": 1.4457,
+      "step": 3265
+    },
+    {
+      "epoch": 0.5634434572586906,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.640723371044398e-05,
+      "loss": 1.4314,
+      "step": 3266
+    },
+    {
+      "epoch": 0.5636159751574226,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.640514369120112e-05,
+      "loss": 1.3798,
+      "step": 3267
+    },
+    {
+      "epoch": 0.5637884930561545,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.640305319742519e-05,
+      "loss": 1.5154,
+      "step": 3268
+    },
+    {
+      "epoch": 0.5639610109548866,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.6400962229271074e-05,
+      "loss": 1.3999,
+      "step": 3269
+    },
+    {
+      "epoch": 0.5641335288536186,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.639887078689368e-05,
+      "loss": 1.473,
+      "step": 3270
+    },
+    {
+      "epoch": 0.5643060467523505,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.639677887044796e-05,
+      "loss": 1.3974,
+      "step": 3271
+    },
+    {
+      "epoch": 0.5644785646510826,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.639468648008889e-05,
+      "loss": 1.5242,
+      "step": 3272
+    },
+    {
+      "epoch": 0.5646510825498146,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.6392593615971487e-05,
+      "loss": 1.5228,
+      "step": 3273
+    },
+    {
+      "epoch": 0.5648236004485465,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.6390500278250806e-05,
+      "loss": 1.4572,
+      "step": 3274
+    },
+    {
+      "epoch": 0.5649961183472785,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.6388406467081936e-05,
+      "loss": 1.5359,
+      "step": 3275
+    },
+    {
+      "epoch": 0.5651686362460105,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.6386312182619997e-05,
+      "loss": 1.4999,
+      "step": 3276
+    },
+    {
+      "epoch": 0.5653411541447425,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.638421742502015e-05,
+      "loss": 1.4103,
+      "step": 3277
+    },
+    {
+      "epoch": 0.5655136720434745,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.638212219443759e-05,
+      "loss": 1.3701,
+      "step": 3278
+    },
+    {
+      "epoch": 0.5656861899422065,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.6380026491027537e-05,
+      "loss": 1.5241,
+      "step": 3279
+    },
+    {
+      "epoch": 0.5658587078409385,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.6377930314945254e-05,
+      "loss": 1.45,
+      "step": 3280
+    },
+    {
+      "epoch": 0.5660312257396705,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.6375833666346048e-05,
+      "loss": 1.4207,
+      "step": 3281
+    },
+    {
+      "epoch": 0.5662037436384025,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.637373654538524e-05,
+      "loss": 1.5209,
+      "step": 3282
+    },
+    {
+      "epoch": 0.5663762615371345,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.637163895221821e-05,
+      "loss": 1.5028,
+      "step": 3283
+    },
+    {
+      "epoch": 0.5665487794358665,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.636954088700035e-05,
+      "loss": 1.4265,
+      "step": 3284
+    },
+    {
+      "epoch": 0.5667212973345984,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.63674423498871e-05,
+      "loss": 1.4944,
+      "step": 3285
+    },
+    {
+      "epoch": 0.5668938152333305,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.636534334103394e-05,
+      "loss": 1.4467,
+      "step": 3286
+    },
+    {
+      "epoch": 0.5670663331320625,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.636324386059637e-05,
+      "loss": 1.4922,
+      "step": 3287
+    },
+    {
+      "epoch": 0.5672388510307944,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.636114390872994e-05,
+      "loss": 1.3776,
+      "step": 3288
+    },
+    {
+      "epoch": 0.5674113689295265,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.6359043485590217e-05,
+      "loss": 1.5595,
+      "step": 3289
+    },
+    {
+      "epoch": 0.5675838868282584,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.635694259133282e-05,
+      "loss": 1.552,
+      "step": 3290
+    },
+    {
+      "epoch": 0.5677564047269904,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.63548412261134e-05,
+      "loss": 1.4487,
+      "step": 3291
+    },
+    {
+      "epoch": 0.5679289226257224,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.635273939008763e-05,
+      "loss": 1.4105,
+      "step": 3292
+    },
+    {
+      "epoch": 0.5681014405244544,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.6350637083411238e-05,
+      "loss": 1.4208,
+      "step": 3293
+    },
+    {
+      "epoch": 0.5682739584231864,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.634853430623997e-05,
+      "loss": 1.4819,
+      "step": 3294
+    },
+    {
+      "epoch": 0.5684464763219184,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.6346431058729607e-05,
+      "loss": 1.4907,
+      "step": 3295
+    },
+    {
+      "epoch": 0.5686189942206504,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.6344327341035982e-05,
+      "loss": 1.4593,
+      "step": 3296
+    },
+    {
+      "epoch": 0.5687915121193824,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.6342223153314946e-05,
+      "loss": 1.4974,
+      "step": 3297
+    },
+    {
+      "epoch": 0.5689640300181144,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.634011849572239e-05,
+      "loss": 1.52,
+      "step": 3298
+    },
+    {
+      "epoch": 0.5691365479168464,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.6338013368414237e-05,
+      "loss": 1.4899,
+      "step": 3299
+    },
+    {
+      "epoch": 0.5693090658155784,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.6335907771546458e-05,
+      "loss": 1.4335,
+      "step": 3300
+    },
+    {
+      "epoch": 0.5693090658155784,
+      "eval_loss": 1.4357802867889404,
+      "eval_runtime": 10.7789,
+      "eval_samples_per_second": 95.0,
+      "eval_steps_per_second": 23.75,
+      "step": 3300
+    },
+    {
+      "epoch": 0.5694815837143103,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.633380170527504e-05,
+      "loss": 1.4017,
+      "step": 3301
+    },
+    {
+      "epoch": 0.5696541016130423,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.6331695169756015e-05,
+      "loss": 1.4695,
+      "step": 3302
+    },
+    {
+      "epoch": 0.5698266195117744,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.6329588165145452e-05,
+      "loss": 1.4122,
+      "step": 3303
+    },
+    {
+      "epoch": 0.5699991374105063,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.6327480691599448e-05,
+      "loss": 1.4928,
+      "step": 3304
+    },
+    {
+      "epoch": 0.5701716553092383,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.632537274927414e-05,
+      "loss": 1.3496,
+      "step": 3305
+    },
+    {
+      "epoch": 0.5703441732079704,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.632326433832569e-05,
+      "loss": 1.4946,
+      "step": 3306
+    },
+    {
+      "epoch": 0.5705166911067023,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.6321155458910314e-05,
+      "loss": 1.4751,
+      "step": 3307
+    },
+    {
+      "epoch": 0.5706892090054343,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.631904611118424e-05,
+      "loss": 1.5496,
+      "step": 3308
+    },
+    {
+      "epoch": 0.5708617269041663,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.6316936295303754e-05,
+      "loss": 1.4039,
+      "step": 3309
+    },
+    {
+      "epoch": 0.5710342448028983,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.6314826011425153e-05,
+      "loss": 1.5043,
+      "step": 3310
+    },
+    {
+      "epoch": 0.5712067627016303,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.6312715259704786e-05,
+      "loss": 1.5031,
+      "step": 3311
+    },
+    {
+      "epoch": 0.5713792806003622,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.6310604040299026e-05,
+      "loss": 1.4799,
+      "step": 3312
+    },
+    {
+      "epoch": 0.5715517984990943,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.630849235336429e-05,
+      "loss": 1.4857,
+      "step": 3313
+    },
+    {
+      "epoch": 0.5717243163978263,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.6306380199057024e-05,
+      "loss": 1.4491,
+      "step": 3314
+    },
+    {
+      "epoch": 0.5718968342965582,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.6304267577533706e-05,
+      "loss": 1.4579,
+      "step": 3315
+    },
+    {
+      "epoch": 0.5720693521952903,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.630215448895086e-05,
+      "loss": 1.4994,
+      "step": 3316
+    },
+    {
+      "epoch": 0.5722418700940223,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.6300040933465025e-05,
+      "loss": 1.4655,
+      "step": 3317
+    },
+    {
+      "epoch": 0.5724143879927542,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.6297926911232796e-05,
+      "loss": 1.4915,
+      "step": 3318
+    },
+    {
+      "epoch": 0.5725869058914862,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.6295812422410794e-05,
+      "loss": 1.5393,
+      "step": 3319
+    },
+    {
+      "epoch": 0.5727594237902183,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.6293697467155667e-05,
+      "loss": 1.5414,
+      "step": 3320
+    },
+    {
+      "epoch": 0.5729319416889502,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.6291582045624107e-05,
+      "loss": 1.4912,
+      "step": 3321
+    },
+    {
+      "epoch": 0.5731044595876822,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.6289466157972835e-05,
+      "loss": 1.3418,
+      "step": 3322
+    },
+    {
+      "epoch": 0.5732769774864143,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.628734980435861e-05,
+      "loss": 1.5682,
+      "step": 3323
+    },
+    {
+      "epoch": 0.5734494953851462,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.6285232984938234e-05,
+      "loss": 1.4293,
+      "step": 3324
+    },
+    {
+      "epoch": 0.5736220132838782,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.6283115699868522e-05,
+      "loss": 1.4405,
+      "step": 3325
+    },
+    {
+      "epoch": 0.5737945311826101,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.6280997949306342e-05,
+      "loss": 1.3624,
+      "step": 3326
+    },
+    {
+      "epoch": 0.5739670490813422,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.6278879733408587e-05,
+      "loss": 1.4126,
+      "step": 3327
+    },
+    {
+      "epoch": 0.5741395669800742,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.627676105233219e-05,
+      "loss": 1.429,
+      "step": 3328
+    },
+    {
+      "epoch": 0.5743120848788061,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.6274641906234113e-05,
+      "loss": 1.4055,
+      "step": 3329
+    },
+    {
+      "epoch": 0.5744846027775382,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.627252229527136e-05,
+      "loss": 1.4173,
+      "step": 3330
+    },
+    {
+      "epoch": 0.5746571206762702,
+      "grad_norm": 2.0,
+      "learning_rate": 1.627040221960096e-05,
+      "loss": 1.5033,
+      "step": 3331
+    },
+    {
+      "epoch": 0.5748296385750021,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.6268281679379983e-05,
+      "loss": 1.4267,
+      "step": 3332
+    },
+    {
+      "epoch": 0.5750021564737342,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.6266160674765542e-05,
+      "loss": 1.4609,
+      "step": 3333
+    },
+    {
+      "epoch": 0.5751746743724662,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.626403920591476e-05,
+      "loss": 1.4585,
+      "step": 3334
+    },
+    {
+      "epoch": 0.5753471922711981,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.626191727298481e-05,
+      "loss": 1.4806,
+      "step": 3335
+    },
+    {
+      "epoch": 0.5755197101699301,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.625979487613291e-05,
+      "loss": 1.4952,
+      "step": 3336
+    },
+    {
+      "epoch": 0.5756922280686622,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.6257672015516287e-05,
+      "loss": 1.463,
+      "step": 3337
+    },
+    {
+      "epoch": 0.5758647459673941,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.6255548691292223e-05,
+      "loss": 1.4367,
+      "step": 3338
+    },
+    {
+      "epoch": 0.5760372638661261,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.6253424903618023e-05,
+      "loss": 1.4492,
+      "step": 3339
+    },
+    {
+      "epoch": 0.5762097817648582,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.6251300652651037e-05,
+      "loss": 1.4922,
+      "step": 3340
+    },
+    {
+      "epoch": 0.5763822996635901,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.6249175938548635e-05,
+      "loss": 1.4813,
+      "step": 3341
+    },
+    {
+      "epoch": 0.5765548175623221,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.6247050761468236e-05,
+      "loss": 1.4617,
+      "step": 3342
+    },
+    {
+      "epoch": 0.576727335461054,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.6244925121567283e-05,
+      "loss": 1.4464,
+      "step": 3343
+    },
+    {
+      "epoch": 0.5768998533597861,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.6242799019003256e-05,
+      "loss": 1.4271,
+      "step": 3344
+    },
+    {
+      "epoch": 0.5770723712585181,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.6240672453933673e-05,
+      "loss": 1.4269,
+      "step": 3345
+    },
+    {
+      "epoch": 0.57724488915725,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.623854542651608e-05,
+      "loss": 1.326,
+      "step": 3346
+    },
+    {
+      "epoch": 0.5774174070559821,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.6236417936908058e-05,
+      "loss": 1.4414,
+      "step": 3347
+    },
+    {
+      "epoch": 0.5775899249547141,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.623428998526723e-05,
+      "loss": 1.5371,
+      "step": 3348
+    },
+    {
+      "epoch": 0.577762442853446,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.6232161571751248e-05,
+      "loss": 1.5019,
+      "step": 3349
+    },
+    {
+      "epoch": 0.577934960752178,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.6230032696517794e-05,
+      "loss": 1.475,
+      "step": 3350
+    },
+    {
+      "epoch": 0.5781074786509101,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.6227903359724595e-05,
+      "loss": 1.4802,
+      "step": 3351
+    },
+    {
+      "epoch": 0.578279996549642,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.62257735615294e-05,
+      "loss": 1.4384,
+      "step": 3352
+    },
+    {
+      "epoch": 0.578452514448374,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.6223643302089994e-05,
+      "loss": 1.4694,
+      "step": 3353
+    },
+    {
+      "epoch": 0.578625032347106,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.622151258156421e-05,
+      "loss": 1.4337,
+      "step": 3354
+    },
+    {
+      "epoch": 0.578797550245838,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.6219381400109898e-05,
+      "loss": 1.6329,
+      "step": 3355
+    },
+    {
+      "epoch": 0.57897006814457,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.6217249757884954e-05,
+      "loss": 1.5241,
+      "step": 3356
+    },
+    {
+      "epoch": 0.579142586043302,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.62151176550473e-05,
+      "loss": 1.4375,
+      "step": 3357
+    },
+    {
+      "epoch": 0.579315103942034,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.6212985091754893e-05,
+      "loss": 1.4399,
+      "step": 3358
+    },
+    {
+      "epoch": 0.579487621840766,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.621085206816573e-05,
+      "loss": 1.5598,
+      "step": 3359
+    },
+    {
+      "epoch": 0.5796601397394979,
+      "grad_norm": 2.375,
+      "learning_rate": 1.6208718584437845e-05,
+      "loss": 1.465,
+      "step": 3360
+    },
+    {
+      "epoch": 0.57983265763823,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.620658464072929e-05,
+      "loss": 1.4114,
+      "step": 3361
+    },
+    {
+      "epoch": 0.580005175536962,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.620445023719816e-05,
+      "loss": 1.4664,
+      "step": 3362
+    },
+    {
+      "epoch": 0.5801776934356939,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.6202315374002594e-05,
+      "loss": 1.2958,
+      "step": 3363
+    },
+    {
+      "epoch": 0.580350211334426,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.620018005130075e-05,
+      "loss": 1.4864,
+      "step": 3364
+    },
+    {
+      "epoch": 0.580522729233158,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.619804426925083e-05,
+      "loss": 1.5509,
+      "step": 3365
+    },
+    {
+      "epoch": 0.5806952471318899,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.619590802801106e-05,
+      "loss": 1.3632,
+      "step": 3366
+    },
+    {
+      "epoch": 0.5808677650306219,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.6193771327739712e-05,
+      "loss": 1.4981,
+      "step": 3367
+    },
+    {
+      "epoch": 0.581040282929354,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.619163416859508e-05,
+      "loss": 1.3906,
+      "step": 3368
+    },
+    {
+      "epoch": 0.5812128008280859,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.6189496550735503e-05,
+      "loss": 1.4839,
+      "step": 3369
+    },
+    {
+      "epoch": 0.5813853187268179,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.6187358474319347e-05,
+      "loss": 1.4431,
+      "step": 3370
+    },
+    {
+      "epoch": 0.5815578366255499,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.6185219939505016e-05,
+      "loss": 1.4361,
+      "step": 3371
+    },
+    {
+      "epoch": 0.5817303545242819,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.6183080946450945e-05,
+      "loss": 1.3862,
+      "step": 3372
+    },
+    {
+      "epoch": 0.5819028724230139,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.6180941495315606e-05,
+      "loss": 1.3674,
+      "step": 3373
+    },
+    {
+      "epoch": 0.5820753903217459,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.6178801586257494e-05,
+      "loss": 1.4896,
+      "step": 3374
+    },
+    {
+      "epoch": 0.5822479082204779,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.617666121943516e-05,
+      "loss": 1.3988,
+      "step": 3375
+    },
+    {
+      "epoch": 0.5824204261192099,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.6174520395007168e-05,
+      "loss": 1.3873,
+      "step": 3376
+    },
+    {
+      "epoch": 0.5825929440179418,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.6172379113132123e-05,
+      "loss": 1.52,
+      "step": 3377
+    },
+    {
+      "epoch": 0.5827654619166739,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.617023737396867e-05,
+      "loss": 1.46,
+      "step": 3378
+    },
+    {
+      "epoch": 0.5829379798154058,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.6168095177675476e-05,
+      "loss": 1.5046,
+      "step": 3379
+    },
+    {
+      "epoch": 0.5831104977141378,
+      "grad_norm": 0.625,
+      "learning_rate": 1.6165952524411253e-05,
+      "loss": 1.4513,
+      "step": 3380
+    },
+    {
+      "epoch": 0.5832830156128699,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.616380941433474e-05,
+      "loss": 1.547,
+      "step": 3381
+    },
+    {
+      "epoch": 0.5834555335116018,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.616166584760471e-05,
+      "loss": 1.5419,
+      "step": 3382
+    },
+    {
+      "epoch": 0.5836280514103338,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.6159521824379977e-05,
+      "loss": 1.4538,
+      "step": 3383
+    },
+    {
+      "epoch": 0.5838005693090658,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.6157377344819377e-05,
+      "loss": 1.5394,
+      "step": 3384
+    },
+    {
+      "epoch": 0.5839730872077978,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.6155232409081794e-05,
+      "loss": 1.4276,
+      "step": 3385
+    },
+    {
+      "epoch": 0.5841456051065298,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.615308701732613e-05,
+      "loss": 1.5032,
+      "step": 3386
+    },
+    {
+      "epoch": 0.5843181230052618,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.615094116971134e-05,
+      "loss": 1.4876,
+      "step": 3387
+    },
+    {
+      "epoch": 0.5844906409039938,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.6148794866396392e-05,
+      "loss": 1.5329,
+      "step": 3388
+    },
+    {
+      "epoch": 0.5846631588027258,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.6146648107540303e-05,
+      "loss": 1.4204,
+      "step": 3389
+    },
+    {
+      "epoch": 0.5848356767014578,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.614450089330211e-05,
+      "loss": 1.4151,
+      "step": 3390
+    },
+    {
+      "epoch": 0.5850081946001898,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.6142353223840902e-05,
+      "loss": 1.467,
+      "step": 3391
+    },
+    {
+      "epoch": 0.5851807124989218,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.6140205099315787e-05,
+      "loss": 1.4764,
+      "step": 3392
+    },
+    {
+      "epoch": 0.5853532303976537,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.6138056519885916e-05,
+      "loss": 1.4966,
+      "step": 3393
+    },
+    {
+      "epoch": 0.5855257482963857,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.6135907485710462e-05,
+      "loss": 1.4324,
+      "step": 3394
+    },
+    {
+      "epoch": 0.5856982661951178,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.6133757996948645e-05,
+      "loss": 1.4643,
+      "step": 3395
+    },
+    {
+      "epoch": 0.5858707840938497,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.613160805375971e-05,
+      "loss": 1.3835,
+      "step": 3396
+    },
+    {
+      "epoch": 0.5860433019925817,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.6129457656302935e-05,
+      "loss": 1.4762,
+      "step": 3397
+    },
+    {
+      "epoch": 0.5862158198913138,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.612730680473764e-05,
+      "loss": 1.4536,
+      "step": 3398
+    },
+    {
+      "epoch": 0.5863883377900457,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.6125155499223174e-05,
+      "loss": 1.5047,
+      "step": 3399
+    },
+    {
+      "epoch": 0.5865608556887777,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.6123003739918915e-05,
+      "loss": 1.5773,
+      "step": 3400
+    },
+    {
+      "epoch": 0.5865608556887777,
+      "eval_loss": 1.433887004852295,
+      "eval_runtime": 10.8057,
+      "eval_samples_per_second": 94.765,
+      "eval_steps_per_second": 23.691,
+      "step": 3400
+    },
+    {
+      "epoch": 0.5867333735875097,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.6120851526984282e-05,
+      "loss": 1.5089,
+      "step": 3401
+    },
+    {
+      "epoch": 0.5869058914862417,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.611869886057872e-05,
+      "loss": 1.3859,
+      "step": 3402
+    },
+    {
+      "epoch": 0.5870784093849737,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.6116545740861718e-05,
+      "loss": 1.4056,
+      "step": 3403
+    },
+    {
+      "epoch": 0.5872509272837056,
+      "grad_norm": 0.625,
+      "learning_rate": 1.6114392167992784e-05,
+      "loss": 1.5003,
+      "step": 3404
+    },
+    {
+      "epoch": 0.5874234451824377,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.611223814213148e-05,
+      "loss": 1.5606,
+      "step": 3405
+    },
+    {
+      "epoch": 0.5875959630811697,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.611008366343738e-05,
+      "loss": 1.4371,
+      "step": 3406
+    },
+    {
+      "epoch": 0.5877684809799016,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.6107928732070107e-05,
+      "loss": 1.4536,
+      "step": 3407
+    },
+    {
+      "epoch": 0.5879409988786336,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.6105773348189312e-05,
+      "loss": 1.3856,
+      "step": 3408
+    },
+    {
+      "epoch": 0.5881135167773657,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.610361751195467e-05,
+      "loss": 1.4591,
+      "step": 3409
+    },
+    {
+      "epoch": 0.5882860346760976,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.6101461223525908e-05,
+      "loss": 1.4714,
+      "step": 3410
+    },
+    {
+      "epoch": 0.5884585525748296,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.6099304483062776e-05,
+      "loss": 1.5012,
+      "step": 3411
+    },
+    {
+      "epoch": 0.5886310704735617,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.609714729072506e-05,
+      "loss": 1.4031,
+      "step": 3412
+    },
+    {
+      "epoch": 0.5888035883722936,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.6094989646672573e-05,
+      "loss": 1.3807,
+      "step": 3413
+    },
+    {
+      "epoch": 0.5889761062710256,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.609283155106517e-05,
+      "loss": 1.476,
+      "step": 3414
+    },
+    {
+      "epoch": 0.5891486241697577,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.6090673004062734e-05,
+      "loss": 1.4646,
+      "step": 3415
+    },
+    {
+      "epoch": 0.5893211420684896,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.608851400582519e-05,
+      "loss": 1.4054,
+      "step": 3416
+    },
+    {
+      "epoch": 0.5894936599672216,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.608635455651248e-05,
+      "loss": 1.4973,
+      "step": 3417
+    },
+    {
+      "epoch": 0.5896661778659535,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.6084194656284598e-05,
+      "loss": 1.3328,
+      "step": 3418
+    },
+    {
+      "epoch": 0.5898386957646856,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.6082034305301563e-05,
+      "loss": 1.5572,
+      "step": 3419
+    },
+    {
+      "epoch": 0.5900112136634176,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.607987350372342e-05,
+      "loss": 1.4634,
+      "step": 3420
+    },
+    {
+      "epoch": 0.5901837315621495,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.607771225171026e-05,
+      "loss": 1.4979,
+      "step": 3421
+    },
+    {
+      "epoch": 0.5903562494608816,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.6075550549422204e-05,
+      "loss": 1.3774,
+      "step": 3422
+    },
+    {
+      "epoch": 0.5905287673596136,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.6073388397019397e-05,
+      "loss": 1.4982,
+      "step": 3423
+    },
+    {
+      "epoch": 0.5907012852583455,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.6071225794662033e-05,
+      "loss": 1.6353,
+      "step": 3424
+    },
+    {
+      "epoch": 0.5908738031570775,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.6069062742510326e-05,
+      "loss": 1.4391,
+      "step": 3425
+    },
+    {
+      "epoch": 0.5910463210558096,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.606689924072453e-05,
+      "loss": 1.3626,
+      "step": 3426
+    },
+    {
+      "epoch": 0.5912188389545415,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.606473528946493e-05,
+      "loss": 1.4261,
+      "step": 3427
+    },
+    {
+      "epoch": 0.5913913568532735,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.6062570888891847e-05,
+      "loss": 1.4534,
+      "step": 3428
+    },
+    {
+      "epoch": 0.5915638747520056,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.6060406039165627e-05,
+      "loss": 1.5243,
+      "step": 3429
+    },
+    {
+      "epoch": 0.5917363926507375,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.6058240740446666e-05,
+      "loss": 1.458,
+      "step": 3430
+    },
+    {
+      "epoch": 0.5919089105494695,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.6056074992895378e-05,
+      "loss": 1.4456,
+      "step": 3431
+    },
+    {
+      "epoch": 0.5920814284482016,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.605390879667221e-05,
+      "loss": 1.4622,
+      "step": 3432
+    },
+    {
+      "epoch": 0.5922539463469335,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.6051742151937655e-05,
+      "loss": 1.4636,
+      "step": 3433
+    },
+    {
+      "epoch": 0.5924264642456655,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.6049575058852223e-05,
+      "loss": 1.3617,
+      "step": 3434
+    },
+    {
+      "epoch": 0.5925989821443974,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.6047407517576478e-05,
+      "loss": 1.5374,
+      "step": 3435
+    },
+    {
+      "epoch": 0.5927715000431295,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.6045239528270993e-05,
+      "loss": 1.3766,
+      "step": 3436
+    },
+    {
+      "epoch": 0.5929440179418615,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.6043071091096397e-05,
+      "loss": 1.3774,
+      "step": 3437
+    },
+    {
+      "epoch": 0.5931165358405934,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.6040902206213332e-05,
+      "loss": 1.5118,
+      "step": 3438
+    },
+    {
+      "epoch": 0.5932890537393255,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.6038732873782485e-05,
+      "loss": 1.4876,
+      "step": 3439
+    },
+    {
+      "epoch": 0.5934615716380575,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.6036563093964577e-05,
+      "loss": 1.4963,
+      "step": 3440
+    },
+    {
+      "epoch": 0.5936340895367894,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.6034392866920354e-05,
+      "loss": 1.5198,
+      "step": 3441
+    },
+    {
+      "epoch": 0.5938066074355214,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.6032222192810607e-05,
+      "loss": 1.3917,
+      "step": 3442
+    },
+    {
+      "epoch": 0.5939791253342535,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.6030051071796146e-05,
+      "loss": 1.4103,
+      "step": 3443
+    },
+    {
+      "epoch": 0.5941516432329854,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.6027879504037826e-05,
+      "loss": 1.4817,
+      "step": 3444
+    },
+    {
+      "epoch": 0.5943241611317174,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.6025707489696527e-05,
+      "loss": 1.5543,
+      "step": 3445
+    },
+    {
+      "epoch": 0.5944966790304494,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.6023535028933167e-05,
+      "loss": 1.5289,
+      "step": 3446
+    },
+    {
+      "epoch": 0.5946691969291814,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.6021362121908697e-05,
+      "loss": 1.5147,
+      "step": 3447
+    },
+    {
+      "epoch": 0.5948417148279134,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.60191887687841e-05,
+      "loss": 1.5132,
+      "step": 3448
+    },
+    {
+      "epoch": 0.5950142327266454,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.6017014969720382e-05,
+      "loss": 1.5291,
+      "step": 3449
+    },
+    {
+      "epoch": 0.5951867506253774,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.6014840724878603e-05,
+      "loss": 1.3743,
+      "step": 3450
+    },
+    {
+      "epoch": 0.5953592685241094,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.6012666034419838e-05,
+      "loss": 1.3932,
+      "step": 3451
+    },
+    {
+      "epoch": 0.5955317864228413,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.601049089850521e-05,
+      "loss": 1.4492,
+      "step": 3452
+    },
+    {
+      "epoch": 0.5957043043215734,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.6008315317295857e-05,
+      "loss": 1.4471,
+      "step": 3453
+    },
+    {
+      "epoch": 0.5958768222203054,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.6006139290952964e-05,
+      "loss": 1.453,
+      "step": 3454
+    },
+    {
+      "epoch": 0.5960493401190373,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.6003962819637744e-05,
+      "loss": 1.5166,
+      "step": 3455
+    },
+    {
+      "epoch": 0.5962218580177694,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.6001785903511446e-05,
+      "loss": 1.5024,
+      "step": 3456
+    },
+    {
+      "epoch": 0.5963943759165014,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.5999608542735345e-05,
+      "loss": 1.4908,
+      "step": 3457
+    },
+    {
+      "epoch": 0.5965668938152333,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.599743073747076e-05,
+      "loss": 1.3899,
+      "step": 3458
+    },
+    {
+      "epoch": 0.5967394117139653,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.5995252487879025e-05,
+      "loss": 1.4486,
+      "step": 3459
+    },
+    {
+      "epoch": 0.5969119296126973,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.599307379412153e-05,
+      "loss": 1.5181,
+      "step": 3460
+    },
+    {
+      "epoch": 0.5970844475114293,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.599089465635968e-05,
+      "loss": 1.4457,
+      "step": 3461
+    },
+    {
+      "epoch": 0.5972569654101613,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.5988715074754923e-05,
+      "loss": 1.3981,
+      "step": 3462
+    },
+    {
+      "epoch": 0.5974294833088933,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.5986535049468736e-05,
+      "loss": 1.4419,
+      "step": 3463
+    },
+    {
+      "epoch": 0.5976020012076253,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.5984354580662623e-05,
+      "loss": 1.389,
+      "step": 3464
+    },
+    {
+      "epoch": 0.5977745191063573,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.5982173668498135e-05,
+      "loss": 1.4545,
+      "step": 3465
+    },
+    {
+      "epoch": 0.5979470370050893,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.5979992313136836e-05,
+      "loss": 1.5201,
+      "step": 3466
+    },
+    {
+      "epoch": 0.5981195549038213,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.597781051474035e-05,
+      "loss": 1.4027,
+      "step": 3467
+    },
+    {
+      "epoch": 0.5982920728025533,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.5975628273470303e-05,
+      "loss": 1.439,
+      "step": 3468
+    },
+    {
+      "epoch": 0.5984645907012852,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.597344558948838e-05,
+      "loss": 1.5406,
+      "step": 3469
+    },
+    {
+      "epoch": 0.5986371086000173,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.5971262462956283e-05,
+      "loss": 1.425,
+      "step": 3470
+    },
+    {
+      "epoch": 0.5988096264987492,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.5969078894035753e-05,
+      "loss": 1.5376,
+      "step": 3471
+    },
+    {
+      "epoch": 0.5989821443974812,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.596689488288856e-05,
+      "loss": 1.3337,
+      "step": 3472
+    },
+    {
+      "epoch": 0.5991546622962133,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.5964710429676514e-05,
+      "loss": 1.4247,
+      "step": 3473
+    },
+    {
+      "epoch": 0.5993271801949452,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.596252553456145e-05,
+      "loss": 1.4202,
+      "step": 3474
+    },
+    {
+      "epoch": 0.5994996980936772,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.5960340197705235e-05,
+      "loss": 1.4148,
+      "step": 3475
+    },
+    {
+      "epoch": 0.5996722159924092,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.5958154419269782e-05,
+      "loss": 1.4566,
+      "step": 3476
+    },
+    {
+      "epoch": 0.5998447338911412,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.5955968199417018e-05,
+      "loss": 1.4067,
+      "step": 3477
+    },
+    {
+      "epoch": 0.6000172517898732,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.5953781538308914e-05,
+      "loss": 1.4475,
+      "step": 3478
+    },
+    {
+      "epoch": 0.6001897696886052,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.5951594436107475e-05,
+      "loss": 1.4135,
+      "step": 3479
+    },
+    {
+      "epoch": 0.6003622875873372,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.5949406892974733e-05,
+      "loss": 1.459,
+      "step": 3480
+    },
+    {
+      "epoch": 0.6005348054860692,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.5947218909072755e-05,
+      "loss": 1.4852,
+      "step": 3481
+    },
+    {
+      "epoch": 0.6007073233848012,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.594503048456364e-05,
+      "loss": 1.3822,
+      "step": 3482
+    },
+    {
+      "epoch": 0.6008798412835331,
+      "grad_norm": 0.625,
+      "learning_rate": 1.5942841619609515e-05,
+      "loss": 1.4175,
+      "step": 3483
+    },
+    {
+      "epoch": 0.6010523591822652,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.5940652314372558e-05,
+      "loss": 1.4183,
+      "step": 3484
+    },
+    {
+      "epoch": 0.6012248770809971,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.5938462569014954e-05,
+      "loss": 1.4889,
+      "step": 3485
+    },
+    {
+      "epoch": 0.6013973949797291,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.5936272383698937e-05,
+      "loss": 1.473,
+      "step": 3486
+    },
+    {
+      "epoch": 0.6015699128784612,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.5934081758586775e-05,
+      "loss": 1.45,
+      "step": 3487
+    },
+    {
+      "epoch": 0.6017424307771931,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.5931890693840755e-05,
+      "loss": 1.3931,
+      "step": 3488
+    },
+    {
+      "epoch": 0.6019149486759251,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.592969918962321e-05,
+      "loss": 1.5415,
+      "step": 3489
+    },
+    {
+      "epoch": 0.6020874665746572,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.5927507246096497e-05,
+      "loss": 1.4022,
+      "step": 3490
+    },
+    {
+      "epoch": 0.6022599844733891,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.592531486342301e-05,
+      "loss": 1.4198,
+      "step": 3491
+    },
+    {
+      "epoch": 0.6024325023721211,
+      "grad_norm": 0.625,
+      "learning_rate": 1.592312204176518e-05,
+      "loss": 1.3621,
+      "step": 3492
+    },
+    {
+      "epoch": 0.602605020270853,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.5920928781285456e-05,
+      "loss": 1.414,
+      "step": 3493
+    },
+    {
+      "epoch": 0.6027775381695851,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.591873508214633e-05,
+      "loss": 1.4441,
+      "step": 3494
+    },
+    {
+      "epoch": 0.6029500560683171,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.591654094451033e-05,
+      "loss": 1.4143,
+      "step": 3495
+    },
+    {
+      "epoch": 0.603122573967049,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.5914346368540007e-05,
+      "loss": 1.5217,
+      "step": 3496
+    },
+    {
+      "epoch": 0.6032950918657811,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.591215135439795e-05,
+      "loss": 1.4534,
+      "step": 3497
+    },
+    {
+      "epoch": 0.6034676097645131,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.5909955902246782e-05,
+      "loss": 1.4631,
+      "step": 3498
+    },
+    {
+      "epoch": 0.603640127663245,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.5907760012249148e-05,
+      "loss": 1.4084,
+      "step": 3499
+    },
+    {
+      "epoch": 0.603812645561977,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.590556368456775e-05,
+      "loss": 1.4021,
+      "step": 3500
+    },
+    {
+      "epoch": 0.603812645561977,
+      "eval_loss": 1.4319883584976196,
+      "eval_runtime": 10.882,
+      "eval_samples_per_second": 94.101,
+      "eval_steps_per_second": 23.525,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6039851634607091,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.5903366919365283e-05,
+      "loss": 1.4274,
+      "step": 3501
+    },
+    {
+      "epoch": 0.604157681359441,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.5901169716804516e-05,
+      "loss": 1.6006,
+      "step": 3502
+    },
+    {
+      "epoch": 0.604330199258173,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.589897207704822e-05,
+      "loss": 1.4748,
+      "step": 3503
+    },
+    {
+      "epoch": 0.6045027171569051,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.5896774000259218e-05,
+      "loss": 1.4845,
+      "step": 3504
+    },
+    {
+      "epoch": 0.604675235055637,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.5894575486600354e-05,
+      "loss": 1.5827,
+      "step": 3505
+    },
+    {
+      "epoch": 0.604847752954369,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.58923765362345e-05,
+      "loss": 1.4438,
+      "step": 3506
+    },
+    {
+      "epoch": 0.6050202708531011,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.5890177149324583e-05,
+      "loss": 1.363,
+      "step": 3507
+    },
+    {
+      "epoch": 0.605192788751833,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.5887977326033533e-05,
+      "loss": 1.4079,
+      "step": 3508
+    },
+    {
+      "epoch": 0.605365306650565,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.5885777066524335e-05,
+      "loss": 1.4491,
+      "step": 3509
+    },
+    {
+      "epoch": 0.6055378245492969,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.588357637096e-05,
+      "loss": 1.4404,
+      "step": 3510
+    },
+    {
+      "epoch": 0.605710342448029,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.5881375239503558e-05,
+      "loss": 1.4348,
+      "step": 3511
+    },
+    {
+      "epoch": 0.605882860346761,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.5879173672318095e-05,
+      "loss": 1.3538,
+      "step": 3512
+    },
+    {
+      "epoch": 0.6060553782454929,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.587697166956671e-05,
+      "loss": 1.4551,
+      "step": 3513
+    },
+    {
+      "epoch": 0.606227896144225,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.587476923141254e-05,
+      "loss": 1.3779,
+      "step": 3514
+    },
+    {
+      "epoch": 0.606400414042957,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.587256635801876e-05,
+      "loss": 1.4017,
+      "step": 3515
+    },
+    {
+      "epoch": 0.6065729319416889,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.5870363049548573e-05,
+      "loss": 1.3796,
+      "step": 3516
+    },
+    {
+      "epoch": 0.6067454498404209,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.5868159306165208e-05,
+      "loss": 1.4247,
+      "step": 3517
+    },
+    {
+      "epoch": 0.606917967739153,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.5865955128031937e-05,
+      "loss": 1.4333,
+      "step": 3518
+    },
+    {
+      "epoch": 0.6070904856378849,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.586375051531206e-05,
+      "loss": 1.4746,
+      "step": 3519
+    },
+    {
+      "epoch": 0.6072630035366169,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.5861545468168907e-05,
+      "loss": 1.4033,
+      "step": 3520
+    },
+    {
+      "epoch": 0.607435521435349,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.5859339986765837e-05,
+      "loss": 1.4576,
+      "step": 3521
+    },
+    {
+      "epoch": 0.6076080393340809,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.5857134071266253e-05,
+      "loss": 1.4398,
+      "step": 3522
+    },
+    {
+      "epoch": 0.6077805572328129,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.5854927721833577e-05,
+      "loss": 1.4282,
+      "step": 3523
+    },
+    {
+      "epoch": 0.607953075131545,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.5852720938631273e-05,
+      "loss": 1.3739,
+      "step": 3524
+    },
+    {
+      "epoch": 0.6081255930302769,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.5850513721822832e-05,
+      "loss": 1.4447,
+      "step": 3525
+    },
+    {
+      "epoch": 0.6082981109290089,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.5848306071571782e-05,
+      "loss": 1.5001,
+      "step": 3526
+    },
+    {
+      "epoch": 0.6084706288277408,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.5846097988041672e-05,
+      "loss": 1.4358,
+      "step": 3527
+    },
+    {
+      "epoch": 0.6086431467264729,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.58438894713961e-05,
+      "loss": 1.3999,
+      "step": 3528
+    },
+    {
+      "epoch": 0.6088156646252049,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.584168052179868e-05,
+      "loss": 1.5327,
+      "step": 3529
+    },
+    {
+      "epoch": 0.6089881825239368,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.5839471139413065e-05,
+      "loss": 1.4658,
+      "step": 3530
+    },
+    {
+      "epoch": 0.6091607004226689,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.583726132440294e-05,
+      "loss": 1.5247,
+      "step": 3531
+    },
+    {
+      "epoch": 0.6093332183214009,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.5835051076932028e-05,
+      "loss": 1.528,
+      "step": 3532
+    },
+    {
+      "epoch": 0.6095057362201328,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.5832840397164067e-05,
+      "loss": 1.4172,
+      "step": 3533
+    },
+    {
+      "epoch": 0.6096782541188648,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.5830629285262848e-05,
+      "loss": 1.5032,
+      "step": 3534
+    },
+    {
+      "epoch": 0.6098507720175969,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.582841774139218e-05,
+      "loss": 1.4678,
+      "step": 3535
+    },
+    {
+      "epoch": 0.6100232899163288,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.5826205765715905e-05,
+      "loss": 1.4701,
+      "step": 3536
+    },
+    {
+      "epoch": 0.6101958078150608,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.5823993358397906e-05,
+      "loss": 1.4376,
+      "step": 3537
+    },
+    {
+      "epoch": 0.6103683257137928,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.582178051960209e-05,
+      "loss": 1.4783,
+      "step": 3538
+    },
+    {
+      "epoch": 0.6105408436125248,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.5819567249492392e-05,
+      "loss": 1.4107,
+      "step": 3539
+    },
+    {
+      "epoch": 0.6107133615112568,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.5817353548232795e-05,
+      "loss": 1.3483,
+      "step": 3540
+    },
+    {
+      "epoch": 0.6108858794099888,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.5815139415987296e-05,
+      "loss": 1.3945,
+      "step": 3541
+    },
+    {
+      "epoch": 0.6110583973087208,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.5812924852919936e-05,
+      "loss": 1.3612,
+      "step": 3542
+    },
+    {
+      "epoch": 0.6112309152074528,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.581070985919478e-05,
+      "loss": 1.4229,
+      "step": 3543
+    },
+    {
+      "epoch": 0.6114034331061847,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.580849443497593e-05,
+      "loss": 1.4903,
+      "step": 3544
+    },
+    {
+      "epoch": 0.6115759510049168,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.5806278580427523e-05,
+      "loss": 1.4872,
+      "step": 3545
+    },
+    {
+      "epoch": 0.6117484689036488,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.580406229571372e-05,
+      "loss": 1.4751,
+      "step": 3546
+    },
+    {
+      "epoch": 0.6119209868023807,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.5801845580998708e-05,
+      "loss": 1.3846,
+      "step": 3547
+    },
+    {
+      "epoch": 0.6120935047011128,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.5799628436446733e-05,
+      "loss": 1.5369,
+      "step": 3548
+    },
+    {
+      "epoch": 0.6122660225998448,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.5797410862222043e-05,
+      "loss": 1.4508,
+      "step": 3549
+    },
+    {
+      "epoch": 0.6124385404985767,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.579519285848893e-05,
+      "loss": 1.3334,
+      "step": 3550
+    },
+    {
+      "epoch": 0.6126110583973087,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.579297442541172e-05,
+      "loss": 1.4288,
+      "step": 3551
+    },
+    {
+      "epoch": 0.6127835762960407,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.5790755563154773e-05,
+      "loss": 1.4861,
+      "step": 3552
+    },
+    {
+      "epoch": 0.6129560941947727,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.578853627188247e-05,
+      "loss": 1.5465,
+      "step": 3553
+    },
+    {
+      "epoch": 0.6131286120935047,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.5786316551759232e-05,
+      "loss": 1.3731,
+      "step": 3554
+    },
+    {
+      "epoch": 0.6133011299922367,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.5784096402949507e-05,
+      "loss": 1.4603,
+      "step": 3555
+    },
+    {
+      "epoch": 0.6134736478909687,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.578187582561778e-05,
+      "loss": 1.3989,
+      "step": 3556
+    },
+    {
+      "epoch": 0.6136461657897007,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.5779654819928565e-05,
+      "loss": 1.5056,
+      "step": 3557
+    },
+    {
+      "epoch": 0.6138186836884326,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.577743338604641e-05,
+      "loss": 1.5337,
+      "step": 3558
+    },
+    {
+      "epoch": 0.6139912015871647,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.5775211524135888e-05,
+      "loss": 1.3694,
+      "step": 3559
+    },
+    {
+      "epoch": 0.6141637194858967,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.5772989234361612e-05,
+      "loss": 1.4417,
+      "step": 3560
+    },
+    {
+      "epoch": 0.6143362373846286,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.5770766516888224e-05,
+      "loss": 1.5343,
+      "step": 3561
+    },
+    {
+      "epoch": 0.6145087552833607,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.5768543371880394e-05,
+      "loss": 1.4724,
+      "step": 3562
+    },
+    {
+      "epoch": 0.6146812731820926,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.5766319799502822e-05,
+      "loss": 1.4003,
+      "step": 3563
+    },
+    {
+      "epoch": 0.6148537910808246,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.5764095799920256e-05,
+      "loss": 1.4674,
+      "step": 3564
+    },
+    {
+      "epoch": 0.6150263089795567,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.576187137329746e-05,
+      "loss": 1.4818,
+      "step": 3565
+    },
+    {
+      "epoch": 0.6151988268782886,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.5759646519799227e-05,
+      "loss": 1.4898,
+      "step": 3566
+    },
+    {
+      "epoch": 0.6153713447770206,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.5757421239590388e-05,
+      "loss": 1.4803,
+      "step": 3567
+    },
+    {
+      "epoch": 0.6155438626757526,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.5755195532835814e-05,
+      "loss": 1.5081,
+      "step": 3568
+    },
+    {
+      "epoch": 0.6157163805744846,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.5752969399700396e-05,
+      "loss": 1.4798,
+      "step": 3569
+    },
+    {
+      "epoch": 0.6158888984732166,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.5750742840349054e-05,
+      "loss": 1.4868,
+      "step": 3570
+    },
+    {
+      "epoch": 0.6160614163719486,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.5748515854946753e-05,
+      "loss": 1.4558,
+      "step": 3571
+    },
+    {
+      "epoch": 0.6162339342706806,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.574628844365848e-05,
+      "loss": 1.4448,
+      "step": 3572
+    },
+    {
+      "epoch": 0.6164064521694126,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.5744060606649253e-05,
+      "loss": 1.4433,
+      "step": 3573
+    },
+    {
+      "epoch": 0.6165789700681445,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.5741832344084126e-05,
+      "loss": 1.4628,
+      "step": 3574
+    },
+    {
+      "epoch": 0.6167514879668765,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.5739603656128183e-05,
+      "loss": 1.4946,
+      "step": 3575
+    },
+    {
+      "epoch": 0.6169240058656086,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.573737454294654e-05,
+      "loss": 1.5086,
+      "step": 3576
+    },
+    {
+      "epoch": 0.6170965237643405,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.573514500470434e-05,
+      "loss": 1.4041,
+      "step": 3577
+    },
+    {
+      "epoch": 0.6172690416630725,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.5732915041566764e-05,
+      "loss": 1.4129,
+      "step": 3578
+    },
+    {
+      "epoch": 0.6174415595618046,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.5730684653699017e-05,
+      "loss": 1.4428,
+      "step": 3579
+    },
+    {
+      "epoch": 0.6176140774605365,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.5728453841266345e-05,
+      "loss": 1.4511,
+      "step": 3580
+    },
+    {
+      "epoch": 0.6177865953592685,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.5726222604434023e-05,
+      "loss": 1.4721,
+      "step": 3581
+    },
+    {
+      "epoch": 0.6179591132580006,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.572399094336735e-05,
+      "loss": 1.4831,
+      "step": 3582
+    },
+    {
+      "epoch": 0.6181316311567325,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.5721758858231662e-05,
+      "loss": 1.4647,
+      "step": 3583
+    },
+    {
+      "epoch": 0.6183041490554645,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.571952634919233e-05,
+      "loss": 1.4295,
+      "step": 3584
+    },
+    {
+      "epoch": 0.6184766669541965,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.5717293416414743e-05,
+      "loss": 1.4365,
+      "step": 3585
+    },
+    {
+      "epoch": 0.6186491848529285,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.571506006006434e-05,
+      "loss": 1.3541,
+      "step": 3586
+    },
+    {
+      "epoch": 0.6188217027516605,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.5712826280306578e-05,
+      "loss": 1.3913,
+      "step": 3587
+    },
+    {
+      "epoch": 0.6189942206503924,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.571059207730695e-05,
+      "loss": 1.4749,
+      "step": 3588
+    },
+    {
+      "epoch": 0.6191667385491245,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.5708357451230978e-05,
+      "loss": 1.5027,
+      "step": 3589
+    },
+    {
+      "epoch": 0.6193392564478565,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.570612240224422e-05,
+      "loss": 1.3575,
+      "step": 3590
+    },
+    {
+      "epoch": 0.6195117743465884,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.570388693051226e-05,
+      "loss": 1.4667,
+      "step": 3591
+    },
+    {
+      "epoch": 0.6196842922453204,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.570165103620072e-05,
+      "loss": 1.4951,
+      "step": 3592
+    },
+    {
+      "epoch": 0.6198568101440525,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.5699414719475243e-05,
+      "loss": 1.4742,
+      "step": 3593
+    },
+    {
+      "epoch": 0.6200293280427844,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.5697177980501507e-05,
+      "loss": 1.4339,
+      "step": 3594
+    },
+    {
+      "epoch": 0.6202018459415164,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.5694940819445234e-05,
+      "loss": 1.505,
+      "step": 3595
+    },
+    {
+      "epoch": 0.6203743638402485,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.5692703236472162e-05,
+      "loss": 1.483,
+      "step": 3596
+    },
+    {
+      "epoch": 0.6205468817389804,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.5690465231748064e-05,
+      "loss": 1.5035,
+      "step": 3597
+    },
+    {
+      "epoch": 0.6207193996377124,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.5688226805438745e-05,
+      "loss": 1.3519,
+      "step": 3598
+    },
+    {
+      "epoch": 0.6208919175364445,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.5685987957710043e-05,
+      "loss": 1.4842,
+      "step": 3599
+    },
+    {
+      "epoch": 0.6210644354351764,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.5683748688727827e-05,
+      "loss": 1.3299,
+      "step": 3600
+    },
+    {
+      "epoch": 0.6210644354351764,
+      "eval_loss": 1.4302067756652832,
+      "eval_runtime": 10.8016,
+      "eval_samples_per_second": 94.801,
+      "eval_steps_per_second": 23.7,
+      "step": 3600
+    },
+    {
+      "epoch": 0.6212369533339084,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.568150899865799e-05,
+      "loss": 1.4975,
+      "step": 3601
+    },
+    {
+      "epoch": 0.6214094712326403,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.567926888766647e-05,
+      "loss": 1.406,
+      "step": 3602
+    },
+    {
+      "epoch": 0.6215819891313724,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.5677028355919225e-05,
+      "loss": 1.4159,
+      "step": 3603
+    },
+    {
+      "epoch": 0.6217545070301044,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.5674787403582247e-05,
+      "loss": 1.3803,
+      "step": 3604
+    },
+    {
+      "epoch": 0.6219270249288363,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.567254603082156e-05,
+      "loss": 1.4789,
+      "step": 3605
+    },
+    {
+      "epoch": 0.6220995428275684,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.567030423780322e-05,
+      "loss": 1.4876,
+      "step": 3606
+    },
+    {
+      "epoch": 0.6222720607263004,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.5668062024693314e-05,
+      "loss": 1.4734,
+      "step": 3607
+    },
+    {
+      "epoch": 0.6224445786250323,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.5665819391657955e-05,
+      "loss": 1.3378,
+      "step": 3608
+    },
+    {
+      "epoch": 0.6226170965237643,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.5663576338863294e-05,
+      "loss": 1.4462,
+      "step": 3609
+    },
+    {
+      "epoch": 0.6227896144224964,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.5661332866475514e-05,
+      "loss": 1.4788,
+      "step": 3610
+    },
+    {
+      "epoch": 0.6229621323212283,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.5659088974660813e-05,
+      "loss": 1.5331,
+      "step": 3611
+    },
+    {
+      "epoch": 0.6231346502199603,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.5656844663585448e-05,
+      "loss": 1.4245,
+      "step": 3612
+    },
+    {
+      "epoch": 0.6233071681186924,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.5654599933415683e-05,
+      "loss": 1.3844,
+      "step": 3613
+    },
+    {
+      "epoch": 0.6234796860174243,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.565235478431783e-05,
+      "loss": 1.3807,
+      "step": 3614
+    },
+    {
+      "epoch": 0.6236522039161563,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.5650109216458208e-05,
+      "loss": 1.4553,
+      "step": 3615
+    },
+    {
+      "epoch": 0.6238247218148883,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.5647863230003193e-05,
+      "loss": 1.398,
+      "step": 3616
+    },
+    {
+      "epoch": 0.6239972397136203,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.564561682511918e-05,
+      "loss": 1.4887,
+      "step": 3617
+    },
+    {
+      "epoch": 0.6241697576123523,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.5643370001972602e-05,
+      "loss": 1.4707,
+      "step": 3618
+    },
+    {
+      "epoch": 0.6243422755110842,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.564112276072991e-05,
+      "loss": 1.4133,
+      "step": 3619
+    },
+    {
+      "epoch": 0.6245147934098163,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.5638875101557596e-05,
+      "loss": 1.4796,
+      "step": 3620
+    },
+    {
+      "epoch": 0.6246873113085483,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.5636627024622183e-05,
+      "loss": 1.3614,
+      "step": 3621
+    },
+    {
+      "epoch": 0.6248598292072802,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.5634378530090217e-05,
+      "loss": 1.3653,
+      "step": 3622
+    },
+    {
+      "epoch": 0.6250323471060123,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.5632129618128285e-05,
+      "loss": 1.4674,
+      "step": 3623
+    },
+    {
+      "epoch": 0.6252048650047443,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.5629880288903002e-05,
+      "loss": 1.427,
+      "step": 3624
+    },
+    {
+      "epoch": 0.6253773829034762,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.5627630542581013e-05,
+      "loss": 1.4733,
+      "step": 3625
+    },
+    {
+      "epoch": 0.6255499008022082,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.5625380379328984e-05,
+      "loss": 1.5152,
+      "step": 3626
+    },
+    {
+      "epoch": 0.6257224187009403,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.562312979931363e-05,
+      "loss": 1.482,
+      "step": 3627
+    },
+    {
+      "epoch": 0.6258949365996722,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.5620878802701687e-05,
+      "loss": 1.4838,
+      "step": 3628
+    },
+    {
+      "epoch": 0.6260674544984042,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.561862738965992e-05,
+      "loss": 1.4425,
+      "step": 3629
+    },
+    {
+      "epoch": 0.6262399723971362,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.561637556035513e-05,
+      "loss": 1.5229,
+      "step": 3630
+    },
+    {
+      "epoch": 0.6264124902958682,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.5614123314954144e-05,
+      "loss": 1.457,
+      "step": 3631
+    },
+    {
+      "epoch": 0.6265850081946002,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.5611870653623826e-05,
+      "loss": 1.5144,
+      "step": 3632
+    },
+    {
+      "epoch": 0.6267575260933321,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.5609617576531065e-05,
+      "loss": 1.4189,
+      "step": 3633
+    },
+    {
+      "epoch": 0.6269300439920642,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.560736408384278e-05,
+      "loss": 1.5909,
+      "step": 3634
+    },
+    {
+      "epoch": 0.6271025618907962,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.560511017572593e-05,
+      "loss": 1.4118,
+      "step": 3635
+    },
+    {
+      "epoch": 0.6272750797895281,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.56028558523475e-05,
+      "loss": 1.4832,
+      "step": 3636
+    },
+    {
+      "epoch": 0.6274475976882602,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.5600601113874497e-05,
+      "loss": 1.4742,
+      "step": 3637
+    },
+    {
+      "epoch": 0.6276201155869922,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.5598345960473965e-05,
+      "loss": 1.4938,
+      "step": 3638
+    },
+    {
+      "epoch": 0.6277926334857241,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.559609039231299e-05,
+      "loss": 1.4624,
+      "step": 3639
+    },
+    {
+      "epoch": 0.6279651513844562,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.559383440955867e-05,
+      "loss": 1.4947,
+      "step": 3640
+    },
+    {
+      "epoch": 0.6281376692831881,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.5591578012378148e-05,
+      "loss": 1.4709,
+      "step": 3641
+    },
+    {
+      "epoch": 0.6283101871819201,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.5589321200938584e-05,
+      "loss": 1.4289,
+      "step": 3642
+    },
+    {
+      "epoch": 0.6284827050806521,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.5587063975407185e-05,
+      "loss": 1.4556,
+      "step": 3643
+    },
+    {
+      "epoch": 0.6286552229793841,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.5584806335951177e-05,
+      "loss": 1.5221,
+      "step": 3644
+    },
+    {
+      "epoch": 0.6288277408781161,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.558254828273782e-05,
+      "loss": 1.407,
+      "step": 3645
+    },
+    {
+      "epoch": 0.6290002587768481,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.55802898159344e-05,
+      "loss": 1.5986,
+      "step": 3646
+    },
+    {
+      "epoch": 0.6291727766755801,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.557803093570825e-05,
+      "loss": 1.4878,
+      "step": 3647
+    },
+    {
+      "epoch": 0.6293452945743121,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.5575771642226715e-05,
+      "loss": 1.45,
+      "step": 3648
+    },
+    {
+      "epoch": 0.6295178124730441,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.5573511935657174e-05,
+      "loss": 1.3846,
+      "step": 3649
+    },
+    {
+      "epoch": 0.629690330371776,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.5571251816167047e-05,
+      "loss": 1.4499,
+      "step": 3650
+    },
+    {
+      "epoch": 0.6298628482705081,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.5568991283923772e-05,
+      "loss": 1.5331,
+      "step": 3651
+    },
+    {
+      "epoch": 0.63003536616924,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.5566730339094827e-05,
+      "loss": 1.4142,
+      "step": 3652
+    },
+    {
+      "epoch": 0.630207884067972,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.5564468981847716e-05,
+      "loss": 1.3816,
+      "step": 3653
+    },
+    {
+      "epoch": 0.6303804019667041,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.5562207212349975e-05,
+      "loss": 1.3857,
+      "step": 3654
+    },
+    {
+      "epoch": 0.630552919865436,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.555994503076917e-05,
+      "loss": 1.5437,
+      "step": 3655
+    },
+    {
+      "epoch": 0.630725437764168,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.5557682437272898e-05,
+      "loss": 1.4104,
+      "step": 3656
+    },
+    {
+      "epoch": 0.6308979556629001,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.555541943202878e-05,
+      "loss": 1.54,
+      "step": 3657
+    },
+    {
+      "epoch": 0.631070473561632,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.555315601520448e-05,
+      "loss": 1.4845,
+      "step": 3658
+    },
+    {
+      "epoch": 0.631242991460364,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.5550892186967685e-05,
+      "loss": 1.4748,
+      "step": 3659
+    },
+    {
+      "epoch": 0.631415509359096,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.5548627947486114e-05,
+      "loss": 1.3876,
+      "step": 3660
+    },
+    {
+      "epoch": 0.631588027257828,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.5546363296927518e-05,
+      "loss": 1.4506,
+      "step": 3661
+    },
+    {
+      "epoch": 0.63176054515656,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.5544098235459673e-05,
+      "loss": 1.4053,
+      "step": 3662
+    },
+    {
+      "epoch": 0.631933063055292,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.5541832763250386e-05,
+      "loss": 1.4933,
+      "step": 3663
+    },
+    {
+      "epoch": 0.632105580954024,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.5539566880467502e-05,
+      "loss": 1.5495,
+      "step": 3664
+    },
+    {
+      "epoch": 0.632278098852756,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.553730058727889e-05,
+      "loss": 1.3917,
+      "step": 3665
+    },
+    {
+      "epoch": 0.632450616751488,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.5535033883852456e-05,
+      "loss": 1.521,
+      "step": 3666
+    },
+    {
+      "epoch": 0.6326231346502199,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.5532766770356125e-05,
+      "loss": 1.5505,
+      "step": 3667
+    },
+    {
+      "epoch": 0.632795652548952,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.553049924695786e-05,
+      "loss": 1.563,
+      "step": 3668
+    },
+    {
+      "epoch": 0.6329681704476839,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.552823131382566e-05,
+      "loss": 1.4697,
+      "step": 3669
+    },
+    {
+      "epoch": 0.6331406883464159,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.5525962971127536e-05,
+      "loss": 1.3944,
+      "step": 3670
+    },
+    {
+      "epoch": 0.633313206245148,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.5523694219031548e-05,
+      "loss": 1.4171,
+      "step": 3671
+    },
+    {
+      "epoch": 0.6334857241438799,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.552142505770578e-05,
+      "loss": 1.4392,
+      "step": 3672
+    },
+    {
+      "epoch": 0.6336582420426119,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.5519155487318345e-05,
+      "loss": 1.4225,
+      "step": 3673
+    },
+    {
+      "epoch": 0.633830759941344,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.5516885508037388e-05,
+      "loss": 1.438,
+      "step": 3674
+    },
+    {
+      "epoch": 0.6340032778400759,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.5514615120031077e-05,
+      "loss": 1.448,
+      "step": 3675
+    },
+    {
+      "epoch": 0.6341757957388079,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.551234432346762e-05,
+      "loss": 1.5713,
+      "step": 3676
+    },
+    {
+      "epoch": 0.6343483136375399,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.551007311851526e-05,
+      "loss": 1.4049,
+      "step": 3677
+    },
+    {
+      "epoch": 0.6345208315362719,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.550780150534225e-05,
+      "loss": 1.4776,
+      "step": 3678
+    },
+    {
+      "epoch": 0.6346933494350039,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.550552948411689e-05,
+      "loss": 1.3499,
+      "step": 3679
+    },
+    {
+      "epoch": 0.6348658673337358,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.5503257055007502e-05,
+      "loss": 1.4269,
+      "step": 3680
+    },
+    {
+      "epoch": 0.6350383852324679,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.5500984218182452e-05,
+      "loss": 1.5338,
+      "step": 3681
+    },
+    {
+      "epoch": 0.6352109031311999,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.549871097381012e-05,
+      "loss": 1.5733,
+      "step": 3682
+    },
+    {
+      "epoch": 0.6353834210299318,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.5496437322058915e-05,
+      "loss": 1.518,
+      "step": 3683
+    },
+    {
+      "epoch": 0.6355559389286638,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.5494163263097294e-05,
+      "loss": 1.4033,
+      "step": 3684
+    },
+    {
+      "epoch": 0.6357284568273959,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.549188879709373e-05,
+      "loss": 1.5143,
+      "step": 3685
+    },
+    {
+      "epoch": 0.6359009747261278,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.5489613924216728e-05,
+      "loss": 1.4921,
+      "step": 3686
+    },
+    {
+      "epoch": 0.6360734926248598,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.5487338644634825e-05,
+      "loss": 1.5659,
+      "step": 3687
+    },
+    {
+      "epoch": 0.6362460105235919,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.548506295851659e-05,
+      "loss": 1.5478,
+      "step": 3688
+    },
+    {
+      "epoch": 0.6364185284223238,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.5482786866030618e-05,
+      "loss": 1.5006,
+      "step": 3689
+    },
+    {
+      "epoch": 0.6365910463210558,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.5480510367345537e-05,
+      "loss": 1.4564,
+      "step": 3690
+    },
+    {
+      "epoch": 0.6367635642197877,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.547823346263001e-05,
+      "loss": 1.4481,
+      "step": 3691
+    },
+    {
+      "epoch": 0.6369360821185198,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.5475956152052706e-05,
+      "loss": 1.5064,
+      "step": 3692
+    },
+    {
+      "epoch": 0.6371086000172518,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.5473678435782365e-05,
+      "loss": 1.4855,
+      "step": 3693
+    },
+    {
+      "epoch": 0.6372811179159837,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.547140031398772e-05,
+      "loss": 1.3907,
+      "step": 3694
+    },
+    {
+      "epoch": 0.6374536358147158,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.5469121786837553e-05,
+      "loss": 1.3983,
+      "step": 3695
+    },
+    {
+      "epoch": 0.6376261537134478,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.546684285450067e-05,
+      "loss": 1.3785,
+      "step": 3696
+    },
+    {
+      "epoch": 0.6377986716121797,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.5464563517145916e-05,
+      "loss": 1.5285,
+      "step": 3697
+    },
+    {
+      "epoch": 0.6379711895109118,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.546228377494215e-05,
+      "loss": 1.4713,
+      "step": 3698
+    },
+    {
+      "epoch": 0.6381437074096438,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.546000362805827e-05,
+      "loss": 1.5182,
+      "step": 3699
+    },
+    {
+      "epoch": 0.6383162253083757,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.5457723076663206e-05,
+      "loss": 1.5226,
+      "step": 3700
+    },
+    {
+      "epoch": 0.6383162253083757,
+      "eval_loss": 1.4285202026367188,
+      "eval_runtime": 10.826,
+      "eval_samples_per_second": 94.587,
+      "eval_steps_per_second": 23.647,
+      "step": 3700
+    },
+    {
+      "epoch": 0.6384887432071077,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.5455442120925916e-05,
+      "loss": 1.393,
+      "step": 3701
+    },
+    {
+      "epoch": 0.6386612611058398,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.5453160761015386e-05,
+      "loss": 1.4783,
+      "step": 3702
+    },
+    {
+      "epoch": 0.6388337790045717,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.5450878997100634e-05,
+      "loss": 1.4916,
+      "step": 3703
+    },
+    {
+      "epoch": 0.6390062969033037,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.5448596829350706e-05,
+      "loss": 1.4923,
+      "step": 3704
+    },
+    {
+      "epoch": 0.6391788148020358,
+      "grad_norm": 0.625,
+      "learning_rate": 1.544631425793468e-05,
+      "loss": 1.4529,
+      "step": 3705
+    },
+    {
+      "epoch": 0.6393513327007677,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.5444031283021668e-05,
+      "loss": 1.4464,
+      "step": 3706
+    },
+    {
+      "epoch": 0.6395238505994997,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.54417479047808e-05,
+      "loss": 1.4512,
+      "step": 3707
+    },
+    {
+      "epoch": 0.6396963684982316,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.5439464123381243e-05,
+      "loss": 1.4565,
+      "step": 3708
+    },
+    {
+      "epoch": 0.6398688863969637,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.54371799389922e-05,
+      "loss": 1.4812,
+      "step": 3709
+    },
+    {
+      "epoch": 0.6400414042956957,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.543489535178289e-05,
+      "loss": 1.4413,
+      "step": 3710
+    },
+    {
+      "epoch": 0.6402139221944276,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.5432610361922578e-05,
+      "loss": 1.3908,
+      "step": 3711
+    },
+    {
+      "epoch": 0.6403864400931597,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.5430324969580545e-05,
+      "loss": 1.4154,
+      "step": 3712
+    },
+    {
+      "epoch": 0.6405589579918917,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.5428039174926106e-05,
+      "loss": 1.4293,
+      "step": 3713
+    },
+    {
+      "epoch": 0.6407314758906236,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.5425752978128612e-05,
+      "loss": 1.4786,
+      "step": 3714
+    },
+    {
+      "epoch": 0.6409039937893557,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.5423466379357433e-05,
+      "loss": 1.4136,
+      "step": 3715
+    },
+    {
+      "epoch": 0.6410765116880877,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.542117937878198e-05,
+      "loss": 1.3674,
+      "step": 3716
+    },
+    {
+      "epoch": 0.6412490295868196,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.5418891976571682e-05,
+      "loss": 1.3893,
+      "step": 3717
+    },
+    {
+      "epoch": 0.6414215474855516,
+      "grad_norm": 4.875,
+      "learning_rate": 1.541660417289601e-05,
+      "loss": 1.524,
+      "step": 3718
+    },
+    {
+      "epoch": 0.6415940653842837,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.5414315967924454e-05,
+      "loss": 1.422,
+      "step": 3719
+    },
+    {
+      "epoch": 0.6417665832830156,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.5412027361826544e-05,
+      "loss": 1.486,
+      "step": 3720
+    },
+    {
+      "epoch": 0.6419391011817476,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.5409738354771832e-05,
+      "loss": 1.443,
+      "step": 3721
+    },
+    {
+      "epoch": 0.6421116190804796,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.5407448946929898e-05,
+      "loss": 1.4336,
+      "step": 3722
+    },
+    {
+      "epoch": 0.6422841369792116,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.540515913847036e-05,
+      "loss": 1.4494,
+      "step": 3723
+    },
+    {
+      "epoch": 0.6424566548779436,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.5402868929562857e-05,
+      "loss": 1.4959,
+      "step": 3724
+    },
+    {
+      "epoch": 0.6426291727766755,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.540057832037707e-05,
+      "loss": 1.3744,
+      "step": 3725
+    },
+    {
+      "epoch": 0.6428016906754076,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.539828731108269e-05,
+      "loss": 1.4017,
+      "step": 3726
+    },
+    {
+      "epoch": 0.6429742085741396,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.539599590184946e-05,
+      "loss": 1.5099,
+      "step": 3727
+    },
+    {
+      "epoch": 0.6431467264728715,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.5393704092847143e-05,
+      "loss": 1.4387,
+      "step": 3728
+    },
+    {
+      "epoch": 0.6433192443716036,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.5391411884245517e-05,
+      "loss": 1.3447,
+      "step": 3729
+    },
+    {
+      "epoch": 0.6434917622703356,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.5389119276214415e-05,
+      "loss": 1.5693,
+      "step": 3730
+    },
+    {
+      "epoch": 0.6436642801690675,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.5386826268923685e-05,
+      "loss": 1.4577,
+      "step": 3731
+    },
+    {
+      "epoch": 0.6438367980677996,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.5384532862543207e-05,
+      "loss": 1.4509,
+      "step": 3732
+    },
+    {
+      "epoch": 0.6440093159665315,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.538223905724289e-05,
+      "loss": 1.3712,
+      "step": 3733
+    },
+    {
+      "epoch": 0.6441818338652635,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.537994485319267e-05,
+      "loss": 1.4282,
+      "step": 3734
+    },
+    {
+      "epoch": 0.6443543517639955,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.537765025056252e-05,
+      "loss": 1.4045,
+      "step": 3735
+    },
+    {
+      "epoch": 0.6445268696627275,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.5375355249522444e-05,
+      "loss": 1.4643,
+      "step": 3736
+    },
+    {
+      "epoch": 0.6446993875614595,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.537305985024246e-05,
+      "loss": 1.5169,
+      "step": 3737
+    },
+    {
+      "epoch": 0.6448719054601915,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.5370764052892634e-05,
+      "loss": 1.4313,
+      "step": 3738
+    },
+    {
+      "epoch": 0.6450444233589235,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.5368467857643045e-05,
+      "loss": 1.534,
+      "step": 3739
+    },
+    {
+      "epoch": 0.6452169412576555,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.5366171264663816e-05,
+      "loss": 1.4798,
+      "step": 3740
+    },
+    {
+      "epoch": 0.6453894591563875,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.5363874274125086e-05,
+      "loss": 1.4276,
+      "step": 3741
+    },
+    {
+      "epoch": 0.6455619770551194,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.5361576886197037e-05,
+      "loss": 1.478,
+      "step": 3742
+    },
+    {
+      "epoch": 0.6457344949538515,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.535927910104987e-05,
+      "loss": 1.516,
+      "step": 3743
+    },
+    {
+      "epoch": 0.6459070128525835,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.5356980918853827e-05,
+      "loss": 1.4182,
+      "step": 3744
+    },
+    {
+      "epoch": 0.6460795307513154,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.535468233977916e-05,
+      "loss": 1.5294,
+      "step": 3745
+    },
+    {
+      "epoch": 0.6462520486500475,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.5352383363996172e-05,
+      "loss": 1.518,
+      "step": 3746
+    },
+    {
+      "epoch": 0.6464245665487794,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.5350083991675177e-05,
+      "loss": 1.4986,
+      "step": 3747
+    },
+    {
+      "epoch": 0.6465970844475114,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.534778422298653e-05,
+      "loss": 1.4611,
+      "step": 3748
+    },
+    {
+      "epoch": 0.6467696023462435,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.534548405810062e-05,
+      "loss": 1.5083,
+      "step": 3749
+    },
+    {
+      "epoch": 0.6469421202449754,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.5343183497187844e-05,
+      "loss": 1.5194,
+      "step": 3750
+    },
+    {
+      "epoch": 0.6471146381437074,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.534088254041865e-05,
+      "loss": 1.4361,
+      "step": 3751
+    },
+    {
+      "epoch": 0.6472871560424394,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.533858118796351e-05,
+      "loss": 1.4349,
+      "step": 3752
+    },
+    {
+      "epoch": 0.6474596739411714,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.5336279439992918e-05,
+      "loss": 1.4821,
+      "step": 3753
+    },
+    {
+      "epoch": 0.6476321918399034,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.53339772966774e-05,
+      "loss": 1.3555,
+      "step": 3754
+    },
+    {
+      "epoch": 0.6478047097386354,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.5331674758187516e-05,
+      "loss": 1.3945,
+      "step": 3755
+    },
+    {
+      "epoch": 0.6479772276373674,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.532937182469385e-05,
+      "loss": 1.4914,
+      "step": 3756
+    },
+    {
+      "epoch": 0.6481497455360994,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.5327068496367023e-05,
+      "loss": 1.5468,
+      "step": 3757
+    },
+    {
+      "epoch": 0.6483222634348313,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.5324764773377677e-05,
+      "loss": 1.471,
+      "step": 3758
+    },
+    {
+      "epoch": 0.6484947813335633,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.532246065589648e-05,
+      "loss": 1.468,
+      "step": 3759
+    },
+    {
+      "epoch": 0.6486672992322954,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.532015614409415e-05,
+      "loss": 1.4026,
+      "step": 3760
+    },
+    {
+      "epoch": 0.6488398171310273,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.5317851238141406e-05,
+      "loss": 1.5021,
+      "step": 3761
+    },
+    {
+      "epoch": 0.6490123350297593,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.5315545938209016e-05,
+      "loss": 1.3502,
+      "step": 3762
+    },
+    {
+      "epoch": 0.6491848529284914,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.531324024446777e-05,
+      "loss": 1.4444,
+      "step": 3763
+    },
+    {
+      "epoch": 0.6493573708272233,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.531093415708849e-05,
+      "loss": 1.4432,
+      "step": 3764
+    },
+    {
+      "epoch": 0.6495298887259553,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.530862767624202e-05,
+      "loss": 1.4107,
+      "step": 3765
+    },
+    {
+      "epoch": 0.6497024066246873,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.5306320802099243e-05,
+      "loss": 1.4785,
+      "step": 3766
+    },
+    {
+      "epoch": 0.6498749245234193,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.5304013534831064e-05,
+      "loss": 1.4472,
+      "step": 3767
+    },
+    {
+      "epoch": 0.6500474424221513,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.5301705874608423e-05,
+      "loss": 1.4185,
+      "step": 3768
+    },
+    {
+      "epoch": 0.6502199603208833,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.5299397821602284e-05,
+      "loss": 1.5577,
+      "step": 3769
+    },
+    {
+      "epoch": 0.6503924782196153,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.5297089375983644e-05,
+      "loss": 1.4669,
+      "step": 3770
+    },
+    {
+      "epoch": 0.6505649961183473,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.5294780537923523e-05,
+      "loss": 1.5208,
+      "step": 3771
+    },
+    {
+      "epoch": 0.6507375140170792,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.5292471307592975e-05,
+      "loss": 1.4643,
+      "step": 3772
+    },
+    {
+      "epoch": 0.6509100319158113,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.5290161685163086e-05,
+      "loss": 1.4022,
+      "step": 3773
+    },
+    {
+      "epoch": 0.6510825498145433,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.5287851670804963e-05,
+      "loss": 1.4932,
+      "step": 3774
+    },
+    {
+      "epoch": 0.6512550677132752,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.528554126468975e-05,
+      "loss": 1.4705,
+      "step": 3775
+    },
+    {
+      "epoch": 0.6514275856120072,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.5283230466988615e-05,
+      "loss": 1.3752,
+      "step": 3776
+    },
+    {
+      "epoch": 0.6516001035107393,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.5280919277872753e-05,
+      "loss": 1.4108,
+      "step": 3777
+    },
+    {
+      "epoch": 0.6517726214094712,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.5278607697513396e-05,
+      "loss": 1.4798,
+      "step": 3778
+    },
+    {
+      "epoch": 0.6519451393082032,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.52762957260818e-05,
+      "loss": 1.4351,
+      "step": 3779
+    },
+    {
+      "epoch": 0.6521176572069353,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.5273983363749246e-05,
+      "loss": 1.4795,
+      "step": 3780
+    },
+    {
+      "epoch": 0.6522901751056672,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.5271670610687058e-05,
+      "loss": 1.4506,
+      "step": 3781
+    },
+    {
+      "epoch": 0.6524626930043992,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.5269357467066566e-05,
+      "loss": 1.4128,
+      "step": 3782
+    },
+    {
+      "epoch": 0.6526352109031311,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.5267043933059147e-05,
+      "loss": 1.5003,
+      "step": 3783
+    },
+    {
+      "epoch": 0.6528077288018632,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.5264730008836205e-05,
+      "loss": 1.4195,
+      "step": 3784
+    },
+    {
+      "epoch": 0.6529802467005952,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.526241569456917e-05,
+      "loss": 1.4756,
+      "step": 3785
+    },
+    {
+      "epoch": 0.6531527645993271,
+      "grad_norm": 0.625,
+      "learning_rate": 1.52601009904295e-05,
+      "loss": 1.4688,
+      "step": 3786
+    },
+    {
+      "epoch": 0.6533252824980592,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.525778589658868e-05,
+      "loss": 1.4437,
+      "step": 3787
+    },
+    {
+      "epoch": 0.6534978003967912,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.5255470413218228e-05,
+      "loss": 1.485,
+      "step": 3788
+    },
+    {
+      "epoch": 0.6536703182955231,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.525315454048969e-05,
+      "loss": 1.3897,
+      "step": 3789
+    },
+    {
+      "epoch": 0.6538428361942552,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.525083827857464e-05,
+      "loss": 1.5565,
+      "step": 3790
+    },
+    {
+      "epoch": 0.6540153540929872,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.5248521627644684e-05,
+      "loss": 1.3609,
+      "step": 3791
+    },
+    {
+      "epoch": 0.6541878719917191,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.524620458787145e-05,
+      "loss": 1.4696,
+      "step": 3792
+    },
+    {
+      "epoch": 0.6543603898904511,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.5243887159426603e-05,
+      "loss": 1.4728,
+      "step": 3793
+    },
+    {
+      "epoch": 0.6545329077891832,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.5241569342481826e-05,
+      "loss": 1.4916,
+      "step": 3794
+    },
+    {
+      "epoch": 0.6547054256879151,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.5239251137208844e-05,
+      "loss": 1.3874,
+      "step": 3795
+    },
+    {
+      "epoch": 0.6548779435866471,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.52369325437794e-05,
+      "loss": 1.4444,
+      "step": 3796
+    },
+    {
+      "epoch": 0.6550504614853792,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.5234613562365272e-05,
+      "loss": 1.3956,
+      "step": 3797
+    },
+    {
+      "epoch": 0.6552229793841111,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.5232294193138264e-05,
+      "loss": 1.357,
+      "step": 3798
+    },
+    {
+      "epoch": 0.6553954972828431,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.5229974436270207e-05,
+      "loss": 1.4616,
+      "step": 3799
+    },
+    {
+      "epoch": 0.655568015181575,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.5227654291932967e-05,
+      "loss": 1.4532,
+      "step": 3800
+    },
+    {
+      "epoch": 0.655568015181575,
+      "eval_loss": 1.4269640445709229,
+      "eval_runtime": 10.8642,
+      "eval_samples_per_second": 94.254,
+      "eval_steps_per_second": 23.564,
+      "step": 3800
+    },
+    {
+      "epoch": 0.6557405330803071,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.5225333760298435e-05,
+      "loss": 1.5473,
+      "step": 3801
+    },
+    {
+      "epoch": 0.6559130509790391,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.5223012841538527e-05,
+      "loss": 1.5213,
+      "step": 3802
+    },
+    {
+      "epoch": 0.656085568877771,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.5220691535825194e-05,
+      "loss": 1.4789,
+      "step": 3803
+    },
+    {
+      "epoch": 0.6562580867765031,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.5218369843330409e-05,
+      "loss": 1.5483,
+      "step": 3804
+    },
+    {
+      "epoch": 0.6564306046752351,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.5216047764226183e-05,
+      "loss": 1.3859,
+      "step": 3805
+    },
+    {
+      "epoch": 0.656603122573967,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.5213725298684546e-05,
+      "loss": 1.4533,
+      "step": 3806
+    },
+    {
+      "epoch": 0.6567756404726991,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.5211402446877561e-05,
+      "loss": 1.417,
+      "step": 3807
+    },
+    {
+      "epoch": 0.6569481583714311,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.5209079208977322e-05,
+      "loss": 1.5,
+      "step": 3808
+    },
+    {
+      "epoch": 0.657120676270163,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.5206755585155946e-05,
+      "loss": 1.494,
+      "step": 3809
+    },
+    {
+      "epoch": 0.657293194168895,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.5204431575585586e-05,
+      "loss": 1.4712,
+      "step": 3810
+    },
+    {
+      "epoch": 0.657465712067627,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.520210718043841e-05,
+      "loss": 1.355,
+      "step": 3811
+    },
+    {
+      "epoch": 0.657638229966359,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.5199782399886633e-05,
+      "loss": 1.4093,
+      "step": 3812
+    },
+    {
+      "epoch": 0.657810747865091,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.519745723410249e-05,
+      "loss": 1.4952,
+      "step": 3813
+    },
+    {
+      "epoch": 0.657983265763823,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.5195131683258235e-05,
+      "loss": 1.5056,
+      "step": 3814
+    },
+    {
+      "epoch": 0.658155783662555,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.5192805747526168e-05,
+      "loss": 1.4134,
+      "step": 3815
+    },
+    {
+      "epoch": 0.658328301561287,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.5190479427078602e-05,
+      "loss": 1.5335,
+      "step": 3816
+    },
+    {
+      "epoch": 0.6585008194600189,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.518815272208789e-05,
+      "loss": 1.4615,
+      "step": 3817
+    },
+    {
+      "epoch": 0.658673337358751,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.5185825632726405e-05,
+      "loss": 1.3429,
+      "step": 3818
+    },
+    {
+      "epoch": 0.658845855257483,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.5183498159166557e-05,
+      "loss": 1.4464,
+      "step": 3819
+    },
+    {
+      "epoch": 0.6590183731562149,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.5181170301580776e-05,
+      "loss": 1.5144,
+      "step": 3820
+    },
+    {
+      "epoch": 0.659190891054947,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.5178842060141526e-05,
+      "loss": 1.4413,
+      "step": 3821
+    },
+    {
+      "epoch": 0.659363408953679,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.5176513435021297e-05,
+      "loss": 1.4433,
+      "step": 3822
+    },
+    {
+      "epoch": 0.6595359268524109,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.5174184426392609e-05,
+      "loss": 1.5579,
+      "step": 3823
+    },
+    {
+      "epoch": 0.659708444751143,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.5171855034428006e-05,
+      "loss": 1.2912,
+      "step": 3824
+    },
+    {
+      "epoch": 0.659880962649875,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.5169525259300071e-05,
+      "loss": 1.518,
+      "step": 3825
+    },
+    {
+      "epoch": 0.6600534805486069,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.5167195101181405e-05,
+      "loss": 1.4642,
+      "step": 3826
+    },
+    {
+      "epoch": 0.6602259984473389,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.5164864560244636e-05,
+      "loss": 1.3841,
+      "step": 3827
+    },
+    {
+      "epoch": 0.6603985163460709,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.516253363666243e-05,
+      "loss": 1.5328,
+      "step": 3828
+    },
+    {
+      "epoch": 0.6605710342448029,
+      "grad_norm": 0.625,
+      "learning_rate": 1.5160202330607476e-05,
+      "loss": 1.4713,
+      "step": 3829
+    },
+    {
+      "epoch": 0.6607435521435349,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.515787064225249e-05,
+      "loss": 1.4119,
+      "step": 3830
+    },
+    {
+      "epoch": 0.6609160700422669,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.515553857177022e-05,
+      "loss": 1.5059,
+      "step": 3831
+    },
+    {
+      "epoch": 0.6610885879409989,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.5153206119333436e-05,
+      "loss": 1.4441,
+      "step": 3832
+    },
+    {
+      "epoch": 0.6612611058397309,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.5150873285114948e-05,
+      "loss": 1.4158,
+      "step": 3833
+    },
+    {
+      "epoch": 0.6614336237384628,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.5148540069287583e-05,
+      "loss": 1.437,
+      "step": 3834
+    },
+    {
+      "epoch": 0.6616061416371949,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.5146206472024196e-05,
+      "loss": 1.4809,
+      "step": 3835
+    },
+    {
+      "epoch": 0.6617786595359268,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.5143872493497683e-05,
+      "loss": 1.4009,
+      "step": 3836
+    },
+    {
+      "epoch": 0.6619511774346588,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.5141538133880951e-05,
+      "loss": 1.4531,
+      "step": 3837
+    },
+    {
+      "epoch": 0.6621236953333909,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.5139203393346953e-05,
+      "loss": 1.4045,
+      "step": 3838
+    },
+    {
+      "epoch": 0.6622962132321228,
+      "grad_norm": 1.203125,
+      "learning_rate": 1.5136868272068653e-05,
+      "loss": 1.4653,
+      "step": 3839
+    },
+    {
+      "epoch": 0.6624687311308548,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.5134532770219054e-05,
+      "loss": 1.5117,
+      "step": 3840
+    },
+    {
+      "epoch": 0.6626412490295868,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.513219688797119e-05,
+      "loss": 1.4574,
+      "step": 3841
+    },
+    {
+      "epoch": 0.6628137669283188,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.512986062549811e-05,
+      "loss": 1.4325,
+      "step": 3842
+    },
+    {
+      "epoch": 0.6629862848270508,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.51275239829729e-05,
+      "loss": 1.6238,
+      "step": 3843
+    },
+    {
+      "epoch": 0.6631588027257828,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.5125186960568678e-05,
+      "loss": 1.4831,
+      "step": 3844
+    },
+    {
+      "epoch": 0.6633313206245148,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.5122849558458583e-05,
+      "loss": 1.3717,
+      "step": 3845
+    },
+    {
+      "epoch": 0.6635038385232468,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.512051177681578e-05,
+      "loss": 1.3688,
+      "step": 3846
+    },
+    {
+      "epoch": 0.6636763564219788,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.5118173615813474e-05,
+      "loss": 1.4871,
+      "step": 3847
+    },
+    {
+      "epoch": 0.6638488743207108,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.5115835075624885e-05,
+      "loss": 1.4088,
+      "step": 3848
+    },
+    {
+      "epoch": 0.6640213922194428,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.5113496156423271e-05,
+      "loss": 1.3805,
+      "step": 3849
+    },
+    {
+      "epoch": 0.6641939101181747,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.5111156858381906e-05,
+      "loss": 1.531,
+      "step": 3850
+    },
+    {
+      "epoch": 0.6643664280169067,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.510881718167411e-05,
+      "loss": 1.494,
+      "step": 3851
+    },
+    {
+      "epoch": 0.6645389459156388,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.5106477126473209e-05,
+      "loss": 1.4728,
+      "step": 3852
+    },
+    {
+      "epoch": 0.6647114638143707,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.5104136692952582e-05,
+      "loss": 1.4782,
+      "step": 3853
+    },
+    {
+      "epoch": 0.6648839817131027,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.5101795881285614e-05,
+      "loss": 1.4778,
+      "step": 3854
+    },
+    {
+      "epoch": 0.6650564996118348,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.5099454691645731e-05,
+      "loss": 1.4344,
+      "step": 3855
+    },
+    {
+      "epoch": 0.6652290175105667,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.509711312420638e-05,
+      "loss": 1.4528,
+      "step": 3856
+    },
+    {
+      "epoch": 0.6654015354092987,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.5094771179141043e-05,
+      "loss": 1.4104,
+      "step": 3857
+    },
+    {
+      "epoch": 0.6655740533080307,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.5092428856623222e-05,
+      "loss": 1.4662,
+      "step": 3858
+    },
+    {
+      "epoch": 0.6657465712067627,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.509008615682645e-05,
+      "loss": 1.4011,
+      "step": 3859
+    },
+    {
+      "epoch": 0.6659190891054947,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.5087743079924293e-05,
+      "loss": 1.3801,
+      "step": 3860
+    },
+    {
+      "epoch": 0.6660916070042266,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.5085399626090343e-05,
+      "loss": 1.478,
+      "step": 3861
+    },
+    {
+      "epoch": 0.6662641249029587,
+      "grad_norm": 0.625,
+      "learning_rate": 1.5083055795498209e-05,
+      "loss": 1.5616,
+      "step": 3862
+    },
+    {
+      "epoch": 0.6664366428016907,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.5080711588321544e-05,
+      "loss": 1.4164,
+      "step": 3863
+    },
+    {
+      "epoch": 0.6666091607004226,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.5078367004734014e-05,
+      "loss": 1.4833,
+      "step": 3864
+    },
+    {
+      "epoch": 0.6667816785991547,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.507602204490933e-05,
+      "loss": 1.4232,
+      "step": 3865
+    },
+    {
+      "epoch": 0.6669541964978867,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.5073676709021214e-05,
+      "loss": 1.4294,
+      "step": 3866
+    },
+    {
+      "epoch": 0.6671267143966186,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.5071330997243427e-05,
+      "loss": 1.4871,
+      "step": 3867
+    },
+    {
+      "epoch": 0.6672992322953506,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.506898490974975e-05,
+      "loss": 1.3962,
+      "step": 3868
+    },
+    {
+      "epoch": 0.6674717501940827,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.5066638446714002e-05,
+      "loss": 1.452,
+      "step": 3869
+    },
+    {
+      "epoch": 0.6676442680928146,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.5064291608310017e-05,
+      "loss": 1.5522,
+      "step": 3870
+    },
+    {
+      "epoch": 0.6678167859915466,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.5061944394711669e-05,
+      "loss": 1.5069,
+      "step": 3871
+    },
+    {
+      "epoch": 0.6679893038902787,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.505959680609285e-05,
+      "loss": 1.4598,
+      "step": 3872
+    },
+    {
+      "epoch": 0.6681618217890106,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.5057248842627488e-05,
+      "loss": 1.4333,
+      "step": 3873
+    },
+    {
+      "epoch": 0.6683343396877426,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.505490050448953e-05,
+      "loss": 1.4226,
+      "step": 3874
+    },
+    {
+      "epoch": 0.6685068575864745,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.505255179185296e-05,
+      "loss": 1.4309,
+      "step": 3875
+    },
+    {
+      "epoch": 0.6686793754852066,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.5050202704891783e-05,
+      "loss": 1.4525,
+      "step": 3876
+    },
+    {
+      "epoch": 0.6688518933839386,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.5047853243780033e-05,
+      "loss": 1.4938,
+      "step": 3877
+    },
+    {
+      "epoch": 0.6690244112826705,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.5045503408691776e-05,
+      "loss": 1.3831,
+      "step": 3878
+    },
+    {
+      "epoch": 0.6691969291814026,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.5043153199801101e-05,
+      "loss": 1.5183,
+      "step": 3879
+    },
+    {
+      "epoch": 0.6693694470801346,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.5040802617282124e-05,
+      "loss": 1.4216,
+      "step": 3880
+    },
+    {
+      "epoch": 0.6695419649788665,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.5038451661308994e-05,
+      "loss": 1.3953,
+      "step": 3881
+    },
+    {
+      "epoch": 0.6697144828775986,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.5036100332055884e-05,
+      "loss": 1.5554,
+      "step": 3882
+    },
+    {
+      "epoch": 0.6698870007763306,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.5033748629696994e-05,
+      "loss": 1.4274,
+      "step": 3883
+    },
+    {
+      "epoch": 0.6700595186750625,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.5031396554406548e-05,
+      "loss": 1.4097,
+      "step": 3884
+    },
+    {
+      "epoch": 0.6702320365737945,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.5029044106358815e-05,
+      "loss": 1.4944,
+      "step": 3885
+    },
+    {
+      "epoch": 0.6704045544725266,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.5026691285728067e-05,
+      "loss": 1.4382,
+      "step": 3886
+    },
+    {
+      "epoch": 0.6705770723712585,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.5024338092688622e-05,
+      "loss": 1.3788,
+      "step": 3887
+    },
+    {
+      "epoch": 0.6707495902699905,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.502198452741482e-05,
+      "loss": 1.4075,
+      "step": 3888
+    },
+    {
+      "epoch": 0.6709221081687226,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.5019630590081025e-05,
+      "loss": 1.5182,
+      "step": 3889
+    },
+    {
+      "epoch": 0.6710946260674545,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.5017276280861628e-05,
+      "loss": 1.5196,
+      "step": 3890
+    },
+    {
+      "epoch": 0.6712671439661865,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.5014921599931056e-05,
+      "loss": 1.3461,
+      "step": 3891
+    },
+    {
+      "epoch": 0.6714396618649184,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.5012566547463758e-05,
+      "loss": 1.4824,
+      "step": 3892
+    },
+    {
+      "epoch": 0.6716121797636505,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.501021112363421e-05,
+      "loss": 1.5494,
+      "step": 3893
+    },
+    {
+      "epoch": 0.6717846976623825,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.5007855328616917e-05,
+      "loss": 1.4709,
+      "step": 3894
+    },
+    {
+      "epoch": 0.6719572155611144,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.500549916258641e-05,
+      "loss": 1.3718,
+      "step": 3895
+    },
+    {
+      "epoch": 0.6721297334598465,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.500314262571725e-05,
+      "loss": 1.347,
+      "step": 3896
+    },
+    {
+      "epoch": 0.6723022513585785,
+      "grad_norm": 0.75,
+      "learning_rate": 1.5000785718184026e-05,
+      "loss": 1.525,
+      "step": 3897
+    },
+    {
+      "epoch": 0.6724747692573104,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.4998428440161345e-05,
+      "loss": 1.4657,
+      "step": 3898
+    },
+    {
+      "epoch": 0.6726472871560425,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.4996070791823856e-05,
+      "loss": 1.4022,
+      "step": 3899
+    },
+    {
+      "epoch": 0.6728198050547745,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.4993712773346225e-05,
+      "loss": 1.3463,
+      "step": 3900
+    },
+    {
+      "epoch": 0.6728198050547745,
+      "eval_loss": 1.4253987073898315,
+      "eval_runtime": 10.893,
+      "eval_samples_per_second": 94.005,
+      "eval_steps_per_second": 23.501,
+      "step": 3900
+    },
+    {
+      "epoch": 0.6729923229535064,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4991354384903148e-05,
+      "loss": 1.4465,
+      "step": 3901
+    },
+    {
+      "epoch": 0.6731648408522384,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.4988995626669352e-05,
+      "loss": 1.4454,
+      "step": 3902
+    },
+    {
+      "epoch": 0.6733373587509704,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.4986636498819586e-05,
+      "loss": 1.3895,
+      "step": 3903
+    },
+    {
+      "epoch": 0.6735098766497024,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.4984277001528634e-05,
+      "loss": 1.4343,
+      "step": 3904
+    },
+    {
+      "epoch": 0.6736823945484344,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.4981917134971298e-05,
+      "loss": 1.5163,
+      "step": 3905
+    },
+    {
+      "epoch": 0.6738549124471664,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.497955689932241e-05,
+      "loss": 1.4312,
+      "step": 3906
+    },
+    {
+      "epoch": 0.6740274303458984,
+      "grad_norm": 0.75,
+      "learning_rate": 1.4977196294756832e-05,
+      "loss": 1.4852,
+      "step": 3907
+    },
+    {
+      "epoch": 0.6741999482446304,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.4974835321449454e-05,
+      "loss": 1.3581,
+      "step": 3908
+    },
+    {
+      "epoch": 0.6743724661433623,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.4972473979575195e-05,
+      "loss": 1.4458,
+      "step": 3909
+    },
+    {
+      "epoch": 0.6745449840420944,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.4970112269308991e-05,
+      "loss": 1.5073,
+      "step": 3910
+    },
+    {
+      "epoch": 0.6747175019408264,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.4967750190825816e-05,
+      "loss": 1.4186,
+      "step": 3911
+    },
+    {
+      "epoch": 0.6748900198395583,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.4965387744300665e-05,
+      "loss": 1.498,
+      "step": 3912
+    },
+    {
+      "epoch": 0.6750625377382904,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.4963024929908568e-05,
+      "loss": 1.541,
+      "step": 3913
+    },
+    {
+      "epoch": 0.6752350556370224,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.4960661747824575e-05,
+      "loss": 1.451,
+      "step": 3914
+    },
+    {
+      "epoch": 0.6754075735357543,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.495829819822376e-05,
+      "loss": 1.4844,
+      "step": 3915
+    },
+    {
+      "epoch": 0.6755800914344863,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.4955934281281234e-05,
+      "loss": 1.3985,
+      "step": 3916
+    },
+    {
+      "epoch": 0.6757526093332183,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.4953569997172133e-05,
+      "loss": 1.4535,
+      "step": 3917
+    },
+    {
+      "epoch": 0.6759251272319503,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.4951205346071613e-05,
+      "loss": 1.4962,
+      "step": 3918
+    },
+    {
+      "epoch": 0.6760976451306823,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.4948840328154863e-05,
+      "loss": 1.4271,
+      "step": 3919
+    },
+    {
+      "epoch": 0.6762701630294143,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.4946474943597103e-05,
+      "loss": 1.3688,
+      "step": 3920
+    },
+    {
+      "epoch": 0.6764426809281463,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.4944109192573572e-05,
+      "loss": 1.4197,
+      "step": 3921
+    },
+    {
+      "epoch": 0.6766151988268783,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.4941743075259539e-05,
+      "loss": 1.4688,
+      "step": 3922
+    },
+    {
+      "epoch": 0.6767877167256103,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.4939376591830301e-05,
+      "loss": 1.4973,
+      "step": 3923
+    },
+    {
+      "epoch": 0.6769602346243423,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.4937009742461181e-05,
+      "loss": 1.3811,
+      "step": 3924
+    },
+    {
+      "epoch": 0.6771327525230743,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.4934642527327537e-05,
+      "loss": 1.477,
+      "step": 3925
+    },
+    {
+      "epoch": 0.6773052704218062,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.4932274946604736e-05,
+      "loss": 1.3482,
+      "step": 3926
+    },
+    {
+      "epoch": 0.6774777883205383,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.492990700046819e-05,
+      "loss": 1.4613,
+      "step": 3927
+    },
+    {
+      "epoch": 0.6776503062192702,
+      "grad_norm": 0.625,
+      "learning_rate": 1.492753868909333e-05,
+      "loss": 1.4796,
+      "step": 3928
+    },
+    {
+      "epoch": 0.6778228241180022,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.4925170012655614e-05,
+      "loss": 1.4654,
+      "step": 3929
+    },
+    {
+      "epoch": 0.6779953420167343,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.4922800971330533e-05,
+      "loss": 1.4711,
+      "step": 3930
+    },
+    {
+      "epoch": 0.6781678599154662,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.4920431565293596e-05,
+      "loss": 1.5027,
+      "step": 3931
+    },
+    {
+      "epoch": 0.6783403778141982,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.4918061794720342e-05,
+      "loss": 1.4228,
+      "step": 3932
+    },
+    {
+      "epoch": 0.6785128957129302,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.4915691659786342e-05,
+      "loss": 1.4788,
+      "step": 3933
+    },
+    {
+      "epoch": 0.6786854136116622,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.4913321160667189e-05,
+      "loss": 1.5247,
+      "step": 3934
+    },
+    {
+      "epoch": 0.6788579315103942,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.4910950297538505e-05,
+      "loss": 1.4611,
+      "step": 3935
+    },
+    {
+      "epoch": 0.6790304494091262,
+      "grad_norm": 0.625,
+      "learning_rate": 1.4908579070575936e-05,
+      "loss": 1.5271,
+      "step": 3936
+    },
+    {
+      "epoch": 0.6792029673078582,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.4906207479955164e-05,
+      "loss": 1.4239,
+      "step": 3937
+    },
+    {
+      "epoch": 0.6793754852065902,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.4903835525851884e-05,
+      "loss": 1.413,
+      "step": 3938
+    },
+    {
+      "epoch": 0.6795480031053222,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.4901463208441827e-05,
+      "loss": 1.4711,
+      "step": 3939
+    },
+    {
+      "epoch": 0.6797205210040542,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.489909052790075e-05,
+      "loss": 1.4645,
+      "step": 3940
+    },
+    {
+      "epoch": 0.6798930389027862,
+      "grad_norm": 0.75,
+      "learning_rate": 1.4896717484404437e-05,
+      "loss": 1.4735,
+      "step": 3941
+    },
+    {
+      "epoch": 0.6800655568015181,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.4894344078128696e-05,
+      "loss": 1.4367,
+      "step": 3942
+    },
+    {
+      "epoch": 0.6802380747002501,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.4891970309249361e-05,
+      "loss": 1.5428,
+      "step": 3943
+    },
+    {
+      "epoch": 0.6804105925989822,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.4889596177942303e-05,
+      "loss": 1.3409,
+      "step": 3944
+    },
+    {
+      "epoch": 0.6805831104977141,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.4887221684383407e-05,
+      "loss": 1.5084,
+      "step": 3945
+    },
+    {
+      "epoch": 0.6807556283964461,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.4884846828748594e-05,
+      "loss": 1.4532,
+      "step": 3946
+    },
+    {
+      "epoch": 0.6809281462951782,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.4882471611213802e-05,
+      "loss": 1.4367,
+      "step": 3947
+    },
+    {
+      "epoch": 0.6811006641939101,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.488009603195501e-05,
+      "loss": 1.4749,
+      "step": 3948
+    },
+    {
+      "epoch": 0.6812731820926421,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.4877720091148209e-05,
+      "loss": 1.4333,
+      "step": 3949
+    },
+    {
+      "epoch": 0.681445699991374,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.4875343788969426e-05,
+      "loss": 1.3904,
+      "step": 3950
+    },
+    {
+      "epoch": 0.6816182178901061,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.4872967125594713e-05,
+      "loss": 1.4519,
+      "step": 3951
+    },
+    {
+      "epoch": 0.6817907357888381,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.4870590101200148e-05,
+      "loss": 1.4636,
+      "step": 3952
+    },
+    {
+      "epoch": 0.68196325368757,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.4868212715961838e-05,
+      "loss": 1.5242,
+      "step": 3953
+    },
+    {
+      "epoch": 0.6821357715863021,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.486583497005591e-05,
+      "loss": 1.4555,
+      "step": 3954
+    },
+    {
+      "epoch": 0.6823082894850341,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.4863456863658522e-05,
+      "loss": 1.5198,
+      "step": 3955
+    },
+    {
+      "epoch": 0.682480807383766,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.4861078396945865e-05,
+      "loss": 1.3825,
+      "step": 3956
+    },
+    {
+      "epoch": 0.6826533252824981,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.4858699570094144e-05,
+      "loss": 1.3163,
+      "step": 3957
+    },
+    {
+      "epoch": 0.6828258431812301,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.4856320383279603e-05,
+      "loss": 1.3736,
+      "step": 3958
+    },
+    {
+      "epoch": 0.682998361079962,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.4853940836678504e-05,
+      "loss": 1.361,
+      "step": 3959
+    },
+    {
+      "epoch": 0.683170878978694,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.4851560930467137e-05,
+      "loss": 1.4556,
+      "step": 3960
+    },
+    {
+      "epoch": 0.6833433968774261,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.4849180664821822e-05,
+      "loss": 1.5069,
+      "step": 3961
+    },
+    {
+      "epoch": 0.683515914776158,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.4846800039918908e-05,
+      "loss": 1.4745,
+      "step": 3962
+    },
+    {
+      "epoch": 0.68368843267489,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.4844419055934761e-05,
+      "loss": 1.415,
+      "step": 3963
+    },
+    {
+      "epoch": 0.6838609505736221,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.4842037713045778e-05,
+      "loss": 1.3378,
+      "step": 3964
+    },
+    {
+      "epoch": 0.684033468472354,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.483965601142839e-05,
+      "loss": 1.5044,
+      "step": 3965
+    },
+    {
+      "epoch": 0.684205986371086,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.4837273951259044e-05,
+      "loss": 1.4319,
+      "step": 3966
+    },
+    {
+      "epoch": 0.6843785042698179,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.4834891532714218e-05,
+      "loss": 1.3895,
+      "step": 3967
+    },
+    {
+      "epoch": 0.68455102216855,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.4832508755970418e-05,
+      "loss": 1.3703,
+      "step": 3968
+    },
+    {
+      "epoch": 0.684723540067282,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.4830125621204177e-05,
+      "loss": 1.4502,
+      "step": 3969
+    },
+    {
+      "epoch": 0.6848960579660139,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.4827742128592046e-05,
+      "loss": 1.4776,
+      "step": 3970
+    },
+    {
+      "epoch": 0.685068575864746,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.4825358278310615e-05,
+      "loss": 1.4527,
+      "step": 3971
+    },
+    {
+      "epoch": 0.685241093763478,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.4822974070536493e-05,
+      "loss": 1.4456,
+      "step": 3972
+    },
+    {
+      "epoch": 0.6854136116622099,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.4820589505446316e-05,
+      "loss": 1.4968,
+      "step": 3973
+    },
+    {
+      "epoch": 0.6855861295609419,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.4818204583216749e-05,
+      "loss": 1.4718,
+      "step": 3974
+    },
+    {
+      "epoch": 0.685758647459674,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.4815819304024482e-05,
+      "loss": 1.4568,
+      "step": 3975
+    },
+    {
+      "epoch": 0.6859311653584059,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.4813433668046227e-05,
+      "loss": 1.4359,
+      "step": 3976
+    },
+    {
+      "epoch": 0.6861036832571379,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.4811047675458729e-05,
+      "loss": 1.4681,
+      "step": 3977
+    },
+    {
+      "epoch": 0.68627620115587,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.4808661326438765e-05,
+      "loss": 1.4685,
+      "step": 3978
+    },
+    {
+      "epoch": 0.6864487190546019,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.480627462116312e-05,
+      "loss": 1.4094,
+      "step": 3979
+    },
+    {
+      "epoch": 0.6866212369533339,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.4803887559808618e-05,
+      "loss": 1.4911,
+      "step": 3980
+    },
+    {
+      "epoch": 0.686793754852066,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.4801500142552112e-05,
+      "loss": 1.4285,
+      "step": 3981
+    },
+    {
+      "epoch": 0.6869662727507979,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.4799112369570475e-05,
+      "loss": 1.4582,
+      "step": 3982
+    },
+    {
+      "epoch": 0.6871387906495299,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.4796724241040604e-05,
+      "loss": 1.4502,
+      "step": 3983
+    },
+    {
+      "epoch": 0.6873113085482618,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.479433575713943e-05,
+      "loss": 1.5991,
+      "step": 3984
+    },
+    {
+      "epoch": 0.6874838264469939,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.4791946918043911e-05,
+      "loss": 1.5316,
+      "step": 3985
+    },
+    {
+      "epoch": 0.6876563443457259,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.478955772393102e-05,
+      "loss": 1.4841,
+      "step": 3986
+    },
+    {
+      "epoch": 0.6878288622444578,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.4787168174977768e-05,
+      "loss": 1.4967,
+      "step": 3987
+    },
+    {
+      "epoch": 0.6880013801431899,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.4784778271361185e-05,
+      "loss": 1.4086,
+      "step": 3988
+    },
+    {
+      "epoch": 0.6881738980419219,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.478238801325833e-05,
+      "loss": 1.4017,
+      "step": 3989
+    },
+    {
+      "epoch": 0.6883464159406538,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.4779997400846292e-05,
+      "loss": 1.3941,
+      "step": 3990
+    },
+    {
+      "epoch": 0.6885189338393858,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.4777606434302176e-05,
+      "loss": 1.3313,
+      "step": 3991
+    },
+    {
+      "epoch": 0.6886914517381179,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.4775215113803127e-05,
+      "loss": 1.3779,
+      "step": 3992
+    },
+    {
+      "epoch": 0.6888639696368498,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.4772823439526302e-05,
+      "loss": 1.4789,
+      "step": 3993
+    },
+    {
+      "epoch": 0.6890364875355818,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.4770431411648898e-05,
+      "loss": 1.5234,
+      "step": 3994
+    },
+    {
+      "epoch": 0.6892090054343138,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.4768039030348127e-05,
+      "loss": 1.5431,
+      "step": 3995
+    },
+    {
+      "epoch": 0.6893815233330458,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.4765646295801231e-05,
+      "loss": 1.395,
+      "step": 3996
+    },
+    {
+      "epoch": 0.6895540412317778,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.4763253208185482e-05,
+      "loss": 1.391,
+      "step": 3997
+    },
+    {
+      "epoch": 0.6897265591305098,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.476085976767817e-05,
+      "loss": 1.4651,
+      "step": 3998
+    },
+    {
+      "epoch": 0.6898990770292418,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.4758465974456622e-05,
+      "loss": 1.5271,
+      "step": 3999
+    },
+    {
+      "epoch": 0.6900715949279738,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.4756071828698182e-05,
+      "loss": 1.4793,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6900715949279738,
+      "eval_loss": 1.4241275787353516,
+      "eval_runtime": 10.9175,
+      "eval_samples_per_second": 93.794,
+      "eval_steps_per_second": 23.449,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6902441128267057,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.4753677330580223e-05,
+      "loss": 1.4217,
+      "step": 4001
+    },
+    {
+      "epoch": 0.6904166307254378,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.4751282480280147e-05,
+      "loss": 1.4645,
+      "step": 4002
+    },
+    {
+      "epoch": 0.6905891486241698,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.4748887277975376e-05,
+      "loss": 1.4786,
+      "step": 4003
+    },
+    {
+      "epoch": 0.6907616665229017,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.4746491723843364e-05,
+      "loss": 1.4642,
+      "step": 4004
+    },
+    {
+      "epoch": 0.6909341844216338,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.4744095818061586e-05,
+      "loss": 1.4822,
+      "step": 4005
+    },
+    {
+      "epoch": 0.6911067023203658,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.474169956080755e-05,
+      "loss": 1.4705,
+      "step": 4006
+    },
+    {
+      "epoch": 0.6912792202190977,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.473930295225878e-05,
+      "loss": 1.512,
+      "step": 4007
+    },
+    {
+      "epoch": 0.6914517381178297,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.4736905992592837e-05,
+      "loss": 1.5389,
+      "step": 4008
+    },
+    {
+      "epoch": 0.6916242560165617,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.4734508681987296e-05,
+      "loss": 1.306,
+      "step": 4009
+    },
+    {
+      "epoch": 0.6917967739152937,
+      "grad_norm": 0.625,
+      "learning_rate": 1.4732111020619775e-05,
+      "loss": 1.4237,
+      "step": 4010
+    },
+    {
+      "epoch": 0.6919692918140257,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.4729713008667901e-05,
+      "loss": 1.4676,
+      "step": 4011
+    },
+    {
+      "epoch": 0.6921418097127577,
+      "grad_norm": 0.625,
+      "learning_rate": 1.4727314646309333e-05,
+      "loss": 1.5024,
+      "step": 4012
+    },
+    {
+      "epoch": 0.6923143276114897,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.4724915933721758e-05,
+      "loss": 1.3662,
+      "step": 4013
+    },
+    {
+      "epoch": 0.6924868455102217,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.4722516871082886e-05,
+      "loss": 1.4298,
+      "step": 4014
+    },
+    {
+      "epoch": 0.6926593634089537,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.472011745857046e-05,
+      "loss": 1.3987,
+      "step": 4015
+    },
+    {
+      "epoch": 0.6928318813076857,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.471771769636224e-05,
+      "loss": 1.4815,
+      "step": 4016
+    },
+    {
+      "epoch": 0.6930043992064177,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.4715317584636012e-05,
+      "loss": 1.524,
+      "step": 4017
+    },
+    {
+      "epoch": 0.6931769171051496,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.47129171235696e-05,
+      "loss": 1.4881,
+      "step": 4018
+    },
+    {
+      "epoch": 0.6933494350038817,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.4710516313340836e-05,
+      "loss": 1.5024,
+      "step": 4019
+    },
+    {
+      "epoch": 0.6935219529026136,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.470811515412759e-05,
+      "loss": 1.4709,
+      "step": 4020
+    },
+    {
+      "epoch": 0.6936944708013456,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.470571364610776e-05,
+      "loss": 1.4479,
+      "step": 4021
+    },
+    {
+      "epoch": 0.6938669887000777,
+      "grad_norm": 0.75,
+      "learning_rate": 1.4703311789459261e-05,
+      "loss": 1.4529,
+      "step": 4022
+    },
+    {
+      "epoch": 0.6940395065988096,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.4700909584360032e-05,
+      "loss": 1.3779,
+      "step": 4023
+    },
+    {
+      "epoch": 0.6942120244975416,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.4698507030988049e-05,
+      "loss": 1.482,
+      "step": 4024
+    },
+    {
+      "epoch": 0.6943845423962736,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.469610412952131e-05,
+      "loss": 1.5018,
+      "step": 4025
+    },
+    {
+      "epoch": 0.6945570602950056,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.4693700880137835e-05,
+      "loss": 1.4349,
+      "step": 4026
+    },
+    {
+      "epoch": 0.6947295781937376,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.4691297283015669e-05,
+      "loss": 1.4857,
+      "step": 4027
+    },
+    {
+      "epoch": 0.6949020960924696,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.4688893338332888e-05,
+      "loss": 1.4458,
+      "step": 4028
+    },
+    {
+      "epoch": 0.6950746139912016,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.4686489046267591e-05,
+      "loss": 1.3835,
+      "step": 4029
+    },
+    {
+      "epoch": 0.6952471318899336,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.4684084406997903e-05,
+      "loss": 1.4679,
+      "step": 4030
+    },
+    {
+      "epoch": 0.6954196497886656,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.4681679420701972e-05,
+      "loss": 1.4534,
+      "step": 4031
+    },
+    {
+      "epoch": 0.6955921676873976,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.4679274087557981e-05,
+      "loss": 1.4716,
+      "step": 4032
+    },
+    {
+      "epoch": 0.6957646855861296,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.4676868407744126e-05,
+      "loss": 1.4279,
+      "step": 4033
+    },
+    {
+      "epoch": 0.6959372034848615,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.467446238143864e-05,
+      "loss": 1.3526,
+      "step": 4034
+    },
+    {
+      "epoch": 0.6961097213835935,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.467205600881977e-05,
+      "loss": 1.4274,
+      "step": 4035
+    },
+    {
+      "epoch": 0.6962822392823256,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.4669649290065801e-05,
+      "loss": 1.3703,
+      "step": 4036
+    },
+    {
+      "epoch": 0.6964547571810575,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.4667242225355034e-05,
+      "loss": 1.4281,
+      "step": 4037
+    },
+    {
+      "epoch": 0.6966272750797895,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.4664834814865802e-05,
+      "loss": 1.5386,
+      "step": 4038
+    },
+    {
+      "epoch": 0.6967997929785216,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.4662427058776459e-05,
+      "loss": 1.5138,
+      "step": 4039
+    },
+    {
+      "epoch": 0.6969723108772535,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.4660018957265386e-05,
+      "loss": 1.4275,
+      "step": 4040
+    },
+    {
+      "epoch": 0.6971448287759855,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.4657610510510992e-05,
+      "loss": 1.415,
+      "step": 4041
+    },
+    {
+      "epoch": 0.6973173466747175,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.4655201718691712e-05,
+      "loss": 1.5508,
+      "step": 4042
+    },
+    {
+      "epoch": 0.6974898645734495,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.4652792581985997e-05,
+      "loss": 1.4285,
+      "step": 4043
+    },
+    {
+      "epoch": 0.6976623824721815,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.4650383100572338e-05,
+      "loss": 1.4697,
+      "step": 4044
+    },
+    {
+      "epoch": 0.6978349003709134,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.464797327462924e-05,
+      "loss": 1.4522,
+      "step": 4045
+    },
+    {
+      "epoch": 0.6980074182696455,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.4645563104335243e-05,
+      "loss": 1.5441,
+      "step": 4046
+    },
+    {
+      "epoch": 0.6981799361683775,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.4643152589868904e-05,
+      "loss": 1.499,
+      "step": 4047
+    },
+    {
+      "epoch": 0.6983524540671094,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.4640741731408805e-05,
+      "loss": 1.4008,
+      "step": 4048
+    },
+    {
+      "epoch": 0.6985249719658414,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.4638330529133566e-05,
+      "loss": 1.3955,
+      "step": 4049
+    },
+    {
+      "epoch": 0.6986974898645735,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.4635918983221823e-05,
+      "loss": 1.4477,
+      "step": 4050
+    },
+    {
+      "epoch": 0.6988700077633054,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.4633507093852229e-05,
+      "loss": 1.5475,
+      "step": 4051
+    },
+    {
+      "epoch": 0.6990425256620374,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.4631094861203478e-05,
+      "loss": 1.3691,
+      "step": 4052
+    },
+    {
+      "epoch": 0.6992150435607695,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.4628682285454288e-05,
+      "loss": 1.4156,
+      "step": 4053
+    },
+    {
+      "epoch": 0.6993875614595014,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.462626936678339e-05,
+      "loss": 1.4583,
+      "step": 4054
+    },
+    {
+      "epoch": 0.6995600793582334,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.4623856105369552e-05,
+      "loss": 1.4267,
+      "step": 4055
+    },
+    {
+      "epoch": 0.6997325972569655,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.462144250139156e-05,
+      "loss": 1.4379,
+      "step": 4056
+    },
+    {
+      "epoch": 0.6999051151556974,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.4619028555028234e-05,
+      "loss": 1.4115,
+      "step": 4057
+    },
+    {
+      "epoch": 0.7000776330544294,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.4616614266458413e-05,
+      "loss": 1.4709,
+      "step": 4058
+    },
+    {
+      "epoch": 0.7002501509531613,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.4614199635860958e-05,
+      "loss": 1.4998,
+      "step": 4059
+    },
+    {
+      "epoch": 0.7004226688518934,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.4611784663414765e-05,
+      "loss": 1.5,
+      "step": 4060
+    },
+    {
+      "epoch": 0.7005951867506254,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.4609369349298745e-05,
+      "loss": 1.4265,
+      "step": 4061
+    },
+    {
+      "epoch": 0.7007677046493573,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.4606953693691848e-05,
+      "loss": 1.4047,
+      "step": 4062
+    },
+    {
+      "epoch": 0.7009402225480894,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.4604537696773031e-05,
+      "loss": 1.4074,
+      "step": 4063
+    },
+    {
+      "epoch": 0.7011127404468214,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.4602121358721295e-05,
+      "loss": 1.5996,
+      "step": 4064
+    },
+    {
+      "epoch": 0.7012852583455533,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.4599704679715649e-05,
+      "loss": 1.448,
+      "step": 4065
+    },
+    {
+      "epoch": 0.7014577762442853,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.4597287659935146e-05,
+      "loss": 1.46,
+      "step": 4066
+    },
+    {
+      "epoch": 0.7016302941430174,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.4594870299558842e-05,
+      "loss": 1.5046,
+      "step": 4067
+    },
+    {
+      "epoch": 0.7018028120417493,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.459245259876584e-05,
+      "loss": 1.5037,
+      "step": 4068
+    },
+    {
+      "epoch": 0.7019753299404813,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.4590034557735253e-05,
+      "loss": 1.3754,
+      "step": 4069
+    },
+    {
+      "epoch": 0.7021478478392134,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.4587616176646229e-05,
+      "loss": 1.3533,
+      "step": 4070
+    },
+    {
+      "epoch": 0.7023203657379453,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.458519745567793e-05,
+      "loss": 1.4342,
+      "step": 4071
+    },
+    {
+      "epoch": 0.7024928836366773,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.4582778395009556e-05,
+      "loss": 1.408,
+      "step": 4072
+    },
+    {
+      "epoch": 0.7026654015354094,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.4580358994820322e-05,
+      "loss": 1.423,
+      "step": 4073
+    },
+    {
+      "epoch": 0.7028379194341413,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.4577939255289477e-05,
+      "loss": 1.4098,
+      "step": 4074
+    },
+    {
+      "epoch": 0.7030104373328733,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.4575519176596286e-05,
+      "loss": 1.4431,
+      "step": 4075
+    },
+    {
+      "epoch": 0.7031829552316052,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.4573098758920046e-05,
+      "loss": 1.408,
+      "step": 4076
+    },
+    {
+      "epoch": 0.7033554731303373,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.457067800244007e-05,
+      "loss": 1.4398,
+      "step": 4077
+    },
+    {
+      "epoch": 0.7035279910290693,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.4568256907335717e-05,
+      "loss": 1.4929,
+      "step": 4078
+    },
+    {
+      "epoch": 0.7037005089278012,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.4565835473786345e-05,
+      "loss": 1.4999,
+      "step": 4079
+    },
+    {
+      "epoch": 0.7038730268265333,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.4563413701971354e-05,
+      "loss": 1.432,
+      "step": 4080
+    },
+    {
+      "epoch": 0.7040455447252653,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.4560991592070159e-05,
+      "loss": 1.4448,
+      "step": 4081
+    },
+    {
+      "epoch": 0.7042180626239972,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.4558569144262213e-05,
+      "loss": 1.3809,
+      "step": 4082
+    },
+    {
+      "epoch": 0.7043905805227292,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.455614635872698e-05,
+      "loss": 1.4151,
+      "step": 4083
+    },
+    {
+      "epoch": 0.7045630984214613,
+      "grad_norm": 0.625,
+      "learning_rate": 1.4553723235643955e-05,
+      "loss": 1.4452,
+      "step": 4084
+    },
+    {
+      "epoch": 0.7047356163201932,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.455129977519266e-05,
+      "loss": 1.3651,
+      "step": 4085
+    },
+    {
+      "epoch": 0.7049081342189252,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.4548875977552643e-05,
+      "loss": 1.3227,
+      "step": 4086
+    },
+    {
+      "epoch": 0.7050806521176572,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.4546451842903468e-05,
+      "loss": 1.4218,
+      "step": 4087
+    },
+    {
+      "epoch": 0.7052531700163892,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.4544027371424732e-05,
+      "loss": 1.4026,
+      "step": 4088
+    },
+    {
+      "epoch": 0.7054256879151212,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.4541602563296058e-05,
+      "loss": 1.3546,
+      "step": 4089
+    },
+    {
+      "epoch": 0.7055982058138532,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.453917741869709e-05,
+      "loss": 1.4975,
+      "step": 4090
+    },
+    {
+      "epoch": 0.7057707237125852,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.4536751937807493e-05,
+      "loss": 1.4868,
+      "step": 4091
+    },
+    {
+      "epoch": 0.7059432416113172,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.4534326120806968e-05,
+      "loss": 1.4914,
+      "step": 4092
+    },
+    {
+      "epoch": 0.7061157595100491,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.4531899967875229e-05,
+      "loss": 1.4802,
+      "step": 4093
+    },
+    {
+      "epoch": 0.7062882774087812,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.4529473479192027e-05,
+      "loss": 1.442,
+      "step": 4094
+    },
+    {
+      "epoch": 0.7064607953075132,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.4527046654937128e-05,
+      "loss": 1.5061,
+      "step": 4095
+    },
+    {
+      "epoch": 0.7066333132062451,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.4524619495290324e-05,
+      "loss": 1.3328,
+      "step": 4096
+    },
+    {
+      "epoch": 0.7068058311049772,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.4522192000431439e-05,
+      "loss": 1.5061,
+      "step": 4097
+    },
+    {
+      "epoch": 0.7069783490037092,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.4519764170540315e-05,
+      "loss": 1.4898,
+      "step": 4098
+    },
+    {
+      "epoch": 0.7071508669024411,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.451733600579682e-05,
+      "loss": 1.4641,
+      "step": 4099
+    },
+    {
+      "epoch": 0.7073233848011731,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.451490750638085e-05,
+      "loss": 1.3865,
+      "step": 4100
+    },
+    {
+      "epoch": 0.7073233848011731,
+      "eval_loss": 1.4227409362792969,
+      "eval_runtime": 10.9937,
+      "eval_samples_per_second": 93.145,
+      "eval_steps_per_second": 23.286,
+      "step": 4100
+    },
+    {
+      "epoch": 0.7074959026999051,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.4512478672472318e-05,
+      "loss": 1.4121,
+      "step": 4101
+    },
+    {
+      "epoch": 0.7076684205986371,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.4510049504251174e-05,
+      "loss": 1.3894,
+      "step": 4102
+    },
+    {
+      "epoch": 0.7078409384973691,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.450762000189738e-05,
+      "loss": 1.4696,
+      "step": 4103
+    },
+    {
+      "epoch": 0.7080134563961011,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.4505190165590932e-05,
+      "loss": 1.4553,
+      "step": 4104
+    },
+    {
+      "epoch": 0.7081859742948331,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.4502759995511845e-05,
+      "loss": 1.4934,
+      "step": 4105
+    },
+    {
+      "epoch": 0.7083584921935651,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.4500329491840165e-05,
+      "loss": 1.5194,
+      "step": 4106
+    },
+    {
+      "epoch": 0.7085310100922971,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.4497898654755957e-05,
+      "loss": 1.4484,
+      "step": 4107
+    },
+    {
+      "epoch": 0.7087035279910291,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.4495467484439312e-05,
+      "loss": 1.5005,
+      "step": 4108
+    },
+    {
+      "epoch": 0.708876045889761,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.4493035981070347e-05,
+      "loss": 1.4317,
+      "step": 4109
+    },
+    {
+      "epoch": 0.709048563788493,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.4490604144829204e-05,
+      "loss": 1.4461,
+      "step": 4110
+    },
+    {
+      "epoch": 0.7092210816872251,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.4488171975896044e-05,
+      "loss": 1.5005,
+      "step": 4111
+    },
+    {
+      "epoch": 0.709393599585957,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.4485739474451061e-05,
+      "loss": 1.5018,
+      "step": 4112
+    },
+    {
+      "epoch": 0.709566117484689,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.4483306640674473e-05,
+      "loss": 1.4708,
+      "step": 4113
+    },
+    {
+      "epoch": 0.7097386353834211,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.4480873474746515e-05,
+      "loss": 1.3709,
+      "step": 4114
+    },
+    {
+      "epoch": 0.709911153282153,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.4478439976847452e-05,
+      "loss": 1.4258,
+      "step": 4115
+    },
+    {
+      "epoch": 0.710083671180885,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.4476006147157571e-05,
+      "loss": 1.4294,
+      "step": 4116
+    },
+    {
+      "epoch": 0.710256189079617,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.4473571985857189e-05,
+      "loss": 1.4314,
+      "step": 4117
+    },
+    {
+      "epoch": 0.710428706978349,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.4471137493126643e-05,
+      "loss": 1.4131,
+      "step": 4118
+    },
+    {
+      "epoch": 0.710601224877081,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.4468702669146292e-05,
+      "loss": 1.3848,
+      "step": 4119
+    },
+    {
+      "epoch": 0.710773742775813,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.4466267514096527e-05,
+      "loss": 1.4524,
+      "step": 4120
+    },
+    {
+      "epoch": 0.710946260674545,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.4463832028157758e-05,
+      "loss": 1.5744,
+      "step": 4121
+    },
+    {
+      "epoch": 0.711118778573277,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.4461396211510421e-05,
+      "loss": 1.5232,
+      "step": 4122
+    },
+    {
+      "epoch": 0.711291296472009,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.4458960064334977e-05,
+      "loss": 1.4307,
+      "step": 4123
+    },
+    {
+      "epoch": 0.7114638143707409,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.4456523586811911e-05,
+      "loss": 1.4285,
+      "step": 4124
+    },
+    {
+      "epoch": 0.711636332269473,
+      "grad_norm": 0.625,
+      "learning_rate": 1.445408677912173e-05,
+      "loss": 1.3565,
+      "step": 4125
+    },
+    {
+      "epoch": 0.7118088501682049,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.445164964144497e-05,
+      "loss": 1.4898,
+      "step": 4126
+    },
+    {
+      "epoch": 0.7119813680669369,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.444921217396219e-05,
+      "loss": 1.4365,
+      "step": 4127
+    },
+    {
+      "epoch": 0.712153885965669,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.4446774376853973e-05,
+      "loss": 1.4543,
+      "step": 4128
+    },
+    {
+      "epoch": 0.7123264038644009,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.4444336250300926e-05,
+      "loss": 1.4361,
+      "step": 4129
+    },
+    {
+      "epoch": 0.7124989217631329,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.4441897794483679e-05,
+      "loss": 1.4084,
+      "step": 4130
+    },
+    {
+      "epoch": 0.712671439661865,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.443945900958289e-05,
+      "loss": 1.4301,
+      "step": 4131
+    },
+    {
+      "epoch": 0.7128439575605969,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.4437019895779235e-05,
+      "loss": 1.6045,
+      "step": 4132
+    },
+    {
+      "epoch": 0.7130164754593289,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.4434580453253426e-05,
+      "loss": 1.4778,
+      "step": 4133
+    },
+    {
+      "epoch": 0.7131889933580609,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.4432140682186192e-05,
+      "loss": 1.4129,
+      "step": 4134
+    },
+    {
+      "epoch": 0.7133615112567929,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.4429700582758276e-05,
+      "loss": 1.4951,
+      "step": 4135
+    },
+    {
+      "epoch": 0.7135340291555249,
+      "grad_norm": 1.1640625,
+      "learning_rate": 1.4427260155150466e-05,
+      "loss": 1.4966,
+      "step": 4136
+    },
+    {
+      "epoch": 0.7137065470542568,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.4424819399543559e-05,
+      "loss": 1.3954,
+      "step": 4137
+    },
+    {
+      "epoch": 0.7138790649529889,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.4422378316118384e-05,
+      "loss": 1.4021,
+      "step": 4138
+    },
+    {
+      "epoch": 0.7140515828517209,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.4419936905055794e-05,
+      "loss": 1.4526,
+      "step": 4139
+    },
+    {
+      "epoch": 0.7142241007504528,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.4417495166536659e-05,
+      "loss": 1.3207,
+      "step": 4140
+    },
+    {
+      "epoch": 0.7143966186491848,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.4415053100741879e-05,
+      "loss": 1.4194,
+      "step": 4141
+    },
+    {
+      "epoch": 0.7145691365479169,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.4412610707852378e-05,
+      "loss": 1.4431,
+      "step": 4142
+    },
+    {
+      "epoch": 0.7147416544466488,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.4410167988049106e-05,
+      "loss": 1.5113,
+      "step": 4143
+    },
+    {
+      "epoch": 0.7149141723453808,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.4407724941513035e-05,
+      "loss": 1.5396,
+      "step": 4144
+    },
+    {
+      "epoch": 0.7150866902441129,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.4405281568425158e-05,
+      "loss": 1.5188,
+      "step": 4145
+    },
+    {
+      "epoch": 0.7152592081428448,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.4402837868966498e-05,
+      "loss": 1.3558,
+      "step": 4146
+    },
+    {
+      "epoch": 0.7154317260415768,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.4400393843318097e-05,
+      "loss": 1.4856,
+      "step": 4147
+    },
+    {
+      "epoch": 0.7156042439403089,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.4397949491661027e-05,
+      "loss": 1.3808,
+      "step": 4148
+    },
+    {
+      "epoch": 0.7157767618390408,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.4395504814176376e-05,
+      "loss": 1.3694,
+      "step": 4149
+    },
+    {
+      "epoch": 0.7159492797377728,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.4393059811045267e-05,
+      "loss": 1.3789,
+      "step": 4150
+    },
+    {
+      "epoch": 0.7161217976365047,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.4390614482448837e-05,
+      "loss": 1.4717,
+      "step": 4151
+    },
+    {
+      "epoch": 0.7162943155352368,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.4388168828568252e-05,
+      "loss": 1.4533,
+      "step": 4152
+    },
+    {
+      "epoch": 0.7164668334339688,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.4385722849584702e-05,
+      "loss": 1.4272,
+      "step": 4153
+    },
+    {
+      "epoch": 0.7166393513327007,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.4383276545679398e-05,
+      "loss": 1.4254,
+      "step": 4154
+    },
+    {
+      "epoch": 0.7168118692314328,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.4380829917033585e-05,
+      "loss": 1.5077,
+      "step": 4155
+    },
+    {
+      "epoch": 0.7169843871301648,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.4378382963828515e-05,
+      "loss": 1.3822,
+      "step": 4156
+    },
+    {
+      "epoch": 0.7171569050288967,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.4375935686245477e-05,
+      "loss": 1.5273,
+      "step": 4157
+    },
+    {
+      "epoch": 0.7173294229276287,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.4373488084465783e-05,
+      "loss": 1.4555,
+      "step": 4158
+    },
+    {
+      "epoch": 0.7175019408263608,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.4371040158670763e-05,
+      "loss": 1.5219,
+      "step": 4159
+    },
+    {
+      "epoch": 0.7176744587250927,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.4368591909041778e-05,
+      "loss": 1.4713,
+      "step": 4160
+    },
+    {
+      "epoch": 0.7178469766238247,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.4366143335760208e-05,
+      "loss": 1.5836,
+      "step": 4161
+    },
+    {
+      "epoch": 0.7180194945225568,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.436369443900746e-05,
+      "loss": 1.4414,
+      "step": 4162
+    },
+    {
+      "epoch": 0.7181920124212887,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.4361245218964961e-05,
+      "loss": 1.4173,
+      "step": 4163
+    },
+    {
+      "epoch": 0.7183645303200207,
+      "grad_norm": 0.625,
+      "learning_rate": 1.4358795675814165e-05,
+      "loss": 1.3769,
+      "step": 4164
+    },
+    {
+      "epoch": 0.7185370482187527,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.4356345809736552e-05,
+      "loss": 1.4559,
+      "step": 4165
+    },
+    {
+      "epoch": 0.7187095661174847,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.4353895620913619e-05,
+      "loss": 1.4834,
+      "step": 4166
+    },
+    {
+      "epoch": 0.7188820840162167,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.4351445109526897e-05,
+      "loss": 1.4671,
+      "step": 4167
+    },
+    {
+      "epoch": 0.7190546019149486,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.4348994275757933e-05,
+      "loss": 1.4191,
+      "step": 4168
+    },
+    {
+      "epoch": 0.7192271198136807,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.4346543119788297e-05,
+      "loss": 1.4187,
+      "step": 4169
+    },
+    {
+      "epoch": 0.7193996377124127,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.4344091641799587e-05,
+      "loss": 1.4377,
+      "step": 4170
+    },
+    {
+      "epoch": 0.7195721556111446,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.434163984197343e-05,
+      "loss": 1.512,
+      "step": 4171
+    },
+    {
+      "epoch": 0.7197446735098767,
+      "grad_norm": 0.625,
+      "learning_rate": 1.433918772049146e-05,
+      "loss": 1.5385,
+      "step": 4172
+    },
+    {
+      "epoch": 0.7199171914086087,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.4336735277535355e-05,
+      "loss": 1.4165,
+      "step": 4173
+    },
+    {
+      "epoch": 0.7200897093073406,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.4334282513286799e-05,
+      "loss": 1.5712,
+      "step": 4174
+    },
+    {
+      "epoch": 0.7202622272060726,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.4331829427927518e-05,
+      "loss": 1.4407,
+      "step": 4175
+    },
+    {
+      "epoch": 0.7204347451048047,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.4329376021639244e-05,
+      "loss": 1.5025,
+      "step": 4176
+    },
+    {
+      "epoch": 0.7206072630035366,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.4326922294603743e-05,
+      "loss": 1.4318,
+      "step": 4177
+    },
+    {
+      "epoch": 0.7207797809022686,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.4324468247002802e-05,
+      "loss": 1.3621,
+      "step": 4178
+    },
+    {
+      "epoch": 0.7209522988010006,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.4322013879018233e-05,
+      "loss": 1.4792,
+      "step": 4179
+    },
+    {
+      "epoch": 0.7211248166997326,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.4319559190831872e-05,
+      "loss": 1.3502,
+      "step": 4180
+    },
+    {
+      "epoch": 0.7212973345984646,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.4317104182625573e-05,
+      "loss": 1.5503,
+      "step": 4181
+    },
+    {
+      "epoch": 0.7214698524971965,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.4314648854581225e-05,
+      "loss": 1.429,
+      "step": 4182
+    },
+    {
+      "epoch": 0.7216423703959286,
+      "grad_norm": 0.625,
+      "learning_rate": 1.431219320688073e-05,
+      "loss": 1.3568,
+      "step": 4183
+    },
+    {
+      "epoch": 0.7218148882946606,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.4309737239706019e-05,
+      "loss": 1.4368,
+      "step": 4184
+    },
+    {
+      "epoch": 0.7219874061933925,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.4307280953239044e-05,
+      "loss": 1.4313,
+      "step": 4185
+    },
+    {
+      "epoch": 0.7221599240921246,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.4304824347661783e-05,
+      "loss": 1.399,
+      "step": 4186
+    },
+    {
+      "epoch": 0.7223324419908566,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.4302367423156236e-05,
+      "loss": 1.4066,
+      "step": 4187
+    },
+    {
+      "epoch": 0.7225049598895885,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.4299910179904429e-05,
+      "loss": 1.3609,
+      "step": 4188
+    },
+    {
+      "epoch": 0.7226774777883206,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.4297452618088407e-05,
+      "loss": 1.4241,
+      "step": 4189
+    },
+    {
+      "epoch": 0.7228499956870525,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.4294994737890245e-05,
+      "loss": 1.3455,
+      "step": 4190
+    },
+    {
+      "epoch": 0.7230225135857845,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.4292536539492034e-05,
+      "loss": 1.5229,
+      "step": 4191
+    },
+    {
+      "epoch": 0.7231950314845165,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.4290078023075897e-05,
+      "loss": 1.5029,
+      "step": 4192
+    },
+    {
+      "epoch": 0.7233675493832485,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.4287619188823975e-05,
+      "loss": 1.4714,
+      "step": 4193
+    },
+    {
+      "epoch": 0.7235400672819805,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.4285160036918431e-05,
+      "loss": 1.4747,
+      "step": 4194
+    },
+    {
+      "epoch": 0.7237125851807125,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.4282700567541461e-05,
+      "loss": 1.4821,
+      "step": 4195
+    },
+    {
+      "epoch": 0.7238851030794445,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.428024078087527e-05,
+      "loss": 1.4791,
+      "step": 4196
+    },
+    {
+      "epoch": 0.7240576209781765,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.4277780677102098e-05,
+      "loss": 1.3904,
+      "step": 4197
+    },
+    {
+      "epoch": 0.7242301388769085,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.4275320256404203e-05,
+      "loss": 1.5428,
+      "step": 4198
+    },
+    {
+      "epoch": 0.7244026567756404,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.4272859518963874e-05,
+      "loss": 1.4717,
+      "step": 4199
+    },
+    {
+      "epoch": 0.7245751746743725,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.4270398464963411e-05,
+      "loss": 1.5452,
+      "step": 4200
+    },
+    {
+      "epoch": 0.7245751746743725,
+      "eval_loss": 1.4216216802597046,
+      "eval_runtime": 10.8689,
+      "eval_samples_per_second": 94.213,
+      "eval_steps_per_second": 23.553,
+      "step": 4200
+    },
+    {
+      "epoch": 0.7247476925731045,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.4267937094585148e-05,
+      "loss": 1.445,
+      "step": 4201
+    },
+    {
+      "epoch": 0.7249202104718364,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.4265475408011438e-05,
+      "loss": 1.5107,
+      "step": 4202
+    },
+    {
+      "epoch": 0.7250927283705685,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.4263013405424657e-05,
+      "loss": 1.4799,
+      "step": 4203
+    },
+    {
+      "epoch": 0.7252652462693004,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.4260551087007205e-05,
+      "loss": 1.4503,
+      "step": 4204
+    },
+    {
+      "epoch": 0.7254377641680324,
+      "grad_norm": 0.75,
+      "learning_rate": 1.425808845294151e-05,
+      "loss": 1.4823,
+      "step": 4205
+    },
+    {
+      "epoch": 0.7256102820667645,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.4255625503410015e-05,
+      "loss": 1.3246,
+      "step": 4206
+    },
+    {
+      "epoch": 0.7257827999654964,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.4253162238595192e-05,
+      "loss": 1.4137,
+      "step": 4207
+    },
+    {
+      "epoch": 0.7259553178642284,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.4250698658679535e-05,
+      "loss": 1.359,
+      "step": 4208
+    },
+    {
+      "epoch": 0.7261278357629604,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.4248234763845565e-05,
+      "loss": 1.4221,
+      "step": 4209
+    },
+    {
+      "epoch": 0.7263003536616924,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.4245770554275817e-05,
+      "loss": 1.4439,
+      "step": 4210
+    },
+    {
+      "epoch": 0.7264728715604244,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.424330603015286e-05,
+      "loss": 1.3822,
+      "step": 4211
+    },
+    {
+      "epoch": 0.7266453894591564,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.4240841191659276e-05,
+      "loss": 1.4447,
+      "step": 4212
+    },
+    {
+      "epoch": 0.7268179073578884,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.423837603897768e-05,
+      "loss": 1.4547,
+      "step": 4213
+    },
+    {
+      "epoch": 0.7269904252566204,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.4235910572290704e-05,
+      "loss": 1.5104,
+      "step": 4214
+    },
+    {
+      "epoch": 0.7271629431553523,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.4233444791781005e-05,
+      "loss": 1.4798,
+      "step": 4215
+    },
+    {
+      "epoch": 0.7273354610540843,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.4230978697631266e-05,
+      "loss": 1.4985,
+      "step": 4216
+    },
+    {
+      "epoch": 0.7275079789528164,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.4228512290024185e-05,
+      "loss": 1.3978,
+      "step": 4217
+    },
+    {
+      "epoch": 0.7276804968515483,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.4226045569142497e-05,
+      "loss": 1.4935,
+      "step": 4218
+    },
+    {
+      "epoch": 0.7278530147502803,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.4223578535168947e-05,
+      "loss": 1.4479,
+      "step": 4219
+    },
+    {
+      "epoch": 0.7280255326490124,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.4221111188286307e-05,
+      "loss": 1.4613,
+      "step": 4220
+    },
+    {
+      "epoch": 0.7281980505477443,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.4218643528677377e-05,
+      "loss": 1.4544,
+      "step": 4221
+    },
+    {
+      "epoch": 0.7283705684464763,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.421617555652497e-05,
+      "loss": 1.4306,
+      "step": 4222
+    },
+    {
+      "epoch": 0.7285430863452084,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.421370727201194e-05,
+      "loss": 1.5617,
+      "step": 4223
+    },
+    {
+      "epoch": 0.7287156042439403,
+      "grad_norm": 0.625,
+      "learning_rate": 1.4211238675321143e-05,
+      "loss": 1.3514,
+      "step": 4224
+    },
+    {
+      "epoch": 0.7288881221426723,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.4208769766635475e-05,
+      "loss": 1.5241,
+      "step": 4225
+    },
+    {
+      "epoch": 0.7290606400414043,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.4206300546137844e-05,
+      "loss": 1.554,
+      "step": 4226
+    },
+    {
+      "epoch": 0.7292331579401363,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.4203831014011186e-05,
+      "loss": 1.4732,
+      "step": 4227
+    },
+    {
+      "epoch": 0.7294056758388683,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.4201361170438461e-05,
+      "loss": 1.4524,
+      "step": 4228
+    },
+    {
+      "epoch": 0.7295781937376002,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.4198891015602648e-05,
+      "loss": 1.4721,
+      "step": 4229
+    },
+    {
+      "epoch": 0.7297507116363323,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.419642054968675e-05,
+      "loss": 1.4939,
+      "step": 4230
+    },
+    {
+      "epoch": 0.7299232295350643,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.4193949772873804e-05,
+      "loss": 1.4559,
+      "step": 4231
+    },
+    {
+      "epoch": 0.7300957474337962,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.4191478685346849e-05,
+      "loss": 1.3952,
+      "step": 4232
+    },
+    {
+      "epoch": 0.7302682653325282,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.4189007287288963e-05,
+      "loss": 1.4461,
+      "step": 4233
+    },
+    {
+      "epoch": 0.7304407832312603,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.4186535578883244e-05,
+      "loss": 1.4821,
+      "step": 4234
+    },
+    {
+      "epoch": 0.7306133011299922,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.4184063560312814e-05,
+      "loss": 1.4381,
+      "step": 4235
+    },
+    {
+      "epoch": 0.7307858190287242,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.4181591231760807e-05,
+      "loss": 1.4034,
+      "step": 4236
+    },
+    {
+      "epoch": 0.7309583369274563,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.4179118593410395e-05,
+      "loss": 1.3891,
+      "step": 4237
+    },
+    {
+      "epoch": 0.7311308548261882,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.4176645645444765e-05,
+      "loss": 1.523,
+      "step": 4238
+    },
+    {
+      "epoch": 0.7313033727249202,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.4174172388047132e-05,
+      "loss": 1.4271,
+      "step": 4239
+    },
+    {
+      "epoch": 0.7314758906236523,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.4171698821400724e-05,
+      "loss": 1.4633,
+      "step": 4240
+    },
+    {
+      "epoch": 0.7316484085223842,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.4169224945688801e-05,
+      "loss": 1.4301,
+      "step": 4241
+    },
+    {
+      "epoch": 0.7318209264211162,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.4166750761094646e-05,
+      "loss": 1.3832,
+      "step": 4242
+    },
+    {
+      "epoch": 0.7319934443198481,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.416427626780156e-05,
+      "loss": 1.4208,
+      "step": 4243
+    },
+    {
+      "epoch": 0.7321659622185802,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.4161801465992867e-05,
+      "loss": 1.4036,
+      "step": 4244
+    },
+    {
+      "epoch": 0.7323384801173122,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.4159326355851915e-05,
+      "loss": 1.3705,
+      "step": 4245
+    },
+    {
+      "epoch": 0.7325109980160441,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.4156850937562079e-05,
+      "loss": 1.4796,
+      "step": 4246
+    },
+    {
+      "epoch": 0.7326835159147762,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.4154375211306754e-05,
+      "loss": 1.4343,
+      "step": 4247
+    },
+    {
+      "epoch": 0.7328560338135082,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.4151899177269357e-05,
+      "loss": 1.3781,
+      "step": 4248
+    },
+    {
+      "epoch": 0.7330285517122401,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.4149422835633324e-05,
+      "loss": 1.3809,
+      "step": 4249
+    },
+    {
+      "epoch": 0.7332010696109721,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.414694618658212e-05,
+      "loss": 1.5156,
+      "step": 4250
+    },
+    {
+      "epoch": 0.7333735875097042,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.4144469230299234e-05,
+      "loss": 1.454,
+      "step": 4251
+    },
+    {
+      "epoch": 0.7335461054084361,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.4141991966968169e-05,
+      "loss": 1.4892,
+      "step": 4252
+    },
+    {
+      "epoch": 0.7337186233071681,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.413951439677246e-05,
+      "loss": 1.3443,
+      "step": 4253
+    },
+    {
+      "epoch": 0.7338911412059002,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.4137036519895656e-05,
+      "loss": 1.5462,
+      "step": 4254
+    },
+    {
+      "epoch": 0.7340636591046321,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.4134558336521342e-05,
+      "loss": 1.4905,
+      "step": 4255
+    },
+    {
+      "epoch": 0.7342361770033641,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.413207984683311e-05,
+      "loss": 1.4553,
+      "step": 4256
+    },
+    {
+      "epoch": 0.734408694902096,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.4129601051014586e-05,
+      "loss": 1.4901,
+      "step": 4257
+    },
+    {
+      "epoch": 0.7345812128008281,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.4127121949249412e-05,
+      "loss": 1.4201,
+      "step": 4258
+    },
+    {
+      "epoch": 0.7347537306995601,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.4124642541721258e-05,
+      "loss": 1.5189,
+      "step": 4259
+    },
+    {
+      "epoch": 0.734926248598292,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.4122162828613812e-05,
+      "loss": 1.4489,
+      "step": 4260
+    },
+    {
+      "epoch": 0.7350987664970241,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.4119682810110787e-05,
+      "loss": 1.4045,
+      "step": 4261
+    },
+    {
+      "epoch": 0.7352712843957561,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.4117202486395918e-05,
+      "loss": 1.5362,
+      "step": 4262
+    },
+    {
+      "epoch": 0.735443802294488,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.4114721857652965e-05,
+      "loss": 1.4357,
+      "step": 4263
+    },
+    {
+      "epoch": 0.7356163201932201,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.4112240924065706e-05,
+      "loss": 1.4519,
+      "step": 4264
+    },
+    {
+      "epoch": 0.7357888380919521,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.4109759685817943e-05,
+      "loss": 1.4781,
+      "step": 4265
+    },
+    {
+      "epoch": 0.735961355990684,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.4107278143093505e-05,
+      "loss": 1.505,
+      "step": 4266
+    },
+    {
+      "epoch": 0.736133873889416,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.410479629607624e-05,
+      "loss": 1.3641,
+      "step": 4267
+    },
+    {
+      "epoch": 0.736306391788148,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.4102314144950016e-05,
+      "loss": 1.451,
+      "step": 4268
+    },
+    {
+      "epoch": 0.73647890968688,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.4099831689898728e-05,
+      "loss": 1.441,
+      "step": 4269
+    },
+    {
+      "epoch": 0.736651427585612,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.409734893110629e-05,
+      "loss": 1.4513,
+      "step": 4270
+    },
+    {
+      "epoch": 0.736823945484344,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.4094865868756644e-05,
+      "loss": 1.4197,
+      "step": 4271
+    },
+    {
+      "epoch": 0.736996463383076,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.4092382503033746e-05,
+      "loss": 1.3671,
+      "step": 4272
+    },
+    {
+      "epoch": 0.737168981281808,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.4089898834121583e-05,
+      "loss": 1.3355,
+      "step": 4273
+    },
+    {
+      "epoch": 0.7373414991805399,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.4087414862204161e-05,
+      "loss": 1.4562,
+      "step": 4274
+    },
+    {
+      "epoch": 0.737514017079272,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.4084930587465506e-05,
+      "loss": 1.4651,
+      "step": 4275
+    },
+    {
+      "epoch": 0.737686534978004,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.4082446010089667e-05,
+      "loss": 1.4966,
+      "step": 4276
+    },
+    {
+      "epoch": 0.7378590528767359,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.4079961130260722e-05,
+      "loss": 1.3972,
+      "step": 4277
+    },
+    {
+      "epoch": 0.738031570775468,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.4077475948162762e-05,
+      "loss": 1.4247,
+      "step": 4278
+    },
+    {
+      "epoch": 0.7382040886742,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.407499046397991e-05,
+      "loss": 1.4448,
+      "step": 4279
+    },
+    {
+      "epoch": 0.7383766065729319,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.40725046778963e-05,
+      "loss": 1.4223,
+      "step": 4280
+    },
+    {
+      "epoch": 0.738549124471664,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.4070018590096096e-05,
+      "loss": 1.3966,
+      "step": 4281
+    },
+    {
+      "epoch": 0.738721642370396,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.4067532200763484e-05,
+      "loss": 1.4662,
+      "step": 4282
+    },
+    {
+      "epoch": 0.7388941602691279,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.4065045510082673e-05,
+      "loss": 1.5419,
+      "step": 4283
+    },
+    {
+      "epoch": 0.7390666781678599,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.4062558518237893e-05,
+      "loss": 1.4291,
+      "step": 4284
+    },
+    {
+      "epoch": 0.7392391960665919,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.4060071225413392e-05,
+      "loss": 1.4024,
+      "step": 4285
+    },
+    {
+      "epoch": 0.7394117139653239,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.4057583631793443e-05,
+      "loss": 1.4185,
+      "step": 4286
+    },
+    {
+      "epoch": 0.7395842318640559,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.405509573756235e-05,
+      "loss": 1.4524,
+      "step": 4287
+    },
+    {
+      "epoch": 0.7397567497627879,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.4052607542904427e-05,
+      "loss": 1.5358,
+      "step": 4288
+    },
+    {
+      "epoch": 0.7399292676615199,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.4050119048004012e-05,
+      "loss": 1.4615,
+      "step": 4289
+    },
+    {
+      "epoch": 0.7401017855602519,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.4047630253045475e-05,
+      "loss": 1.3796,
+      "step": 4290
+    },
+    {
+      "epoch": 0.7402743034589838,
+      "grad_norm": 0.625,
+      "learning_rate": 1.4045141158213197e-05,
+      "loss": 1.5141,
+      "step": 4291
+    },
+    {
+      "epoch": 0.7404468213577159,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.4042651763691586e-05,
+      "loss": 1.4644,
+      "step": 4292
+    },
+    {
+      "epoch": 0.7406193392564479,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.4040162069665076e-05,
+      "loss": 1.3469,
+      "step": 4293
+    },
+    {
+      "epoch": 0.7407918571551798,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.4037672076318112e-05,
+      "loss": 1.4975,
+      "step": 4294
+    },
+    {
+      "epoch": 0.7409643750539119,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.4035181783835174e-05,
+      "loss": 1.4385,
+      "step": 4295
+    },
+    {
+      "epoch": 0.7411368929526438,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.4032691192400756e-05,
+      "loss": 1.4352,
+      "step": 4296
+    },
+    {
+      "epoch": 0.7413094108513758,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.4030200302199375e-05,
+      "loss": 1.4521,
+      "step": 4297
+    },
+    {
+      "epoch": 0.7414819287501079,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.4027709113415578e-05,
+      "loss": 1.4853,
+      "step": 4298
+    },
+    {
+      "epoch": 0.7416544466488398,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.4025217626233919e-05,
+      "loss": 1.4171,
+      "step": 4299
+    },
+    {
+      "epoch": 0.7418269645475718,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.402272584083899e-05,
+      "loss": 1.4167,
+      "step": 4300
+    },
+    {
+      "epoch": 0.7418269645475718,
+      "eval_loss": 1.4205090999603271,
+      "eval_runtime": 11.0817,
+      "eval_samples_per_second": 92.405,
+      "eval_steps_per_second": 23.101,
+      "step": 4300
+    },
+    {
+      "epoch": 0.7419994824463038,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.4020233757415396e-05,
+      "loss": 1.4844,
+      "step": 4301
+    },
+    {
+      "epoch": 0.7421720003450358,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.4017741376147762e-05,
+      "loss": 1.4134,
+      "step": 4302
+    },
+    {
+      "epoch": 0.7423445182437678,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.4015248697220746e-05,
+      "loss": 1.4863,
+      "step": 4303
+    },
+    {
+      "epoch": 0.7425170361424998,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.401275572081902e-05,
+      "loss": 1.4767,
+      "step": 4304
+    },
+    {
+      "epoch": 0.7426895540412318,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.4010262447127275e-05,
+      "loss": 1.4214,
+      "step": 4305
+    },
+    {
+      "epoch": 0.7428620719399638,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.400776887633023e-05,
+      "loss": 1.4982,
+      "step": 4306
+    },
+    {
+      "epoch": 0.7430345898386957,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.4005275008612627e-05,
+      "loss": 1.4187,
+      "step": 4307
+    },
+    {
+      "epoch": 0.7432071077374277,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.4002780844159225e-05,
+      "loss": 1.5067,
+      "step": 4308
+    },
+    {
+      "epoch": 0.7433796256361598,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.4000286383154804e-05,
+      "loss": 1.4092,
+      "step": 4309
+    },
+    {
+      "epoch": 0.7435521435348917,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.3997791625784176e-05,
+      "loss": 1.5382,
+      "step": 4310
+    },
+    {
+      "epoch": 0.7437246614336237,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.3995296572232166e-05,
+      "loss": 1.4975,
+      "step": 4311
+    },
+    {
+      "epoch": 0.7438971793323558,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.399280122268362e-05,
+      "loss": 1.4389,
+      "step": 4312
+    },
+    {
+      "epoch": 0.7440696972310877,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.399030557732341e-05,
+      "loss": 1.4836,
+      "step": 4313
+    },
+    {
+      "epoch": 0.7442422151298197,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.3987809636336434e-05,
+      "loss": 1.4865,
+      "step": 4314
+    },
+    {
+      "epoch": 0.7444147330285518,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.39853133999076e-05,
+      "loss": 1.5212,
+      "step": 4315
+    },
+    {
+      "epoch": 0.7445872509272837,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.398281686822185e-05,
+      "loss": 1.4433,
+      "step": 4316
+    },
+    {
+      "epoch": 0.7447597688260157,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.398032004146414e-05,
+      "loss": 1.4844,
+      "step": 4317
+    },
+    {
+      "epoch": 0.7449322867247476,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.3977822919819448e-05,
+      "loss": 1.4672,
+      "step": 4318
+    },
+    {
+      "epoch": 0.7451048046234797,
+      "grad_norm": 0.625,
+      "learning_rate": 1.3975325503472786e-05,
+      "loss": 1.4728,
+      "step": 4319
+    },
+    {
+      "epoch": 0.7452773225222117,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.3972827792609168e-05,
+      "loss": 1.4121,
+      "step": 4320
+    },
+    {
+      "epoch": 0.7454498404209436,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.397032978741364e-05,
+      "loss": 1.4833,
+      "step": 4321
+    },
+    {
+      "epoch": 0.7456223583196757,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.3967831488071279e-05,
+      "loss": 1.4789,
+      "step": 4322
+    },
+    {
+      "epoch": 0.7457948762184077,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.396533289476717e-05,
+      "loss": 1.3527,
+      "step": 4323
+    },
+    {
+      "epoch": 0.7459673941171396,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.3962834007686418e-05,
+      "loss": 1.4939,
+      "step": 4324
+    },
+    {
+      "epoch": 0.7461399120158716,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.3960334827014168e-05,
+      "loss": 1.4151,
+      "step": 4325
+    },
+    {
+      "epoch": 0.7463124299146037,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.3957835352935566e-05,
+      "loss": 1.4769,
+      "step": 4326
+    },
+    {
+      "epoch": 0.7464849478133356,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.395533558563579e-05,
+      "loss": 1.3579,
+      "step": 4327
+    },
+    {
+      "epoch": 0.7466574657120676,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.395283552530004e-05,
+      "loss": 1.4144,
+      "step": 4328
+    },
+    {
+      "epoch": 0.7468299836107997,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.3950335172113539e-05,
+      "loss": 1.4561,
+      "step": 4329
+    },
+    {
+      "epoch": 0.7470025015095316,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.3947834526261521e-05,
+      "loss": 1.4137,
+      "step": 4330
+    },
+    {
+      "epoch": 0.7471750194082636,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.3945333587929257e-05,
+      "loss": 1.5995,
+      "step": 4331
+    },
+    {
+      "epoch": 0.7473475373069955,
+      "grad_norm": 0.75,
+      "learning_rate": 1.394283235730203e-05,
+      "loss": 1.4139,
+      "step": 4332
+    },
+    {
+      "epoch": 0.7475200552057276,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.3940330834565144e-05,
+      "loss": 1.4916,
+      "step": 4333
+    },
+    {
+      "epoch": 0.7476925731044596,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.393782901990393e-05,
+      "loss": 1.4787,
+      "step": 4334
+    },
+    {
+      "epoch": 0.7478650910031915,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.3935326913503737e-05,
+      "loss": 1.341,
+      "step": 4335
+    },
+    {
+      "epoch": 0.7480376089019236,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.3932824515549936e-05,
+      "loss": 1.5001,
+      "step": 4336
+    },
+    {
+      "epoch": 0.7482101268006556,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.3930321826227924e-05,
+      "loss": 1.4086,
+      "step": 4337
+    },
+    {
+      "epoch": 0.7483826446993875,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.3927818845723114e-05,
+      "loss": 1.491,
+      "step": 4338
+    },
+    {
+      "epoch": 0.7485551625981196,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.392531557422094e-05,
+      "loss": 1.4783,
+      "step": 4339
+    },
+    {
+      "epoch": 0.7487276804968516,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.3922812011906862e-05,
+      "loss": 1.403,
+      "step": 4340
+    },
+    {
+      "epoch": 0.7489001983955835,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.3920308158966358e-05,
+      "loss": 1.3879,
+      "step": 4341
+    },
+    {
+      "epoch": 0.7490727162943155,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.3917804015584932e-05,
+      "loss": 1.4304,
+      "step": 4342
+    },
+    {
+      "epoch": 0.7492452341930476,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.3915299581948106e-05,
+      "loss": 1.4423,
+      "step": 4343
+    },
+    {
+      "epoch": 0.7494177520917795,
+      "grad_norm": 0.875,
+      "learning_rate": 1.3912794858241423e-05,
+      "loss": 1.483,
+      "step": 4344
+    },
+    {
+      "epoch": 0.7495902699905115,
+      "grad_norm": 0.75,
+      "learning_rate": 1.3910289844650446e-05,
+      "loss": 1.3615,
+      "step": 4345
+    },
+    {
+      "epoch": 0.7497627878892436,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.3907784541360765e-05,
+      "loss": 1.5126,
+      "step": 4346
+    },
+    {
+      "epoch": 0.7499353057879755,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.390527894855799e-05,
+      "loss": 1.4707,
+      "step": 4347
+    },
+    {
+      "epoch": 0.7501078236867075,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.390277306642775e-05,
+      "loss": 1.385,
+      "step": 4348
+    },
+    {
+      "epoch": 0.7502803415854394,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.3900266895155694e-05,
+      "loss": 1.3342,
+      "step": 4349
+    },
+    {
+      "epoch": 0.7504528594841715,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.3897760434927495e-05,
+      "loss": 1.4733,
+      "step": 4350
+    },
+    {
+      "epoch": 0.7506253773829035,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.3895253685928851e-05,
+      "loss": 1.4812,
+      "step": 4351
+    },
+    {
+      "epoch": 0.7507978952816354,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.3892746648345475e-05,
+      "loss": 1.3713,
+      "step": 4352
+    },
+    {
+      "epoch": 0.7509704131803675,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.3890239322363102e-05,
+      "loss": 1.4003,
+      "step": 4353
+    },
+    {
+      "epoch": 0.7511429310790995,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3887731708167493e-05,
+      "loss": 1.4404,
+      "step": 4354
+    },
+    {
+      "epoch": 0.7513154489778314,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.388522380594443e-05,
+      "loss": 1.4032,
+      "step": 4355
+    },
+    {
+      "epoch": 0.7514879668765635,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.3882715615879709e-05,
+      "loss": 1.398,
+      "step": 4356
+    },
+    {
+      "epoch": 0.7516604847752955,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.3880207138159158e-05,
+      "loss": 1.4063,
+      "step": 4357
+    },
+    {
+      "epoch": 0.7518330026740274,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.3877698372968612e-05,
+      "loss": 1.4327,
+      "step": 4358
+    },
+    {
+      "epoch": 0.7520055205727594,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.3875189320493947e-05,
+      "loss": 1.6136,
+      "step": 4359
+    },
+    {
+      "epoch": 0.7521780384714915,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.3872679980921044e-05,
+      "loss": 1.353,
+      "step": 4360
+    },
+    {
+      "epoch": 0.7523505563702234,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.3870170354435808e-05,
+      "loss": 1.5392,
+      "step": 4361
+    },
+    {
+      "epoch": 0.7525230742689554,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.3867660441224172e-05,
+      "loss": 1.4298,
+      "step": 4362
+    },
+    {
+      "epoch": 0.7526955921676874,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.386515024147208e-05,
+      "loss": 1.4125,
+      "step": 4363
+    },
+    {
+      "epoch": 0.7528681100664194,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.3862639755365512e-05,
+      "loss": 1.5319,
+      "step": 4364
+    },
+    {
+      "epoch": 0.7530406279651514,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.3860128983090455e-05,
+      "loss": 1.3204,
+      "step": 4365
+    },
+    {
+      "epoch": 0.7532131458638833,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.3857617924832923e-05,
+      "loss": 1.468,
+      "step": 4366
+    },
+    {
+      "epoch": 0.7533856637626154,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.385510658077895e-05,
+      "loss": 1.4754,
+      "step": 4367
+    },
+    {
+      "epoch": 0.7535581816613474,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.3852594951114594e-05,
+      "loss": 1.448,
+      "step": 4368
+    },
+    {
+      "epoch": 0.7537306995600793,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.3850083036025934e-05,
+      "loss": 1.3702,
+      "step": 4369
+    },
+    {
+      "epoch": 0.7539032174588114,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.3847570835699066e-05,
+      "loss": 1.557,
+      "step": 4370
+    },
+    {
+      "epoch": 0.7540757353575434,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.3845058350320109e-05,
+      "loss": 1.3292,
+      "step": 4371
+    },
+    {
+      "epoch": 0.7542482532562753,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.3842545580075206e-05,
+      "loss": 1.5168,
+      "step": 4372
+    },
+    {
+      "epoch": 0.7544207711550074,
+      "grad_norm": 0.625,
+      "learning_rate": 1.3840032525150516e-05,
+      "loss": 1.47,
+      "step": 4373
+    },
+    {
+      "epoch": 0.7545932890537393,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.3837519185732222e-05,
+      "loss": 1.3561,
+      "step": 4374
+    },
+    {
+      "epoch": 0.7547658069524713,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.3835005562006529e-05,
+      "loss": 1.3902,
+      "step": 4375
+    },
+    {
+      "epoch": 0.7549383248512033,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.3832491654159663e-05,
+      "loss": 1.4831,
+      "step": 4376
+    },
+    {
+      "epoch": 0.7551108427499353,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.3829977462377868e-05,
+      "loss": 1.4294,
+      "step": 4377
+    },
+    {
+      "epoch": 0.7552833606486673,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.3827462986847411e-05,
+      "loss": 1.4088,
+      "step": 4378
+    },
+    {
+      "epoch": 0.7554558785473993,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.3824948227754583e-05,
+      "loss": 1.4768,
+      "step": 4379
+    },
+    {
+      "epoch": 0.7556283964461313,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.3822433185285691e-05,
+      "loss": 1.4937,
+      "step": 4380
+    },
+    {
+      "epoch": 0.7558009143448633,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.3819917859627063e-05,
+      "loss": 1.3964,
+      "step": 4381
+    },
+    {
+      "epoch": 0.7559734322435953,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.381740225096505e-05,
+      "loss": 1.3588,
+      "step": 4382
+    },
+    {
+      "epoch": 0.7561459501423272,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.3814886359486028e-05,
+      "loss": 1.5209,
+      "step": 4383
+    },
+    {
+      "epoch": 0.7563184680410593,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.3812370185376389e-05,
+      "loss": 1.4714,
+      "step": 4384
+    },
+    {
+      "epoch": 0.7564909859397912,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.3809853728822545e-05,
+      "loss": 1.3945,
+      "step": 4385
+    },
+    {
+      "epoch": 0.7566635038385232,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.3807336990010934e-05,
+      "loss": 1.4341,
+      "step": 4386
+    },
+    {
+      "epoch": 0.7568360217372553,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.3804819969128006e-05,
+      "loss": 1.3775,
+      "step": 4387
+    },
+    {
+      "epoch": 0.7570085396359872,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.3802302666360244e-05,
+      "loss": 1.4069,
+      "step": 4388
+    },
+    {
+      "epoch": 0.7571810575347192,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.379978508189414e-05,
+      "loss": 1.4814,
+      "step": 4389
+    },
+    {
+      "epoch": 0.7573535754334513,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.3797267215916215e-05,
+      "loss": 1.364,
+      "step": 4390
+    },
+    {
+      "epoch": 0.7575260933321832,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.3794749068613009e-05,
+      "loss": 1.4798,
+      "step": 4391
+    },
+    {
+      "epoch": 0.7576986112309152,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.3792230640171085e-05,
+      "loss": 1.4018,
+      "step": 4392
+    },
+    {
+      "epoch": 0.7578711291296472,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.3789711930777017e-05,
+      "loss": 1.4674,
+      "step": 4393
+    },
+    {
+      "epoch": 0.7580436470283792,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.3787192940617408e-05,
+      "loss": 1.394,
+      "step": 4394
+    },
+    {
+      "epoch": 0.7582161649271112,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.3784673669878883e-05,
+      "loss": 1.4453,
+      "step": 4395
+    },
+    {
+      "epoch": 0.7583886828258432,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.378215411874809e-05,
+      "loss": 1.5361,
+      "step": 4396
+    },
+    {
+      "epoch": 0.7585612007245752,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.3779634287411683e-05,
+      "loss": 1.4041,
+      "step": 4397
+    },
+    {
+      "epoch": 0.7587337186233072,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.3777114176056353e-05,
+      "loss": 1.4944,
+      "step": 4398
+    },
+    {
+      "epoch": 0.7589062365220391,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.3774593784868802e-05,
+      "loss": 1.4541,
+      "step": 4399
+    },
+    {
+      "epoch": 0.7590787544207711,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.3772073114035762e-05,
+      "loss": 1.3346,
+      "step": 4400
+    },
+    {
+      "epoch": 0.7590787544207711,
+      "eval_loss": 1.4194252490997314,
+      "eval_runtime": 10.7734,
+      "eval_samples_per_second": 95.049,
+      "eval_steps_per_second": 23.762,
+      "step": 4400
+    },
+    {
+      "epoch": 0.7592512723195032,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.3769552163743974e-05,
+      "loss": 1.5201,
+      "step": 4401
+    },
+    {
+      "epoch": 0.7594237902182351,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.376703093418021e-05,
+      "loss": 1.4532,
+      "step": 4402
+    },
+    {
+      "epoch": 0.7595963081169671,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.3764509425531256e-05,
+      "loss": 1.4176,
+      "step": 4403
+    },
+    {
+      "epoch": 0.7597688260156992,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.3761987637983924e-05,
+      "loss": 1.4204,
+      "step": 4404
+    },
+    {
+      "epoch": 0.7599413439144311,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.375946557172504e-05,
+      "loss": 1.4231,
+      "step": 4405
+    },
+    {
+      "epoch": 0.7601138618131631,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.3756943226941458e-05,
+      "loss": 1.3763,
+      "step": 4406
+    },
+    {
+      "epoch": 0.760286379711895,
+      "grad_norm": 0.625,
+      "learning_rate": 1.3754420603820045e-05,
+      "loss": 1.4556,
+      "step": 4407
+    },
+    {
+      "epoch": 0.7604588976106271,
+      "grad_norm": 0.75,
+      "learning_rate": 1.3751897702547698e-05,
+      "loss": 1.4294,
+      "step": 4408
+    },
+    {
+      "epoch": 0.7606314155093591,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.3749374523311325e-05,
+      "loss": 1.4873,
+      "step": 4409
+    },
+    {
+      "epoch": 0.760803933408091,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.374685106629786e-05,
+      "loss": 1.3729,
+      "step": 4410
+    },
+    {
+      "epoch": 0.7609764513068231,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.3744327331694254e-05,
+      "loss": 1.419,
+      "step": 4411
+    },
+    {
+      "epoch": 0.7611489692055551,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.3741803319687488e-05,
+      "loss": 1.4207,
+      "step": 4412
+    },
+    {
+      "epoch": 0.761321487104287,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.373927903046455e-05,
+      "loss": 1.4772,
+      "step": 4413
+    },
+    {
+      "epoch": 0.7614940050030191,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.3736754464212456e-05,
+      "loss": 1.4847,
+      "step": 4414
+    },
+    {
+      "epoch": 0.7616665229017511,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.3734229621118243e-05,
+      "loss": 1.3985,
+      "step": 4415
+    },
+    {
+      "epoch": 0.761839040800483,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.373170450136897e-05,
+      "loss": 1.4265,
+      "step": 4416
+    },
+    {
+      "epoch": 0.762011558699215,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.3729179105151708e-05,
+      "loss": 1.4387,
+      "step": 4417
+    },
+    {
+      "epoch": 0.7621840765979471,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.3726653432653559e-05,
+      "loss": 1.514,
+      "step": 4418
+    },
+    {
+      "epoch": 0.762356594496679,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.3724127484061635e-05,
+      "loss": 1.5758,
+      "step": 4419
+    },
+    {
+      "epoch": 0.762529112395411,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.3721601259563083e-05,
+      "loss": 1.4226,
+      "step": 4420
+    },
+    {
+      "epoch": 0.7627016302941431,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.3719074759345052e-05,
+      "loss": 1.4775,
+      "step": 4421
+    },
+    {
+      "epoch": 0.762874148192875,
+      "grad_norm": 0.75,
+      "learning_rate": 1.3716547983594726e-05,
+      "loss": 1.5176,
+      "step": 4422
+    },
+    {
+      "epoch": 0.763046666091607,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.3714020932499303e-05,
+      "loss": 1.3939,
+      "step": 4423
+    },
+    {
+      "epoch": 0.763219183990339,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.3711493606246005e-05,
+      "loss": 1.4731,
+      "step": 4424
+    },
+    {
+      "epoch": 0.763391701889071,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.370896600502207e-05,
+      "loss": 1.5322,
+      "step": 4425
+    },
+    {
+      "epoch": 0.763564219787803,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.3706438129014756e-05,
+      "loss": 1.4391,
+      "step": 4426
+    },
+    {
+      "epoch": 0.7637367376865349,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.3703909978411348e-05,
+      "loss": 1.3965,
+      "step": 4427
+    },
+    {
+      "epoch": 0.763909255585267,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.3701381553399147e-05,
+      "loss": 1.4971,
+      "step": 4428
+    },
+    {
+      "epoch": 0.764081773483999,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.369885285416547e-05,
+      "loss": 1.4536,
+      "step": 4429
+    },
+    {
+      "epoch": 0.7642542913827309,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.3696323880897664e-05,
+      "loss": 1.4337,
+      "step": 4430
+    },
+    {
+      "epoch": 0.764426809281463,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.3693794633783087e-05,
+      "loss": 1.3907,
+      "step": 4431
+    },
+    {
+      "epoch": 0.764599327180195,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.3691265113009126e-05,
+      "loss": 1.5093,
+      "step": 4432
+    },
+    {
+      "epoch": 0.7647718450789269,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.3688735318763183e-05,
+      "loss": 1.374,
+      "step": 4433
+    },
+    {
+      "epoch": 0.7649443629776589,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.3686205251232676e-05,
+      "loss": 1.4475,
+      "step": 4434
+    },
+    {
+      "epoch": 0.765116880876391,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.3683674910605053e-05,
+      "loss": 1.3778,
+      "step": 4435
+    },
+    {
+      "epoch": 0.7652893987751229,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.3681144297067777e-05,
+      "loss": 1.3342,
+      "step": 4436
+    },
+    {
+      "epoch": 0.7654619166738549,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.367861341080833e-05,
+      "loss": 1.3892,
+      "step": 4437
+    },
+    {
+      "epoch": 0.765634434572587,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.3676082252014213e-05,
+      "loss": 1.4011,
+      "step": 4438
+    },
+    {
+      "epoch": 0.7658069524713189,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.3673550820872957e-05,
+      "loss": 1.465,
+      "step": 4439
+    },
+    {
+      "epoch": 0.7659794703700509,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.3671019117572104e-05,
+      "loss": 1.3866,
+      "step": 4440
+    },
+    {
+      "epoch": 0.7661519882687828,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.3668487142299217e-05,
+      "loss": 1.527,
+      "step": 4441
+    },
+    {
+      "epoch": 0.7663245061675149,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.3665954895241877e-05,
+      "loss": 1.4585,
+      "step": 4442
+    },
+    {
+      "epoch": 0.7664970240662469,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.3663422376587695e-05,
+      "loss": 1.4673,
+      "step": 4443
+    },
+    {
+      "epoch": 0.7666695419649788,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.3660889586524295e-05,
+      "loss": 1.4605,
+      "step": 4444
+    },
+    {
+      "epoch": 0.7668420598637109,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.3658356525239316e-05,
+      "loss": 1.4475,
+      "step": 4445
+    },
+    {
+      "epoch": 0.7670145777624429,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.365582319292043e-05,
+      "loss": 1.5515,
+      "step": 4446
+    },
+    {
+      "epoch": 0.7671870956611748,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.3653289589755314e-05,
+      "loss": 1.4148,
+      "step": 4447
+    },
+    {
+      "epoch": 0.7673596135599069,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.3650755715931685e-05,
+      "loss": 1.4041,
+      "step": 4448
+    },
+    {
+      "epoch": 0.7675321314586389,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.3648221571637259e-05,
+      "loss": 1.4792,
+      "step": 4449
+    },
+    {
+      "epoch": 0.7677046493573708,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.3645687157059783e-05,
+      "loss": 1.3566,
+      "step": 4450
+    },
+    {
+      "epoch": 0.7678771672561028,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.3643152472387024e-05,
+      "loss": 1.3547,
+      "step": 4451
+    },
+    {
+      "epoch": 0.7680496851548348,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.364061751780677e-05,
+      "loss": 1.4479,
+      "step": 4452
+    },
+    {
+      "epoch": 0.7682222030535668,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.3638082293506818e-05,
+      "loss": 1.4658,
+      "step": 4453
+    },
+    {
+      "epoch": 0.7683947209522988,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.3635546799674999e-05,
+      "loss": 1.4645,
+      "step": 4454
+    },
+    {
+      "epoch": 0.7685672388510308,
+      "grad_norm": 0.625,
+      "learning_rate": 1.3633011036499158e-05,
+      "loss": 1.4945,
+      "step": 4455
+    },
+    {
+      "epoch": 0.7687397567497628,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.3630475004167159e-05,
+      "loss": 1.4176,
+      "step": 4456
+    },
+    {
+      "epoch": 0.7689122746484948,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.3627938702866885e-05,
+      "loss": 1.3811,
+      "step": 4457
+    },
+    {
+      "epoch": 0.7690847925472267,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.3625402132786247e-05,
+      "loss": 1.5251,
+      "step": 4458
+    },
+    {
+      "epoch": 0.7692573104459588,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.3622865294113164e-05,
+      "loss": 1.5809,
+      "step": 4459
+    },
+    {
+      "epoch": 0.7694298283446908,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.3620328187035585e-05,
+      "loss": 1.4454,
+      "step": 4460
+    },
+    {
+      "epoch": 0.7696023462434227,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.3617790811741473e-05,
+      "loss": 1.5201,
+      "step": 4461
+    },
+    {
+      "epoch": 0.7697748641421548,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.3615253168418811e-05,
+      "loss": 1.5174,
+      "step": 4462
+    },
+    {
+      "epoch": 0.7699473820408868,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.3612715257255604e-05,
+      "loss": 1.5593,
+      "step": 4463
+    },
+    {
+      "epoch": 0.7701198999396187,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.3610177078439882e-05,
+      "loss": 1.4046,
+      "step": 4464
+    },
+    {
+      "epoch": 0.7702924178383507,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.3607638632159681e-05,
+      "loss": 1.5045,
+      "step": 4465
+    },
+    {
+      "epoch": 0.7704649357370827,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.3605099918603069e-05,
+      "loss": 1.3441,
+      "step": 4466
+    },
+    {
+      "epoch": 0.7706374536358147,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.360256093795813e-05,
+      "loss": 1.3589,
+      "step": 4467
+    },
+    {
+      "epoch": 0.7708099715345467,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.3600021690412968e-05,
+      "loss": 1.3647,
+      "step": 4468
+    },
+    {
+      "epoch": 0.7709824894332787,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.3597482176155705e-05,
+      "loss": 1.5231,
+      "step": 4469
+    },
+    {
+      "epoch": 0.7711550073320107,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.3594942395374482e-05,
+      "loss": 1.3733,
+      "step": 4470
+    },
+    {
+      "epoch": 0.7713275252307427,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.3592402348257465e-05,
+      "loss": 1.4228,
+      "step": 4471
+    },
+    {
+      "epoch": 0.7715000431294747,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.3589862034992838e-05,
+      "loss": 1.5187,
+      "step": 4472
+    },
+    {
+      "epoch": 0.7716725610282067,
+      "grad_norm": 0.625,
+      "learning_rate": 1.3587321455768798e-05,
+      "loss": 1.5008,
+      "step": 4473
+    },
+    {
+      "epoch": 0.7718450789269387,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.3584780610773572e-05,
+      "loss": 1.5196,
+      "step": 4474
+    },
+    {
+      "epoch": 0.7720175968256706,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.3582239500195399e-05,
+      "loss": 1.3848,
+      "step": 4475
+    },
+    {
+      "epoch": 0.7721901147244027,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.357969812422254e-05,
+      "loss": 1.4315,
+      "step": 4476
+    },
+    {
+      "epoch": 0.7723626326231346,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.3577156483043278e-05,
+      "loss": 1.4297,
+      "step": 4477
+    },
+    {
+      "epoch": 0.7725351505218666,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3574614576845912e-05,
+      "loss": 1.5183,
+      "step": 4478
+    },
+    {
+      "epoch": 0.7727076684205987,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.3572072405818762e-05,
+      "loss": 1.3574,
+      "step": 4479
+    },
+    {
+      "epoch": 0.7728801863193306,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.356952997015017e-05,
+      "loss": 1.3782,
+      "step": 4480
+    },
+    {
+      "epoch": 0.7730527042180626,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.3566987270028495e-05,
+      "loss": 1.3915,
+      "step": 4481
+    },
+    {
+      "epoch": 0.7732252221167946,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.3564444305642116e-05,
+      "loss": 1.4268,
+      "step": 4482
+    },
+    {
+      "epoch": 0.7733977400155266,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.3561901077179429e-05,
+      "loss": 1.4138,
+      "step": 4483
+    },
+    {
+      "epoch": 0.7735702579142586,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.3559357584828857e-05,
+      "loss": 1.384,
+      "step": 4484
+    },
+    {
+      "epoch": 0.7737427758129906,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.3556813828778833e-05,
+      "loss": 1.454,
+      "step": 4485
+    },
+    {
+      "epoch": 0.7739152937117226,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.3554269809217817e-05,
+      "loss": 1.3827,
+      "step": 4486
+    },
+    {
+      "epoch": 0.7740878116104546,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.3551725526334286e-05,
+      "loss": 1.4953,
+      "step": 4487
+    },
+    {
+      "epoch": 0.7742603295091866,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.3549180980316737e-05,
+      "loss": 1.3566,
+      "step": 4488
+    },
+    {
+      "epoch": 0.7744328474079186,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.3546636171353681e-05,
+      "loss": 1.4721,
+      "step": 4489
+    },
+    {
+      "epoch": 0.7746053653066506,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.354409109963366e-05,
+      "loss": 1.4763,
+      "step": 4490
+    },
+    {
+      "epoch": 0.7747778832053825,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.3541545765345222e-05,
+      "loss": 1.4729,
+      "step": 4491
+    },
+    {
+      "epoch": 0.7749504011041145,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.353900016867695e-05,
+      "loss": 1.4975,
+      "step": 4492
+    },
+    {
+      "epoch": 0.7751229190028466,
+      "grad_norm": 0.625,
+      "learning_rate": 1.3536454309817433e-05,
+      "loss": 1.4912,
+      "step": 4493
+    },
+    {
+      "epoch": 0.7752954369015785,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.3533908188955281e-05,
+      "loss": 1.4514,
+      "step": 4494
+    },
+    {
+      "epoch": 0.7754679548003105,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.3531361806279128e-05,
+      "loss": 1.5305,
+      "step": 4495
+    },
+    {
+      "epoch": 0.7756404726990426,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.352881516197763e-05,
+      "loss": 1.4676,
+      "step": 4496
+    },
+    {
+      "epoch": 0.7758129905977745,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.3526268256239456e-05,
+      "loss": 1.4794,
+      "step": 4497
+    },
+    {
+      "epoch": 0.7759855084965065,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.3523721089253296e-05,
+      "loss": 1.4701,
+      "step": 4498
+    },
+    {
+      "epoch": 0.7761580263952385,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.352117366120786e-05,
+      "loss": 1.483,
+      "step": 4499
+    },
+    {
+      "epoch": 0.7763305442939705,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.351862597229188e-05,
+      "loss": 1.4353,
+      "step": 4500
+    },
+    {
+      "epoch": 0.7763305442939705,
+      "eval_loss": 1.418400526046753,
+      "eval_runtime": 10.861,
+      "eval_samples_per_second": 94.282,
+      "eval_steps_per_second": 23.571,
+      "step": 4500
+    },
+    {
+      "epoch": 0.7765030621927025,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.35160780226941e-05,
+      "loss": 1.5238,
+      "step": 4501
+    },
+    {
+      "epoch": 0.7766755800914344,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.351352981260329e-05,
+      "loss": 1.4664,
+      "step": 4502
+    },
+    {
+      "epoch": 0.7768480979901665,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.351098134220824e-05,
+      "loss": 1.379,
+      "step": 4503
+    },
+    {
+      "epoch": 0.7770206158888985,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.3508432611697755e-05,
+      "loss": 1.3869,
+      "step": 4504
+    },
+    {
+      "epoch": 0.7771931337876304,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.350588362126066e-05,
+      "loss": 1.5221,
+      "step": 4505
+    },
+    {
+      "epoch": 0.7773656516863625,
+      "grad_norm": 0.625,
+      "learning_rate": 1.35033343710858e-05,
+      "loss": 1.3187,
+      "step": 4506
+    },
+    {
+      "epoch": 0.7775381695850945,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.350078486136204e-05,
+      "loss": 1.3784,
+      "step": 4507
+    },
+    {
+      "epoch": 0.7777106874838264,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.3498235092278263e-05,
+      "loss": 1.4881,
+      "step": 4508
+    },
+    {
+      "epoch": 0.7778832053825584,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.3495685064023374e-05,
+      "loss": 1.3369,
+      "step": 4509
+    },
+    {
+      "epoch": 0.7780557232812905,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.3493134776786292e-05,
+      "loss": 1.4683,
+      "step": 4510
+    },
+    {
+      "epoch": 0.7782282411800224,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.3490584230755956e-05,
+      "loss": 1.4639,
+      "step": 4511
+    },
+    {
+      "epoch": 0.7784007590787544,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.3488033426121336e-05,
+      "loss": 1.3748,
+      "step": 4512
+    },
+    {
+      "epoch": 0.7785732769774865,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.3485482363071402e-05,
+      "loss": 1.3758,
+      "step": 4513
+    },
+    {
+      "epoch": 0.7787457948762184,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.3482931041795158e-05,
+      "loss": 1.3721,
+      "step": 4514
+    },
+    {
+      "epoch": 0.7789183127749504,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.3480379462481619e-05,
+      "loss": 1.4502,
+      "step": 4515
+    },
+    {
+      "epoch": 0.7790908306736823,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.3477827625319826e-05,
+      "loss": 1.4347,
+      "step": 4516
+    },
+    {
+      "epoch": 0.7792633485724144,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.3475275530498828e-05,
+      "loss": 1.4338,
+      "step": 4517
+    },
+    {
+      "epoch": 0.7794358664711464,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.3472723178207708e-05,
+      "loss": 1.4674,
+      "step": 4518
+    },
+    {
+      "epoch": 0.7796083843698783,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.3470170568635552e-05,
+      "loss": 1.51,
+      "step": 4519
+    },
+    {
+      "epoch": 0.7797809022686104,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.346761770197148e-05,
+      "loss": 1.4031,
+      "step": 4520
+    },
+    {
+      "epoch": 0.7799534201673424,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.3465064578404623e-05,
+      "loss": 1.4428,
+      "step": 4521
+    },
+    {
+      "epoch": 0.7801259380660743,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.346251119812413e-05,
+      "loss": 1.5229,
+      "step": 4522
+    },
+    {
+      "epoch": 0.7802984559648064,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.3459957561319175e-05,
+      "loss": 1.4638,
+      "step": 4523
+    },
+    {
+      "epoch": 0.7804709738635384,
+      "grad_norm": 0.625,
+      "learning_rate": 1.3457403668178941e-05,
+      "loss": 1.384,
+      "step": 4524
+    },
+    {
+      "epoch": 0.7806434917622703,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.3454849518892644e-05,
+      "loss": 1.4915,
+      "step": 4525
+    },
+    {
+      "epoch": 0.7808160096610023,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.3452295113649505e-05,
+      "loss": 1.4184,
+      "step": 4526
+    },
+    {
+      "epoch": 0.7809885275597344,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.3449740452638773e-05,
+      "loss": 1.3411,
+      "step": 4527
+    },
+    {
+      "epoch": 0.7811610454584663,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.3447185536049716e-05,
+      "loss": 1.4498,
+      "step": 4528
+    },
+    {
+      "epoch": 0.7813335633571983,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.3444630364071614e-05,
+      "loss": 1.4373,
+      "step": 4529
+    },
+    {
+      "epoch": 0.7815060812559304,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.344207493689377e-05,
+      "loss": 1.5779,
+      "step": 4530
+    },
+    {
+      "epoch": 0.7816785991546623,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.3439519254705509e-05,
+      "loss": 1.395,
+      "step": 4531
+    },
+    {
+      "epoch": 0.7818511170533943,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.3436963317696172e-05,
+      "loss": 1.498,
+      "step": 4532
+    },
+    {
+      "epoch": 0.7820236349521262,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.3434407126055115e-05,
+      "loss": 1.4939,
+      "step": 4533
+    },
+    {
+      "epoch": 0.7821961528508583,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.343185067997172e-05,
+      "loss": 1.411,
+      "step": 4534
+    },
+    {
+      "epoch": 0.7823686707495903,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.3429293979635382e-05,
+      "loss": 1.4517,
+      "step": 4535
+    },
+    {
+      "epoch": 0.7825411886483222,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.3426737025235521e-05,
+      "loss": 1.4377,
+      "step": 4536
+    },
+    {
+      "epoch": 0.7827137065470543,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.3424179816961572e-05,
+      "loss": 1.4659,
+      "step": 4537
+    },
+    {
+      "epoch": 0.7828862244457863,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.3421622355002981e-05,
+      "loss": 1.5162,
+      "step": 4538
+    },
+    {
+      "epoch": 0.7830587423445182,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.3419064639549232e-05,
+      "loss": 1.4428,
+      "step": 4539
+    },
+    {
+      "epoch": 0.7832312602432502,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.3416506670789807e-05,
+      "loss": 1.454,
+      "step": 4540
+    },
+    {
+      "epoch": 0.7834037781419823,
+      "grad_norm": 0.625,
+      "learning_rate": 1.3413948448914223e-05,
+      "loss": 1.4996,
+      "step": 4541
+    },
+    {
+      "epoch": 0.7835762960407142,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.3411389974112006e-05,
+      "loss": 1.4176,
+      "step": 4542
+    },
+    {
+      "epoch": 0.7837488139394462,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.3408831246572703e-05,
+      "loss": 1.4108,
+      "step": 4543
+    },
+    {
+      "epoch": 0.7839213318381782,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.3406272266485882e-05,
+      "loss": 1.4055,
+      "step": 4544
+    },
+    {
+      "epoch": 0.7840938497369102,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.340371303404113e-05,
+      "loss": 1.4123,
+      "step": 4545
+    },
+    {
+      "epoch": 0.7842663676356422,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.3401153549428048e-05,
+      "loss": 1.4729,
+      "step": 4546
+    },
+    {
+      "epoch": 0.7844388855343742,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.3398593812836259e-05,
+      "loss": 1.5255,
+      "step": 4547
+    },
+    {
+      "epoch": 0.7846114034331062,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.3396033824455408e-05,
+      "loss": 1.4295,
+      "step": 4548
+    },
+    {
+      "epoch": 0.7847839213318382,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.339347358447515e-05,
+      "loss": 1.3872,
+      "step": 4549
+    },
+    {
+      "epoch": 0.7849564392305701,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.3390913093085165e-05,
+      "loss": 1.3967,
+      "step": 4550
+    },
+    {
+      "epoch": 0.7851289571293022,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.338835235047515e-05,
+      "loss": 1.4759,
+      "step": 4551
+    },
+    {
+      "epoch": 0.7853014750280342,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.3385791356834821e-05,
+      "loss": 1.3507,
+      "step": 4552
+    },
+    {
+      "epoch": 0.7854739929267661,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.3383230112353917e-05,
+      "loss": 1.3755,
+      "step": 4553
+    },
+    {
+      "epoch": 0.7856465108254982,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.3380668617222183e-05,
+      "loss": 1.4981,
+      "step": 4554
+    },
+    {
+      "epoch": 0.7858190287242302,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.3378106871629395e-05,
+      "loss": 1.4185,
+      "step": 4555
+    },
+    {
+      "epoch": 0.7859915466229621,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.3375544875765343e-05,
+      "loss": 1.5073,
+      "step": 4556
+    },
+    {
+      "epoch": 0.7861640645216941,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.3372982629819837e-05,
+      "loss": 1.4959,
+      "step": 4557
+    },
+    {
+      "epoch": 0.7863365824204261,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.33704201339827e-05,
+      "loss": 1.4126,
+      "step": 4558
+    },
+    {
+      "epoch": 0.7865091003191581,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.336785738844378e-05,
+      "loss": 1.456,
+      "step": 4559
+    },
+    {
+      "epoch": 0.7866816182178901,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.3365294393392942e-05,
+      "loss": 1.545,
+      "step": 4560
+    },
+    {
+      "epoch": 0.7868541361166221,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.336273114902007e-05,
+      "loss": 1.4324,
+      "step": 4561
+    },
+    {
+      "epoch": 0.7870266540153541,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.336016765551506e-05,
+      "loss": 1.4027,
+      "step": 4562
+    },
+    {
+      "epoch": 0.7871991719140861,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.3357603913067836e-05,
+      "loss": 1.4484,
+      "step": 4563
+    },
+    {
+      "epoch": 0.7873716898128181,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.3355039921868334e-05,
+      "loss": 1.447,
+      "step": 4564
+    },
+    {
+      "epoch": 0.7875442077115501,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.3352475682106515e-05,
+      "loss": 1.5331,
+      "step": 4565
+    },
+    {
+      "epoch": 0.787716725610282,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.3349911193972345e-05,
+      "loss": 1.5616,
+      "step": 4566
+    },
+    {
+      "epoch": 0.787889243509014,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.3347346457655826e-05,
+      "loss": 1.4837,
+      "step": 4567
+    },
+    {
+      "epoch": 0.7880617614077461,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.3344781473346964e-05,
+      "loss": 1.4843,
+      "step": 4568
+    },
+    {
+      "epoch": 0.788234279306478,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.3342216241235794e-05,
+      "loss": 1.3826,
+      "step": 4569
+    },
+    {
+      "epoch": 0.78840679720521,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.3339650761512361e-05,
+      "loss": 1.4971,
+      "step": 4570
+    },
+    {
+      "epoch": 0.7885793151039421,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.333708503436673e-05,
+      "loss": 1.4259,
+      "step": 4571
+    },
+    {
+      "epoch": 0.788751833002674,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.333451905998899e-05,
+      "loss": 1.5063,
+      "step": 4572
+    },
+    {
+      "epoch": 0.788924350901406,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.3331952838569246e-05,
+      "loss": 1.4997,
+      "step": 4573
+    },
+    {
+      "epoch": 0.789096868800138,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.3329386370297615e-05,
+      "loss": 1.4358,
+      "step": 4574
+    },
+    {
+      "epoch": 0.78926938669887,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.3326819655364236e-05,
+      "loss": 1.4828,
+      "step": 4575
+    },
+    {
+      "epoch": 0.789441904597602,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.3324252693959271e-05,
+      "loss": 1.4323,
+      "step": 4576
+    },
+    {
+      "epoch": 0.789614422496334,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.3321685486272898e-05,
+      "loss": 1.5302,
+      "step": 4577
+    },
+    {
+      "epoch": 0.789786940395066,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.3319118032495306e-05,
+      "loss": 1.5162,
+      "step": 4578
+    },
+    {
+      "epoch": 0.789959458293798,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.3316550332816713e-05,
+      "loss": 1.5533,
+      "step": 4579
+    },
+    {
+      "epoch": 0.79013197619253,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.3313982387427346e-05,
+      "loss": 1.3622,
+      "step": 4580
+    },
+    {
+      "epoch": 0.790304494091262,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.3311414196517462e-05,
+      "loss": 1.3823,
+      "step": 4581
+    },
+    {
+      "epoch": 0.790477011989994,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.3308845760277322e-05,
+      "loss": 1.4654,
+      "step": 4582
+    },
+    {
+      "epoch": 0.7906495298887259,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.3306277078897211e-05,
+      "loss": 1.3795,
+      "step": 4583
+    },
+    {
+      "epoch": 0.7908220477874579,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.3303708152567439e-05,
+      "loss": 1.4473,
+      "step": 4584
+    },
+    {
+      "epoch": 0.79099456568619,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.3301138981478322e-05,
+      "loss": 1.4669,
+      "step": 4585
+    },
+    {
+      "epoch": 0.7911670835849219,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.3298569565820205e-05,
+      "loss": 1.4592,
+      "step": 4586
+    },
+    {
+      "epoch": 0.7913396014836539,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.3295999905783444e-05,
+      "loss": 1.463,
+      "step": 4587
+    },
+    {
+      "epoch": 0.791512119382386,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.3293430001558411e-05,
+      "loss": 1.4708,
+      "step": 4588
+    },
+    {
+      "epoch": 0.7916846372811179,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.3290859853335512e-05,
+      "loss": 1.4816,
+      "step": 4589
+    },
+    {
+      "epoch": 0.7918571551798499,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.3288289461305149e-05,
+      "loss": 1.4373,
+      "step": 4590
+    },
+    {
+      "epoch": 0.7920296730785819,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.3285718825657754e-05,
+      "loss": 1.4675,
+      "step": 4591
+    },
+    {
+      "epoch": 0.7922021909773139,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.3283147946583781e-05,
+      "loss": 1.3418,
+      "step": 4592
+    },
+    {
+      "epoch": 0.7923747088760459,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.3280576824273694e-05,
+      "loss": 1.4304,
+      "step": 4593
+    },
+    {
+      "epoch": 0.7925472267747778,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.3278005458917978e-05,
+      "loss": 1.4713,
+      "step": 4594
+    },
+    {
+      "epoch": 0.7927197446735099,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.3275433850707136e-05,
+      "loss": 1.3984,
+      "step": 4595
+    },
+    {
+      "epoch": 0.7928922625722419,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.3272861999831688e-05,
+      "loss": 1.3734,
+      "step": 4596
+    },
+    {
+      "epoch": 0.7930647804709738,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.3270289906482174e-05,
+      "loss": 1.4992,
+      "step": 4597
+    },
+    {
+      "epoch": 0.7932372983697059,
+      "grad_norm": 0.625,
+      "learning_rate": 1.326771757084915e-05,
+      "loss": 1.416,
+      "step": 4598
+    },
+    {
+      "epoch": 0.7934098162684379,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.326514499312319e-05,
+      "loss": 1.4168,
+      "step": 4599
+    },
+    {
+      "epoch": 0.7935823341671698,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.3262572173494888e-05,
+      "loss": 1.481,
+      "step": 4600
+    },
+    {
+      "epoch": 0.7935823341671698,
+      "eval_loss": 1.4177677631378174,
+      "eval_runtime": 10.9052,
+      "eval_samples_per_second": 93.9,
+      "eval_steps_per_second": 23.475,
+      "step": 4600
+    },
+    {
+      "epoch": 0.7937548520659018,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.3259999112154854e-05,
+      "loss": 1.446,
+      "step": 4601
+    },
+    {
+      "epoch": 0.7939273699646339,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.3257425809293714e-05,
+      "loss": 1.5257,
+      "step": 4602
+    },
+    {
+      "epoch": 0.7940998878633658,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.3254852265102118e-05,
+      "loss": 1.4193,
+      "step": 4603
+    },
+    {
+      "epoch": 0.7942724057620978,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.325227847977073e-05,
+      "loss": 1.4235,
+      "step": 4604
+    },
+    {
+      "epoch": 0.7944449236608299,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.324970445349023e-05,
+      "loss": 1.3991,
+      "step": 4605
+    },
+    {
+      "epoch": 0.7946174415595618,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.3247130186451321e-05,
+      "loss": 1.4472,
+      "step": 4606
+    },
+    {
+      "epoch": 0.7947899594582938,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.3244555678844717e-05,
+      "loss": 1.4479,
+      "step": 4607
+    },
+    {
+      "epoch": 0.7949624773570257,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.3241980930861153e-05,
+      "loss": 1.3988,
+      "step": 4608
+    },
+    {
+      "epoch": 0.7951349952557578,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.3239405942691388e-05,
+      "loss": 1.5663,
+      "step": 4609
+    },
+    {
+      "epoch": 0.7953075131544898,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.323683071452619e-05,
+      "loss": 1.4208,
+      "step": 4610
+    },
+    {
+      "epoch": 0.7954800310532217,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.3234255246556347e-05,
+      "loss": 1.3979,
+      "step": 4611
+    },
+    {
+      "epoch": 0.7956525489519538,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.3231679538972668e-05,
+      "loss": 1.4996,
+      "step": 4612
+    },
+    {
+      "epoch": 0.7958250668506858,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.3229103591965977e-05,
+      "loss": 1.4751,
+      "step": 4613
+    },
+    {
+      "epoch": 0.7959975847494177,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.3226527405727115e-05,
+      "loss": 1.4927,
+      "step": 4614
+    },
+    {
+      "epoch": 0.7961701026481497,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.322395098044694e-05,
+      "loss": 1.5126,
+      "step": 4615
+    },
+    {
+      "epoch": 0.7963426205468818,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.3221374316316334e-05,
+      "loss": 1.4083,
+      "step": 4616
+    },
+    {
+      "epoch": 0.7965151384456137,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.321879741352619e-05,
+      "loss": 1.4313,
+      "step": 4617
+    },
+    {
+      "epoch": 0.7966876563443457,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.3216220272267421e-05,
+      "loss": 1.4374,
+      "step": 4618
+    },
+    {
+      "epoch": 0.7968601742430778,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.321364289273096e-05,
+      "loss": 1.4172,
+      "step": 4619
+    },
+    {
+      "epoch": 0.7970326921418097,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.3211065275107754e-05,
+      "loss": 1.469,
+      "step": 4620
+    },
+    {
+      "epoch": 0.7972052100405417,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.3208487419588768e-05,
+      "loss": 1.4686,
+      "step": 4621
+    },
+    {
+      "epoch": 0.7973777279392738,
+      "grad_norm": 0.75,
+      "learning_rate": 1.3205909326364986e-05,
+      "loss": 1.5068,
+      "step": 4622
+    },
+    {
+      "epoch": 0.7975502458380057,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.320333099562741e-05,
+      "loss": 1.4224,
+      "step": 4623
+    },
+    {
+      "epoch": 0.7977227637367377,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.3200752427567056e-05,
+      "loss": 1.4457,
+      "step": 4624
+    },
+    {
+      "epoch": 0.7978952816354696,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.3198173622374966e-05,
+      "loss": 1.3616,
+      "step": 4625
+    },
+    {
+      "epoch": 0.7980677995342017,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.319559458024219e-05,
+      "loss": 1.4188,
+      "step": 4626
+    },
+    {
+      "epoch": 0.7982403174329337,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.31930153013598e-05,
+      "loss": 1.4014,
+      "step": 4627
+    },
+    {
+      "epoch": 0.7984128353316656,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.3190435785918887e-05,
+      "loss": 1.4106,
+      "step": 4628
+    },
+    {
+      "epoch": 0.7985853532303977,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.3187856034110555e-05,
+      "loss": 1.4869,
+      "step": 4629
+    },
+    {
+      "epoch": 0.7987578711291297,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.318527604612593e-05,
+      "loss": 1.5005,
+      "step": 4630
+    },
+    {
+      "epoch": 0.7989303890278616,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.3182695822156153e-05,
+      "loss": 1.5098,
+      "step": 4631
+    },
+    {
+      "epoch": 0.7991029069265936,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.3180115362392383e-05,
+      "loss": 1.5083,
+      "step": 4632
+    },
+    {
+      "epoch": 0.7992754248253257,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.31775346670258e-05,
+      "loss": 1.4761,
+      "step": 4633
+    },
+    {
+      "epoch": 0.7994479427240576,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.3174953736247589e-05,
+      "loss": 1.4419,
+      "step": 4634
+    },
+    {
+      "epoch": 0.7996204606227896,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.317237257024897e-05,
+      "loss": 1.3964,
+      "step": 4635
+    },
+    {
+      "epoch": 0.7997929785215216,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.3169791169221168e-05,
+      "loss": 1.3349,
+      "step": 4636
+    },
+    {
+      "epoch": 0.7999654964202536,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.3167209533355432e-05,
+      "loss": 1.3971,
+      "step": 4637
+    },
+    {
+      "epoch": 0.8001380143189856,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.3164627662843024e-05,
+      "loss": 1.4125,
+      "step": 4638
+    },
+    {
+      "epoch": 0.8003105322177176,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.3162045557875223e-05,
+      "loss": 1.4942,
+      "step": 4639
+    },
+    {
+      "epoch": 0.8004830501164496,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.315946321864333e-05,
+      "loss": 1.469,
+      "step": 4640
+    },
+    {
+      "epoch": 0.8006555680151816,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.3156880645338663e-05,
+      "loss": 1.4833,
+      "step": 4641
+    },
+    {
+      "epoch": 0.8008280859139135,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.315429783815255e-05,
+      "loss": 1.5352,
+      "step": 4642
+    },
+    {
+      "epoch": 0.8010006038126456,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.3151714797276345e-05,
+      "loss": 1.4352,
+      "step": 4643
+    },
+    {
+      "epoch": 0.8011731217113776,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.3149131522901417e-05,
+      "loss": 1.367,
+      "step": 4644
+    },
+    {
+      "epoch": 0.8013456396101095,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.3146548015219147e-05,
+      "loss": 1.3862,
+      "step": 4645
+    },
+    {
+      "epoch": 0.8015181575088416,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.3143964274420941e-05,
+      "loss": 1.3852,
+      "step": 4646
+    },
+    {
+      "epoch": 0.8016906754075735,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.3141380300698217e-05,
+      "loss": 1.4443,
+      "step": 4647
+    },
+    {
+      "epoch": 0.8018631933063055,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.313879609424241e-05,
+      "loss": 1.4288,
+      "step": 4648
+    },
+    {
+      "epoch": 0.8020357112050375,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.3136211655244979e-05,
+      "loss": 1.4431,
+      "step": 4649
+    },
+    {
+      "epoch": 0.8022082291037695,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.3133626983897392e-05,
+      "loss": 1.4145,
+      "step": 4650
+    },
+    {
+      "epoch": 0.8023807470025015,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.3131042080391136e-05,
+      "loss": 1.3694,
+      "step": 4651
+    },
+    {
+      "epoch": 0.8025532649012335,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.312845694491772e-05,
+      "loss": 1.4015,
+      "step": 4652
+    },
+    {
+      "epoch": 0.8027257827999655,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.3125871577668665e-05,
+      "loss": 1.4519,
+      "step": 4653
+    },
+    {
+      "epoch": 0.8028983006986975,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.3123285978835517e-05,
+      "loss": 1.4086,
+      "step": 4654
+    },
+    {
+      "epoch": 0.8030708185974295,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.3120700148609824e-05,
+      "loss": 1.4072,
+      "step": 4655
+    },
+    {
+      "epoch": 0.8032433364961615,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.3118114087183164e-05,
+      "loss": 1.3822,
+      "step": 4656
+    },
+    {
+      "epoch": 0.8034158543948935,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.3115527794747134e-05,
+      "loss": 1.3953,
+      "step": 4657
+    },
+    {
+      "epoch": 0.8035883722936255,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.3112941271493336e-05,
+      "loss": 1.4183,
+      "step": 4658
+    },
+    {
+      "epoch": 0.8037608901923574,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.3110354517613396e-05,
+      "loss": 1.4297,
+      "step": 4659
+    },
+    {
+      "epoch": 0.8039334080910895,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.310776753329896e-05,
+      "loss": 1.5198,
+      "step": 4660
+    },
+    {
+      "epoch": 0.8041059259898214,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.3105180318741691e-05,
+      "loss": 1.3614,
+      "step": 4661
+    },
+    {
+      "epoch": 0.8042784438885534,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.3102592874133257e-05,
+      "loss": 1.3834,
+      "step": 4662
+    },
+    {
+      "epoch": 0.8044509617872855,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.310000519966536e-05,
+      "loss": 1.5917,
+      "step": 4663
+    },
+    {
+      "epoch": 0.8046234796860174,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.3097417295529706e-05,
+      "loss": 1.5054,
+      "step": 4664
+    },
+    {
+      "epoch": 0.8047959975847494,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.3094829161918028e-05,
+      "loss": 1.5035,
+      "step": 4665
+    },
+    {
+      "epoch": 0.8049685154834814,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.3092240799022065e-05,
+      "loss": 1.3778,
+      "step": 4666
+    },
+    {
+      "epoch": 0.8051410333822134,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.3089652207033583e-05,
+      "loss": 1.5439,
+      "step": 4667
+    },
+    {
+      "epoch": 0.8053135512809454,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.3087063386144361e-05,
+      "loss": 1.491,
+      "step": 4668
+    },
+    {
+      "epoch": 0.8054860691796774,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.3084474336546196e-05,
+      "loss": 1.4339,
+      "step": 4669
+    },
+    {
+      "epoch": 0.8056585870784094,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.3081885058430899e-05,
+      "loss": 1.4916,
+      "step": 4670
+    },
+    {
+      "epoch": 0.8058311049771414,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.30792955519903e-05,
+      "loss": 1.4565,
+      "step": 4671
+    },
+    {
+      "epoch": 0.8060036228758733,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.3076705817416242e-05,
+      "loss": 1.5596,
+      "step": 4672
+    },
+    {
+      "epoch": 0.8061761407746053,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.3074115854900598e-05,
+      "loss": 1.4778,
+      "step": 4673
+    },
+    {
+      "epoch": 0.8063486586733374,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.3071525664635241e-05,
+      "loss": 1.4988,
+      "step": 4674
+    },
+    {
+      "epoch": 0.8065211765720693,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.306893524681207e-05,
+      "loss": 1.3927,
+      "step": 4675
+    },
+    {
+      "epoch": 0.8066936944708013,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.3066344601623004e-05,
+      "loss": 1.4619,
+      "step": 4676
+    },
+    {
+      "epoch": 0.8068662123695334,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.3063753729259972e-05,
+      "loss": 1.4088,
+      "step": 4677
+    },
+    {
+      "epoch": 0.8070387302682653,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.3061162629914917e-05,
+      "loss": 1.491,
+      "step": 4678
+    },
+    {
+      "epoch": 0.8072112481669973,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.3058571303779806e-05,
+      "loss": 1.4972,
+      "step": 4679
+    },
+    {
+      "epoch": 0.8073837660657294,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.3055979751046624e-05,
+      "loss": 1.4425,
+      "step": 4680
+    },
+    {
+      "epoch": 0.8075562839644613,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.3053387971907368e-05,
+      "loss": 1.4029,
+      "step": 4681
+    },
+    {
+      "epoch": 0.8077288018631933,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.3050795966554051e-05,
+      "loss": 1.4645,
+      "step": 4682
+    },
+    {
+      "epoch": 0.8079013197619253,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.3048203735178709e-05,
+      "loss": 1.4078,
+      "step": 4683
+    },
+    {
+      "epoch": 0.8080738376606573,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.3045611277973385e-05,
+      "loss": 1.5524,
+      "step": 4684
+    },
+    {
+      "epoch": 0.8082463555593893,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.3043018595130148e-05,
+      "loss": 1.4131,
+      "step": 4685
+    },
+    {
+      "epoch": 0.8084188734581212,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.3040425686841083e-05,
+      "loss": 1.4879,
+      "step": 4686
+    },
+    {
+      "epoch": 0.8085913913568533,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.3037832553298282e-05,
+      "loss": 1.3294,
+      "step": 4687
+    },
+    {
+      "epoch": 0.8087639092555853,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.3035239194693865e-05,
+      "loss": 1.4486,
+      "step": 4688
+    },
+    {
+      "epoch": 0.8089364271543172,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.3032645611219965e-05,
+      "loss": 1.4143,
+      "step": 4689
+    },
+    {
+      "epoch": 0.8091089450530492,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.3030051803068729e-05,
+      "loss": 1.4405,
+      "step": 4690
+    },
+    {
+      "epoch": 0.8092814629517813,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.3027457770432322e-05,
+      "loss": 1.4022,
+      "step": 4691
+    },
+    {
+      "epoch": 0.8094539808505132,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.3024863513502926e-05,
+      "loss": 1.4181,
+      "step": 4692
+    },
+    {
+      "epoch": 0.8096264987492452,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.3022269032472745e-05,
+      "loss": 1.5205,
+      "step": 4693
+    },
+    {
+      "epoch": 0.8097990166479773,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.3019674327533984e-05,
+      "loss": 1.4996,
+      "step": 4694
+    },
+    {
+      "epoch": 0.8099715345467092,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.3017079398878884e-05,
+      "loss": 1.3902,
+      "step": 4695
+    },
+    {
+      "epoch": 0.8101440524454412,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.3014484246699693e-05,
+      "loss": 1.4089,
+      "step": 4696
+    },
+    {
+      "epoch": 0.8103165703441733,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.3011888871188669e-05,
+      "loss": 1.4944,
+      "step": 4697
+    },
+    {
+      "epoch": 0.8104890882429052,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.3009293272538104e-05,
+      "loss": 1.438,
+      "step": 4698
+    },
+    {
+      "epoch": 0.8106616061416372,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.3006697450940284e-05,
+      "loss": 1.4211,
+      "step": 4699
+    },
+    {
+      "epoch": 0.8108341240403691,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.3004101406587535e-05,
+      "loss": 1.3598,
+      "step": 4700
+    },
+    {
+      "epoch": 0.8108341240403691,
+      "eval_loss": 1.4166769981384277,
+      "eval_runtime": 10.9127,
+      "eval_samples_per_second": 93.836,
+      "eval_steps_per_second": 23.459,
+      "step": 4700
+    },
+    {
+      "epoch": 0.8110066419391012,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.300150513967218e-05,
+      "loss": 1.4421,
+      "step": 4701
+    },
+    {
+      "epoch": 0.8111791598378332,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.2998908650386573e-05,
+      "loss": 1.4649,
+      "step": 4702
+    },
+    {
+      "epoch": 0.8113516777365651,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.2996311938923074e-05,
+      "loss": 1.5004,
+      "step": 4703
+    },
+    {
+      "epoch": 0.8115241956352972,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.299371500547406e-05,
+      "loss": 1.3757,
+      "step": 4704
+    },
+    {
+      "epoch": 0.8116967135340292,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.299111785023194e-05,
+      "loss": 1.471,
+      "step": 4705
+    },
+    {
+      "epoch": 0.8118692314327611,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.2988520473389117e-05,
+      "loss": 1.4475,
+      "step": 4706
+    },
+    {
+      "epoch": 0.8120417493314931,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.2985922875138025e-05,
+      "loss": 1.417,
+      "step": 4707
+    },
+    {
+      "epoch": 0.8122142672302252,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.2983325055671108e-05,
+      "loss": 1.4515,
+      "step": 4708
+    },
+    {
+      "epoch": 0.8123867851289571,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.2980727015180833e-05,
+      "loss": 1.4563,
+      "step": 4709
+    },
+    {
+      "epoch": 0.8125593030276891,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.2978128753859674e-05,
+      "loss": 1.4191,
+      "step": 4710
+    },
+    {
+      "epoch": 0.8127318209264212,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.2975530271900127e-05,
+      "loss": 1.3702,
+      "step": 4711
+    },
+    {
+      "epoch": 0.8129043388251531,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.2972931569494707e-05,
+      "loss": 1.4901,
+      "step": 4712
+    },
+    {
+      "epoch": 0.8130768567238851,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.2970332646835942e-05,
+      "loss": 1.4811,
+      "step": 4713
+    },
+    {
+      "epoch": 0.8132493746226171,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.2967733504116375e-05,
+      "loss": 1.4553,
+      "step": 4714
+    },
+    {
+      "epoch": 0.8134218925213491,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.2965134141528565e-05,
+      "loss": 1.4019,
+      "step": 4715
+    },
+    {
+      "epoch": 0.8135944104200811,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.2962534559265092e-05,
+      "loss": 1.435,
+      "step": 4716
+    },
+    {
+      "epoch": 0.813766928318813,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.2959934757518548e-05,
+      "loss": 1.4376,
+      "step": 4717
+    },
+    {
+      "epoch": 0.8139394462175451,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.2957334736481544e-05,
+      "loss": 1.4521,
+      "step": 4718
+    },
+    {
+      "epoch": 0.8141119641162771,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.2954734496346704e-05,
+      "loss": 1.4023,
+      "step": 4719
+    },
+    {
+      "epoch": 0.814284482015009,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.2952134037306667e-05,
+      "loss": 1.4733,
+      "step": 4720
+    },
+    {
+      "epoch": 0.8144569999137411,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.2949533359554099e-05,
+      "loss": 1.4591,
+      "step": 4721
+    },
+    {
+      "epoch": 0.8146295178124731,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.2946932463281667e-05,
+      "loss": 1.4012,
+      "step": 4722
+    },
+    {
+      "epoch": 0.814802035711205,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.2944331348682067e-05,
+      "loss": 1.476,
+      "step": 4723
+    },
+    {
+      "epoch": 0.814974553609937,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.2941730015948001e-05,
+      "loss": 1.4007,
+      "step": 4724
+    },
+    {
+      "epoch": 0.815147071508669,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.2939128465272197e-05,
+      "loss": 1.3761,
+      "step": 4725
+    },
+    {
+      "epoch": 0.815319589407401,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.2936526696847392e-05,
+      "loss": 1.3495,
+      "step": 4726
+    },
+    {
+      "epoch": 0.815492107306133,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.2933924710866342e-05,
+      "loss": 1.4033,
+      "step": 4727
+    },
+    {
+      "epoch": 0.815664625204865,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.2931322507521813e-05,
+      "loss": 1.456,
+      "step": 4728
+    },
+    {
+      "epoch": 0.815837143103597,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.29287200870066e-05,
+      "loss": 1.3995,
+      "step": 4729
+    },
+    {
+      "epoch": 0.816009661002329,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.2926117449513505e-05,
+      "loss": 1.4896,
+      "step": 4730
+    },
+    {
+      "epoch": 0.816182178901061,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.2923514595235345e-05,
+      "loss": 1.4255,
+      "step": 4731
+    },
+    {
+      "epoch": 0.816354696799793,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.2920911524364954e-05,
+      "loss": 1.4791,
+      "step": 4732
+    },
+    {
+      "epoch": 0.816527214698525,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.291830823709519e-05,
+      "loss": 1.4537,
+      "step": 4733
+    },
+    {
+      "epoch": 0.8166997325972569,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.2915704733618917e-05,
+      "loss": 1.5038,
+      "step": 4734
+    },
+    {
+      "epoch": 0.816872250495989,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.2913101014129017e-05,
+      "loss": 1.5189,
+      "step": 4735
+    },
+    {
+      "epoch": 0.817044768394721,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.291049707881839e-05,
+      "loss": 1.3664,
+      "step": 4736
+    },
+    {
+      "epoch": 0.8172172862934529,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.2907892927879959e-05,
+      "loss": 1.3988,
+      "step": 4737
+    },
+    {
+      "epoch": 0.817389804192185,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.2905288561506649e-05,
+      "loss": 1.4323,
+      "step": 4738
+    },
+    {
+      "epoch": 0.817562322090917,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.2902683979891406e-05,
+      "loss": 1.4768,
+      "step": 4739
+    },
+    {
+      "epoch": 0.8177348399896489,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.29000791832272e-05,
+      "loss": 1.4621,
+      "step": 4740
+    },
+    {
+      "epoch": 0.8179073578883809,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.2897474171707008e-05,
+      "loss": 1.3043,
+      "step": 4741
+    },
+    {
+      "epoch": 0.8180798757871129,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.2894868945523826e-05,
+      "loss": 1.4202,
+      "step": 4742
+    },
+    {
+      "epoch": 0.8182523936858449,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.2892263504870662e-05,
+      "loss": 1.4528,
+      "step": 4743
+    },
+    {
+      "epoch": 0.8184249115845769,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.2889657849940547e-05,
+      "loss": 1.4524,
+      "step": 4744
+    },
+    {
+      "epoch": 0.8185974294833089,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.2887051980926522e-05,
+      "loss": 1.4295,
+      "step": 4745
+    },
+    {
+      "epoch": 0.8187699473820409,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.2884445898021649e-05,
+      "loss": 1.4305,
+      "step": 4746
+    },
+    {
+      "epoch": 0.8189424652807729,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.2881839601419e-05,
+      "loss": 1.5478,
+      "step": 4747
+    },
+    {
+      "epoch": 0.8191149831795048,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.2879233091311667e-05,
+      "loss": 1.6081,
+      "step": 4748
+    },
+    {
+      "epoch": 0.8192875010782369,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.2876626367892754e-05,
+      "loss": 1.4978,
+      "step": 4749
+    },
+    {
+      "epoch": 0.8194600189769689,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.2874019431355392e-05,
+      "loss": 1.4072,
+      "step": 4750
+    },
+    {
+      "epoch": 0.8196325368757008,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.2871412281892705e-05,
+      "loss": 1.4681,
+      "step": 4751
+    },
+    {
+      "epoch": 0.8198050547744329,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.2868804919697858e-05,
+      "loss": 1.4266,
+      "step": 4752
+    },
+    {
+      "epoch": 0.8199775726731648,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.286619734496402e-05,
+      "loss": 1.4385,
+      "step": 4753
+    },
+    {
+      "epoch": 0.8201500905718968,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.2863589557884371e-05,
+      "loss": 1.5308,
+      "step": 4754
+    },
+    {
+      "epoch": 0.8203226084706289,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.2860981558652114e-05,
+      "loss": 1.3169,
+      "step": 4755
+    },
+    {
+      "epoch": 0.8204951263693608,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.2858373347460469e-05,
+      "loss": 1.4768,
+      "step": 4756
+    },
+    {
+      "epoch": 0.8206676442680928,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.2855764924502665e-05,
+      "loss": 1.4356,
+      "step": 4757
+    },
+    {
+      "epoch": 0.8208401621668248,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.2853156289971955e-05,
+      "loss": 1.5222,
+      "step": 4758
+    },
+    {
+      "epoch": 0.8210126800655568,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.2850547444061597e-05,
+      "loss": 1.523,
+      "step": 4759
+    },
+    {
+      "epoch": 0.8211851979642888,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.2847938386964871e-05,
+      "loss": 1.4683,
+      "step": 4760
+    },
+    {
+      "epoch": 0.8213577158630208,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.2845329118875079e-05,
+      "loss": 1.4199,
+      "step": 4761
+    },
+    {
+      "epoch": 0.8215302337617528,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.2842719639985525e-05,
+      "loss": 1.4601,
+      "step": 4762
+    },
+    {
+      "epoch": 0.8217027516604848,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.2840109950489538e-05,
+      "loss": 1.3938,
+      "step": 4763
+    },
+    {
+      "epoch": 0.8218752695592167,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.2837500050580463e-05,
+      "loss": 1.3915,
+      "step": 4764
+    },
+    {
+      "epoch": 0.8220477874579487,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.2834889940451652e-05,
+      "loss": 1.3742,
+      "step": 4765
+    },
+    {
+      "epoch": 0.8222203053566808,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.2832279620296481e-05,
+      "loss": 1.4963,
+      "step": 4766
+    },
+    {
+      "epoch": 0.8223928232554127,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.2829669090308339e-05,
+      "loss": 1.4451,
+      "step": 4767
+    },
+    {
+      "epoch": 0.8225653411541447,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.2827058350680632e-05,
+      "loss": 1.4309,
+      "step": 4768
+    },
+    {
+      "epoch": 0.8227378590528768,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.2824447401606776e-05,
+      "loss": 1.5286,
+      "step": 4769
+    },
+    {
+      "epoch": 0.8229103769516087,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.2821836243280209e-05,
+      "loss": 1.41,
+      "step": 4770
+    },
+    {
+      "epoch": 0.8230828948503407,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.281922487589438e-05,
+      "loss": 1.5432,
+      "step": 4771
+    },
+    {
+      "epoch": 0.8232554127490728,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.2816613299642758e-05,
+      "loss": 1.4329,
+      "step": 4772
+    },
+    {
+      "epoch": 0.8234279306478047,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.2814001514718824e-05,
+      "loss": 1.4705,
+      "step": 4773
+    },
+    {
+      "epoch": 0.8236004485465367,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.2811389521316077e-05,
+      "loss": 1.4188,
+      "step": 4774
+    },
+    {
+      "epoch": 0.8237729664452687,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.2808777319628025e-05,
+      "loss": 1.4128,
+      "step": 4775
+    },
+    {
+      "epoch": 0.8239454843440007,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.28061649098482e-05,
+      "loss": 1.4184,
+      "step": 4776
+    },
+    {
+      "epoch": 0.8241180022427327,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.2803552292170145e-05,
+      "loss": 1.4397,
+      "step": 4777
+    },
+    {
+      "epoch": 0.8242905201414646,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.280093946678742e-05,
+      "loss": 1.5588,
+      "step": 4778
+    },
+    {
+      "epoch": 0.8244630380401967,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.2798326433893598e-05,
+      "loss": 1.5087,
+      "step": 4779
+    },
+    {
+      "epoch": 0.8246355559389287,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.2795713193682266e-05,
+      "loss": 1.4846,
+      "step": 4780
+    },
+    {
+      "epoch": 0.8248080738376606,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.2793099746347034e-05,
+      "loss": 1.448,
+      "step": 4781
+    },
+    {
+      "epoch": 0.8249805917363926,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.2790486092081522e-05,
+      "loss": 1.4624,
+      "step": 4782
+    },
+    {
+      "epoch": 0.8251531096351247,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.2787872231079363e-05,
+      "loss": 1.4425,
+      "step": 4783
+    },
+    {
+      "epoch": 0.8253256275338566,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.2785258163534211e-05,
+      "loss": 1.4563,
+      "step": 4784
+    },
+    {
+      "epoch": 0.8254981454325886,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.2782643889639727e-05,
+      "loss": 1.472,
+      "step": 4785
+    },
+    {
+      "epoch": 0.8256706633313207,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.2780029409589603e-05,
+      "loss": 1.4048,
+      "step": 4786
+    },
+    {
+      "epoch": 0.8258431812300526,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.2777414723577527e-05,
+      "loss": 1.5023,
+      "step": 4787
+    },
+    {
+      "epoch": 0.8260156991287846,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.2774799831797214e-05,
+      "loss": 1.4089,
+      "step": 4788
+    },
+    {
+      "epoch": 0.8261882170275167,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.277218473444239e-05,
+      "loss": 1.4833,
+      "step": 4789
+    },
+    {
+      "epoch": 0.8263607349262486,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.2769569431706804e-05,
+      "loss": 1.4398,
+      "step": 4790
+    },
+    {
+      "epoch": 0.8265332528249806,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.2766953923784207e-05,
+      "loss": 1.4302,
+      "step": 4791
+    },
+    {
+      "epoch": 0.8267057707237125,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.2764338210868372e-05,
+      "loss": 1.4379,
+      "step": 4792
+    },
+    {
+      "epoch": 0.8268782886224446,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.2761722293153094e-05,
+      "loss": 1.3314,
+      "step": 4793
+    },
+    {
+      "epoch": 0.8270508065211766,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.2759106170832172e-05,
+      "loss": 1.4575,
+      "step": 4794
+    },
+    {
+      "epoch": 0.8272233244199085,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.2756489844099424e-05,
+      "loss": 1.4598,
+      "step": 4795
+    },
+    {
+      "epoch": 0.8273958423186406,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.2753873313148683e-05,
+      "loss": 1.3463,
+      "step": 4796
+    },
+    {
+      "epoch": 0.8275683602173726,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.2751256578173803e-05,
+      "loss": 1.4494,
+      "step": 4797
+    },
+    {
+      "epoch": 0.8277408781161045,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.2748639639368645e-05,
+      "loss": 1.5213,
+      "step": 4798
+    },
+    {
+      "epoch": 0.8279133960148365,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.2746022496927086e-05,
+      "loss": 1.3503,
+      "step": 4799
+    },
+    {
+      "epoch": 0.8280859139135686,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.2743405151043025e-05,
+      "loss": 1.4791,
+      "step": 4800
+    },
+    {
+      "epoch": 0.8280859139135686,
+      "eval_loss": 1.4158532619476318,
+      "eval_runtime": 10.8421,
+      "eval_samples_per_second": 94.447,
+      "eval_steps_per_second": 23.612,
+      "step": 4800
+    },
+    {
+      "epoch": 0.8282584318123005,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.2740787601910365e-05,
+      "loss": 1.495,
+      "step": 4801
+    },
+    {
+      "epoch": 0.8284309497110325,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.2738169849723039e-05,
+      "loss": 1.428,
+      "step": 4802
+    },
+    {
+      "epoch": 0.8286034676097646,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.2735551894674978e-05,
+      "loss": 1.4787,
+      "step": 4803
+    },
+    {
+      "epoch": 0.8287759855084965,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.2732933736960143e-05,
+      "loss": 1.4055,
+      "step": 4804
+    },
+    {
+      "epoch": 0.8289485034072285,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.2730315376772498e-05,
+      "loss": 1.3993,
+      "step": 4805
+    },
+    {
+      "epoch": 0.8291210213059605,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.2727696814306034e-05,
+      "loss": 1.343,
+      "step": 4806
+    },
+    {
+      "epoch": 0.8292935392046925,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.2725078049754742e-05,
+      "loss": 1.4253,
+      "step": 4807
+    },
+    {
+      "epoch": 0.8294660571034245,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.2722459083312645e-05,
+      "loss": 1.3983,
+      "step": 4808
+    },
+    {
+      "epoch": 0.8296385750021564,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.2719839915173764e-05,
+      "loss": 1.3923,
+      "step": 4809
+    },
+    {
+      "epoch": 0.8298110929008885,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.2717220545532151e-05,
+      "loss": 1.3616,
+      "step": 4810
+    },
+    {
+      "epoch": 0.8299836107996205,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.271460097458186e-05,
+      "loss": 1.3973,
+      "step": 4811
+    },
+    {
+      "epoch": 0.8301561286983524,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.271198120251697e-05,
+      "loss": 1.3502,
+      "step": 4812
+    },
+    {
+      "epoch": 0.8303286465970845,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.2709361229531565e-05,
+      "loss": 1.3808,
+      "step": 4813
+    },
+    {
+      "epoch": 0.8305011644958165,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.2706741055819753e-05,
+      "loss": 1.4364,
+      "step": 4814
+    },
+    {
+      "epoch": 0.8306736823945484,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.270412068157565e-05,
+      "loss": 1.3211,
+      "step": 4815
+    },
+    {
+      "epoch": 0.8308462002932804,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.2701500106993389e-05,
+      "loss": 1.4494,
+      "step": 4816
+    },
+    {
+      "epoch": 0.8310187181920125,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.269887933226712e-05,
+      "loss": 1.4341,
+      "step": 4817
+    },
+    {
+      "epoch": 0.8311912360907444,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.269625835759101e-05,
+      "loss": 1.3866,
+      "step": 4818
+    },
+    {
+      "epoch": 0.8313637539894764,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.2693637183159231e-05,
+      "loss": 1.4771,
+      "step": 4819
+    },
+    {
+      "epoch": 0.8315362718882084,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.269101580916598e-05,
+      "loss": 1.4753,
+      "step": 4820
+    },
+    {
+      "epoch": 0.8317087897869404,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.2688394235805466e-05,
+      "loss": 1.4644,
+      "step": 4821
+    },
+    {
+      "epoch": 0.8318813076856724,
+      "grad_norm": 0.625,
+      "learning_rate": 1.2685772463271909e-05,
+      "loss": 1.4329,
+      "step": 4822
+    },
+    {
+      "epoch": 0.8320538255844043,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.2683150491759544e-05,
+      "loss": 1.4185,
+      "step": 4823
+    },
+    {
+      "epoch": 0.8322263434831364,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.2680528321462624e-05,
+      "loss": 1.4757,
+      "step": 4824
+    },
+    {
+      "epoch": 0.8323988613818684,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.2677905952575417e-05,
+      "loss": 1.4213,
+      "step": 4825
+    },
+    {
+      "epoch": 0.8325713792806003,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.2675283385292212e-05,
+      "loss": 1.4002,
+      "step": 4826
+    },
+    {
+      "epoch": 0.8327438971793324,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.2672660619807291e-05,
+      "loss": 1.3573,
+      "step": 4827
+    },
+    {
+      "epoch": 0.8329164150780644,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.2670037656314973e-05,
+      "loss": 1.3828,
+      "step": 4828
+    },
+    {
+      "epoch": 0.8330889329767963,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.2667414495009583e-05,
+      "loss": 1.394,
+      "step": 4829
+    },
+    {
+      "epoch": 0.8332614508755284,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.2664791136085462e-05,
+      "loss": 1.4278,
+      "step": 4830
+    },
+    {
+      "epoch": 0.8334339687742603,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.2662167579736961e-05,
+      "loss": 1.4386,
+      "step": 4831
+    },
+    {
+      "epoch": 0.8336064866729923,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.2659543826158451e-05,
+      "loss": 1.3753,
+      "step": 4832
+    },
+    {
+      "epoch": 0.8337790045717243,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.2656919875544316e-05,
+      "loss": 1.3979,
+      "step": 4833
+    },
+    {
+      "epoch": 0.8339515224704563,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.2654295728088959e-05,
+      "loss": 1.5161,
+      "step": 4834
+    },
+    {
+      "epoch": 0.8341240403691883,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.2651671383986788e-05,
+      "loss": 1.4717,
+      "step": 4835
+    },
+    {
+      "epoch": 0.8342965582679203,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.2649046843432232e-05,
+      "loss": 1.4729,
+      "step": 4836
+    },
+    {
+      "epoch": 0.8344690761666523,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.2646422106619733e-05,
+      "loss": 1.4337,
+      "step": 4837
+    },
+    {
+      "epoch": 0.8346415940653843,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.2643797173743753e-05,
+      "loss": 1.3985,
+      "step": 4838
+    },
+    {
+      "epoch": 0.8348141119641163,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.2641172044998754e-05,
+      "loss": 1.3862,
+      "step": 4839
+    },
+    {
+      "epoch": 0.8349866298628482,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.263854672057923e-05,
+      "loss": 1.2768,
+      "step": 4840
+    },
+    {
+      "epoch": 0.8351591477615803,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.2635921200679677e-05,
+      "loss": 1.4262,
+      "step": 4841
+    },
+    {
+      "epoch": 0.8353316656603123,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.2633295485494614e-05,
+      "loss": 1.3765,
+      "step": 4842
+    },
+    {
+      "epoch": 0.8355041835590442,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.2630669575218568e-05,
+      "loss": 1.3913,
+      "step": 4843
+    },
+    {
+      "epoch": 0.8356767014577763,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.2628043470046078e-05,
+      "loss": 1.4059,
+      "step": 4844
+    },
+    {
+      "epoch": 0.8358492193565082,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.2625417170171712e-05,
+      "loss": 1.4799,
+      "step": 4845
+    },
+    {
+      "epoch": 0.8360217372552402,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.2622790675790033e-05,
+      "loss": 1.3965,
+      "step": 4846
+    },
+    {
+      "epoch": 0.8361942551539723,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.2620163987095637e-05,
+      "loss": 1.4537,
+      "step": 4847
+    },
+    {
+      "epoch": 0.8363667730527042,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.2617537104283119e-05,
+      "loss": 1.3701,
+      "step": 4848
+    },
+    {
+      "epoch": 0.8365392909514362,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.2614910027547096e-05,
+      "loss": 1.3818,
+      "step": 4849
+    },
+    {
+      "epoch": 0.8367118088501682,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.2612282757082204e-05,
+      "loss": 1.5375,
+      "step": 4850
+    },
+    {
+      "epoch": 0.8368843267489002,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.2609655293083079e-05,
+      "loss": 1.4179,
+      "step": 4851
+    },
+    {
+      "epoch": 0.8370568446476322,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.2607027635744384e-05,
+      "loss": 1.3935,
+      "step": 4852
+    },
+    {
+      "epoch": 0.8372293625463642,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.2604399785260794e-05,
+      "loss": 1.4925,
+      "step": 4853
+    },
+    {
+      "epoch": 0.8374018804450962,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.2601771741826996e-05,
+      "loss": 1.4307,
+      "step": 4854
+    },
+    {
+      "epoch": 0.8375743983438282,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.259914350563769e-05,
+      "loss": 1.3686,
+      "step": 4855
+    },
+    {
+      "epoch": 0.8377469162425601,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.259651507688759e-05,
+      "loss": 1.3841,
+      "step": 4856
+    },
+    {
+      "epoch": 0.8379194341412921,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.259388645577143e-05,
+      "loss": 1.3666,
+      "step": 4857
+    },
+    {
+      "epoch": 0.8380919520400242,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.2591257642483959e-05,
+      "loss": 1.532,
+      "step": 4858
+    },
+    {
+      "epoch": 0.8382644699387561,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.2588628637219927e-05,
+      "loss": 1.4264,
+      "step": 4859
+    },
+    {
+      "epoch": 0.8384369878374881,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.258599944017411e-05,
+      "loss": 1.4054,
+      "step": 4860
+    },
+    {
+      "epoch": 0.8386095057362202,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.2583370051541298e-05,
+      "loss": 1.3527,
+      "step": 4861
+    },
+    {
+      "epoch": 0.8387820236349521,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.258074047151629e-05,
+      "loss": 1.4412,
+      "step": 4862
+    },
+    {
+      "epoch": 0.8389545415336841,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.2578110700293906e-05,
+      "loss": 1.6089,
+      "step": 4863
+    },
+    {
+      "epoch": 0.8391270594324162,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.2575480738068971e-05,
+      "loss": 1.4927,
+      "step": 4864
+    },
+    {
+      "epoch": 0.8392995773311481,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.2572850585036325e-05,
+      "loss": 1.3671,
+      "step": 4865
+    },
+    {
+      "epoch": 0.8394720952298801,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.2570220241390838e-05,
+      "loss": 1.3574,
+      "step": 4866
+    },
+    {
+      "epoch": 0.839644613128612,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.2567589707327375e-05,
+      "loss": 1.4596,
+      "step": 4867
+    },
+    {
+      "epoch": 0.8398171310273441,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.256495898304082e-05,
+      "loss": 1.5785,
+      "step": 4868
+    },
+    {
+      "epoch": 0.8399896489260761,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.2562328068726081e-05,
+      "loss": 1.3818,
+      "step": 4869
+    },
+    {
+      "epoch": 0.840162166824808,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.2559696964578068e-05,
+      "loss": 1.4781,
+      "step": 4870
+    },
+    {
+      "epoch": 0.8403346847235401,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.2557065670791708e-05,
+      "loss": 1.3785,
+      "step": 4871
+    },
+    {
+      "epoch": 0.8405072026222721,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.2554434187561949e-05,
+      "loss": 1.4297,
+      "step": 4872
+    },
+    {
+      "epoch": 0.840679720521004,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.2551802515083742e-05,
+      "loss": 1.5078,
+      "step": 4873
+    },
+    {
+      "epoch": 0.840852238419736,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.2549170653552062e-05,
+      "loss": 1.6393,
+      "step": 4874
+    },
+    {
+      "epoch": 0.8410247563184681,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.254653860316189e-05,
+      "loss": 1.4881,
+      "step": 4875
+    },
+    {
+      "epoch": 0.8411972742172,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.2543906364108227e-05,
+      "loss": 1.4819,
+      "step": 4876
+    },
+    {
+      "epoch": 0.841369792115932,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.2541273936586088e-05,
+      "loss": 1.288,
+      "step": 4877
+    },
+    {
+      "epoch": 0.8415423100146641,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.2538641320790494e-05,
+      "loss": 1.5058,
+      "step": 4878
+    },
+    {
+      "epoch": 0.841714827913396,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.2536008516916491e-05,
+      "loss": 1.5025,
+      "step": 4879
+    },
+    {
+      "epoch": 0.841887345812128,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.253337552515913e-05,
+      "loss": 1.5237,
+      "step": 4880
+    },
+    {
+      "epoch": 0.8420598637108601,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.253074234571348e-05,
+      "loss": 1.3913,
+      "step": 4881
+    },
+    {
+      "epoch": 0.842232381609592,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.2528108978774627e-05,
+      "loss": 1.4599,
+      "step": 4882
+    },
+    {
+      "epoch": 0.842404899508324,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.2525475424537664e-05,
+      "loss": 1.468,
+      "step": 4883
+    },
+    {
+      "epoch": 0.8425774174070559,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.25228416831977e-05,
+      "loss": 1.5137,
+      "step": 4884
+    },
+    {
+      "epoch": 0.842749935305788,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.2520207754949861e-05,
+      "loss": 1.3381,
+      "step": 4885
+    },
+    {
+      "epoch": 0.84292245320452,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.2517573639989284e-05,
+      "loss": 1.449,
+      "step": 4886
+    },
+    {
+      "epoch": 0.8430949711032519,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.2514939338511123e-05,
+      "loss": 1.486,
+      "step": 4887
+    },
+    {
+      "epoch": 0.843267489001984,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.2512304850710542e-05,
+      "loss": 1.5363,
+      "step": 4888
+    },
+    {
+      "epoch": 0.843440006900716,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.2509670176782718e-05,
+      "loss": 1.3917,
+      "step": 4889
+    },
+    {
+      "epoch": 0.8436125247994479,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.2507035316922847e-05,
+      "loss": 1.4611,
+      "step": 4890
+    },
+    {
+      "epoch": 0.8437850426981799,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.2504400271326136e-05,
+      "loss": 1.3827,
+      "step": 4891
+    },
+    {
+      "epoch": 0.843957560596912,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.2501765040187804e-05,
+      "loss": 1.4403,
+      "step": 4892
+    },
+    {
+      "epoch": 0.8441300784956439,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.2499129623703086e-05,
+      "loss": 1.4834,
+      "step": 4893
+    },
+    {
+      "epoch": 0.8443025963943759,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.249649402206723e-05,
+      "loss": 1.4435,
+      "step": 4894
+    },
+    {
+      "epoch": 0.844475114293108,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.2493858235475497e-05,
+      "loss": 1.4356,
+      "step": 4895
+    },
+    {
+      "epoch": 0.8446476321918399,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.2491222264123166e-05,
+      "loss": 1.4481,
+      "step": 4896
+    },
+    {
+      "epoch": 0.8448201500905719,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.2488586108205521e-05,
+      "loss": 1.497,
+      "step": 4897
+    },
+    {
+      "epoch": 0.8449926679893038,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.2485949767917869e-05,
+      "loss": 1.3873,
+      "step": 4898
+    },
+    {
+      "epoch": 0.8451651858880359,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.2483313243455526e-05,
+      "loss": 1.4457,
+      "step": 4899
+    },
+    {
+      "epoch": 0.8453377037867679,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.2480676535013821e-05,
+      "loss": 1.3509,
+      "step": 4900
+    },
+    {
+      "epoch": 0.8453377037867679,
+      "eval_loss": 1.4150875806808472,
+      "eval_runtime": 10.8763,
+      "eval_samples_per_second": 94.15,
+      "eval_steps_per_second": 23.537,
+      "step": 4900
+    },
+    {
+      "epoch": 0.8455102216854998,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.2478039642788097e-05,
+      "loss": 1.3508,
+      "step": 4901
+    },
+    {
+      "epoch": 0.8456827395842319,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.2475402566973715e-05,
+      "loss": 1.431,
+      "step": 4902
+    },
+    {
+      "epoch": 0.8458552574829639,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.2472765307766045e-05,
+      "loss": 1.5606,
+      "step": 4903
+    },
+    {
+      "epoch": 0.8460277753816958,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.2470127865360467e-05,
+      "loss": 1.4353,
+      "step": 4904
+    },
+    {
+      "epoch": 0.8462002932804279,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.2467490239952382e-05,
+      "loss": 1.3749,
+      "step": 4905
+    },
+    {
+      "epoch": 0.8463728111791599,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.2464852431737205e-05,
+      "loss": 1.4545,
+      "step": 4906
+    },
+    {
+      "epoch": 0.8465453290778918,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.2462214440910359e-05,
+      "loss": 1.4226,
+      "step": 4907
+    },
+    {
+      "epoch": 0.8467178469766238,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.2459576267667281e-05,
+      "loss": 1.3848,
+      "step": 4908
+    },
+    {
+      "epoch": 0.8468903648753558,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.2456937912203426e-05,
+      "loss": 1.4049,
+      "step": 4909
+    },
+    {
+      "epoch": 0.8470628827740878,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.2454299374714258e-05,
+      "loss": 1.4062,
+      "step": 4910
+    },
+    {
+      "epoch": 0.8472354006728198,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.2451660655395258e-05,
+      "loss": 1.4654,
+      "step": 4911
+    },
+    {
+      "epoch": 0.8474079185715518,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.2449021754441919e-05,
+      "loss": 1.4948,
+      "step": 4912
+    },
+    {
+      "epoch": 0.8475804364702838,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.2446382672049741e-05,
+      "loss": 1.4434,
+      "step": 4913
+    },
+    {
+      "epoch": 0.8477529543690158,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.2443743408414256e-05,
+      "loss": 1.4044,
+      "step": 4914
+    },
+    {
+      "epoch": 0.8479254722677477,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.2441103963730985e-05,
+      "loss": 1.4474,
+      "step": 4915
+    },
+    {
+      "epoch": 0.8480979901664798,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.2438464338195481e-05,
+      "loss": 1.4127,
+      "step": 4916
+    },
+    {
+      "epoch": 0.8482705080652118,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.2435824532003304e-05,
+      "loss": 1.4637,
+      "step": 4917
+    },
+    {
+      "epoch": 0.8484430259639437,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.2433184545350026e-05,
+      "loss": 1.4083,
+      "step": 4918
+    },
+    {
+      "epoch": 0.8486155438626758,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.2430544378431233e-05,
+      "loss": 1.4099,
+      "step": 4919
+    },
+    {
+      "epoch": 0.8487880617614078,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.2427904031442526e-05,
+      "loss": 1.3869,
+      "step": 4920
+    },
+    {
+      "epoch": 0.8489605796601397,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.2425263504579517e-05,
+      "loss": 1.439,
+      "step": 4921
+    },
+    {
+      "epoch": 0.8491330975588718,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.2422622798037833e-05,
+      "loss": 1.3557,
+      "step": 4922
+    },
+    {
+      "epoch": 0.8493056154576037,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.2419981912013116e-05,
+      "loss": 1.5707,
+      "step": 4923
+    },
+    {
+      "epoch": 0.8494781333563357,
+      "grad_norm": 0.625,
+      "learning_rate": 1.241734084670102e-05,
+      "loss": 1.4085,
+      "step": 4924
+    },
+    {
+      "epoch": 0.8496506512550677,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.2414699602297206e-05,
+      "loss": 1.4046,
+      "step": 4925
+    },
+    {
+      "epoch": 0.8498231691537997,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.2412058178997362e-05,
+      "loss": 1.34,
+      "step": 4926
+    },
+    {
+      "epoch": 0.8499956870525317,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.2409416576997176e-05,
+      "loss": 1.3569,
+      "step": 4927
+    },
+    {
+      "epoch": 0.8501682049512637,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.2406774796492353e-05,
+      "loss": 1.4416,
+      "step": 4928
+    },
+    {
+      "epoch": 0.8503407228499957,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.2404132837678613e-05,
+      "loss": 1.4531,
+      "step": 4929
+    },
+    {
+      "epoch": 0.8505132407487277,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.2401490700751695e-05,
+      "loss": 1.4516,
+      "step": 4930
+    },
+    {
+      "epoch": 0.8506857586474597,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.239884838590734e-05,
+      "loss": 1.45,
+      "step": 4931
+    },
+    {
+      "epoch": 0.8508582765461916,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.2396205893341306e-05,
+      "loss": 1.5007,
+      "step": 4932
+    },
+    {
+      "epoch": 0.8510307944449237,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.2393563223249367e-05,
+      "loss": 1.4125,
+      "step": 4933
+    },
+    {
+      "epoch": 0.8512033123436556,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.239092037582731e-05,
+      "loss": 1.4413,
+      "step": 4934
+    },
+    {
+      "epoch": 0.8513758302423876,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.2388277351270936e-05,
+      "loss": 1.5108,
+      "step": 4935
+    },
+    {
+      "epoch": 0.8515483481411197,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.2385634149776049e-05,
+      "loss": 1.5181,
+      "step": 4936
+    },
+    {
+      "epoch": 0.8517208660398516,
+      "grad_norm": 0.625,
+      "learning_rate": 1.2382990771538479e-05,
+      "loss": 1.4072,
+      "step": 4937
+    },
+    {
+      "epoch": 0.8518933839385836,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.238034721675406e-05,
+      "loss": 1.4797,
+      "step": 4938
+    },
+    {
+      "epoch": 0.8520659018373157,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.237770348561865e-05,
+      "loss": 1.4337,
+      "step": 4939
+    },
+    {
+      "epoch": 0.8522384197360476,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.237505957832811e-05,
+      "loss": 1.4196,
+      "step": 4940
+    },
+    {
+      "epoch": 0.8524109376347796,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.2372415495078314e-05,
+      "loss": 1.515,
+      "step": 4941
+    },
+    {
+      "epoch": 0.8525834555335116,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.2369771236065154e-05,
+      "loss": 1.361,
+      "step": 4942
+    },
+    {
+      "epoch": 0.8527559734322436,
+      "grad_norm": 0.875,
+      "learning_rate": 1.2367126801484537e-05,
+      "loss": 1.4598,
+      "step": 4943
+    },
+    {
+      "epoch": 0.8529284913309756,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.2364482191532371e-05,
+      "loss": 1.4902,
+      "step": 4944
+    },
+    {
+      "epoch": 0.8531010092297076,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.2361837406404592e-05,
+      "loss": 1.4504,
+      "step": 4945
+    },
+    {
+      "epoch": 0.8532735271284396,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.2359192446297143e-05,
+      "loss": 1.4503,
+      "step": 4946
+    },
+    {
+      "epoch": 0.8534460450271716,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.2356547311405977e-05,
+      "loss": 1.4309,
+      "step": 4947
+    },
+    {
+      "epoch": 0.8536185629259035,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.2353902001927058e-05,
+      "loss": 1.3899,
+      "step": 4948
+    },
+    {
+      "epoch": 0.8537910808246355,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.2351256518056373e-05,
+      "loss": 1.371,
+      "step": 4949
+    },
+    {
+      "epoch": 0.8539635987233676,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.2348610859989913e-05,
+      "loss": 1.4957,
+      "step": 4950
+    },
+    {
+      "epoch": 0.8541361166220995,
+      "grad_norm": 0.625,
+      "learning_rate": 1.234596502792369e-05,
+      "loss": 1.351,
+      "step": 4951
+    },
+    {
+      "epoch": 0.8543086345208315,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.2343319022053715e-05,
+      "loss": 1.4467,
+      "step": 4952
+    },
+    {
+      "epoch": 0.8544811524195636,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.2340672842576025e-05,
+      "loss": 1.4147,
+      "step": 4953
+    },
+    {
+      "epoch": 0.8546536703182955,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.2338026489686668e-05,
+      "loss": 1.4445,
+      "step": 4954
+    },
+    {
+      "epoch": 0.8548261882170275,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.2335379963581699e-05,
+      "loss": 1.4351,
+      "step": 4955
+    },
+    {
+      "epoch": 0.8549987061157595,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.2332733264457188e-05,
+      "loss": 1.5089,
+      "step": 4956
+    },
+    {
+      "epoch": 0.8551712240144915,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.2330086392509224e-05,
+      "loss": 1.5096,
+      "step": 4957
+    },
+    {
+      "epoch": 0.8553437419132235,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.2327439347933901e-05,
+      "loss": 1.3495,
+      "step": 4958
+    },
+    {
+      "epoch": 0.8555162598119554,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.2324792130927328e-05,
+      "loss": 1.3313,
+      "step": 4959
+    },
+    {
+      "epoch": 0.8556887777106875,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.2322144741685627e-05,
+      "loss": 1.4532,
+      "step": 4960
+    },
+    {
+      "epoch": 0.8558612956094195,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.231949718040493e-05,
+      "loss": 1.5288,
+      "step": 4961
+    },
+    {
+      "epoch": 0.8560338135081514,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.2316849447281395e-05,
+      "loss": 1.4836,
+      "step": 4962
+    },
+    {
+      "epoch": 0.8562063314068835,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.2314201542511176e-05,
+      "loss": 1.4735,
+      "step": 4963
+    },
+    {
+      "epoch": 0.8563788493056155,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.2311553466290447e-05,
+      "loss": 1.2235,
+      "step": 4964
+    },
+    {
+      "epoch": 0.8565513672043474,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.2308905218815392e-05,
+      "loss": 1.4499,
+      "step": 4965
+    },
+    {
+      "epoch": 0.8567238851030794,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.2306256800282213e-05,
+      "loss": 1.4595,
+      "step": 4966
+    },
+    {
+      "epoch": 0.8568964030018115,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.2303608210887121e-05,
+      "loss": 1.4724,
+      "step": 4967
+    },
+    {
+      "epoch": 0.8570689209005434,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.230095945082634e-05,
+      "loss": 1.5173,
+      "step": 4968
+    },
+    {
+      "epoch": 0.8572414387992754,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.2298310520296101e-05,
+      "loss": 1.5255,
+      "step": 4969
+    },
+    {
+      "epoch": 0.8574139566980075,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.2295661419492662e-05,
+      "loss": 1.4696,
+      "step": 4970
+    },
+    {
+      "epoch": 0.8575864745967394,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.2293012148612281e-05,
+      "loss": 1.484,
+      "step": 4971
+    },
+    {
+      "epoch": 0.8577589924954714,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.229036270785123e-05,
+      "loss": 1.4593,
+      "step": 4972
+    },
+    {
+      "epoch": 0.8579315103942033,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.2287713097405802e-05,
+      "loss": 1.4657,
+      "step": 4973
+    },
+    {
+      "epoch": 0.8581040282929354,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.2285063317472293e-05,
+      "loss": 1.3391,
+      "step": 4974
+    },
+    {
+      "epoch": 0.8582765461916674,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.2282413368247015e-05,
+      "loss": 1.3832,
+      "step": 4975
+    },
+    {
+      "epoch": 0.8584490640903993,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.2279763249926293e-05,
+      "loss": 1.5026,
+      "step": 4976
+    },
+    {
+      "epoch": 0.8586215819891314,
+      "grad_norm": 0.875,
+      "learning_rate": 1.2277112962706463e-05,
+      "loss": 1.4334,
+      "step": 4977
+    },
+    {
+      "epoch": 0.8587940998878634,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.2274462506783877e-05,
+      "loss": 1.515,
+      "step": 4978
+    },
+    {
+      "epoch": 0.8589666177865953,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.22718118823549e-05,
+      "loss": 1.4328,
+      "step": 4979
+    },
+    {
+      "epoch": 0.8591391356853274,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.2269161089615902e-05,
+      "loss": 1.442,
+      "step": 4980
+    },
+    {
+      "epoch": 0.8593116535840594,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.2266510128763271e-05,
+      "loss": 1.4677,
+      "step": 4981
+    },
+    {
+      "epoch": 0.8594841714827913,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.226385899999341e-05,
+      "loss": 1.3962,
+      "step": 4982
+    },
+    {
+      "epoch": 0.8596566893815233,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.2261207703502731e-05,
+      "loss": 1.4035,
+      "step": 4983
+    },
+    {
+      "epoch": 0.8598292072802554,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.2258556239487654e-05,
+      "loss": 1.4359,
+      "step": 4984
+    },
+    {
+      "epoch": 0.8600017251789873,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.2255904608144618e-05,
+      "loss": 1.3858,
+      "step": 4985
+    },
+    {
+      "epoch": 0.8601742430777193,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.2253252809670074e-05,
+      "loss": 1.4108,
+      "step": 4986
+    },
+    {
+      "epoch": 0.8603467609764514,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.2250600844260482e-05,
+      "loss": 1.4605,
+      "step": 4987
+    },
+    {
+      "epoch": 0.8605192788751833,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.2247948712112318e-05,
+      "loss": 1.419,
+      "step": 4988
+    },
+    {
+      "epoch": 0.8606917967739153,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.2245296413422067e-05,
+      "loss": 1.3961,
+      "step": 4989
+    },
+    {
+      "epoch": 0.8608643146726472,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.2242643948386231e-05,
+      "loss": 1.494,
+      "step": 4990
+    },
+    {
+      "epoch": 0.8610368325713793,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.2239991317201316e-05,
+      "loss": 1.4661,
+      "step": 4991
+    },
+    {
+      "epoch": 0.8612093504701113,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.2237338520063848e-05,
+      "loss": 1.4005,
+      "step": 4992
+    },
+    {
+      "epoch": 0.8613818683688432,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.2234685557170365e-05,
+      "loss": 1.491,
+      "step": 4993
+    },
+    {
+      "epoch": 0.8615543862675753,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.2232032428717408e-05,
+      "loss": 1.3907,
+      "step": 4994
+    },
+    {
+      "epoch": 0.8617269041663073,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.2229379134901546e-05,
+      "loss": 1.3841,
+      "step": 4995
+    },
+    {
+      "epoch": 0.8618994220650392,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.2226725675919349e-05,
+      "loss": 1.52,
+      "step": 4996
+    },
+    {
+      "epoch": 0.8620719399637713,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.2224072051967398e-05,
+      "loss": 1.3878,
+      "step": 4997
+    },
+    {
+      "epoch": 0.8622444578625033,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.2221418263242292e-05,
+      "loss": 1.4099,
+      "step": 4998
+    },
+    {
+      "epoch": 0.8624169757612352,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.2218764309940647e-05,
+      "loss": 1.4386,
+      "step": 4999
+    },
+    {
+      "epoch": 0.8625894936599672,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.2216110192259075e-05,
+      "loss": 1.4912,
+      "step": 5000
+    },
+    {
+      "epoch": 0.8625894936599672,
+      "eval_loss": 1.414391040802002,
+      "eval_runtime": 10.8891,
+      "eval_samples_per_second": 94.039,
+      "eval_steps_per_second": 23.51,
+      "step": 5000
+    },
+    {
+      "epoch": 0.8627620115586992,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.221345591039421e-05,
+      "loss": 1.4483,
+      "step": 5001
+    },
+    {
+      "epoch": 0.8629345294574312,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.2210801464542704e-05,
+      "loss": 1.4031,
+      "step": 5002
+    },
+    {
+      "epoch": 0.8631070473561632,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.2208146854901213e-05,
+      "loss": 1.5115,
+      "step": 5003
+    },
+    {
+      "epoch": 0.8632795652548952,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.2205492081666404e-05,
+      "loss": 1.4754,
+      "step": 5004
+    },
+    {
+      "epoch": 0.8634520831536272,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.220283714503496e-05,
+      "loss": 1.4836,
+      "step": 5005
+    },
+    {
+      "epoch": 0.8636246010523592,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.2200182045203576e-05,
+      "loss": 1.4632,
+      "step": 5006
+    },
+    {
+      "epoch": 0.8637971189510911,
+      "grad_norm": 0.625,
+      "learning_rate": 1.2197526782368962e-05,
+      "loss": 1.3843,
+      "step": 5007
+    },
+    {
+      "epoch": 0.8639696368498232,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.219487135672783e-05,
+      "loss": 1.4724,
+      "step": 5008
+    },
+    {
+      "epoch": 0.8641421547485552,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.2192215768476915e-05,
+      "loss": 1.4989,
+      "step": 5009
+    },
+    {
+      "epoch": 0.8643146726472871,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.2189560017812955e-05,
+      "loss": 1.4918,
+      "step": 5010
+    },
+    {
+      "epoch": 0.8644871905460192,
+      "grad_norm": 1.0,
+      "learning_rate": 1.2186904104932716e-05,
+      "loss": 1.3991,
+      "step": 5011
+    },
+    {
+      "epoch": 0.8646597084447512,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.218424803003295e-05,
+      "loss": 1.4472,
+      "step": 5012
+    },
+    {
+      "epoch": 0.8648322263434831,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.2181591793310444e-05,
+      "loss": 1.2846,
+      "step": 5013
+    },
+    {
+      "epoch": 0.8650047442422152,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.2178935394961985e-05,
+      "loss": 1.4859,
+      "step": 5014
+    },
+    {
+      "epoch": 0.8651772621409471,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.2176278835184381e-05,
+      "loss": 1.4976,
+      "step": 5015
+    },
+    {
+      "epoch": 0.8653497800396791,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.2173622114174439e-05,
+      "loss": 1.3716,
+      "step": 5016
+    },
+    {
+      "epoch": 0.8655222979384111,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.2170965232128991e-05,
+      "loss": 1.4576,
+      "step": 5017
+    },
+    {
+      "epoch": 0.8656948158371431,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.2168308189244872e-05,
+      "loss": 1.4493,
+      "step": 5018
+    },
+    {
+      "epoch": 0.8658673337358751,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.216565098571894e-05,
+      "loss": 1.5467,
+      "step": 5019
+    },
+    {
+      "epoch": 0.8660398516346071,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.2162993621748048e-05,
+      "loss": 1.4384,
+      "step": 5020
+    },
+    {
+      "epoch": 0.8662123695333391,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.216033609752907e-05,
+      "loss": 1.4942,
+      "step": 5021
+    },
+    {
+      "epoch": 0.8663848874320711,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.2157678413258898e-05,
+      "loss": 1.5091,
+      "step": 5022
+    },
+    {
+      "epoch": 0.866557405330803,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.2155020569134428e-05,
+      "loss": 1.4179,
+      "step": 5023
+    },
+    {
+      "epoch": 0.866729923229535,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.2152362565352571e-05,
+      "loss": 1.4006,
+      "step": 5024
+    },
+    {
+      "epoch": 0.8669024411282671,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.2149704402110243e-05,
+      "loss": 1.431,
+      "step": 5025
+    },
+    {
+      "epoch": 0.867074959026999,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.2147046079604378e-05,
+      "loss": 1.517,
+      "step": 5026
+    },
+    {
+      "epoch": 0.867247476925731,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.2144387598031933e-05,
+      "loss": 1.5725,
+      "step": 5027
+    },
+    {
+      "epoch": 0.8674199948244631,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.2141728957589848e-05,
+      "loss": 1.4621,
+      "step": 5028
+    },
+    {
+      "epoch": 0.867592512723195,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.2139070158475104e-05,
+      "loss": 1.4756,
+      "step": 5029
+    },
+    {
+      "epoch": 0.867765030621927,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.2136411200884676e-05,
+      "loss": 1.4426,
+      "step": 5030
+    },
+    {
+      "epoch": 0.867937548520659,
+      "grad_norm": 0.625,
+      "learning_rate": 1.2133752085015558e-05,
+      "loss": 1.4562,
+      "step": 5031
+    },
+    {
+      "epoch": 0.868110066419391,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.2131092811064753e-05,
+      "loss": 1.4278,
+      "step": 5032
+    },
+    {
+      "epoch": 0.868282584318123,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.2128433379229276e-05,
+      "loss": 1.5037,
+      "step": 5033
+    },
+    {
+      "epoch": 0.868455102216855,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.2125773789706155e-05,
+      "loss": 1.4705,
+      "step": 5034
+    },
+    {
+      "epoch": 0.868627620115587,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.2123114042692432e-05,
+      "loss": 1.4946,
+      "step": 5035
+    },
+    {
+      "epoch": 0.868800138014319,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.212045413838515e-05,
+      "loss": 1.4562,
+      "step": 5036
+    },
+    {
+      "epoch": 0.868972655913051,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.2117794076981381e-05,
+      "loss": 1.4539,
+      "step": 5037
+    },
+    {
+      "epoch": 0.869145173811783,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.2115133858678192e-05,
+      "loss": 1.4397,
+      "step": 5038
+    },
+    {
+      "epoch": 0.869317691710515,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.2112473483672671e-05,
+      "loss": 1.3853,
+      "step": 5039
+    },
+    {
+      "epoch": 0.869490209609247,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.2109812952161916e-05,
+      "loss": 1.5076,
+      "step": 5040
+    },
+    {
+      "epoch": 0.8696627275079789,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.2107152264343033e-05,
+      "loss": 1.4681,
+      "step": 5041
+    },
+    {
+      "epoch": 0.869835245406711,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.2104491420413143e-05,
+      "loss": 1.4733,
+      "step": 5042
+    },
+    {
+      "epoch": 0.8700077633054429,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.2101830420569382e-05,
+      "loss": 1.3826,
+      "step": 5043
+    },
+    {
+      "epoch": 0.8701802812041749,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.209916926500889e-05,
+      "loss": 1.454,
+      "step": 5044
+    },
+    {
+      "epoch": 0.870352799102907,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.2096507953928823e-05,
+      "loss": 1.4646,
+      "step": 5045
+    },
+    {
+      "epoch": 0.8705253170016389,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.2093846487526344e-05,
+      "loss": 1.5317,
+      "step": 5046
+    },
+    {
+      "epoch": 0.8706978349003709,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.209118486599864e-05,
+      "loss": 1.4352,
+      "step": 5047
+    },
+    {
+      "epoch": 0.8708703527991029,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.208852308954289e-05,
+      "loss": 1.4748,
+      "step": 5048
+    },
+    {
+      "epoch": 0.8710428706978349,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.20858611583563e-05,
+      "loss": 1.5263,
+      "step": 5049
+    },
+    {
+      "epoch": 0.8712153885965669,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.2083199072636084e-05,
+      "loss": 1.4742,
+      "step": 5050
+    },
+    {
+      "epoch": 0.8713879064952988,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.2080536832579466e-05,
+      "loss": 1.4146,
+      "step": 5051
+    },
+    {
+      "epoch": 0.8715604243940309,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.2077874438383676e-05,
+      "loss": 1.3773,
+      "step": 5052
+    },
+    {
+      "epoch": 0.8717329422927629,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.2075211890245965e-05,
+      "loss": 1.4945,
+      "step": 5053
+    },
+    {
+      "epoch": 0.8719054601914948,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.2072549188363594e-05,
+      "loss": 1.5134,
+      "step": 5054
+    },
+    {
+      "epoch": 0.8720779780902269,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.2069886332933824e-05,
+      "loss": 1.4979,
+      "step": 5055
+    },
+    {
+      "epoch": 0.8722504959889589,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.2067223324153947e-05,
+      "loss": 1.4189,
+      "step": 5056
+    },
+    {
+      "epoch": 0.8724230138876908,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.2064560162221246e-05,
+      "loss": 1.4177,
+      "step": 5057
+    },
+    {
+      "epoch": 0.8725955317864228,
+      "grad_norm": 0.625,
+      "learning_rate": 1.2061896847333025e-05,
+      "loss": 1.4413,
+      "step": 5058
+    },
+    {
+      "epoch": 0.8727680496851549,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.205923337968661e-05,
+      "loss": 1.4782,
+      "step": 5059
+    },
+    {
+      "epoch": 0.8729405675838868,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.2056569759479315e-05,
+      "loss": 1.4539,
+      "step": 5060
+    },
+    {
+      "epoch": 0.8731130854826188,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.205390598690848e-05,
+      "loss": 1.4785,
+      "step": 5061
+    },
+    {
+      "epoch": 0.8732856033813509,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.205124206217146e-05,
+      "loss": 1.4304,
+      "step": 5062
+    },
+    {
+      "epoch": 0.8734581212800828,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.2048577985465613e-05,
+      "loss": 1.4521,
+      "step": 5063
+    },
+    {
+      "epoch": 0.8736306391788148,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.2045913756988305e-05,
+      "loss": 1.4276,
+      "step": 5064
+    },
+    {
+      "epoch": 0.8738031570775467,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.2043249376936923e-05,
+      "loss": 1.4619,
+      "step": 5065
+    },
+    {
+      "epoch": 0.8739756749762788,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.204058484550886e-05,
+      "loss": 1.475,
+      "step": 5066
+    },
+    {
+      "epoch": 0.8741481928750108,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.2037920162901522e-05,
+      "loss": 1.4353,
+      "step": 5067
+    },
+    {
+      "epoch": 0.8743207107737427,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.2035255329312325e-05,
+      "loss": 1.5358,
+      "step": 5068
+    },
+    {
+      "epoch": 0.8744932286724748,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.2032590344938697e-05,
+      "loss": 1.4719,
+      "step": 5069
+    },
+    {
+      "epoch": 0.8746657465712068,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.2029925209978075e-05,
+      "loss": 1.4687,
+      "step": 5070
+    },
+    {
+      "epoch": 0.8748382644699387,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.202725992462791e-05,
+      "loss": 1.4727,
+      "step": 5071
+    },
+    {
+      "epoch": 0.8750107823686708,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.2024594489085665e-05,
+      "loss": 1.4396,
+      "step": 5072
+    },
+    {
+      "epoch": 0.8751833002674028,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.2021928903548807e-05,
+      "loss": 1.3646,
+      "step": 5073
+    },
+    {
+      "epoch": 0.8753558181661347,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.2019263168214822e-05,
+      "loss": 1.5268,
+      "step": 5074
+    },
+    {
+      "epoch": 0.8755283360648667,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.2016597283281209e-05,
+      "loss": 1.4303,
+      "step": 5075
+    },
+    {
+      "epoch": 0.8757008539635988,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.2013931248945463e-05,
+      "loss": 1.5223,
+      "step": 5076
+    },
+    {
+      "epoch": 0.8758733718623307,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.2011265065405108e-05,
+      "loss": 1.5449,
+      "step": 5077
+    },
+    {
+      "epoch": 0.8760458897610627,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.2008598732857673e-05,
+      "loss": 1.4209,
+      "step": 5078
+    },
+    {
+      "epoch": 0.8762184076597948,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.2005932251500693e-05,
+      "loss": 1.4707,
+      "step": 5079
+    },
+    {
+      "epoch": 0.8763909255585267,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.2003265621531716e-05,
+      "loss": 1.4294,
+      "step": 5080
+    },
+    {
+      "epoch": 0.8765634434572587,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.2000598843148306e-05,
+      "loss": 1.404,
+      "step": 5081
+    },
+    {
+      "epoch": 0.8767359613559906,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.1997931916548034e-05,
+      "loss": 1.4738,
+      "step": 5082
+    },
+    {
+      "epoch": 0.8769084792547227,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.199526484192848e-05,
+      "loss": 1.4521,
+      "step": 5083
+    },
+    {
+      "epoch": 0.8770809971534547,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.1992597619487242e-05,
+      "loss": 1.4946,
+      "step": 5084
+    },
+    {
+      "epoch": 0.8772535150521866,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.198993024942192e-05,
+      "loss": 1.3803,
+      "step": 5085
+    },
+    {
+      "epoch": 0.8774260329509187,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.1987262731930132e-05,
+      "loss": 1.4693,
+      "step": 5086
+    },
+    {
+      "epoch": 0.8775985508496507,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.1984595067209503e-05,
+      "loss": 1.3118,
+      "step": 5087
+    },
+    {
+      "epoch": 0.8777710687483826,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.1981927255457675e-05,
+      "loss": 1.4328,
+      "step": 5088
+    },
+    {
+      "epoch": 0.8779435866471147,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.1979259296872287e-05,
+      "loss": 1.3952,
+      "step": 5089
+    },
+    {
+      "epoch": 0.8781161045458467,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.1976591191651003e-05,
+      "loss": 1.3956,
+      "step": 5090
+    },
+    {
+      "epoch": 0.8782886224445786,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.19739229399915e-05,
+      "loss": 1.3836,
+      "step": 5091
+    },
+    {
+      "epoch": 0.8784611403433106,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.1971254542091446e-05,
+      "loss": 1.411,
+      "step": 5092
+    },
+    {
+      "epoch": 0.8786336582420426,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.1968585998148541e-05,
+      "loss": 1.3845,
+      "step": 5093
+    },
+    {
+      "epoch": 0.8788061761407746,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.1965917308360484e-05,
+      "loss": 1.3995,
+      "step": 5094
+    },
+    {
+      "epoch": 0.8789786940395066,
+      "grad_norm": 0.625,
+      "learning_rate": 1.196324847292499e-05,
+      "loss": 1.426,
+      "step": 5095
+    },
+    {
+      "epoch": 0.8791512119382386,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.1960579492039783e-05,
+      "loss": 1.5251,
+      "step": 5096
+    },
+    {
+      "epoch": 0.8793237298369706,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.1957910365902594e-05,
+      "loss": 1.4423,
+      "step": 5097
+    },
+    {
+      "epoch": 0.8794962477357026,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.1955241094711174e-05,
+      "loss": 1.4563,
+      "step": 5098
+    },
+    {
+      "epoch": 0.8796687656344345,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.1952571678663274e-05,
+      "loss": 1.3736,
+      "step": 5099
+    },
+    {
+      "epoch": 0.8798412835331666,
+      "grad_norm": 0.625,
+      "learning_rate": 1.194990211795667e-05,
+      "loss": 1.5295,
+      "step": 5100
+    },
+    {
+      "epoch": 0.8798412835331666,
+      "eval_loss": 1.4137705564498901,
+      "eval_runtime": 10.9003,
+      "eval_samples_per_second": 93.943,
+      "eval_steps_per_second": 23.486,
+      "step": 5100
+    },
+    {
+      "epoch": 0.8800138014318986,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.1947232412789127e-05,
+      "loss": 1.4976,
+      "step": 5101
+    },
+    {
+      "epoch": 0.8801863193306305,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.1944562563358442e-05,
+      "loss": 1.3986,
+      "step": 5102
+    },
+    {
+      "epoch": 0.8803588372293626,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.1941892569862413e-05,
+      "loss": 1.4513,
+      "step": 5103
+    },
+    {
+      "epoch": 0.8805313551280946,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.1939222432498849e-05,
+      "loss": 1.4324,
+      "step": 5104
+    },
+    {
+      "epoch": 0.8807038730268265,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.193655215146557e-05,
+      "loss": 1.4404,
+      "step": 5105
+    },
+    {
+      "epoch": 0.8808763909255585,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.1933881726960403e-05,
+      "loss": 1.4431,
+      "step": 5106
+    },
+    {
+      "epoch": 0.8810489088242905,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.1931211159181201e-05,
+      "loss": 1.4725,
+      "step": 5107
+    },
+    {
+      "epoch": 0.8812214267230225,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.1928540448325807e-05,
+      "loss": 1.4504,
+      "step": 5108
+    },
+    {
+      "epoch": 0.8813939446217545,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.1925869594592086e-05,
+      "loss": 1.3811,
+      "step": 5109
+    },
+    {
+      "epoch": 0.8815664625204865,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.1923198598177912e-05,
+      "loss": 1.4481,
+      "step": 5110
+    },
+    {
+      "epoch": 0.8817389804192185,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.192052745928117e-05,
+      "loss": 1.4472,
+      "step": 5111
+    },
+    {
+      "epoch": 0.8819114983179505,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.1917856178099756e-05,
+      "loss": 1.4421,
+      "step": 5112
+    },
+    {
+      "epoch": 0.8820840162166825,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.191518475483157e-05,
+      "loss": 1.4525,
+      "step": 5113
+    },
+    {
+      "epoch": 0.8822565341154145,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.191251318967453e-05,
+      "loss": 1.4304,
+      "step": 5114
+    },
+    {
+      "epoch": 0.8824290520141465,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.1909841482826564e-05,
+      "loss": 1.4208,
+      "step": 5115
+    },
+    {
+      "epoch": 0.8826015699128784,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.190716963448561e-05,
+      "loss": 1.5275,
+      "step": 5116
+    },
+    {
+      "epoch": 0.8827740878116105,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.1904497644849611e-05,
+      "loss": 1.4372,
+      "step": 5117
+    },
+    {
+      "epoch": 0.8829466057103424,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.1901825514116526e-05,
+      "loss": 1.4102,
+      "step": 5118
+    },
+    {
+      "epoch": 0.8831191236090744,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.1899153242484322e-05,
+      "loss": 1.4247,
+      "step": 5119
+    },
+    {
+      "epoch": 0.8832916415078065,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.1896480830150985e-05,
+      "loss": 1.4263,
+      "step": 5120
+    },
+    {
+      "epoch": 0.8834641594065384,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.1893808277314494e-05,
+      "loss": 1.3491,
+      "step": 5121
+    },
+    {
+      "epoch": 0.8836366773052704,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.189113558417285e-05,
+      "loss": 1.4428,
+      "step": 5122
+    },
+    {
+      "epoch": 0.8838091952040024,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.1888462750924072e-05,
+      "loss": 1.4553,
+      "step": 5123
+    },
+    {
+      "epoch": 0.8839817131027344,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.188578977776617e-05,
+      "loss": 1.4678,
+      "step": 5124
+    },
+    {
+      "epoch": 0.8841542310014664,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.1883116664897179e-05,
+      "loss": 1.3806,
+      "step": 5125
+    },
+    {
+      "epoch": 0.8843267489001984,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.1880443412515138e-05,
+      "loss": 1.4011,
+      "step": 5126
+    },
+    {
+      "epoch": 0.8844992667989304,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.1877770020818103e-05,
+      "loss": 1.4743,
+      "step": 5127
+    },
+    {
+      "epoch": 0.8846717846976624,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.187509649000413e-05,
+      "loss": 1.4523,
+      "step": 5128
+    },
+    {
+      "epoch": 0.8848443025963943,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.1872422820271294e-05,
+      "loss": 1.4416,
+      "step": 5129
+    },
+    {
+      "epoch": 0.8850168204951264,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.1869749011817675e-05,
+      "loss": 1.4134,
+      "step": 5130
+    },
+    {
+      "epoch": 0.8851893383938584,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.1867075064841365e-05,
+      "loss": 1.3652,
+      "step": 5131
+    },
+    {
+      "epoch": 0.8853618562925903,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.1864400979540472e-05,
+      "loss": 1.4375,
+      "step": 5132
+    },
+    {
+      "epoch": 0.8855343741913223,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.1861726756113101e-05,
+      "loss": 1.4964,
+      "step": 5133
+    },
+    {
+      "epoch": 0.8857068920900544,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.185905239475738e-05,
+      "loss": 1.4896,
+      "step": 5134
+    },
+    {
+      "epoch": 0.8858794099887863,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.1856377895671442e-05,
+      "loss": 1.4705,
+      "step": 5135
+    },
+    {
+      "epoch": 0.8860519278875183,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.1853703259053436e-05,
+      "loss": 1.4255,
+      "step": 5136
+    },
+    {
+      "epoch": 0.8862244457862504,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.1851028485101504e-05,
+      "loss": 1.4481,
+      "step": 5137
+    },
+    {
+      "epoch": 0.8863969636849823,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.1848353574013813e-05,
+      "loss": 1.4347,
+      "step": 5138
+    },
+    {
+      "epoch": 0.8865694815837143,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.1845678525988547e-05,
+      "loss": 1.4899,
+      "step": 5139
+    },
+    {
+      "epoch": 0.8867419994824463,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.184300334122388e-05,
+      "loss": 1.5167,
+      "step": 5140
+    },
+    {
+      "epoch": 0.8869145173811783,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.1840328019918011e-05,
+      "loss": 1.3896,
+      "step": 5141
+    },
+    {
+      "epoch": 0.8870870352799103,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.1837652562269141e-05,
+      "loss": 1.4038,
+      "step": 5142
+    },
+    {
+      "epoch": 0.8872595531786422,
+      "grad_norm": 0.75,
+      "learning_rate": 1.1834976968475488e-05,
+      "loss": 1.4514,
+      "step": 5143
+    },
+    {
+      "epoch": 0.8874320710773743,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.183230123873528e-05,
+      "loss": 1.4866,
+      "step": 5144
+    },
+    {
+      "epoch": 0.8876045889761063,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.1829625373246745e-05,
+      "loss": 1.4061,
+      "step": 5145
+    },
+    {
+      "epoch": 0.8877771068748382,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.1826949372208128e-05,
+      "loss": 1.4768,
+      "step": 5146
+    },
+    {
+      "epoch": 0.8879496247735703,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.1824273235817687e-05,
+      "loss": 1.3845,
+      "step": 5147
+    },
+    {
+      "epoch": 0.8881221426723023,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.1821596964273689e-05,
+      "loss": 1.6053,
+      "step": 5148
+    },
+    {
+      "epoch": 0.8882946605710342,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.1818920557774402e-05,
+      "loss": 1.44,
+      "step": 5149
+    },
+    {
+      "epoch": 0.8884671784697662,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.1816244016518117e-05,
+      "loss": 1.4575,
+      "step": 5150
+    },
+    {
+      "epoch": 0.8886396963684983,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.1813567340703128e-05,
+      "loss": 1.4115,
+      "step": 5151
+    },
+    {
+      "epoch": 0.8888122142672302,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.1810890530527738e-05,
+      "loss": 1.4147,
+      "step": 5152
+    },
+    {
+      "epoch": 0.8889847321659622,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.180821358619026e-05,
+      "loss": 1.3576,
+      "step": 5153
+    },
+    {
+      "epoch": 0.8891572500646943,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.1805536507889021e-05,
+      "loss": 1.4656,
+      "step": 5154
+    },
+    {
+      "epoch": 0.8893297679634262,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.1802859295822358e-05,
+      "loss": 1.4959,
+      "step": 5155
+    },
+    {
+      "epoch": 0.8895022858621582,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.1800181950188617e-05,
+      "loss": 1.4124,
+      "step": 5156
+    },
+    {
+      "epoch": 0.8896748037608901,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.1797504471186146e-05,
+      "loss": 1.4673,
+      "step": 5157
+    },
+    {
+      "epoch": 0.8898473216596222,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.179482685901331e-05,
+      "loss": 1.4117,
+      "step": 5158
+    },
+    {
+      "epoch": 0.8900198395583542,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.1792149113868488e-05,
+      "loss": 1.5153,
+      "step": 5159
+    },
+    {
+      "epoch": 0.8901923574570861,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.1789471235950062e-05,
+      "loss": 1.4547,
+      "step": 5160
+    },
+    {
+      "epoch": 0.8903648753558182,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.1786793225456428e-05,
+      "loss": 1.4885,
+      "step": 5161
+    },
+    {
+      "epoch": 0.8905373932545502,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.1784115082585982e-05,
+      "loss": 1.4485,
+      "step": 5162
+    },
+    {
+      "epoch": 0.8907099111532821,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.1781436807537148e-05,
+      "loss": 1.4793,
+      "step": 5163
+    },
+    {
+      "epoch": 0.8908824290520142,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.1778758400508343e-05,
+      "loss": 1.485,
+      "step": 5164
+    },
+    {
+      "epoch": 0.8910549469507462,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.1776079861698e-05,
+      "loss": 1.4215,
+      "step": 5165
+    },
+    {
+      "epoch": 0.8912274648494781,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.1773401191304564e-05,
+      "loss": 1.4582,
+      "step": 5166
+    },
+    {
+      "epoch": 0.8913999827482101,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.1770722389526487e-05,
+      "loss": 1.4509,
+      "step": 5167
+    },
+    {
+      "epoch": 0.8915725006469422,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.1768043456562233e-05,
+      "loss": 1.5415,
+      "step": 5168
+    },
+    {
+      "epoch": 0.8917450185456741,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.176536439261027e-05,
+      "loss": 1.4549,
+      "step": 5169
+    },
+    {
+      "epoch": 0.8919175364444061,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.176268519786908e-05,
+      "loss": 1.4992,
+      "step": 5170
+    },
+    {
+      "epoch": 0.8920900543431382,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.1760005872537161e-05,
+      "loss": 1.4561,
+      "step": 5171
+    },
+    {
+      "epoch": 0.8922625722418701,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.175732641681301e-05,
+      "loss": 1.4543,
+      "step": 5172
+    },
+    {
+      "epoch": 0.8924350901406021,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.1754646830895138e-05,
+      "loss": 1.4101,
+      "step": 5173
+    },
+    {
+      "epoch": 0.892607608039334,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.1751967114982063e-05,
+      "loss": 1.4677,
+      "step": 5174
+    },
+    {
+      "epoch": 0.8927801259380661,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.1749287269272318e-05,
+      "loss": 1.4727,
+      "step": 5175
+    },
+    {
+      "epoch": 0.8929526438367981,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.1746607293964446e-05,
+      "loss": 1.4093,
+      "step": 5176
+    },
+    {
+      "epoch": 0.89312516173553,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.1743927189256988e-05,
+      "loss": 1.569,
+      "step": 5177
+    },
+    {
+      "epoch": 0.8932976796342621,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.1741246955348506e-05,
+      "loss": 1.5247,
+      "step": 5178
+    },
+    {
+      "epoch": 0.8934701975329941,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.1738566592437573e-05,
+      "loss": 1.3986,
+      "step": 5179
+    },
+    {
+      "epoch": 0.893642715431726,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.1735886100722764e-05,
+      "loss": 1.3666,
+      "step": 5180
+    },
+    {
+      "epoch": 0.893815233330458,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.1733205480402663e-05,
+      "loss": 1.5267,
+      "step": 5181
+    },
+    {
+      "epoch": 0.89398775122919,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.1730524731675872e-05,
+      "loss": 1.3956,
+      "step": 5182
+    },
+    {
+      "epoch": 0.894160269127922,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.1727843854740997e-05,
+      "loss": 1.4123,
+      "step": 5183
+    },
+    {
+      "epoch": 0.894332787026654,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.1725162849796653e-05,
+      "loss": 1.4602,
+      "step": 5184
+    },
+    {
+      "epoch": 0.894505304925386,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.1722481717041467e-05,
+      "loss": 1.3849,
+      "step": 5185
+    },
+    {
+      "epoch": 0.894677822824118,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.171980045667407e-05,
+      "loss": 1.454,
+      "step": 5186
+    },
+    {
+      "epoch": 0.89485034072285,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.1717119068893108e-05,
+      "loss": 1.4808,
+      "step": 5187
+    },
+    {
+      "epoch": 0.895022858621582,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.171443755389724e-05,
+      "loss": 1.5093,
+      "step": 5188
+    },
+    {
+      "epoch": 0.895195376520314,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.1711755911885126e-05,
+      "loss": 1.3697,
+      "step": 5189
+    },
+    {
+      "epoch": 0.895367894419046,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.1709074143055435e-05,
+      "loss": 1.3519,
+      "step": 5190
+    },
+    {
+      "epoch": 0.8955404123177779,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.1706392247606855e-05,
+      "loss": 1.4844,
+      "step": 5191
+    },
+    {
+      "epoch": 0.89571293021651,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.1703710225738077e-05,
+      "loss": 1.4182,
+      "step": 5192
+    },
+    {
+      "epoch": 0.895885448115242,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.1701028077647798e-05,
+      "loss": 1.4381,
+      "step": 5193
+    },
+    {
+      "epoch": 0.8960579660139739,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.169834580353473e-05,
+      "loss": 1.3898,
+      "step": 5194
+    },
+    {
+      "epoch": 0.896230483912706,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.1695663403597592e-05,
+      "loss": 1.3194,
+      "step": 5195
+    },
+    {
+      "epoch": 0.896403001811438,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.1692980878035119e-05,
+      "loss": 1.4205,
+      "step": 5196
+    },
+    {
+      "epoch": 0.8965755197101699,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.1690298227046041e-05,
+      "loss": 1.48,
+      "step": 5197
+    },
+    {
+      "epoch": 0.8967480376089019,
+      "grad_norm": 1.0703125,
+      "learning_rate": 1.1687615450829108e-05,
+      "loss": 1.4633,
+      "step": 5198
+    },
+    {
+      "epoch": 0.8969205555076339,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.1684932549583079e-05,
+      "loss": 1.3481,
+      "step": 5199
+    },
+    {
+      "epoch": 0.8970930734063659,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.1682249523506721e-05,
+      "loss": 1.4053,
+      "step": 5200
+    },
+    {
+      "epoch": 0.8970930734063659,
+      "eval_loss": 1.41326105594635,
+      "eval_runtime": 10.8591,
+      "eval_samples_per_second": 94.299,
+      "eval_steps_per_second": 23.575,
+      "step": 5200
+    },
+    {
+      "epoch": 0.8972655913050979,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.1679566372798803e-05,
+      "loss": 1.4101,
+      "step": 5201
+    },
+    {
+      "epoch": 0.8974381092038299,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.1676883097658117e-05,
+      "loss": 1.43,
+      "step": 5202
+    },
+    {
+      "epoch": 0.8976106271025619,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.1674199698283448e-05,
+      "loss": 1.3728,
+      "step": 5203
+    },
+    {
+      "epoch": 0.8977831450012939,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.167151617487361e-05,
+      "loss": 1.4642,
+      "step": 5204
+    },
+    {
+      "epoch": 0.8979556629000259,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.1668832527627407e-05,
+      "loss": 1.4412,
+      "step": 5205
+    },
+    {
+      "epoch": 0.8981281807987579,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.1666148756743665e-05,
+      "loss": 1.5616,
+      "step": 5206
+    },
+    {
+      "epoch": 0.8983006986974899,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.166346486242121e-05,
+      "loss": 1.444,
+      "step": 5207
+    },
+    {
+      "epoch": 0.8984732165962218,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.1660780844858887e-05,
+      "loss": 1.4412,
+      "step": 5208
+    },
+    {
+      "epoch": 0.8986457344949539,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.1658096704255542e-05,
+      "loss": 1.3195,
+      "step": 5209
+    },
+    {
+      "epoch": 0.8988182523936858,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.165541244081003e-05,
+      "loss": 1.371,
+      "step": 5210
+    },
+    {
+      "epoch": 0.8989907702924178,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.1652728054721223e-05,
+      "loss": 1.4934,
+      "step": 5211
+    },
+    {
+      "epoch": 0.8991632881911499,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.1650043546187994e-05,
+      "loss": 1.412,
+      "step": 5212
+    },
+    {
+      "epoch": 0.8993358060898818,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.1647358915409231e-05,
+      "loss": 1.4778,
+      "step": 5213
+    },
+    {
+      "epoch": 0.8995083239886138,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.1644674162583825e-05,
+      "loss": 1.5873,
+      "step": 5214
+    },
+    {
+      "epoch": 0.8996808418873458,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.164198928791068e-05,
+      "loss": 1.5613,
+      "step": 5215
+    },
+    {
+      "epoch": 0.8998533597860778,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.1639304291588708e-05,
+      "loss": 1.4511,
+      "step": 5216
+    },
+    {
+      "epoch": 0.9000258776848098,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.1636619173816834e-05,
+      "loss": 1.406,
+      "step": 5217
+    },
+    {
+      "epoch": 0.9001983955835418,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.1633933934793984e-05,
+      "loss": 1.5064,
+      "step": 5218
+    },
+    {
+      "epoch": 0.9003709134822738,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.1631248574719098e-05,
+      "loss": 1.3637,
+      "step": 5219
+    },
+    {
+      "epoch": 0.9005434313810058,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.1628563093791128e-05,
+      "loss": 1.501,
+      "step": 5220
+    },
+    {
+      "epoch": 0.9007159492797377,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.1625877492209025e-05,
+      "loss": 1.523,
+      "step": 5221
+    },
+    {
+      "epoch": 0.9008884671784698,
+      "grad_norm": 0.75,
+      "learning_rate": 1.1623191770171761e-05,
+      "loss": 1.4947,
+      "step": 5222
+    },
+    {
+      "epoch": 0.9010609850772018,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.1620505927878305e-05,
+      "loss": 1.5754,
+      "step": 5223
+    },
+    {
+      "epoch": 0.9012335029759337,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.161781996552765e-05,
+      "loss": 1.3632,
+      "step": 5224
+    },
+    {
+      "epoch": 0.9014060208746657,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.1615133883318778e-05,
+      "loss": 1.4967,
+      "step": 5225
+    },
+    {
+      "epoch": 0.9015785387733978,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.1612447681450697e-05,
+      "loss": 1.3705,
+      "step": 5226
+    },
+    {
+      "epoch": 0.9017510566721297,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.160976136012242e-05,
+      "loss": 1.4213,
+      "step": 5227
+    },
+    {
+      "epoch": 0.9019235745708617,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.1607074919532964e-05,
+      "loss": 1.487,
+      "step": 5228
+    },
+    {
+      "epoch": 0.9020960924695938,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.160438835988135e-05,
+      "loss": 1.4603,
+      "step": 5229
+    },
+    {
+      "epoch": 0.9022686103683257,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.1601701681366625e-05,
+      "loss": 1.4169,
+      "step": 5230
+    },
+    {
+      "epoch": 0.9024411282670577,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.1599014884187834e-05,
+      "loss": 1.4951,
+      "step": 5231
+    },
+    {
+      "epoch": 0.9026136461657897,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.1596327968544025e-05,
+      "loss": 1.4138,
+      "step": 5232
+    },
+    {
+      "epoch": 0.9027861640645217,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.1593640934634272e-05,
+      "loss": 1.4008,
+      "step": 5233
+    },
+    {
+      "epoch": 0.9029586819632537,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.1590953782657635e-05,
+      "loss": 1.5109,
+      "step": 5234
+    },
+    {
+      "epoch": 0.9031311998619856,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.15882665128132e-05,
+      "loss": 1.4836,
+      "step": 5235
+    },
+    {
+      "epoch": 0.9033037177607177,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.1585579125300063e-05,
+      "loss": 1.4583,
+      "step": 5236
+    },
+    {
+      "epoch": 0.9034762356594497,
+      "grad_norm": 0.75,
+      "learning_rate": 1.1582891620317316e-05,
+      "loss": 1.3559,
+      "step": 5237
+    },
+    {
+      "epoch": 0.9036487535581816,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.1580203998064066e-05,
+      "loss": 1.51,
+      "step": 5238
+    },
+    {
+      "epoch": 0.9038212714569136,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.157751625873943e-05,
+      "loss": 1.4308,
+      "step": 5239
+    },
+    {
+      "epoch": 0.9039937893556457,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.1574828402542535e-05,
+      "loss": 1.4897,
+      "step": 5240
+    },
+    {
+      "epoch": 0.9041663072543776,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.157214042967251e-05,
+      "loss": 1.4364,
+      "step": 5241
+    },
+    {
+      "epoch": 0.9043388251531096,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.1569452340328497e-05,
+      "loss": 1.4573,
+      "step": 5242
+    },
+    {
+      "epoch": 0.9045113430518417,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.1566764134709652e-05,
+      "loss": 1.4314,
+      "step": 5243
+    },
+    {
+      "epoch": 0.9046838609505736,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.156407581301513e-05,
+      "loss": 1.532,
+      "step": 5244
+    },
+    {
+      "epoch": 0.9048563788493056,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.1561387375444098e-05,
+      "loss": 1.4244,
+      "step": 5245
+    },
+    {
+      "epoch": 0.9050288967480377,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.155869882219573e-05,
+      "loss": 1.4332,
+      "step": 5246
+    },
+    {
+      "epoch": 0.9052014146467696,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.1556010153469219e-05,
+      "loss": 1.3592,
+      "step": 5247
+    },
+    {
+      "epoch": 0.9053739325455016,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.155332136946375e-05,
+      "loss": 1.4865,
+      "step": 5248
+    },
+    {
+      "epoch": 0.9055464504442335,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.155063247037853e-05,
+      "loss": 1.5179,
+      "step": 5249
+    },
+    {
+      "epoch": 0.9057189683429656,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.154794345641277e-05,
+      "loss": 1.4468,
+      "step": 5250
+    },
+    {
+      "epoch": 0.9058914862416976,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.154525432776568e-05,
+      "loss": 1.4827,
+      "step": 5251
+    },
+    {
+      "epoch": 0.9060640041404295,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.15425650846365e-05,
+      "loss": 1.4675,
+      "step": 5252
+    },
+    {
+      "epoch": 0.9062365220391616,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.153987572722446e-05,
+      "loss": 1.46,
+      "step": 5253
+    },
+    {
+      "epoch": 0.9064090399378936,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.1537186255728803e-05,
+      "loss": 1.3107,
+      "step": 5254
+    },
+    {
+      "epoch": 0.9065815578366255,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.1534496670348783e-05,
+      "loss": 1.5258,
+      "step": 5255
+    },
+    {
+      "epoch": 0.9067540757353575,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.1531806971283663e-05,
+      "loss": 1.4652,
+      "step": 5256
+    },
+    {
+      "epoch": 0.9069265936340896,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.152911715873271e-05,
+      "loss": 1.5179,
+      "step": 5257
+    },
+    {
+      "epoch": 0.9070991115328215,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.1526427232895205e-05,
+      "loss": 1.493,
+      "step": 5258
+    },
+    {
+      "epoch": 0.9072716294315535,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.1523737193970432e-05,
+      "loss": 1.4812,
+      "step": 5259
+    },
+    {
+      "epoch": 0.9074441473302856,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.1521047042157684e-05,
+      "loss": 1.4551,
+      "step": 5260
+    },
+    {
+      "epoch": 0.9076166652290175,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.1518356777656274e-05,
+      "loss": 1.3966,
+      "step": 5261
+    },
+    {
+      "epoch": 0.9077891831277495,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.1515666400665504e-05,
+      "loss": 1.3979,
+      "step": 5262
+    },
+    {
+      "epoch": 0.9079617010264815,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.1512975911384695e-05,
+      "loss": 1.3519,
+      "step": 5263
+    },
+    {
+      "epoch": 0.9081342189252135,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.1510285310013176e-05,
+      "loss": 1.4521,
+      "step": 5264
+    },
+    {
+      "epoch": 0.9083067368239455,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.1507594596750288e-05,
+      "loss": 1.445,
+      "step": 5265
+    },
+    {
+      "epoch": 0.9084792547226774,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.150490377179537e-05,
+      "loss": 1.3316,
+      "step": 5266
+    },
+    {
+      "epoch": 0.9086517726214095,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.1502212835347776e-05,
+      "loss": 1.4873,
+      "step": 5267
+    },
+    {
+      "epoch": 0.9088242905201415,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.1499521787606874e-05,
+      "loss": 1.446,
+      "step": 5268
+    },
+    {
+      "epoch": 0.9089968084188734,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.1496830628772025e-05,
+      "loss": 1.4837,
+      "step": 5269
+    },
+    {
+      "epoch": 0.9091693263176055,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.1494139359042612e-05,
+      "loss": 1.5096,
+      "step": 5270
+    },
+    {
+      "epoch": 0.9093418442163375,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.1491447978618015e-05,
+      "loss": 1.4487,
+      "step": 5271
+    },
+    {
+      "epoch": 0.9095143621150694,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.148875648769764e-05,
+      "loss": 1.4519,
+      "step": 5272
+    },
+    {
+      "epoch": 0.9096868800138014,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.1486064886480876e-05,
+      "loss": 1.4319,
+      "step": 5273
+    },
+    {
+      "epoch": 0.9098593979125335,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.1483373175167142e-05,
+      "loss": 1.3451,
+      "step": 5274
+    },
+    {
+      "epoch": 0.9100319158112654,
+      "grad_norm": 0.625,
+      "learning_rate": 1.1480681353955856e-05,
+      "loss": 1.4576,
+      "step": 5275
+    },
+    {
+      "epoch": 0.9102044337099974,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.1477989423046442e-05,
+      "loss": 1.4853,
+      "step": 5276
+    },
+    {
+      "epoch": 0.9103769516087294,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.147529738263834e-05,
+      "loss": 1.4957,
+      "step": 5277
+    },
+    {
+      "epoch": 0.9105494695074614,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.1472605232930985e-05,
+      "loss": 1.4662,
+      "step": 5278
+    },
+    {
+      "epoch": 0.9107219874061934,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.1469912974123835e-05,
+      "loss": 1.3621,
+      "step": 5279
+    },
+    {
+      "epoch": 0.9108945053049254,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.1467220606416348e-05,
+      "loss": 1.4839,
+      "step": 5280
+    },
+    {
+      "epoch": 0.9110670232036574,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.1464528130007992e-05,
+      "loss": 1.3379,
+      "step": 5281
+    },
+    {
+      "epoch": 0.9112395411023894,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.146183554509824e-05,
+      "loss": 1.4744,
+      "step": 5282
+    },
+    {
+      "epoch": 0.9114120590011213,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.1459142851886573e-05,
+      "loss": 1.4277,
+      "step": 5283
+    },
+    {
+      "epoch": 0.9115845768998534,
+      "grad_norm": 0.625,
+      "learning_rate": 1.1456450050572491e-05,
+      "loss": 1.5622,
+      "step": 5284
+    },
+    {
+      "epoch": 0.9117570947985854,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.1453757141355488e-05,
+      "loss": 1.4224,
+      "step": 5285
+    },
+    {
+      "epoch": 0.9119296126973173,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.1451064124435072e-05,
+      "loss": 1.4353,
+      "step": 5286
+    },
+    {
+      "epoch": 0.9121021305960494,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.1448371000010758e-05,
+      "loss": 1.3761,
+      "step": 5287
+    },
+    {
+      "epoch": 0.9122746484947813,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.1445677768282073e-05,
+      "loss": 1.4288,
+      "step": 5288
+    },
+    {
+      "epoch": 0.9124471663935133,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.1442984429448545e-05,
+      "loss": 1.4021,
+      "step": 5289
+    },
+    {
+      "epoch": 0.9126196842922453,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.1440290983709715e-05,
+      "loss": 1.4077,
+      "step": 5290
+    },
+    {
+      "epoch": 0.9127922021909773,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.1437597431265126e-05,
+      "loss": 1.403,
+      "step": 5291
+    },
+    {
+      "epoch": 0.9129647200897093,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.143490377231434e-05,
+      "loss": 1.37,
+      "step": 5292
+    },
+    {
+      "epoch": 0.9131372379884413,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.1432210007056919e-05,
+      "loss": 1.4122,
+      "step": 5293
+    },
+    {
+      "epoch": 0.9133097558871733,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.1429516135692427e-05,
+      "loss": 1.4618,
+      "step": 5294
+    },
+    {
+      "epoch": 0.9134822737859053,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.1426822158420449e-05,
+      "loss": 1.3283,
+      "step": 5295
+    },
+    {
+      "epoch": 0.9136547916846373,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.1424128075440572e-05,
+      "loss": 1.438,
+      "step": 5296
+    },
+    {
+      "epoch": 0.9138273095833693,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.142143388695239e-05,
+      "loss": 1.4426,
+      "step": 5297
+    },
+    {
+      "epoch": 0.9139998274821013,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.14187395931555e-05,
+      "loss": 1.4818,
+      "step": 5298
+    },
+    {
+      "epoch": 0.9141723453808333,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.1416045194249517e-05,
+      "loss": 1.4444,
+      "step": 5299
+    },
+    {
+      "epoch": 0.9143448632795652,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.141335069043406e-05,
+      "loss": 1.4915,
+      "step": 5300
+    },
+    {
+      "epoch": 0.9143448632795652,
+      "eval_loss": 1.4127678871154785,
+      "eval_runtime": 10.8242,
+      "eval_samples_per_second": 94.603,
+      "eval_steps_per_second": 23.651,
+      "step": 5300
+    },
+    {
+      "epoch": 0.9145173811782973,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.1410656081908754e-05,
+      "loss": 1.3751,
+      "step": 5301
+    },
+    {
+      "epoch": 0.9146898990770292,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.1407961368873226e-05,
+      "loss": 1.4008,
+      "step": 5302
+    },
+    {
+      "epoch": 0.9148624169757612,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.1405266551527126e-05,
+      "loss": 1.4782,
+      "step": 5303
+    },
+    {
+      "epoch": 0.9150349348744933,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.1402571630070098e-05,
+      "loss": 1.4454,
+      "step": 5304
+    },
+    {
+      "epoch": 0.9152074527732252,
+      "grad_norm": 0.625,
+      "learning_rate": 1.1399876604701805e-05,
+      "loss": 1.3988,
+      "step": 5305
+    },
+    {
+      "epoch": 0.9153799706719572,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.1397181475621901e-05,
+      "loss": 1.3682,
+      "step": 5306
+    },
+    {
+      "epoch": 0.9155524885706892,
+      "grad_norm": 0.625,
+      "learning_rate": 1.1394486243030066e-05,
+      "loss": 1.438,
+      "step": 5307
+    },
+    {
+      "epoch": 0.9157250064694212,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.1391790907125975e-05,
+      "loss": 1.4038,
+      "step": 5308
+    },
+    {
+      "epoch": 0.9158975243681532,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.138909546810932e-05,
+      "loss": 1.4956,
+      "step": 5309
+    },
+    {
+      "epoch": 0.9160700422668852,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.138639992617979e-05,
+      "loss": 1.5037,
+      "step": 5310
+    },
+    {
+      "epoch": 0.9162425601656172,
+      "grad_norm": 0.625,
+      "learning_rate": 1.138370428153709e-05,
+      "loss": 1.4534,
+      "step": 5311
+    },
+    {
+      "epoch": 0.9164150780643492,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.1381008534380933e-05,
+      "loss": 1.4813,
+      "step": 5312
+    },
+    {
+      "epoch": 0.9165875959630811,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.1378312684911036e-05,
+      "loss": 1.4618,
+      "step": 5313
+    },
+    {
+      "epoch": 0.9167601138618131,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.1375616733327125e-05,
+      "loss": 1.3733,
+      "step": 5314
+    },
+    {
+      "epoch": 0.9169326317605452,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.1372920679828922e-05,
+      "loss": 1.4316,
+      "step": 5315
+    },
+    {
+      "epoch": 0.9171051496592771,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.1370224524616187e-05,
+      "loss": 1.419,
+      "step": 5316
+    },
+    {
+      "epoch": 0.9172776675580091,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.1367528267888653e-05,
+      "loss": 1.4067,
+      "step": 5317
+    },
+    {
+      "epoch": 0.9174501854567412,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.136483190984608e-05,
+      "loss": 1.4496,
+      "step": 5318
+    },
+    {
+      "epoch": 0.9176227033554731,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.1362135450688232e-05,
+      "loss": 1.4289,
+      "step": 5319
+    },
+    {
+      "epoch": 0.9177952212542051,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.1359438890614878e-05,
+      "loss": 1.4319,
+      "step": 5320
+    },
+    {
+      "epoch": 0.9179677391529372,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.13567422298258e-05,
+      "loss": 1.4672,
+      "step": 5321
+    },
+    {
+      "epoch": 0.9181402570516691,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.1354045468520777e-05,
+      "loss": 1.4084,
+      "step": 5322
+    },
+    {
+      "epoch": 0.9183127749504011,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.1351348606899605e-05,
+      "loss": 1.3973,
+      "step": 5323
+    },
+    {
+      "epoch": 0.918485292849133,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.1348651645162088e-05,
+      "loss": 1.4132,
+      "step": 5324
+    },
+    {
+      "epoch": 0.9186578107478651,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.1345954583508028e-05,
+      "loss": 1.3457,
+      "step": 5325
+    },
+    {
+      "epoch": 0.9188303286465971,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.1343257422137244e-05,
+      "loss": 1.3226,
+      "step": 5326
+    },
+    {
+      "epoch": 0.919002846545329,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.1340560161249554e-05,
+      "loss": 1.3878,
+      "step": 5327
+    },
+    {
+      "epoch": 0.9191753644440611,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.1337862801044792e-05,
+      "loss": 1.4944,
+      "step": 5328
+    },
+    {
+      "epoch": 0.9193478823427931,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.1335165341722798e-05,
+      "loss": 1.4164,
+      "step": 5329
+    },
+    {
+      "epoch": 0.919520400241525,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.133246778348341e-05,
+      "loss": 1.5615,
+      "step": 5330
+    },
+    {
+      "epoch": 0.919692918140257,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.132977012652648e-05,
+      "loss": 1.3541,
+      "step": 5331
+    },
+    {
+      "epoch": 0.9198654360389891,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.1327072371051873e-05,
+      "loss": 1.3902,
+      "step": 5332
+    },
+    {
+      "epoch": 0.920037953937721,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.1324374517259455e-05,
+      "loss": 1.4545,
+      "step": 5333
+    },
+    {
+      "epoch": 0.920210471836453,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.1321676565349096e-05,
+      "loss": 1.4162,
+      "step": 5334
+    },
+    {
+      "epoch": 0.9203829897351851,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.1318978515520678e-05,
+      "loss": 1.3777,
+      "step": 5335
+    },
+    {
+      "epoch": 0.920555507633917,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.1316280367974091e-05,
+      "loss": 1.4113,
+      "step": 5336
+    },
+    {
+      "epoch": 0.920728025532649,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.131358212290923e-05,
+      "loss": 1.5264,
+      "step": 5337
+    },
+    {
+      "epoch": 0.9209005434313811,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.1310883780525996e-05,
+      "loss": 1.4073,
+      "step": 5338
+    },
+    {
+      "epoch": 0.921073061330113,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.1308185341024303e-05,
+      "loss": 1.3889,
+      "step": 5339
+    },
+    {
+      "epoch": 0.921245579228845,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.1305486804604065e-05,
+      "loss": 1.4394,
+      "step": 5340
+    },
+    {
+      "epoch": 0.9214180971275769,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.1302788171465208e-05,
+      "loss": 1.2919,
+      "step": 5341
+    },
+    {
+      "epoch": 0.921590615026309,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.1300089441807664e-05,
+      "loss": 1.4688,
+      "step": 5342
+    },
+    {
+      "epoch": 0.921763132925041,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.129739061583137e-05,
+      "loss": 1.5077,
+      "step": 5343
+    },
+    {
+      "epoch": 0.9219356508237729,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.129469169373627e-05,
+      "loss": 1.5521,
+      "step": 5344
+    },
+    {
+      "epoch": 0.922108168722505,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.1291992675722325e-05,
+      "loss": 1.55,
+      "step": 5345
+    },
+    {
+      "epoch": 0.922280686621237,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.1289293561989486e-05,
+      "loss": 1.3981,
+      "step": 5346
+    },
+    {
+      "epoch": 0.9224532045199689,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.1286594352737723e-05,
+      "loss": 1.4134,
+      "step": 5347
+    },
+    {
+      "epoch": 0.9226257224187009,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.1283895048167013e-05,
+      "loss": 1.3881,
+      "step": 5348
+    },
+    {
+      "epoch": 0.922798240317433,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.1281195648477336e-05,
+      "loss": 1.4075,
+      "step": 5349
+    },
+    {
+      "epoch": 0.9229707582161649,
+      "grad_norm": 0.625,
+      "learning_rate": 1.1278496153868681e-05,
+      "loss": 1.3477,
+      "step": 5350
+    },
+    {
+      "epoch": 0.9231432761148969,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.127579656454104e-05,
+      "loss": 1.4386,
+      "step": 5351
+    },
+    {
+      "epoch": 0.923315794013629,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.1273096880694419e-05,
+      "loss": 1.4019,
+      "step": 5352
+    },
+    {
+      "epoch": 0.9234883119123609,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.127039710252883e-05,
+      "loss": 1.4122,
+      "step": 5353
+    },
+    {
+      "epoch": 0.9236608298110929,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.1267697230244281e-05,
+      "loss": 1.4157,
+      "step": 5354
+    },
+    {
+      "epoch": 0.923833347709825,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.1264997264040802e-05,
+      "loss": 1.392,
+      "step": 5355
+    },
+    {
+      "epoch": 0.9240058656085569,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.1262297204118422e-05,
+      "loss": 1.4343,
+      "step": 5356
+    },
+    {
+      "epoch": 0.9241783835072889,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.1259597050677178e-05,
+      "loss": 1.4186,
+      "step": 5357
+    },
+    {
+      "epoch": 0.9243509014060208,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.1256896803917115e-05,
+      "loss": 1.3255,
+      "step": 5358
+    },
+    {
+      "epoch": 0.9245234193047529,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.1254196464038281e-05,
+      "loss": 1.4194,
+      "step": 5359
+    },
+    {
+      "epoch": 0.9246959372034849,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.1251496031240736e-05,
+      "loss": 1.5005,
+      "step": 5360
+    },
+    {
+      "epoch": 0.9248684551022168,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.1248795505724548e-05,
+      "loss": 1.3214,
+      "step": 5361
+    },
+    {
+      "epoch": 0.9250409730009489,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.1246094887689784e-05,
+      "loss": 1.4798,
+      "step": 5362
+    },
+    {
+      "epoch": 0.9252134908996809,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.1243394177336524e-05,
+      "loss": 1.4729,
+      "step": 5363
+    },
+    {
+      "epoch": 0.9253860087984128,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.1240693374864854e-05,
+      "loss": 1.5883,
+      "step": 5364
+    },
+    {
+      "epoch": 0.9255585266971448,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.123799248047487e-05,
+      "loss": 1.4788,
+      "step": 5365
+    },
+    {
+      "epoch": 0.9257310445958769,
+      "grad_norm": 0.625,
+      "learning_rate": 1.1235291494366668e-05,
+      "loss": 1.4362,
+      "step": 5366
+    },
+    {
+      "epoch": 0.9259035624946088,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.123259041674035e-05,
+      "loss": 1.4446,
+      "step": 5367
+    },
+    {
+      "epoch": 0.9260760803933408,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.1229889247796033e-05,
+      "loss": 1.416,
+      "step": 5368
+    },
+    {
+      "epoch": 0.9262485982920728,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.1227187987733838e-05,
+      "loss": 1.4744,
+      "step": 5369
+    },
+    {
+      "epoch": 0.9264211161908048,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.1224486636753888e-05,
+      "loss": 1.3834,
+      "step": 5370
+    },
+    {
+      "epoch": 0.9265936340895368,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.1221785195056316e-05,
+      "loss": 1.4061,
+      "step": 5371
+    },
+    {
+      "epoch": 0.9267661519882688,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.1219083662841263e-05,
+      "loss": 1.473,
+      "step": 5372
+    },
+    {
+      "epoch": 0.9269386698870008,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.1216382040308877e-05,
+      "loss": 1.3396,
+      "step": 5373
+    },
+    {
+      "epoch": 0.9271111877857328,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.121368032765931e-05,
+      "loss": 1.4057,
+      "step": 5374
+    },
+    {
+      "epoch": 0.9272837056844647,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.1210978525092717e-05,
+      "loss": 1.4684,
+      "step": 5375
+    },
+    {
+      "epoch": 0.9274562235831968,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.120827663280927e-05,
+      "loss": 1.4043,
+      "step": 5376
+    },
+    {
+      "epoch": 0.9276287414819288,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.120557465100914e-05,
+      "loss": 1.4343,
+      "step": 5377
+    },
+    {
+      "epoch": 0.9278012593806607,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.1202872579892507e-05,
+      "loss": 1.4044,
+      "step": 5378
+    },
+    {
+      "epoch": 0.9279737772793928,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.1200170419659558e-05,
+      "loss": 1.3756,
+      "step": 5379
+    },
+    {
+      "epoch": 0.9281462951781247,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.1197468170510483e-05,
+      "loss": 1.4508,
+      "step": 5380
+    },
+    {
+      "epoch": 0.9283188130768567,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.1194765832645489e-05,
+      "loss": 1.5315,
+      "step": 5381
+    },
+    {
+      "epoch": 0.9284913309755887,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.1192063406264772e-05,
+      "loss": 1.4169,
+      "step": 5382
+    },
+    {
+      "epoch": 0.9286638488743207,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.1189360891568551e-05,
+      "loss": 1.483,
+      "step": 5383
+    },
+    {
+      "epoch": 0.9288363667730527,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.1186658288757044e-05,
+      "loss": 1.4337,
+      "step": 5384
+    },
+    {
+      "epoch": 0.9290088846717847,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.1183955598030479e-05,
+      "loss": 1.5479,
+      "step": 5385
+    },
+    {
+      "epoch": 0.9291814025705167,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.1181252819589081e-05,
+      "loss": 1.4495,
+      "step": 5386
+    },
+    {
+      "epoch": 0.9293539204692487,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.1178549953633096e-05,
+      "loss": 1.4111,
+      "step": 5387
+    },
+    {
+      "epoch": 0.9295264383679807,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.1175847000362765e-05,
+      "loss": 1.453,
+      "step": 5388
+    },
+    {
+      "epoch": 0.9296989562667126,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.1173143959978344e-05,
+      "loss": 1.5223,
+      "step": 5389
+    },
+    {
+      "epoch": 0.9298714741654447,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.1170440832680086e-05,
+      "loss": 1.4029,
+      "step": 5390
+    },
+    {
+      "epoch": 0.9300439920641767,
+      "grad_norm": 0.625,
+      "learning_rate": 1.1167737618668258e-05,
+      "loss": 1.3839,
+      "step": 5391
+    },
+    {
+      "epoch": 0.9302165099629086,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.1165034318143129e-05,
+      "loss": 1.4541,
+      "step": 5392
+    },
+    {
+      "epoch": 0.9303890278616407,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.116233093130498e-05,
+      "loss": 1.3738,
+      "step": 5393
+    },
+    {
+      "epoch": 0.9305615457603726,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.1159627458354096e-05,
+      "loss": 1.4244,
+      "step": 5394
+    },
+    {
+      "epoch": 0.9307340636591046,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.1156923899490761e-05,
+      "loss": 1.3742,
+      "step": 5395
+    },
+    {
+      "epoch": 0.9309065815578367,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.115422025491527e-05,
+      "loss": 1.4678,
+      "step": 5396
+    },
+    {
+      "epoch": 0.9310790994565686,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.1151516524827938e-05,
+      "loss": 1.3177,
+      "step": 5397
+    },
+    {
+      "epoch": 0.9312516173553006,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.1148812709429067e-05,
+      "loss": 1.3989,
+      "step": 5398
+    },
+    {
+      "epoch": 0.9314241352540326,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.114610880891897e-05,
+      "loss": 1.3871,
+      "step": 5399
+    },
+    {
+      "epoch": 0.9315966531527646,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.1143404823497969e-05,
+      "loss": 1.4404,
+      "step": 5400
+    },
+    {
+      "epoch": 0.9315966531527646,
+      "eval_loss": 1.4122982025146484,
+      "eval_runtime": 10.8508,
+      "eval_samples_per_second": 94.371,
+      "eval_steps_per_second": 23.593,
+      "step": 5400
+    },
+    {
+      "epoch": 0.9317691710514966,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.1140700753366397e-05,
+      "loss": 1.3977,
+      "step": 5401
+    },
+    {
+      "epoch": 0.9319416889502286,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.1137996598724588e-05,
+      "loss": 1.4804,
+      "step": 5402
+    },
+    {
+      "epoch": 0.9321142068489606,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.1135292359772875e-05,
+      "loss": 1.4298,
+      "step": 5403
+    },
+    {
+      "epoch": 0.9322867247476926,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.1132588036711614e-05,
+      "loss": 1.4478,
+      "step": 5404
+    },
+    {
+      "epoch": 0.9324592426464245,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.1129883629741156e-05,
+      "loss": 1.4497,
+      "step": 5405
+    },
+    {
+      "epoch": 0.9326317605451565,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.1127179139061854e-05,
+      "loss": 1.4567,
+      "step": 5406
+    },
+    {
+      "epoch": 0.9328042784438886,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.112447456487408e-05,
+      "loss": 1.3938,
+      "step": 5407
+    },
+    {
+      "epoch": 0.9329767963426205,
+      "grad_norm": 0.625,
+      "learning_rate": 1.1121769907378206e-05,
+      "loss": 1.5709,
+      "step": 5408
+    },
+    {
+      "epoch": 0.9331493142413525,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.1119065166774608e-05,
+      "loss": 1.4594,
+      "step": 5409
+    },
+    {
+      "epoch": 0.9333218321400846,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.1116360343263671e-05,
+      "loss": 1.3834,
+      "step": 5410
+    },
+    {
+      "epoch": 0.9334943500388165,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.1113655437045782e-05,
+      "loss": 1.5306,
+      "step": 5411
+    },
+    {
+      "epoch": 0.9336668679375485,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.1110950448321338e-05,
+      "loss": 1.3906,
+      "step": 5412
+    },
+    {
+      "epoch": 0.9338393858362806,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.1108245377290748e-05,
+      "loss": 1.4253,
+      "step": 5413
+    },
+    {
+      "epoch": 0.9340119037350125,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.1105540224154413e-05,
+      "loss": 1.5626,
+      "step": 5414
+    },
+    {
+      "epoch": 0.9341844216337445,
+      "grad_norm": 0.625,
+      "learning_rate": 1.1102834989112752e-05,
+      "loss": 1.3905,
+      "step": 5415
+    },
+    {
+      "epoch": 0.9343569395324764,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.1100129672366182e-05,
+      "loss": 1.404,
+      "step": 5416
+    },
+    {
+      "epoch": 0.9345294574312085,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.1097424274115136e-05,
+      "loss": 1.4319,
+      "step": 5417
+    },
+    {
+      "epoch": 0.9347019753299405,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.1094718794560037e-05,
+      "loss": 1.5673,
+      "step": 5418
+    },
+    {
+      "epoch": 0.9348744932286724,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.1092013233901334e-05,
+      "loss": 1.3819,
+      "step": 5419
+    },
+    {
+      "epoch": 0.9350470111274045,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.1089307592339467e-05,
+      "loss": 1.4296,
+      "step": 5420
+    },
+    {
+      "epoch": 0.9352195290261365,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.1086601870074887e-05,
+      "loss": 1.4434,
+      "step": 5421
+    },
+    {
+      "epoch": 0.9353920469248684,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.1083896067308052e-05,
+      "loss": 1.5113,
+      "step": 5422
+    },
+    {
+      "epoch": 0.9355645648236004,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.1081190184239418e-05,
+      "loss": 1.4677,
+      "step": 5423
+    },
+    {
+      "epoch": 0.9357370827223325,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.1078484221069465e-05,
+      "loss": 1.4168,
+      "step": 5424
+    },
+    {
+      "epoch": 0.9359096006210644,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.107577817799866e-05,
+      "loss": 1.3563,
+      "step": 5425
+    },
+    {
+      "epoch": 0.9360821185197964,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.107307205522749e-05,
+      "loss": 1.3891,
+      "step": 5426
+    },
+    {
+      "epoch": 0.9362546364185285,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.1070365852956432e-05,
+      "loss": 1.4826,
+      "step": 5427
+    },
+    {
+      "epoch": 0.9364271543172604,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.1067659571385985e-05,
+      "loss": 1.3225,
+      "step": 5428
+    },
+    {
+      "epoch": 0.9365996722159924,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.1064953210716647e-05,
+      "loss": 1.4668,
+      "step": 5429
+    },
+    {
+      "epoch": 0.9367721901147245,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.1062246771148922e-05,
+      "loss": 1.4022,
+      "step": 5430
+    },
+    {
+      "epoch": 0.9369447080134564,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.1059540252883317e-05,
+      "loss": 1.4116,
+      "step": 5431
+    },
+    {
+      "epoch": 0.9371172259121884,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.105683365612035e-05,
+      "loss": 1.5058,
+      "step": 5432
+    },
+    {
+      "epoch": 0.9372897438109203,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.1054126981060542e-05,
+      "loss": 1.4929,
+      "step": 5433
+    },
+    {
+      "epoch": 0.9374622617096524,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.1051420227904423e-05,
+      "loss": 1.4555,
+      "step": 5434
+    },
+    {
+      "epoch": 0.9376347796083844,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.1048713396852522e-05,
+      "loss": 1.4708,
+      "step": 5435
+    },
+    {
+      "epoch": 0.9378072975071163,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.1046006488105379e-05,
+      "loss": 1.456,
+      "step": 5436
+    },
+    {
+      "epoch": 0.9379798154058484,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.104329950186354e-05,
+      "loss": 1.3934,
+      "step": 5437
+    },
+    {
+      "epoch": 0.9381523333045804,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.1040592438327558e-05,
+      "loss": 1.4241,
+      "step": 5438
+    },
+    {
+      "epoch": 0.9383248512033123,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.1037885297697985e-05,
+      "loss": 1.4923,
+      "step": 5439
+    },
+    {
+      "epoch": 0.9384973691020443,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.1035178080175382e-05,
+      "loss": 1.4042,
+      "step": 5440
+    },
+    {
+      "epoch": 0.9386698870007764,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.1032470785960321e-05,
+      "loss": 1.488,
+      "step": 5441
+    },
+    {
+      "epoch": 0.9388424048995083,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.1029763415253374e-05,
+      "loss": 1.4338,
+      "step": 5442
+    },
+    {
+      "epoch": 0.9390149227982403,
+      "grad_norm": 1.75,
+      "learning_rate": 1.1027055968255116e-05,
+      "loss": 1.3755,
+      "step": 5443
+    },
+    {
+      "epoch": 0.9391874406969724,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.1024348445166133e-05,
+      "loss": 1.472,
+      "step": 5444
+    },
+    {
+      "epoch": 0.9393599585957043,
+      "grad_norm": 0.625,
+      "learning_rate": 1.1021640846187021e-05,
+      "loss": 1.3511,
+      "step": 5445
+    },
+    {
+      "epoch": 0.9395324764944363,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.1018933171518369e-05,
+      "loss": 1.4285,
+      "step": 5446
+    },
+    {
+      "epoch": 0.9397049943931682,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.1016225421360783e-05,
+      "loss": 1.4036,
+      "step": 5447
+    },
+    {
+      "epoch": 0.9398775122919003,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.1013517595914867e-05,
+      "loss": 1.4262,
+      "step": 5448
+    },
+    {
+      "epoch": 0.9400500301906323,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.1010809695381235e-05,
+      "loss": 1.3865,
+      "step": 5449
+    },
+    {
+      "epoch": 0.9402225480893642,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.1008101719960504e-05,
+      "loss": 1.4087,
+      "step": 5450
+    },
+    {
+      "epoch": 0.9403950659880963,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.10053936698533e-05,
+      "loss": 1.3771,
+      "step": 5451
+    },
+    {
+      "epoch": 0.9405675838868283,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.100268554526025e-05,
+      "loss": 1.4645,
+      "step": 5452
+    },
+    {
+      "epoch": 0.9407401017855602,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.099997734638199e-05,
+      "loss": 1.4652,
+      "step": 5453
+    },
+    {
+      "epoch": 0.9409126196842923,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.0997269073419162e-05,
+      "loss": 1.5207,
+      "step": 5454
+    },
+    {
+      "epoch": 0.9410851375830243,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.0994560726572408e-05,
+      "loss": 1.4699,
+      "step": 5455
+    },
+    {
+      "epoch": 0.9412576554817562,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.0991852306042381e-05,
+      "loss": 1.3744,
+      "step": 5456
+    },
+    {
+      "epoch": 0.9414301733804882,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.0989143812029736e-05,
+      "loss": 1.4185,
+      "step": 5457
+    },
+    {
+      "epoch": 0.9416026912792202,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.0986435244735141e-05,
+      "loss": 1.3928,
+      "step": 5458
+    },
+    {
+      "epoch": 0.9417752091779522,
+      "grad_norm": 0.625,
+      "learning_rate": 1.0983726604359257e-05,
+      "loss": 1.422,
+      "step": 5459
+    },
+    {
+      "epoch": 0.9419477270766842,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.0981017891102757e-05,
+      "loss": 1.4217,
+      "step": 5460
+    },
+    {
+      "epoch": 0.9421202449754162,
+      "grad_norm": 0.625,
+      "learning_rate": 1.0978309105166328e-05,
+      "loss": 1.455,
+      "step": 5461
+    },
+    {
+      "epoch": 0.9422927628741482,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.0975600246750644e-05,
+      "loss": 1.4021,
+      "step": 5462
+    },
+    {
+      "epoch": 0.9424652807728802,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.0972891316056397e-05,
+      "loss": 1.4949,
+      "step": 5463
+    },
+    {
+      "epoch": 0.9426377986716121,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.0970182313284283e-05,
+      "loss": 1.4539,
+      "step": 5464
+    },
+    {
+      "epoch": 0.9428103165703442,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.0967473238635005e-05,
+      "loss": 1.4881,
+      "step": 5465
+    },
+    {
+      "epoch": 0.9429828344690762,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.0964764092309261e-05,
+      "loss": 1.423,
+      "step": 5466
+    },
+    {
+      "epoch": 0.9431553523678081,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.0962054874507766e-05,
+      "loss": 1.5349,
+      "step": 5467
+    },
+    {
+      "epoch": 0.9433278702665402,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.0959345585431233e-05,
+      "loss": 1.3661,
+      "step": 5468
+    },
+    {
+      "epoch": 0.9435003881652722,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.0956636225280386e-05,
+      "loss": 1.5621,
+      "step": 5469
+    },
+    {
+      "epoch": 0.9436729060640041,
+      "grad_norm": 0.625,
+      "learning_rate": 1.0953926794255952e-05,
+      "loss": 1.4969,
+      "step": 5470
+    },
+    {
+      "epoch": 0.9438454239627362,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.0951217292558659e-05,
+      "loss": 1.5013,
+      "step": 5471
+    },
+    {
+      "epoch": 0.9440179418614681,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.0948507720389242e-05,
+      "loss": 1.4703,
+      "step": 5472
+    },
+    {
+      "epoch": 0.9441904597602001,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.094579807794845e-05,
+      "loss": 1.4418,
+      "step": 5473
+    },
+    {
+      "epoch": 0.9443629776589321,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.094308836543703e-05,
+      "loss": 1.3773,
+      "step": 5474
+    },
+    {
+      "epoch": 0.9445354955576641,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.0940378583055727e-05,
+      "loss": 1.4028,
+      "step": 5475
+    },
+    {
+      "epoch": 0.9447080134563961,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.09376687310053e-05,
+      "loss": 1.3619,
+      "step": 5476
+    },
+    {
+      "epoch": 0.9448805313551281,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.093495880948652e-05,
+      "loss": 1.4371,
+      "step": 5477
+    },
+    {
+      "epoch": 0.9450530492538601,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.0932248818700146e-05,
+      "loss": 1.5281,
+      "step": 5478
+    },
+    {
+      "epoch": 0.9452255671525921,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.0929538758846956e-05,
+      "loss": 1.4308,
+      "step": 5479
+    },
+    {
+      "epoch": 0.945398085051324,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.0926828630127728e-05,
+      "loss": 1.4224,
+      "step": 5480
+    },
+    {
+      "epoch": 0.945570602950056,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.0924118432743243e-05,
+      "loss": 1.4619,
+      "step": 5481
+    },
+    {
+      "epoch": 0.9457431208487881,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.0921408166894292e-05,
+      "loss": 1.5235,
+      "step": 5482
+    },
+    {
+      "epoch": 0.94591563874752,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.0918697832781663e-05,
+      "loss": 1.3516,
+      "step": 5483
+    },
+    {
+      "epoch": 0.946088156646252,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.0915987430606161e-05,
+      "loss": 1.4879,
+      "step": 5484
+    },
+    {
+      "epoch": 0.9462606745449841,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.0913276960568583e-05,
+      "loss": 1.551,
+      "step": 5485
+    },
+    {
+      "epoch": 0.946433192443716,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.0910566422869748e-05,
+      "loss": 1.4284,
+      "step": 5486
+    },
+    {
+      "epoch": 0.946605710342448,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.0907855817710457e-05,
+      "loss": 1.3764,
+      "step": 5487
+    },
+    {
+      "epoch": 0.9467782282411801,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.0905145145291537e-05,
+      "loss": 1.4533,
+      "step": 5488
+    },
+    {
+      "epoch": 0.946950746139912,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.0902434405813809e-05,
+      "loss": 1.3727,
+      "step": 5489
+    },
+    {
+      "epoch": 0.947123264038644,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.0899723599478103e-05,
+      "loss": 1.4027,
+      "step": 5490
+    },
+    {
+      "epoch": 0.947295781937376,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.089701272648525e-05,
+      "loss": 1.4288,
+      "step": 5491
+    },
+    {
+      "epoch": 0.947468299836108,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.0894301787036085e-05,
+      "loss": 1.3706,
+      "step": 5492
+    },
+    {
+      "epoch": 0.94764081773484,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.0891590781331463e-05,
+      "loss": 1.4171,
+      "step": 5493
+    },
+    {
+      "epoch": 0.947813335633572,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.088887970957222e-05,
+      "loss": 1.3683,
+      "step": 5494
+    },
+    {
+      "epoch": 0.947985853532304,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.0886168571959214e-05,
+      "loss": 1.3923,
+      "step": 5495
+    },
+    {
+      "epoch": 0.948158371431036,
+      "grad_norm": 0.625,
+      "learning_rate": 1.0883457368693307e-05,
+      "loss": 1.4652,
+      "step": 5496
+    },
+    {
+      "epoch": 0.948330889329768,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.0880746099975355e-05,
+      "loss": 1.5098,
+      "step": 5497
+    },
+    {
+      "epoch": 0.9485034072284999,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.087803476600623e-05,
+      "loss": 1.4231,
+      "step": 5498
+    },
+    {
+      "epoch": 0.948675925127232,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.0875323366986803e-05,
+      "loss": 1.4784,
+      "step": 5499
+    },
+    {
+      "epoch": 0.9488484430259639,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.0872611903117951e-05,
+      "loss": 1.4145,
+      "step": 5500
+    },
+    {
+      "epoch": 0.9488484430259639,
+      "eval_loss": 1.4118043184280396,
+      "eval_runtime": 11.0208,
+      "eval_samples_per_second": 92.915,
+      "eval_steps_per_second": 23.229,
+      "step": 5500
+    },
+    {
+      "epoch": 0.9490209609246959,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.0869900374600557e-05,
+      "loss": 1.4202,
+      "step": 5501
+    },
+    {
+      "epoch": 0.949193478823428,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.086718878163551e-05,
+      "loss": 1.4019,
+      "step": 5502
+    },
+    {
+      "epoch": 0.9493659967221599,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.0864477124423698e-05,
+      "loss": 1.4151,
+      "step": 5503
+    },
+    {
+      "epoch": 0.9495385146208919,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.0861765403166018e-05,
+      "loss": 1.471,
+      "step": 5504
+    },
+    {
+      "epoch": 0.949711032519624,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.0859053618063372e-05,
+      "loss": 1.4649,
+      "step": 5505
+    },
+    {
+      "epoch": 0.9498835504183559,
+      "grad_norm": 1.0,
+      "learning_rate": 1.0856341769316672e-05,
+      "loss": 1.426,
+      "step": 5506
+    },
+    {
+      "epoch": 0.9500560683170879,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.0853629857126817e-05,
+      "loss": 1.3961,
+      "step": 5507
+    },
+    {
+      "epoch": 0.9502285862158198,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.085091788169473e-05,
+      "loss": 1.4524,
+      "step": 5508
+    },
+    {
+      "epoch": 0.9504011041145519,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.084820584322133e-05,
+      "loss": 1.4308,
+      "step": 5509
+    },
+    {
+      "epoch": 0.9505736220132839,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.084549374190754e-05,
+      "loss": 1.4239,
+      "step": 5510
+    },
+    {
+      "epoch": 0.9507461399120158,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.0842781577954294e-05,
+      "loss": 1.3743,
+      "step": 5511
+    },
+    {
+      "epoch": 0.9509186578107479,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.0840069351562519e-05,
+      "loss": 1.3684,
+      "step": 5512
+    },
+    {
+      "epoch": 0.9510911757094799,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.0837357062933158e-05,
+      "loss": 1.3119,
+      "step": 5513
+    },
+    {
+      "epoch": 0.9512636936082118,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.0834644712267158e-05,
+      "loss": 1.462,
+      "step": 5514
+    },
+    {
+      "epoch": 0.9514362115069438,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.0831932299765458e-05,
+      "loss": 1.4212,
+      "step": 5515
+    },
+    {
+      "epoch": 0.9516087294056759,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.0829219825629016e-05,
+      "loss": 1.4816,
+      "step": 5516
+    },
+    {
+      "epoch": 0.9517812473044078,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.0826507290058787e-05,
+      "loss": 1.4519,
+      "step": 5517
+    },
+    {
+      "epoch": 0.9519537652031398,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.0823794693255738e-05,
+      "loss": 1.4065,
+      "step": 5518
+    },
+    {
+      "epoch": 0.9521262831018719,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.0821082035420829e-05,
+      "loss": 1.4785,
+      "step": 5519
+    },
+    {
+      "epoch": 0.9522988010006038,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.0818369316755031e-05,
+      "loss": 1.41,
+      "step": 5520
+    },
+    {
+      "epoch": 0.9524713188993358,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.081565653745932e-05,
+      "loss": 1.3799,
+      "step": 5521
+    },
+    {
+      "epoch": 0.9526438367980677,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.081294369773468e-05,
+      "loss": 1.3599,
+      "step": 5522
+    },
+    {
+      "epoch": 0.9528163546967998,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.0810230797782088e-05,
+      "loss": 1.4379,
+      "step": 5523
+    },
+    {
+      "epoch": 0.9529888725955318,
+      "grad_norm": 0.625,
+      "learning_rate": 1.0807517837802535e-05,
+      "loss": 1.447,
+      "step": 5524
+    },
+    {
+      "epoch": 0.9531613904942637,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.0804804817997015e-05,
+      "loss": 1.5195,
+      "step": 5525
+    },
+    {
+      "epoch": 0.9533339083929958,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.080209173856653e-05,
+      "loss": 1.4686,
+      "step": 5526
+    },
+    {
+      "epoch": 0.9535064262917278,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.0799378599712073e-05,
+      "loss": 1.4608,
+      "step": 5527
+    },
+    {
+      "epoch": 0.9536789441904597,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.0796665401634657e-05,
+      "loss": 1.3722,
+      "step": 5528
+    },
+    {
+      "epoch": 0.9538514620891918,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.0793952144535289e-05,
+      "loss": 1.4454,
+      "step": 5529
+    },
+    {
+      "epoch": 0.9540239799879238,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.0791238828614987e-05,
+      "loss": 1.3737,
+      "step": 5530
+    },
+    {
+      "epoch": 0.9541964978866557,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.0788525454074765e-05,
+      "loss": 1.3756,
+      "step": 5531
+    },
+    {
+      "epoch": 0.9543690157853877,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.0785812021115654e-05,
+      "loss": 1.4899,
+      "step": 5532
+    },
+    {
+      "epoch": 0.9545415336841198,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.0783098529938675e-05,
+      "loss": 1.3998,
+      "step": 5533
+    },
+    {
+      "epoch": 0.9547140515828517,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.078038498074487e-05,
+      "loss": 1.437,
+      "step": 5534
+    },
+    {
+      "epoch": 0.9548865694815837,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.0777671373735266e-05,
+      "loss": 1.4801,
+      "step": 5535
+    },
+    {
+      "epoch": 0.9550590873803158,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.0774957709110905e-05,
+      "loss": 1.4471,
+      "step": 5536
+    },
+    {
+      "epoch": 0.9552316052790477,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.0772243987072838e-05,
+      "loss": 1.4805,
+      "step": 5537
+    },
+    {
+      "epoch": 0.9554041231777797,
+      "grad_norm": 0.625,
+      "learning_rate": 1.076953020782211e-05,
+      "loss": 1.4201,
+      "step": 5538
+    },
+    {
+      "epoch": 0.9555766410765116,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.0766816371559776e-05,
+      "loss": 1.4411,
+      "step": 5539
+    },
+    {
+      "epoch": 0.9557491589752437,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.0764102478486894e-05,
+      "loss": 1.4425,
+      "step": 5540
+    },
+    {
+      "epoch": 0.9559216768739757,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.0761388528804524e-05,
+      "loss": 1.488,
+      "step": 5541
+    },
+    {
+      "epoch": 0.9560941947727076,
+      "grad_norm": 0.625,
+      "learning_rate": 1.0758674522713737e-05,
+      "loss": 1.4224,
+      "step": 5542
+    },
+    {
+      "epoch": 0.9562667126714397,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.07559604604156e-05,
+      "loss": 1.4734,
+      "step": 5543
+    },
+    {
+      "epoch": 0.9564392305701717,
+      "grad_norm": 0.625,
+      "learning_rate": 1.075324634211119e-05,
+      "loss": 1.3959,
+      "step": 5544
+    },
+    {
+      "epoch": 0.9566117484689036,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.0750532168001581e-05,
+      "loss": 1.5127,
+      "step": 5545
+    },
+    {
+      "epoch": 0.9567842663676357,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.0747817938287866e-05,
+      "loss": 1.5154,
+      "step": 5546
+    },
+    {
+      "epoch": 0.9569567842663677,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.074510365317112e-05,
+      "loss": 1.3529,
+      "step": 5547
+    },
+    {
+      "epoch": 0.9571293021650996,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.0742389312852441e-05,
+      "loss": 1.3786,
+      "step": 5548
+    },
+    {
+      "epoch": 0.9573018200638316,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.0739674917532923e-05,
+      "loss": 1.3981,
+      "step": 5549
+    },
+    {
+      "epoch": 0.9574743379625636,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.0736960467413668e-05,
+      "loss": 1.4618,
+      "step": 5550
+    },
+    {
+      "epoch": 0.9576468558612956,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.0734245962695775e-05,
+      "loss": 1.3747,
+      "step": 5551
+    },
+    {
+      "epoch": 0.9578193737600276,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.073153140358035e-05,
+      "loss": 1.3701,
+      "step": 5552
+    },
+    {
+      "epoch": 0.9579918916587596,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.0728816790268513e-05,
+      "loss": 1.4192,
+      "step": 5553
+    },
+    {
+      "epoch": 0.9581644095574916,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.0726102122961373e-05,
+      "loss": 1.4915,
+      "step": 5554
+    },
+    {
+      "epoch": 0.9583369274562236,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.0723387401860048e-05,
+      "loss": 1.4411,
+      "step": 5555
+    },
+    {
+      "epoch": 0.9585094453549555,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.0720672627165665e-05,
+      "loss": 1.4482,
+      "step": 5556
+    },
+    {
+      "epoch": 0.9586819632536876,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.071795779907935e-05,
+      "loss": 1.4209,
+      "step": 5557
+    },
+    {
+      "epoch": 0.9588544811524196,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.071524291780224e-05,
+      "loss": 1.4751,
+      "step": 5558
+    },
+    {
+      "epoch": 0.9590269990511515,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.0712527983535463e-05,
+      "loss": 1.4628,
+      "step": 5559
+    },
+    {
+      "epoch": 0.9591995169498836,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.070981299648016e-05,
+      "loss": 1.3229,
+      "step": 5560
+    },
+    {
+      "epoch": 0.9593720348486156,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.0707097956837475e-05,
+      "loss": 1.3905,
+      "step": 5561
+    },
+    {
+      "epoch": 0.9595445527473475,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.0704382864808558e-05,
+      "loss": 1.4366,
+      "step": 5562
+    },
+    {
+      "epoch": 0.9597170706460796,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.0701667720594555e-05,
+      "loss": 1.4201,
+      "step": 5563
+    },
+    {
+      "epoch": 0.9598895885448115,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.0698952524396621e-05,
+      "loss": 1.4332,
+      "step": 5564
+    },
+    {
+      "epoch": 0.9600621064435435,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.069623727641592e-05,
+      "loss": 1.5507,
+      "step": 5565
+    },
+    {
+      "epoch": 0.9602346243422755,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.0693521976853612e-05,
+      "loss": 1.4058,
+      "step": 5566
+    },
+    {
+      "epoch": 0.9604071422410075,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.0690806625910862e-05,
+      "loss": 1.5297,
+      "step": 5567
+    },
+    {
+      "epoch": 0.9605796601397395,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.068809122378884e-05,
+      "loss": 1.5256,
+      "step": 5568
+    },
+    {
+      "epoch": 0.9607521780384715,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.0685375770688717e-05,
+      "loss": 1.4706,
+      "step": 5569
+    },
+    {
+      "epoch": 0.9609246959372035,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.068266026681168e-05,
+      "loss": 1.4291,
+      "step": 5570
+    },
+    {
+      "epoch": 0.9610972138359355,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.0679944712358903e-05,
+      "loss": 1.3866,
+      "step": 5571
+    },
+    {
+      "epoch": 0.9612697317346675,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.067722910753157e-05,
+      "loss": 1.4339,
+      "step": 5572
+    },
+    {
+      "epoch": 0.9614422496333994,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.0674513452530877e-05,
+      "loss": 1.3447,
+      "step": 5573
+    },
+    {
+      "epoch": 0.9616147675321315,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.0671797747558013e-05,
+      "loss": 1.4081,
+      "step": 5574
+    },
+    {
+      "epoch": 0.9617872854308634,
+      "grad_norm": 0.625,
+      "learning_rate": 1.0669081992814174e-05,
+      "loss": 1.4666,
+      "step": 5575
+    },
+    {
+      "epoch": 0.9619598033295954,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.0666366188500559e-05,
+      "loss": 1.3793,
+      "step": 5576
+    },
+    {
+      "epoch": 0.9621323212283275,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.0663650334818374e-05,
+      "loss": 1.4518,
+      "step": 5577
+    },
+    {
+      "epoch": 0.9623048391270594,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.0660934431968829e-05,
+      "loss": 1.4876,
+      "step": 5578
+    },
+    {
+      "epoch": 0.9624773570257914,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.0658218480153127e-05,
+      "loss": 1.3871,
+      "step": 5579
+    },
+    {
+      "epoch": 0.9626498749245235,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.0655502479572486e-05,
+      "loss": 1.4577,
+      "step": 5580
+    },
+    {
+      "epoch": 0.9628223928232554,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.0652786430428128e-05,
+      "loss": 1.3674,
+      "step": 5581
+    },
+    {
+      "epoch": 0.9629949107219874,
+      "grad_norm": 0.75,
+      "learning_rate": 1.0650070332921275e-05,
+      "loss": 1.4384,
+      "step": 5582
+    },
+    {
+      "epoch": 0.9631674286207194,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.0647354187253145e-05,
+      "loss": 1.3278,
+      "step": 5583
+    },
+    {
+      "epoch": 0.9633399465194514,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.0644637993624973e-05,
+      "loss": 1.4714,
+      "step": 5584
+    },
+    {
+      "epoch": 0.9635124644181834,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.0641921752237992e-05,
+      "loss": 1.4265,
+      "step": 5585
+    },
+    {
+      "epoch": 0.9636849823169154,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.0639205463293436e-05,
+      "loss": 1.4047,
+      "step": 5586
+    },
+    {
+      "epoch": 0.9638575002156474,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.0636489126992548e-05,
+      "loss": 1.4471,
+      "step": 5587
+    },
+    {
+      "epoch": 0.9640300181143794,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.0633772743536563e-05,
+      "loss": 1.438,
+      "step": 5588
+    },
+    {
+      "epoch": 0.9642025360131113,
+      "grad_norm": 0.625,
+      "learning_rate": 1.0631056313126736e-05,
+      "loss": 1.4633,
+      "step": 5589
+    },
+    {
+      "epoch": 0.9643750539118433,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.0628339835964317e-05,
+      "loss": 1.3694,
+      "step": 5590
+    },
+    {
+      "epoch": 0.9645475718105754,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.0625623312250554e-05,
+      "loss": 1.4391,
+      "step": 5591
+    },
+    {
+      "epoch": 0.9647200897093073,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.0622906742186707e-05,
+      "loss": 1.409,
+      "step": 5592
+    },
+    {
+      "epoch": 0.9648926076080393,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.0620190125974036e-05,
+      "loss": 1.5127,
+      "step": 5593
+    },
+    {
+      "epoch": 0.9650651255067714,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.0617473463813808e-05,
+      "loss": 1.4054,
+      "step": 5594
+    },
+    {
+      "epoch": 0.9652376434055033,
+      "grad_norm": 3.453125,
+      "learning_rate": 1.0614756755907284e-05,
+      "loss": 1.462,
+      "step": 5595
+    },
+    {
+      "epoch": 0.9654101613042353,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.0612040002455742e-05,
+      "loss": 1.403,
+      "step": 5596
+    },
+    {
+      "epoch": 0.9655826792029673,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.0609323203660451e-05,
+      "loss": 1.4555,
+      "step": 5597
+    },
+    {
+      "epoch": 0.9657551971016993,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.0606606359722691e-05,
+      "loss": 1.3381,
+      "step": 5598
+    },
+    {
+      "epoch": 0.9659277150004313,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.060388947084374e-05,
+      "loss": 1.2738,
+      "step": 5599
+    },
+    {
+      "epoch": 0.9661002328991632,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.0601172537224881e-05,
+      "loss": 1.3784,
+      "step": 5600
+    },
+    {
+      "epoch": 0.9661002328991632,
+      "eval_loss": 1.411436915397644,
+      "eval_runtime": 10.8561,
+      "eval_samples_per_second": 94.325,
+      "eval_steps_per_second": 23.581,
+      "step": 5600
+    },
+    {
+      "epoch": 0.9662727507978953,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.0598455559067409e-05,
+      "loss": 1.3455,
+      "step": 5601
+    },
+    {
+      "epoch": 0.9664452686966273,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.0595738536572607e-05,
+      "loss": 1.5216,
+      "step": 5602
+    },
+    {
+      "epoch": 0.9666177865953592,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.0593021469941773e-05,
+      "loss": 1.4521,
+      "step": 5603
+    },
+    {
+      "epoch": 0.9667903044940913,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.0590304359376202e-05,
+      "loss": 1.3953,
+      "step": 5604
+    },
+    {
+      "epoch": 0.9669628223928233,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.0587587205077196e-05,
+      "loss": 1.4433,
+      "step": 5605
+    },
+    {
+      "epoch": 0.9671353402915552,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.0584870007246059e-05,
+      "loss": 1.4629,
+      "step": 5606
+    },
+    {
+      "epoch": 0.9673078581902872,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.05821527660841e-05,
+      "loss": 1.4307,
+      "step": 5607
+    },
+    {
+      "epoch": 0.9674803760890193,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.0579435481792621e-05,
+      "loss": 1.3582,
+      "step": 5608
+    },
+    {
+      "epoch": 0.9676528939877512,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.0576718154572944e-05,
+      "loss": 1.372,
+      "step": 5609
+    },
+    {
+      "epoch": 0.9678254118864832,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.0574000784626386e-05,
+      "loss": 1.4447,
+      "step": 5610
+    },
+    {
+      "epoch": 0.9679979297852153,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.057128337215426e-05,
+      "loss": 1.4734,
+      "step": 5611
+    },
+    {
+      "epoch": 0.9681704476839472,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.0568565917357892e-05,
+      "loss": 1.4525,
+      "step": 5612
+    },
+    {
+      "epoch": 0.9683429655826792,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.0565848420438608e-05,
+      "loss": 1.4177,
+      "step": 5613
+    },
+    {
+      "epoch": 0.9685154834814111,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.0563130881597739e-05,
+      "loss": 1.3963,
+      "step": 5614
+    },
+    {
+      "epoch": 0.9686880013801432,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.0560413301036614e-05,
+      "loss": 1.3608,
+      "step": 5615
+    },
+    {
+      "epoch": 0.9688605192788752,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.0557695678956569e-05,
+      "loss": 1.4067,
+      "step": 5616
+    },
+    {
+      "epoch": 0.9690330371776071,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.0554978015558946e-05,
+      "loss": 1.4527,
+      "step": 5617
+    },
+    {
+      "epoch": 0.9692055550763392,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.0552260311045082e-05,
+      "loss": 1.4399,
+      "step": 5618
+    },
+    {
+      "epoch": 0.9693780729750712,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.0549542565616326e-05,
+      "loss": 1.5231,
+      "step": 5619
+    },
+    {
+      "epoch": 0.9695505908738031,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.0546824779474022e-05,
+      "loss": 1.4008,
+      "step": 5620
+    },
+    {
+      "epoch": 0.9697231087725352,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.054410695281952e-05,
+      "loss": 1.4296,
+      "step": 5621
+    },
+    {
+      "epoch": 0.9698956266712672,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.0541389085854177e-05,
+      "loss": 1.4242,
+      "step": 5622
+    },
+    {
+      "epoch": 0.9700681445699991,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.0538671178779346e-05,
+      "loss": 1.3831,
+      "step": 5623
+    },
+    {
+      "epoch": 0.9702406624687311,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.053595323179639e-05,
+      "loss": 1.3795,
+      "step": 5624
+    },
+    {
+      "epoch": 0.9704131803674632,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.0533235245106668e-05,
+      "loss": 1.3784,
+      "step": 5625
+    },
+    {
+      "epoch": 0.9705856982661951,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.053051721891155e-05,
+      "loss": 1.4076,
+      "step": 5626
+    },
+    {
+      "epoch": 0.9707582161649271,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.0527799153412402e-05,
+      "loss": 1.3815,
+      "step": 5627
+    },
+    {
+      "epoch": 0.9709307340636592,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.052508104881059e-05,
+      "loss": 1.4889,
+      "step": 5628
+    },
+    {
+      "epoch": 0.9711032519623911,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.0522362905307497e-05,
+      "loss": 1.4448,
+      "step": 5629
+    },
+    {
+      "epoch": 0.9712757698611231,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.0519644723104494e-05,
+      "loss": 1.3009,
+      "step": 5630
+    },
+    {
+      "epoch": 0.971448287759855,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.0516926502402966e-05,
+      "loss": 1.5176,
+      "step": 5631
+    },
+    {
+      "epoch": 0.9716208056585871,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.0514208243404291e-05,
+      "loss": 1.4149,
+      "step": 5632
+    },
+    {
+      "epoch": 0.9717933235573191,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.0511489946309856e-05,
+      "loss": 1.4443,
+      "step": 5633
+    },
+    {
+      "epoch": 0.971965841456051,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.050877161132105e-05,
+      "loss": 1.4321,
+      "step": 5634
+    },
+    {
+      "epoch": 0.9721383593547831,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.0506053238639267e-05,
+      "loss": 1.4988,
+      "step": 5635
+    },
+    {
+      "epoch": 0.9723108772535151,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.0503334828465895e-05,
+      "loss": 1.4718,
+      "step": 5636
+    },
+    {
+      "epoch": 0.972483395152247,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.0500616381002331e-05,
+      "loss": 1.3965,
+      "step": 5637
+    },
+    {
+      "epoch": 0.9726559130509791,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.0497897896449983e-05,
+      "loss": 1.5349,
+      "step": 5638
+    },
+    {
+      "epoch": 0.972828430949711,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.0495179375010244e-05,
+      "loss": 1.3807,
+      "step": 5639
+    },
+    {
+      "epoch": 0.973000948848443,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.0492460816884524e-05,
+      "loss": 1.5651,
+      "step": 5640
+    },
+    {
+      "epoch": 0.973173466747175,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.0489742222274231e-05,
+      "loss": 1.3553,
+      "step": 5641
+    },
+    {
+      "epoch": 0.973345984645907,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.0487023591380775e-05,
+      "loss": 1.4698,
+      "step": 5642
+    },
+    {
+      "epoch": 0.973518502544639,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.0484304924405566e-05,
+      "loss": 1.4298,
+      "step": 5643
+    },
+    {
+      "epoch": 0.973691020443371,
+      "grad_norm": 6.875,
+      "learning_rate": 1.0481586221550022e-05,
+      "loss": 1.5664,
+      "step": 5644
+    },
+    {
+      "epoch": 0.973863538342103,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.0478867483015563e-05,
+      "loss": 1.4564,
+      "step": 5645
+    },
+    {
+      "epoch": 0.974036056240835,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.0476148709003607e-05,
+      "loss": 1.3828,
+      "step": 5646
+    },
+    {
+      "epoch": 0.974208574139567,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.047342989971558e-05,
+      "loss": 1.4288,
+      "step": 5647
+    },
+    {
+      "epoch": 0.9743810920382989,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.0470711055352912e-05,
+      "loss": 1.5092,
+      "step": 5648
+    },
+    {
+      "epoch": 0.974553609937031,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.0467992176117024e-05,
+      "loss": 1.4224,
+      "step": 5649
+    },
+    {
+      "epoch": 0.974726127835763,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.046527326220935e-05,
+      "loss": 1.5092,
+      "step": 5650
+    },
+    {
+      "epoch": 0.9748986457344949,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.0462554313831327e-05,
+      "loss": 1.459,
+      "step": 5651
+    },
+    {
+      "epoch": 0.975071163633227,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.0459835331184392e-05,
+      "loss": 1.413,
+      "step": 5652
+    },
+    {
+      "epoch": 0.975243681531959,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.045711631446998e-05,
+      "loss": 1.3732,
+      "step": 5653
+    },
+    {
+      "epoch": 0.9754161994306909,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.0454397263889538e-05,
+      "loss": 1.4446,
+      "step": 5654
+    },
+    {
+      "epoch": 0.975588717329423,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.0451678179644504e-05,
+      "loss": 1.4371,
+      "step": 5655
+    },
+    {
+      "epoch": 0.9757612352281549,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.044895906193633e-05,
+      "loss": 1.5077,
+      "step": 5656
+    },
+    {
+      "epoch": 0.9759337531268869,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.0446239910966462e-05,
+      "loss": 1.4009,
+      "step": 5657
+    },
+    {
+      "epoch": 0.9761062710256189,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.0443520726936356e-05,
+      "loss": 1.4608,
+      "step": 5658
+    },
+    {
+      "epoch": 0.9762787889243509,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.0440801510047462e-05,
+      "loss": 1.4192,
+      "step": 5659
+    },
+    {
+      "epoch": 0.9764513068230829,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.0438082260501239e-05,
+      "loss": 1.4076,
+      "step": 5660
+    },
+    {
+      "epoch": 0.9766238247218149,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.0435362978499143e-05,
+      "loss": 1.3499,
+      "step": 5661
+    },
+    {
+      "epoch": 0.9767963426205469,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.043264366424264e-05,
+      "loss": 1.4749,
+      "step": 5662
+    },
+    {
+      "epoch": 0.9769688605192789,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.0429924317933189e-05,
+      "loss": 1.3881,
+      "step": 5663
+    },
+    {
+      "epoch": 0.9771413784180109,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.0427204939772257e-05,
+      "loss": 1.4782,
+      "step": 5664
+    },
+    {
+      "epoch": 0.9773138963167428,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.0424485529961314e-05,
+      "loss": 1.4459,
+      "step": 5665
+    },
+    {
+      "epoch": 0.9774864142154749,
+      "grad_norm": 0.625,
+      "learning_rate": 1.0421766088701832e-05,
+      "loss": 1.3883,
+      "step": 5666
+    },
+    {
+      "epoch": 0.9776589321142068,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.0419046616195285e-05,
+      "loss": 1.5965,
+      "step": 5667
+    },
+    {
+      "epoch": 0.9778314500129388,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.0416327112643143e-05,
+      "loss": 1.3919,
+      "step": 5668
+    },
+    {
+      "epoch": 0.9780039679116709,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.0413607578246886e-05,
+      "loss": 1.4496,
+      "step": 5669
+    },
+    {
+      "epoch": 0.9781764858104028,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.0410888013208001e-05,
+      "loss": 1.4421,
+      "step": 5670
+    },
+    {
+      "epoch": 0.9783490037091348,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.040816841772796e-05,
+      "loss": 1.5079,
+      "step": 5671
+    },
+    {
+      "epoch": 0.9785215216078668,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.0405448792008252e-05,
+      "loss": 1.4715,
+      "step": 5672
+    },
+    {
+      "epoch": 0.9786940395065988,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.0402729136250364e-05,
+      "loss": 1.4881,
+      "step": 5673
+    },
+    {
+      "epoch": 0.9788665574053308,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.0400009450655789e-05,
+      "loss": 1.3597,
+      "step": 5674
+    },
+    {
+      "epoch": 0.9790390753040628,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.0397289735426012e-05,
+      "loss": 1.5159,
+      "step": 5675
+    },
+    {
+      "epoch": 0.9792115932027948,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.0394569990762528e-05,
+      "loss": 1.4996,
+      "step": 5676
+    },
+    {
+      "epoch": 0.9793841111015268,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.0391850216866834e-05,
+      "loss": 1.4435,
+      "step": 5677
+    },
+    {
+      "epoch": 0.9795566290002587,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.038913041394043e-05,
+      "loss": 1.335,
+      "step": 5678
+    },
+    {
+      "epoch": 0.9797291468989908,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.0386410582184813e-05,
+      "loss": 1.3793,
+      "step": 5679
+    },
+    {
+      "epoch": 0.9799016647977228,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.0383690721801485e-05,
+      "loss": 1.4007,
+      "step": 5680
+    },
+    {
+      "epoch": 0.9800741826964547,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.038097083299195e-05,
+      "loss": 1.3936,
+      "step": 5681
+    },
+    {
+      "epoch": 0.9802467005951867,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.0378250915957716e-05,
+      "loss": 1.5647,
+      "step": 5682
+    },
+    {
+      "epoch": 0.9804192184939188,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.0375530970900292e-05,
+      "loss": 1.3682,
+      "step": 5683
+    },
+    {
+      "epoch": 0.9805917363926507,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.0372810998021185e-05,
+      "loss": 1.4399,
+      "step": 5684
+    },
+    {
+      "epoch": 0.9807642542913827,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.0370090997521906e-05,
+      "loss": 1.4918,
+      "step": 5685
+    },
+    {
+      "epoch": 0.9809367721901148,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.0367370969603981e-05,
+      "loss": 1.4271,
+      "step": 5686
+    },
+    {
+      "epoch": 0.9811092900888467,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.0364650914468917e-05,
+      "loss": 1.4676,
+      "step": 5687
+    },
+    {
+      "epoch": 0.9812818079875787,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.0361930832318232e-05,
+      "loss": 1.3725,
+      "step": 5688
+    },
+    {
+      "epoch": 0.9814543258863107,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.0359210723353451e-05,
+      "loss": 1.3957,
+      "step": 5689
+    },
+    {
+      "epoch": 0.9816268437850427,
+      "grad_norm": 1.25,
+      "learning_rate": 1.0356490587776095e-05,
+      "loss": 1.4496,
+      "step": 5690
+    },
+    {
+      "epoch": 0.9817993616837747,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.0353770425787693e-05,
+      "loss": 1.4051,
+      "step": 5691
+    },
+    {
+      "epoch": 0.9819718795825066,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.0351050237589763e-05,
+      "loss": 1.4122,
+      "step": 5692
+    },
+    {
+      "epoch": 0.9821443974812387,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.034833002338384e-05,
+      "loss": 1.4373,
+      "step": 5693
+    },
+    {
+      "epoch": 0.9823169153799707,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.0345609783371448e-05,
+      "loss": 1.4795,
+      "step": 5694
+    },
+    {
+      "epoch": 0.9824894332787026,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.0342889517754131e-05,
+      "loss": 1.4264,
+      "step": 5695
+    },
+    {
+      "epoch": 0.9826619511774347,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.034016922673341e-05,
+      "loss": 1.3468,
+      "step": 5696
+    },
+    {
+      "epoch": 0.9828344690761667,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.033744891051083e-05,
+      "loss": 1.3907,
+      "step": 5697
+    },
+    {
+      "epoch": 0.9830069869748986,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.0334728569287924e-05,
+      "loss": 1.4186,
+      "step": 5698
+    },
+    {
+      "epoch": 0.9831795048736306,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.0332008203266237e-05,
+      "loss": 1.4321,
+      "step": 5699
+    },
+    {
+      "epoch": 0.9833520227723627,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.0329287812647307e-05,
+      "loss": 1.3987,
+      "step": 5700
+    },
+    {
+      "epoch": 0.9833520227723627,
+      "eval_loss": 1.4110112190246582,
+      "eval_runtime": 10.8773,
+      "eval_samples_per_second": 94.141,
+      "eval_steps_per_second": 23.535,
+      "step": 5700
+    },
+    {
+      "epoch": 0.9835245406710946,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.0326567397632675e-05,
+      "loss": 1.449,
+      "step": 5701
+    },
+    {
+      "epoch": 0.9836970585698266,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.0323846958423894e-05,
+      "loss": 1.4914,
+      "step": 5702
+    },
+    {
+      "epoch": 0.9838695764685587,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.0321126495222505e-05,
+      "loss": 1.4824,
+      "step": 5703
+    },
+    {
+      "epoch": 0.9840420943672906,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.031840600823006e-05,
+      "loss": 1.4821,
+      "step": 5704
+    },
+    {
+      "epoch": 0.9842146122660226,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.0315685497648107e-05,
+      "loss": 1.4482,
+      "step": 5705
+    },
+    {
+      "epoch": 0.9843871301647545,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.0312964963678198e-05,
+      "loss": 1.5063,
+      "step": 5706
+    },
+    {
+      "epoch": 0.9845596480634866,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.0310244406521895e-05,
+      "loss": 1.4312,
+      "step": 5707
+    },
+    {
+      "epoch": 0.9847321659622186,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.0307523826380743e-05,
+      "loss": 1.4777,
+      "step": 5708
+    },
+    {
+      "epoch": 0.9849046838609505,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.0304803223456305e-05,
+      "loss": 1.4131,
+      "step": 5709
+    },
+    {
+      "epoch": 0.9850772017596826,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.0302082597950141e-05,
+      "loss": 1.4888,
+      "step": 5710
+    },
+    {
+      "epoch": 0.9852497196584146,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.029936195006381e-05,
+      "loss": 1.5442,
+      "step": 5711
+    },
+    {
+      "epoch": 0.9854222375571465,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.0296641279998876e-05,
+      "loss": 1.4095,
+      "step": 5712
+    },
+    {
+      "epoch": 0.9855947554558786,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.02939205879569e-05,
+      "loss": 1.4457,
+      "step": 5713
+    },
+    {
+      "epoch": 0.9857672733546106,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.0291199874139453e-05,
+      "loss": 1.4679,
+      "step": 5714
+    },
+    {
+      "epoch": 0.9859397912533425,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.02884791387481e-05,
+      "loss": 1.4651,
+      "step": 5715
+    },
+    {
+      "epoch": 0.9861123091520745,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.0285758381984408e-05,
+      "loss": 1.5139,
+      "step": 5716
+    },
+    {
+      "epoch": 0.9862848270508066,
+      "grad_norm": 0.75,
+      "learning_rate": 1.0283037604049948e-05,
+      "loss": 1.4568,
+      "step": 5717
+    },
+    {
+      "epoch": 0.9864573449495385,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.0280316805146295e-05,
+      "loss": 1.4287,
+      "step": 5718
+    },
+    {
+      "epoch": 0.9866298628482705,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.0277595985475024e-05,
+      "loss": 1.5319,
+      "step": 5719
+    },
+    {
+      "epoch": 0.9868023807470025,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.0274875145237706e-05,
+      "loss": 1.4455,
+      "step": 5720
+    },
+    {
+      "epoch": 0.9869748986457345,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.027215428463592e-05,
+      "loss": 1.4713,
+      "step": 5721
+    },
+    {
+      "epoch": 0.9871474165444665,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.0269433403871246e-05,
+      "loss": 1.4784,
+      "step": 5722
+    },
+    {
+      "epoch": 0.9873199344431984,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.0266712503145262e-05,
+      "loss": 1.5312,
+      "step": 5723
+    },
+    {
+      "epoch": 0.9874924523419305,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.0263991582659547e-05,
+      "loss": 1.5036,
+      "step": 5724
+    },
+    {
+      "epoch": 0.9876649702406625,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.0261270642615687e-05,
+      "loss": 1.3068,
+      "step": 5725
+    },
+    {
+      "epoch": 0.9878374881393944,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.0258549683215262e-05,
+      "loss": 1.4302,
+      "step": 5726
+    },
+    {
+      "epoch": 0.9880100060381265,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.0255828704659868e-05,
+      "loss": 1.4892,
+      "step": 5727
+    },
+    {
+      "epoch": 0.9881825239368585,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.025310770715108e-05,
+      "loss": 1.4428,
+      "step": 5728
+    },
+    {
+      "epoch": 0.9883550418355904,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.0250386690890493e-05,
+      "loss": 1.4245,
+      "step": 5729
+    },
+    {
+      "epoch": 0.9885275597343224,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.0247665656079692e-05,
+      "loss": 1.4714,
+      "step": 5730
+    },
+    {
+      "epoch": 0.9887000776330545,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.0244944602920277e-05,
+      "loss": 1.3732,
+      "step": 5731
+    },
+    {
+      "epoch": 0.9888725955317864,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.0242223531613834e-05,
+      "loss": 1.3682,
+      "step": 5732
+    },
+    {
+      "epoch": 0.9890451134305184,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.0239502442361955e-05,
+      "loss": 1.433,
+      "step": 5733
+    },
+    {
+      "epoch": 0.9892176313292504,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.0236781335366239e-05,
+      "loss": 1.4126,
+      "step": 5734
+    },
+    {
+      "epoch": 0.9893901492279824,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.0234060210828288e-05,
+      "loss": 1.3825,
+      "step": 5735
+    },
+    {
+      "epoch": 0.9895626671267144,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.0231339068949688e-05,
+      "loss": 1.4591,
+      "step": 5736
+    },
+    {
+      "epoch": 0.9897351850254464,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.0228617909932046e-05,
+      "loss": 1.3771,
+      "step": 5737
+    },
+    {
+      "epoch": 0.9899077029241784,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.0225896733976962e-05,
+      "loss": 1.4394,
+      "step": 5738
+    },
+    {
+      "epoch": 0.9900802208229104,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.0223175541286036e-05,
+      "loss": 1.4656,
+      "step": 5739
+    },
+    {
+      "epoch": 0.9902527387216423,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.0220454332060872e-05,
+      "loss": 1.5557,
+      "step": 5740
+    },
+    {
+      "epoch": 0.9904252566203744,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.0217733106503072e-05,
+      "loss": 1.5004,
+      "step": 5741
+    },
+    {
+      "epoch": 0.9905977745191064,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.0215011864814244e-05,
+      "loss": 1.4148,
+      "step": 5742
+    },
+    {
+      "epoch": 0.9907702924178383,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.0212290607195997e-05,
+      "loss": 1.4913,
+      "step": 5743
+    },
+    {
+      "epoch": 0.9909428103165704,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.020956933384993e-05,
+      "loss": 1.4065,
+      "step": 5744
+    },
+    {
+      "epoch": 0.9911153282153023,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.0206848044977658e-05,
+      "loss": 1.4257,
+      "step": 5745
+    },
+    {
+      "epoch": 0.9912878461140343,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.0204126740780791e-05,
+      "loss": 1.4675,
+      "step": 5746
+    },
+    {
+      "epoch": 0.9914603640127663,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.0201405421460942e-05,
+      "loss": 1.508,
+      "step": 5747
+    },
+    {
+      "epoch": 0.9916328819114983,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.0198684087219718e-05,
+      "loss": 1.472,
+      "step": 5748
+    },
+    {
+      "epoch": 0.9918053998102303,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.0195962738258736e-05,
+      "loss": 1.4882,
+      "step": 5749
+    },
+    {
+      "epoch": 0.9919779177089623,
+      "grad_norm": 0.625,
+      "learning_rate": 1.0193241374779607e-05,
+      "loss": 1.4459,
+      "step": 5750
+    },
+    {
+      "epoch": 0.9921504356076943,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.0190519996983956e-05,
+      "loss": 1.4201,
+      "step": 5751
+    },
+    {
+      "epoch": 0.9923229535064263,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.0187798605073389e-05,
+      "loss": 1.4262,
+      "step": 5752
+    },
+    {
+      "epoch": 0.9924954714051583,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.018507719924953e-05,
+      "loss": 1.3847,
+      "step": 5753
+    },
+    {
+      "epoch": 0.9926679893038903,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.0182355779713992e-05,
+      "loss": 1.4505,
+      "step": 5754
+    },
+    {
+      "epoch": 0.9928405072026223,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.0179634346668406e-05,
+      "loss": 1.4877,
+      "step": 5755
+    },
+    {
+      "epoch": 0.9930130251013543,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.0176912900314378e-05,
+      "loss": 1.4688,
+      "step": 5756
+    },
+    {
+      "epoch": 0.9931855430000862,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.0174191440853541e-05,
+      "loss": 1.5153,
+      "step": 5757
+    },
+    {
+      "epoch": 0.9933580608988183,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.017146996848751e-05,
+      "loss": 1.5,
+      "step": 5758
+    },
+    {
+      "epoch": 0.9935305787975502,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.0168748483417916e-05,
+      "loss": 1.4298,
+      "step": 5759
+    },
+    {
+      "epoch": 0.9937030966962822,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.0166026985846377e-05,
+      "loss": 1.4473,
+      "step": 5760
+    },
+    {
+      "epoch": 0.9938756145950143,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.0163305475974523e-05,
+      "loss": 1.3173,
+      "step": 5761
+    },
+    {
+      "epoch": 0.9940481324937462,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.0160583954003978e-05,
+      "loss": 1.4909,
+      "step": 5762
+    },
+    {
+      "epoch": 0.9942206503924782,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.0157862420136371e-05,
+      "loss": 1.4453,
+      "step": 5763
+    },
+    {
+      "epoch": 0.9943931682912102,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.015514087457333e-05,
+      "loss": 1.5272,
+      "step": 5764
+    },
+    {
+      "epoch": 0.9945656861899422,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.0152419317516482e-05,
+      "loss": 1.4821,
+      "step": 5765
+    },
+    {
+      "epoch": 0.9947382040886742,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.0149697749167459e-05,
+      "loss": 1.3855,
+      "step": 5766
+    },
+    {
+      "epoch": 0.9949107219874062,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.0146976169727893e-05,
+      "loss": 1.4995,
+      "step": 5767
+    },
+    {
+      "epoch": 0.9950832398861382,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.0144254579399413e-05,
+      "loss": 1.4256,
+      "step": 5768
+    },
+    {
+      "epoch": 0.9952557577848702,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.0141532978383653e-05,
+      "loss": 1.35,
+      "step": 5769
+    },
+    {
+      "epoch": 0.9954282756836021,
+      "grad_norm": 0.625,
+      "learning_rate": 1.0138811366882243e-05,
+      "loss": 1.4445,
+      "step": 5770
+    },
+    {
+      "epoch": 0.9956007935823342,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.0136089745096824e-05,
+      "loss": 1.4354,
+      "step": 5771
+    },
+    {
+      "epoch": 0.9957733114810662,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.0133368113229026e-05,
+      "loss": 1.3907,
+      "step": 5772
+    },
+    {
+      "epoch": 0.9959458293797981,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.013064647148048e-05,
+      "loss": 1.4249,
+      "step": 5773
+    },
+    {
+      "epoch": 0.9961183472785301,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.0127924820052831e-05,
+      "loss": 1.4933,
+      "step": 5774
+    },
+    {
+      "epoch": 0.9962908651772622,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.0125203159147712e-05,
+      "loss": 1.4064,
+      "step": 5775
+    },
+    {
+      "epoch": 0.9964633830759941,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.012248148896676e-05,
+      "loss": 1.3864,
+      "step": 5776
+    },
+    {
+      "epoch": 0.9966359009747261,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.0119759809711614e-05,
+      "loss": 1.3759,
+      "step": 5777
+    },
+    {
+      "epoch": 0.9968084188734582,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.0117038121583911e-05,
+      "loss": 1.4703,
+      "step": 5778
+    },
+    {
+      "epoch": 0.9969809367721901,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.0114316424785295e-05,
+      "loss": 1.3742,
+      "step": 5779
+    },
+    {
+      "epoch": 0.9971534546709221,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.0111594719517406e-05,
+      "loss": 1.4711,
+      "step": 5780
+    },
+    {
+      "epoch": 0.997325972569654,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.0108873005981876e-05,
+      "loss": 1.3791,
+      "step": 5781
+    },
+    {
+      "epoch": 0.9974984904683861,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.0106151284380359e-05,
+      "loss": 1.4406,
+      "step": 5782
+    },
+    {
+      "epoch": 0.9976710083671181,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.010342955491449e-05,
+      "loss": 1.3951,
+      "step": 5783
+    },
+    {
+      "epoch": 0.99784352626585,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.0100707817785915e-05,
+      "loss": 1.3135,
+      "step": 5784
+    },
+    {
+      "epoch": 0.9980160441645821,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.0097986073196273e-05,
+      "loss": 1.4802,
+      "step": 5785
+    },
+    {
+      "epoch": 0.9981885620633141,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.0095264321347212e-05,
+      "loss": 1.5621,
+      "step": 5786
+    },
+    {
+      "epoch": 0.998361079962046,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.0092542562440375e-05,
+      "loss": 1.3868,
+      "step": 5787
+    },
+    {
+      "epoch": 0.9985335978607781,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.0089820796677407e-05,
+      "loss": 1.4492,
+      "step": 5788
+    },
+    {
+      "epoch": 0.9987061157595101,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.0087099024259951e-05,
+      "loss": 1.3685,
+      "step": 5789
+    },
+    {
+      "epoch": 0.998878633658242,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.0084377245389656e-05,
+      "loss": 1.4912,
+      "step": 5790
+    },
+    {
+      "epoch": 0.999051151556974,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.008165546026817e-05,
+      "loss": 1.463,
+      "step": 5791
+    },
+    {
+      "epoch": 0.9992236694557061,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.0078933669097135e-05,
+      "loss": 1.4201,
+      "step": 5792
+    },
+    {
+      "epoch": 0.999396187354438,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.00762118720782e-05,
+      "loss": 1.5223,
+      "step": 5793
+    },
+    {
+      "epoch": 0.99956870525317,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.0073490069413014e-05,
+      "loss": 1.4212,
+      "step": 5794
+    },
+    {
+      "epoch": 0.9997412231519021,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.0070768261303226e-05,
+      "loss": 1.4884,
+      "step": 5795
+    },
+    {
+      "epoch": 0.999913741050634,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.006804644795048e-05,
+      "loss": 1.4197,
+      "step": 5796
+    },
+    {
+      "epoch": 1.000086258949366,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.006532462955643e-05,
+      "loss": 1.4797,
+      "step": 5797
+    },
+    {
+      "epoch": 1.000258776848098,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.006260280632272e-05,
+      "loss": 1.4291,
+      "step": 5798
+    },
+    {
+      "epoch": 1.00043129474683,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.0059880978451009e-05,
+      "loss": 1.4776,
+      "step": 5799
+    },
+    {
+      "epoch": 1.000603812645562,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.0057159146142937e-05,
+      "loss": 1.466,
+      "step": 5800
+    },
+    {
+      "epoch": 1.000603812645562,
+      "eval_loss": 1.4107141494750977,
+      "eval_runtime": 10.9354,
+      "eval_samples_per_second": 93.641,
+      "eval_steps_per_second": 23.41,
+      "step": 5800
+    },
+    {
+      "epoch": 1.000776330544294,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.0054437309600159e-05,
+      "loss": 1.5335,
+      "step": 5801
+    },
+    {
+      "epoch": 1.000948848443026,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.0051715469024325e-05,
+      "loss": 1.3836,
+      "step": 5802
+    },
+    {
+      "epoch": 1.001121366341758,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.0048993624617087e-05,
+      "loss": 1.5043,
+      "step": 5803
+    },
+    {
+      "epoch": 1.00129388424049,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.0046271776580094e-05,
+      "loss": 1.4079,
+      "step": 5804
+    },
+    {
+      "epoch": 1.001466402139222,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.0043549925115e-05,
+      "loss": 1.4617,
+      "step": 5805
+    },
+    {
+      "epoch": 1.0016389200379539,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.0040828070423451e-05,
+      "loss": 1.3933,
+      "step": 5806
+    },
+    {
+      "epoch": 1.001811437936686,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.0038106212707108e-05,
+      "loss": 1.4361,
+      "step": 5807
+    },
+    {
+      "epoch": 1.001983955835418,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.0035384352167619e-05,
+      "loss": 1.5073,
+      "step": 5808
+    },
+    {
+      "epoch": 1.00215647373415,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.0032662489006634e-05,
+      "loss": 1.3779,
+      "step": 5809
+    },
+    {
+      "epoch": 1.002328991632882,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.0029940623425807e-05,
+      "loss": 1.4819,
+      "step": 5810
+    },
+    {
+      "epoch": 1.0025015095316139,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.0027218755626793e-05,
+      "loss": 1.3348,
+      "step": 5811
+    },
+    {
+      "epoch": 1.0026740274303458,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.0024496885811245e-05,
+      "loss": 1.3513,
+      "step": 5812
+    },
+    {
+      "epoch": 1.0028465453290778,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.002177501418081e-05,
+      "loss": 1.4489,
+      "step": 5813
+    },
+    {
+      "epoch": 1.00301906322781,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.001905314093715e-05,
+      "loss": 1.4161,
+      "step": 5814
+    },
+    {
+      "epoch": 1.003191581126542,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.0016331266281913e-05,
+      "loss": 1.3771,
+      "step": 5815
+    },
+    {
+      "epoch": 1.003364099025274,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.0013609390416753e-05,
+      "loss": 1.2211,
+      "step": 5816
+    },
+    {
+      "epoch": 1.0035366169240059,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.0010887513543325e-05,
+      "loss": 1.5178,
+      "step": 5817
+    },
+    {
+      "epoch": 1.0037091348227378,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.0008165635863283e-05,
+      "loss": 1.5569,
+      "step": 5818
+    },
+    {
+      "epoch": 1.0038816527214698,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.0005443757578279e-05,
+      "loss": 1.3907,
+      "step": 5819
+    },
+    {
+      "epoch": 1.0040541706202017,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.0002721878889967e-05,
+      "loss": 1.4906,
+      "step": 5820
+    },
+    {
+      "epoch": 1.004226688518934,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1e-05,
+      "loss": 1.4273,
+      "step": 5821
+    },
+    {
+      "epoch": 1.004399206417666,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.997278121110033e-06,
+      "loss": 1.5296,
+      "step": 5822
+    },
+    {
+      "epoch": 1.0045717243163979,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.994556242421723e-06,
+      "loss": 1.3968,
+      "step": 5823
+    },
+    {
+      "epoch": 1.0047442422151298,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.991834364136719e-06,
+      "loss": 1.5477,
+      "step": 5824
+    },
+    {
+      "epoch": 1.0049167601138618,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.989112486456677e-06,
+      "loss": 1.5011,
+      "step": 5825
+    },
+    {
+      "epoch": 1.0050892780125937,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.986390609583246e-06,
+      "loss": 1.5186,
+      "step": 5826
+    },
+    {
+      "epoch": 1.0052617959113257,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.983668733718089e-06,
+      "loss": 1.5327,
+      "step": 5827
+    },
+    {
+      "epoch": 1.0054343138100579,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.980946859062852e-06,
+      "loss": 1.4618,
+      "step": 5828
+    },
+    {
+      "epoch": 1.0056068317087898,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.978224985819193e-06,
+      "loss": 1.3387,
+      "step": 5829
+    },
+    {
+      "epoch": 1.0057793496075218,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.975503114188758e-06,
+      "loss": 1.4009,
+      "step": 5830
+    },
+    {
+      "epoch": 1.0059518675062538,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.97278124437321e-06,
+      "loss": 1.3932,
+      "step": 5831
+    },
+    {
+      "epoch": 1.0061243854049857,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.970059376574195e-06,
+      "loss": 1.4174,
+      "step": 5832
+    },
+    {
+      "epoch": 1.0062969033037177,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.96733751099337e-06,
+      "loss": 1.4151,
+      "step": 5833
+    },
+    {
+      "epoch": 1.0064694212024499,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.964615647832384e-06,
+      "loss": 1.4401,
+      "step": 5834
+    },
+    {
+      "epoch": 1.0066419391011818,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.961893787292895e-06,
+      "loss": 1.4587,
+      "step": 5835
+    },
+    {
+      "epoch": 1.0068144569999138,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.95917192957655e-06,
+      "loss": 1.4193,
+      "step": 5836
+    },
+    {
+      "epoch": 1.0069869748986457,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.956450074885007e-06,
+      "loss": 1.4307,
+      "step": 5837
+    },
+    {
+      "epoch": 1.0071594927973777,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.953728223419908e-06,
+      "loss": 1.397,
+      "step": 5838
+    },
+    {
+      "epoch": 1.0073320106961097,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.951006375382915e-06,
+      "loss": 1.5258,
+      "step": 5839
+    },
+    {
+      "epoch": 1.0075045285948416,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.948284530975678e-06,
+      "loss": 1.3965,
+      "step": 5840
+    },
+    {
+      "epoch": 1.0076770464935738,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.945562690399841e-06,
+      "loss": 1.4365,
+      "step": 5841
+    },
+    {
+      "epoch": 1.0078495643923058,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.942840853857065e-06,
+      "loss": 1.5303,
+      "step": 5842
+    },
+    {
+      "epoch": 1.0080220822910377,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.940119021548994e-06,
+      "loss": 1.45,
+      "step": 5843
+    },
+    {
+      "epoch": 1.0081946001897697,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.937397193677281e-06,
+      "loss": 1.4291,
+      "step": 5844
+    },
+    {
+      "epoch": 1.0083671180885017,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.934675370443571e-06,
+      "loss": 1.3639,
+      "step": 5845
+    },
+    {
+      "epoch": 1.0085396359872336,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.931953552049524e-06,
+      "loss": 1.316,
+      "step": 5846
+    },
+    {
+      "epoch": 1.0087121538859656,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.929231738696779e-06,
+      "loss": 1.377,
+      "step": 5847
+    },
+    {
+      "epoch": 1.0088846717846978,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.926509930586991e-06,
+      "loss": 1.4196,
+      "step": 5848
+    },
+    {
+      "epoch": 1.0090571896834297,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.923788127921801e-06,
+      "loss": 1.386,
+      "step": 5849
+    },
+    {
+      "epoch": 1.0092297075821617,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.92106633090287e-06,
+      "loss": 1.3856,
+      "step": 5850
+    },
+    {
+      "epoch": 1.0094022254808936,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.918344539731832e-06,
+      "loss": 1.4346,
+      "step": 5851
+    },
+    {
+      "epoch": 1.0095747433796256,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.915622754610349e-06,
+      "loss": 1.4435,
+      "step": 5852
+    },
+    {
+      "epoch": 1.0097472612783576,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.912900975740052e-06,
+      "loss": 1.3264,
+      "step": 5853
+    },
+    {
+      "epoch": 1.0099197791770895,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.910179203322595e-06,
+      "loss": 1.3786,
+      "step": 5854
+    },
+    {
+      "epoch": 1.0100922970758217,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.907457437559626e-06,
+      "loss": 1.3746,
+      "step": 5855
+    },
+    {
+      "epoch": 1.0102648149745537,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.90473567865279e-06,
+      "loss": 1.5273,
+      "step": 5856
+    },
+    {
+      "epoch": 1.0104373328732856,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.90201392680373e-06,
+      "loss": 1.3937,
+      "step": 5857
+    },
+    {
+      "epoch": 1.0106098507720176,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.899292182214087e-06,
+      "loss": 1.4236,
+      "step": 5858
+    },
+    {
+      "epoch": 1.0107823686707496,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.896570445085511e-06,
+      "loss": 1.391,
+      "step": 5859
+    },
+    {
+      "epoch": 1.0109548865694815,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.893848715619643e-06,
+      "loss": 1.4368,
+      "step": 5860
+    },
+    {
+      "epoch": 1.0111274044682135,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.891126994018126e-06,
+      "loss": 1.4629,
+      "step": 5861
+    },
+    {
+      "epoch": 1.0112999223669457,
+      "grad_norm": 1.015625,
+      "learning_rate": 9.888405280482598e-06,
+      "loss": 1.4976,
+      "step": 5862
+    },
+    {
+      "epoch": 1.0114724402656776,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.885683575214709e-06,
+      "loss": 1.4451,
+      "step": 5863
+    },
+    {
+      "epoch": 1.0116449581644096,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.88296187841609e-06,
+      "loss": 1.4385,
+      "step": 5864
+    },
+    {
+      "epoch": 1.0118174760631415,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.880240190288391e-06,
+      "loss": 1.4685,
+      "step": 5865
+    },
+    {
+      "epoch": 1.0119899939618735,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.877518511033243e-06,
+      "loss": 1.4252,
+      "step": 5866
+    },
+    {
+      "epoch": 1.0121625118606055,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.874796840852292e-06,
+      "loss": 1.416,
+      "step": 5867
+    },
+    {
+      "epoch": 1.0123350297593374,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.87207517994717e-06,
+      "loss": 1.3435,
+      "step": 5868
+    },
+    {
+      "epoch": 1.0125075476580696,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.869353528519523e-06,
+      "loss": 1.4463,
+      "step": 5869
+    },
+    {
+      "epoch": 1.0126800655568016,
+      "grad_norm": 0.8203125,
+      "learning_rate": 9.866631886770979e-06,
+      "loss": 1.389,
+      "step": 5870
+    },
+    {
+      "epoch": 1.0128525834555335,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.86391025490318e-06,
+      "loss": 1.3701,
+      "step": 5871
+    },
+    {
+      "epoch": 1.0130251013542655,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.861188633117758e-06,
+      "loss": 1.4432,
+      "step": 5872
+    },
+    {
+      "epoch": 1.0131976192529975,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.858467021616349e-06,
+      "loss": 1.3332,
+      "step": 5873
+    },
+    {
+      "epoch": 1.0133701371517294,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.855745420600589e-06,
+      "loss": 1.3983,
+      "step": 5874
+    },
+    {
+      "epoch": 1.0135426550504616,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.85302383027211e-06,
+      "loss": 1.4667,
+      "step": 5875
+    },
+    {
+      "epoch": 1.0137151729491936,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.850302250832544e-06,
+      "loss": 1.3457,
+      "step": 5876
+    },
+    {
+      "epoch": 1.0138876908479255,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.84758068248352e-06,
+      "loss": 1.4847,
+      "step": 5877
+    },
+    {
+      "epoch": 1.0140602087466575,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.844859125426674e-06,
+      "loss": 1.4181,
+      "step": 5878
+    },
+    {
+      "epoch": 1.0142327266453894,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.842137579863632e-06,
+      "loss": 1.3819,
+      "step": 5879
+    },
+    {
+      "epoch": 1.0144052445441214,
+      "grad_norm": 0.76953125,
+      "learning_rate": 9.839416045996027e-06,
+      "loss": 1.4222,
+      "step": 5880
+    },
+    {
+      "epoch": 1.0145777624428534,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.83669452402548e-06,
+      "loss": 1.551,
+      "step": 5881
+    },
+    {
+      "epoch": 1.0147502803415855,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.833973014153628e-06,
+      "loss": 1.5052,
+      "step": 5882
+    },
+    {
+      "epoch": 1.0149227982403175,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.831251516582087e-06,
+      "loss": 1.4471,
+      "step": 5883
+    },
+    {
+      "epoch": 1.0150953161390495,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.828530031512493e-06,
+      "loss": 1.426,
+      "step": 5884
+    },
+    {
+      "epoch": 1.0152678340377814,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.825808559146464e-06,
+      "loss": 1.3441,
+      "step": 5885
+    },
+    {
+      "epoch": 1.0154403519365134,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.823087099685622e-06,
+      "loss": 1.48,
+      "step": 5886
+    },
+    {
+      "epoch": 1.0156128698352453,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.820365653331599e-06,
+      "loss": 1.5439,
+      "step": 5887
+    },
+    {
+      "epoch": 1.0157853877339773,
+      "grad_norm": 0.625,
+      "learning_rate": 9.817644220286006e-06,
+      "loss": 1.381,
+      "step": 5888
+    },
+    {
+      "epoch": 1.0159579056327095,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.814922800750474e-06,
+      "loss": 1.4234,
+      "step": 5889
+    },
+    {
+      "epoch": 1.0161304235314415,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.81220139492661e-06,
+      "loss": 1.5136,
+      "step": 5890
+    },
+    {
+      "epoch": 1.0163029414301734,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.809480003016046e-06,
+      "loss": 1.3156,
+      "step": 5891
+    },
+    {
+      "epoch": 1.0164754593289054,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.806758625220391e-06,
+      "loss": 1.3625,
+      "step": 5892
+    },
+    {
+      "epoch": 1.0166479772276373,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.804037261741269e-06,
+      "loss": 1.3784,
+      "step": 5893
+    },
+    {
+      "epoch": 1.0168204951263693,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.801315912780283e-06,
+      "loss": 1.4707,
+      "step": 5894
+    },
+    {
+      "epoch": 1.0169930130251013,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.798594578539063e-06,
+      "loss": 1.4436,
+      "step": 5895
+    },
+    {
+      "epoch": 1.0171655309238334,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.795873259219212e-06,
+      "loss": 1.4904,
+      "step": 5896
+    },
+    {
+      "epoch": 1.0173380488225654,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.793151955022346e-06,
+      "loss": 1.4208,
+      "step": 5897
+    },
+    {
+      "epoch": 1.0175105667212974,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.790430666150073e-06,
+      "loss": 1.4334,
+      "step": 5898
+    },
+    {
+      "epoch": 1.0176830846200293,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.78770939280401e-06,
+      "loss": 1.4077,
+      "step": 5899
+    },
+    {
+      "epoch": 1.0178556025187613,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.78498813518576e-06,
+      "loss": 1.425,
+      "step": 5900
+    },
+    {
+      "epoch": 1.0178556025187613,
+      "eval_loss": 1.410369634628296,
+      "eval_runtime": 10.8142,
+      "eval_samples_per_second": 94.69,
+      "eval_steps_per_second": 23.673,
+      "step": 5900
+    },
+    {
+      "epoch": 1.0180281204174932,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.782266893496933e-06,
+      "loss": 1.3682,
+      "step": 5901
+    },
+    {
+      "epoch": 1.0182006383162252,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.779545667939132e-06,
+      "loss": 1.4094,
+      "step": 5902
+    },
+    {
+      "epoch": 1.0183731562149574,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.776824458713965e-06,
+      "loss": 1.4804,
+      "step": 5903
+    },
+    {
+      "epoch": 1.0185456741136893,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.774103266023042e-06,
+      "loss": 1.4789,
+      "step": 5904
+    },
+    {
+      "epoch": 1.0187181920124213,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.771382090067954e-06,
+      "loss": 1.4821,
+      "step": 5905
+    },
+    {
+      "epoch": 1.0188907099111533,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.768660931050314e-06,
+      "loss": 1.2989,
+      "step": 5906
+    },
+    {
+      "epoch": 1.0190632278098852,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.765939789171717e-06,
+      "loss": 1.456,
+      "step": 5907
+    },
+    {
+      "epoch": 1.0192357457086172,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.763218664633763e-06,
+      "loss": 1.4041,
+      "step": 5908
+    },
+    {
+      "epoch": 1.0194082636073492,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.760497557638047e-06,
+      "loss": 1.5247,
+      "step": 5909
+    },
+    {
+      "epoch": 1.0195807815060813,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.757776468386171e-06,
+      "loss": 1.3292,
+      "step": 5910
+    },
+    {
+      "epoch": 1.0197532994048133,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.755055397079724e-06,
+      "loss": 1.5492,
+      "step": 5911
+    },
+    {
+      "epoch": 1.0199258173035453,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.75233434392031e-06,
+      "loss": 1.3097,
+      "step": 5912
+    },
+    {
+      "epoch": 1.0200983352022772,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.749613309109512e-06,
+      "loss": 1.3986,
+      "step": 5913
+    },
+    {
+      "epoch": 1.0202708531010092,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.746892292848925e-06,
+      "loss": 1.4641,
+      "step": 5914
+    },
+    {
+      "epoch": 1.0204433709997411,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.744171295340136e-06,
+      "loss": 1.5802,
+      "step": 5915
+    },
+    {
+      "epoch": 1.0206158888984733,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.74145031678474e-06,
+      "loss": 1.4543,
+      "step": 5916
+    },
+    {
+      "epoch": 1.0207884067972053,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.738729357384318e-06,
+      "loss": 1.4212,
+      "step": 5917
+    },
+    {
+      "epoch": 1.0209609246959372,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.736008417340455e-06,
+      "loss": 1.3583,
+      "step": 5918
+    },
+    {
+      "epoch": 1.0211334425946692,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.733287496854743e-06,
+      "loss": 1.5124,
+      "step": 5919
+    },
+    {
+      "epoch": 1.0213059604934012,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.730566596128756e-06,
+      "loss": 1.4214,
+      "step": 5920
+    },
+    {
+      "epoch": 1.0214784783921331,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.727845715364081e-06,
+      "loss": 1.4619,
+      "step": 5921
+    },
+    {
+      "epoch": 1.021650996290865,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.725124854762294e-06,
+      "loss": 1.329,
+      "step": 5922
+    },
+    {
+      "epoch": 1.0218235141895973,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.722404014524978e-06,
+      "loss": 1.4048,
+      "step": 5923
+    },
+    {
+      "epoch": 1.0219960320883292,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.719683194853705e-06,
+      "loss": 1.3779,
+      "step": 5924
+    },
+    {
+      "epoch": 1.0221685499870612,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.716962395950055e-06,
+      "loss": 1.3623,
+      "step": 5925
+    },
+    {
+      "epoch": 1.0223410678857932,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.714241618015596e-06,
+      "loss": 1.3491,
+      "step": 5926
+    },
+    {
+      "epoch": 1.0225135857845251,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.711520861251904e-06,
+      "loss": 1.4558,
+      "step": 5927
+    },
+    {
+      "epoch": 1.022686103683257,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.708800125860552e-06,
+      "loss": 1.3601,
+      "step": 5928
+    },
+    {
+      "epoch": 1.022858621581989,
+      "grad_norm": 1.0546875,
+      "learning_rate": 9.706079412043105e-06,
+      "loss": 1.4023,
+      "step": 5929
+    },
+    {
+      "epoch": 1.0230311394807212,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.703358720001129e-06,
+      "loss": 1.5388,
+      "step": 5930
+    },
+    {
+      "epoch": 1.0232036573794532,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.700638049936194e-06,
+      "loss": 1.4385,
+      "step": 5931
+    },
+    {
+      "epoch": 1.0233761752781851,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.697917402049864e-06,
+      "loss": 1.5253,
+      "step": 5932
+    },
+    {
+      "epoch": 1.023548693176917,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.695196776543701e-06,
+      "loss": 1.4444,
+      "step": 5933
+    },
+    {
+      "epoch": 1.023721211075649,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.69247617361926e-06,
+      "loss": 1.3254,
+      "step": 5934
+    },
+    {
+      "epoch": 1.023893728974381,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.689755593478109e-06,
+      "loss": 1.3835,
+      "step": 5935
+    },
+    {
+      "epoch": 1.024066246873113,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.687035036321804e-06,
+      "loss": 1.3976,
+      "step": 5936
+    },
+    {
+      "epoch": 1.0242387647718452,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.684314502351895e-06,
+      "loss": 1.5053,
+      "step": 5937
+    },
+    {
+      "epoch": 1.0244112826705771,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.681593991769944e-06,
+      "loss": 1.423,
+      "step": 5938
+    },
+    {
+      "epoch": 1.024583800569309,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.678873504777495e-06,
+      "loss": 1.4018,
+      "step": 5939
+    },
+    {
+      "epoch": 1.024756318468041,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.676153041576111e-06,
+      "loss": 1.4337,
+      "step": 5940
+    },
+    {
+      "epoch": 1.024928836366773,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.673432602367325e-06,
+      "loss": 1.4924,
+      "step": 5941
+    },
+    {
+      "epoch": 1.025101354265505,
+      "grad_norm": 0.7890625,
+      "learning_rate": 9.670712187352698e-06,
+      "loss": 1.4626,
+      "step": 5942
+    },
+    {
+      "epoch": 1.025273872164237,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.667991796733764e-06,
+      "loss": 1.5558,
+      "step": 5943
+    },
+    {
+      "epoch": 1.0254463900629691,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.665271430712079e-06,
+      "loss": 1.4,
+      "step": 5944
+    },
+    {
+      "epoch": 1.025618907961701,
+      "grad_norm": 1.359375,
+      "learning_rate": 9.662551089489174e-06,
+      "loss": 1.474,
+      "step": 5945
+    },
+    {
+      "epoch": 1.025791425860433,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.659830773266593e-06,
+      "loss": 1.4282,
+      "step": 5946
+    },
+    {
+      "epoch": 1.025963943759165,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.657110482245872e-06,
+      "loss": 1.4422,
+      "step": 5947
+    },
+    {
+      "epoch": 1.026136461657897,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.654390216628554e-06,
+      "loss": 1.3946,
+      "step": 5948
+    },
+    {
+      "epoch": 1.026308979556629,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.651669976616165e-06,
+      "loss": 1.547,
+      "step": 5949
+    },
+    {
+      "epoch": 1.026481497455361,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.648949762410238e-06,
+      "loss": 1.449,
+      "step": 5950
+    },
+    {
+      "epoch": 1.026654015354093,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.64622957421231e-06,
+      "loss": 1.4005,
+      "step": 5951
+    },
+    {
+      "epoch": 1.026826533252825,
+      "grad_norm": 0.75,
+      "learning_rate": 9.643509412223905e-06,
+      "loss": 1.3911,
+      "step": 5952
+    },
+    {
+      "epoch": 1.026999051151557,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.64078927664655e-06,
+      "loss": 1.5121,
+      "step": 5953
+    },
+    {
+      "epoch": 1.027171569050289,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.638069167681768e-06,
+      "loss": 1.3723,
+      "step": 5954
+    },
+    {
+      "epoch": 1.027344086949021,
+      "grad_norm": 0.703125,
+      "learning_rate": 9.635349085531088e-06,
+      "loss": 1.3944,
+      "step": 5955
+    },
+    {
+      "epoch": 1.0275166048477529,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.63262903039602e-06,
+      "loss": 1.4605,
+      "step": 5956
+    },
+    {
+      "epoch": 1.027689122746485,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.629909002478095e-06,
+      "loss": 1.5529,
+      "step": 5957
+    },
+    {
+      "epoch": 1.027861640645217,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.627189001978818e-06,
+      "loss": 1.5283,
+      "step": 5958
+    },
+    {
+      "epoch": 1.028034158543949,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.624469029099713e-06,
+      "loss": 1.4798,
+      "step": 5959
+    },
+    {
+      "epoch": 1.028206676442681,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.621749084042288e-06,
+      "loss": 1.4375,
+      "step": 5960
+    },
+    {
+      "epoch": 1.028379194341413,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.619029167008056e-06,
+      "loss": 1.4525,
+      "step": 5961
+    },
+    {
+      "epoch": 1.0285517122401449,
+      "grad_norm": 0.7265625,
+      "learning_rate": 9.61630927819852e-06,
+      "loss": 1.4167,
+      "step": 5962
+    },
+    {
+      "epoch": 1.0287242301388768,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.613589417815192e-06,
+      "loss": 1.4118,
+      "step": 5963
+    },
+    {
+      "epoch": 1.028896748037609,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.610869586059574e-06,
+      "loss": 1.4564,
+      "step": 5964
+    },
+    {
+      "epoch": 1.029069265936341,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.608149783133169e-06,
+      "loss": 1.4491,
+      "step": 5965
+    },
+    {
+      "epoch": 1.029241783835073,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.605430009237474e-06,
+      "loss": 1.4347,
+      "step": 5966
+    },
+    {
+      "epoch": 1.0294143017338049,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.602710264573991e-06,
+      "loss": 1.4235,
+      "step": 5967
+    },
+    {
+      "epoch": 1.0295868196325368,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.599990549344216e-06,
+      "loss": 1.3666,
+      "step": 5968
+    },
+    {
+      "epoch": 1.0297593375312688,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.597270863749635e-06,
+      "loss": 1.4757,
+      "step": 5969
+    },
+    {
+      "epoch": 1.0299318554300008,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.59455120799175e-06,
+      "loss": 1.4431,
+      "step": 5970
+    },
+    {
+      "epoch": 1.030104373328733,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.59183158227204e-06,
+      "loss": 1.4592,
+      "step": 5971
+    },
+    {
+      "epoch": 1.030276891227465,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.589111986792004e-06,
+      "loss": 1.5002,
+      "step": 5972
+    },
+    {
+      "epoch": 1.0304494091261969,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.586392421753115e-06,
+      "loss": 1.4541,
+      "step": 5973
+    },
+    {
+      "epoch": 1.0306219270249288,
+      "grad_norm": 1.0625,
+      "learning_rate": 9.583672887356862e-06,
+      "loss": 1.3925,
+      "step": 5974
+    },
+    {
+      "epoch": 1.0307944449236608,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.580953383804718e-06,
+      "loss": 1.5073,
+      "step": 5975
+    },
+    {
+      "epoch": 1.0309669628223928,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.57823391129817e-06,
+      "loss": 1.4615,
+      "step": 5976
+    },
+    {
+      "epoch": 1.0311394807211247,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.575514470038688e-06,
+      "loss": 1.525,
+      "step": 5977
+    },
+    {
+      "epoch": 1.031311998619857,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.572795060227748e-06,
+      "loss": 1.4352,
+      "step": 5978
+    },
+    {
+      "epoch": 1.0314845165185889,
+      "grad_norm": 0.765625,
+      "learning_rate": 9.570075682066815e-06,
+      "loss": 1.4863,
+      "step": 5979
+    },
+    {
+      "epoch": 1.0316570344173208,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.567356335757366e-06,
+      "loss": 1.4522,
+      "step": 5980
+    },
+    {
+      "epoch": 1.0318295523160528,
+      "grad_norm": 0.66015625,
+      "learning_rate": 9.56463702150086e-06,
+      "loss": 1.4746,
+      "step": 5981
+    },
+    {
+      "epoch": 1.0320020702147847,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.561917739498761e-06,
+      "loss": 1.4016,
+      "step": 5982
+    },
+    {
+      "epoch": 1.0321745881135167,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.55919848995254e-06,
+      "loss": 1.4406,
+      "step": 5983
+    },
+    {
+      "epoch": 1.0323471060122489,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.556479273063644e-06,
+      "loss": 1.4135,
+      "step": 5984
+    },
+    {
+      "epoch": 1.0325196239109808,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.55376008903354e-06,
+      "loss": 1.4579,
+      "step": 5985
+    },
+    {
+      "epoch": 1.0326921418097128,
+      "grad_norm": 0.625,
+      "learning_rate": 9.551040938063671e-06,
+      "loss": 1.4233,
+      "step": 5986
+    },
+    {
+      "epoch": 1.0328646597084448,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.548321820355498e-06,
+      "loss": 1.4576,
+      "step": 5987
+    },
+    {
+      "epoch": 1.0330371776071767,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.545602736110467e-06,
+      "loss": 1.4506,
+      "step": 5988
+    },
+    {
+      "epoch": 1.0332096955059087,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.542883685530024e-06,
+      "loss": 1.4305,
+      "step": 5989
+    },
+    {
+      "epoch": 1.0333822134046406,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.54016466881561e-06,
+      "loss": 1.4593,
+      "step": 5990
+    },
+    {
+      "epoch": 1.0335547313033728,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.537445686168676e-06,
+      "loss": 1.3949,
+      "step": 5991
+    },
+    {
+      "epoch": 1.0337272492021048,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.534726737790652e-06,
+      "loss": 1.3817,
+      "step": 5992
+    },
+    {
+      "epoch": 1.0338997671008368,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.532007823882983e-06,
+      "loss": 1.3964,
+      "step": 5993
+    },
+    {
+      "epoch": 1.0340722849995687,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.529288944647093e-06,
+      "loss": 1.4328,
+      "step": 5994
+    },
+    {
+      "epoch": 1.0342448028983007,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.526570100284423e-06,
+      "loss": 1.4962,
+      "step": 5995
+    },
+    {
+      "epoch": 1.0344173207970326,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.523851290996397e-06,
+      "loss": 1.4092,
+      "step": 5996
+    },
+    {
+      "epoch": 1.0345898386957646,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.521132516984442e-06,
+      "loss": 1.3904,
+      "step": 5997
+    },
+    {
+      "epoch": 1.0347623565944968,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.518413778449981e-06,
+      "loss": 1.3421,
+      "step": 5998
+    },
+    {
+      "epoch": 1.0349348744932287,
+      "grad_norm": 0.859375,
+      "learning_rate": 9.515695075594434e-06,
+      "loss": 1.3831,
+      "step": 5999
+    },
+    {
+      "epoch": 1.0351073923919607,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.512976408619227e-06,
+      "loss": 1.4474,
+      "step": 6000
+    },
+    {
+      "epoch": 1.0351073923919607,
+      "eval_loss": 1.4101243019104004,
+      "eval_runtime": 10.873,
+      "eval_samples_per_second": 94.178,
+      "eval_steps_per_second": 23.545,
+      "step": 6000
+    },
+    {
+      "epoch": 1.0352799102906927,
+      "grad_norm": 0.86328125,
+      "learning_rate": 9.510257777725769e-06,
+      "loss": 1.3563,
+      "step": 6001
+    },
+    {
+      "epoch": 1.0354524281894246,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.507539183115479e-06,
+      "loss": 1.3893,
+      "step": 6002
+    },
+    {
+      "epoch": 1.0356249460881566,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.504820624989756e-06,
+      "loss": 1.3792,
+      "step": 6003
+    },
+    {
+      "epoch": 1.0357974639868885,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.50210210355002e-06,
+      "loss": 1.5117,
+      "step": 6004
+    },
+    {
+      "epoch": 1.0359699818856207,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.499383618997669e-06,
+      "loss": 1.3749,
+      "step": 6005
+    },
+    {
+      "epoch": 1.0361424997843527,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.49666517153411e-06,
+      "loss": 1.5192,
+      "step": 6006
+    },
+    {
+      "epoch": 1.0363150176830846,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.493946761360736e-06,
+      "loss": 1.4583,
+      "step": 6007
+    },
+    {
+      "epoch": 1.0364875355818166,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.491228388678952e-06,
+      "loss": 1.4411,
+      "step": 6008
+    },
+    {
+      "epoch": 1.0366600534805486,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.488510053690147e-06,
+      "loss": 1.4219,
+      "step": 6009
+    },
+    {
+      "epoch": 1.0368325713792805,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.485791756595714e-06,
+      "loss": 1.4116,
+      "step": 6010
+    },
+    {
+      "epoch": 1.0370050892780125,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.483073497597037e-06,
+      "loss": 1.4544,
+      "step": 6011
+    },
+    {
+      "epoch": 1.0371776071767447,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.480355276895508e-06,
+      "loss": 1.381,
+      "step": 6012
+    },
+    {
+      "epoch": 1.0373501250754766,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.477637094692505e-06,
+      "loss": 1.5196,
+      "step": 6013
+    },
+    {
+      "epoch": 1.0375226429742086,
+      "grad_norm": 0.734375,
+      "learning_rate": 9.47491895118941e-06,
+      "loss": 1.4163,
+      "step": 6014
+    },
+    {
+      "epoch": 1.0376951608729406,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.472200846587603e-06,
+      "loss": 1.4176,
+      "step": 6015
+    },
+    {
+      "epoch": 1.0378676787716725,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.469482781088451e-06,
+      "loss": 1.435,
+      "step": 6016
+    },
+    {
+      "epoch": 1.0380401966704045,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.466764754893334e-06,
+      "loss": 1.4506,
+      "step": 6017
+    },
+    {
+      "epoch": 1.0382127145691364,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.464046768203611e-06,
+      "loss": 1.3685,
+      "step": 6018
+    },
+    {
+      "epoch": 1.0383852324678686,
+      "grad_norm": 0.81640625,
+      "learning_rate": 9.461328821220657e-06,
+      "loss": 1.4066,
+      "step": 6019
+    },
+    {
+      "epoch": 1.0385577503666006,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.458610914145826e-06,
+      "loss": 1.3749,
+      "step": 6020
+    },
+    {
+      "epoch": 1.0387302682653325,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.455893047180485e-06,
+      "loss": 1.3851,
+      "step": 6021
+    },
+    {
+      "epoch": 1.0389027861640645,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.453175220525981e-06,
+      "loss": 1.4537,
+      "step": 6022
+    },
+    {
+      "epoch": 1.0390753040627965,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.450457434383679e-06,
+      "loss": 1.3407,
+      "step": 6023
+    },
+    {
+      "epoch": 1.0392478219615284,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.44773968895492e-06,
+      "loss": 1.4205,
+      "step": 6024
+    },
+    {
+      "epoch": 1.0394203398602606,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.445021984441059e-06,
+      "loss": 1.4163,
+      "step": 6025
+    },
+    {
+      "epoch": 1.0395928577589926,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.442304321043433e-06,
+      "loss": 1.4961,
+      "step": 6026
+    },
+    {
+      "epoch": 1.0397653756577245,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.43958669896339e-06,
+      "loss": 1.3901,
+      "step": 6027
+    },
+    {
+      "epoch": 1.0399378935564565,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.436869118402265e-06,
+      "loss": 1.4272,
+      "step": 6028
+    },
+    {
+      "epoch": 1.0401104114551885,
+      "grad_norm": 0.68359375,
+      "learning_rate": 9.434151579561397e-06,
+      "loss": 1.4163,
+      "step": 6029
+    },
+    {
+      "epoch": 1.0402829293539204,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.431434082642112e-06,
+      "loss": 1.5161,
+      "step": 6030
+    },
+    {
+      "epoch": 1.0404554472526524,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.428716627845742e-06,
+      "loss": 1.3931,
+      "step": 6031
+    },
+    {
+      "epoch": 1.0406279651513846,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.425999215373617e-06,
+      "loss": 1.4832,
+      "step": 6032
+    },
+    {
+      "epoch": 1.0408004830501165,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.423281845427056e-06,
+      "loss": 1.4823,
+      "step": 6033
+    },
+    {
+      "epoch": 1.0409730009488485,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.420564518207382e-06,
+      "loss": 1.4338,
+      "step": 6034
+    },
+    {
+      "epoch": 1.0411455188475804,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.417847233915902e-06,
+      "loss": 1.5086,
+      "step": 6035
+    },
+    {
+      "epoch": 1.0413180367463124,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.415129992753943e-06,
+      "loss": 1.4712,
+      "step": 6036
+    },
+    {
+      "epoch": 1.0414905546450444,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.412412794922805e-06,
+      "loss": 1.4372,
+      "step": 6037
+    },
+    {
+      "epoch": 1.0416630725437763,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.409695640623802e-06,
+      "loss": 1.3527,
+      "step": 6038
+    },
+    {
+      "epoch": 1.0418355904425085,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.406978530058229e-06,
+      "loss": 1.4659,
+      "step": 6039
+    },
+    {
+      "epoch": 1.0420081083412405,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.404261463427396e-06,
+      "loss": 1.3717,
+      "step": 6040
+    },
+    {
+      "epoch": 1.0421806262399724,
+      "grad_norm": 0.91015625,
+      "learning_rate": 9.401544440932596e-06,
+      "loss": 1.3971,
+      "step": 6041
+    },
+    {
+      "epoch": 1.0423531441387044,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.398827462775122e-06,
+      "loss": 1.408,
+      "step": 6042
+    },
+    {
+      "epoch": 1.0425256620374364,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.396110529156263e-06,
+      "loss": 1.4351,
+      "step": 6043
+    },
+    {
+      "epoch": 1.0426981799361683,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.393393640277316e-06,
+      "loss": 1.4595,
+      "step": 6044
+    },
+    {
+      "epoch": 1.0428706978349003,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.390676796339552e-06,
+      "loss": 1.4187,
+      "step": 6045
+    },
+    {
+      "epoch": 1.0430432157336325,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.387959997544263e-06,
+      "loss": 1.3626,
+      "step": 6046
+    },
+    {
+      "epoch": 1.0432157336323644,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.385243244092717e-06,
+      "loss": 1.6359,
+      "step": 6047
+    },
+    {
+      "epoch": 1.0433882515310964,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.382526536186194e-06,
+      "loss": 1.3945,
+      "step": 6048
+    },
+    {
+      "epoch": 1.0435607694298283,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.379809874025967e-06,
+      "loss": 1.5335,
+      "step": 6049
+    },
+    {
+      "epoch": 1.0437332873285603,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.377093257813293e-06,
+      "loss": 1.3544,
+      "step": 6050
+    },
+    {
+      "epoch": 1.0439058052272923,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.374376687749449e-06,
+      "loss": 1.4395,
+      "step": 6051
+    },
+    {
+      "epoch": 1.0440783231260242,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.371660164035687e-06,
+      "loss": 1.4785,
+      "step": 6052
+    },
+    {
+      "epoch": 1.0442508410247564,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.368943686873266e-06,
+      "loss": 1.4134,
+      "step": 6053
+    },
+    {
+      "epoch": 1.0444233589234884,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.366227256463437e-06,
+      "loss": 1.4454,
+      "step": 6054
+    },
+    {
+      "epoch": 1.0445958768222203,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.363510873007458e-06,
+      "loss": 1.4406,
+      "step": 6055
+    },
+    {
+      "epoch": 1.0447683947209523,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.360794536706566e-06,
+      "loss": 1.5395,
+      "step": 6056
+    },
+    {
+      "epoch": 1.0449409126196842,
+      "grad_norm": 0.7109375,
+      "learning_rate": 9.358078247762012e-06,
+      "loss": 1.4343,
+      "step": 6057
+    },
+    {
+      "epoch": 1.0451134305184162,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.355362006375029e-06,
+      "loss": 1.4209,
+      "step": 6058
+    },
+    {
+      "epoch": 1.0452859484171482,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.35264581274686e-06,
+      "loss": 1.4438,
+      "step": 6059
+    },
+    {
+      "epoch": 1.0454584663158804,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.349929667078729e-06,
+      "loss": 1.4041,
+      "step": 6060
+    },
+    {
+      "epoch": 1.0456309842146123,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.347213569571876e-06,
+      "loss": 1.4805,
+      "step": 6061
+    },
+    {
+      "epoch": 1.0458035021133443,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.344497520427517e-06,
+      "loss": 1.4596,
+      "step": 6062
+    },
+    {
+      "epoch": 1.0459760200120762,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.341781519846875e-06,
+      "loss": 1.4914,
+      "step": 6063
+    },
+    {
+      "epoch": 1.0461485379108082,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.339065568031176e-06,
+      "loss": 1.4308,
+      "step": 6064
+    },
+    {
+      "epoch": 1.0463210558095402,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.336349665181628e-06,
+      "loss": 1.3531,
+      "step": 6065
+    },
+    {
+      "epoch": 1.0464935737082723,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.333633811499444e-06,
+      "loss": 1.4219,
+      "step": 6066
+    },
+    {
+      "epoch": 1.0466660916070043,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.330918007185828e-06,
+      "loss": 1.4653,
+      "step": 6067
+    },
+    {
+      "epoch": 1.0468386095057363,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.328202252441989e-06,
+      "loss": 1.332,
+      "step": 6068
+    },
+    {
+      "epoch": 1.0470111274044682,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.325486547469124e-06,
+      "loss": 1.3493,
+      "step": 6069
+    },
+    {
+      "epoch": 1.0471836453032002,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.322770892468433e-06,
+      "loss": 1.4375,
+      "step": 6070
+    },
+    {
+      "epoch": 1.0473561632019321,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.3200552876411e-06,
+      "loss": 1.3934,
+      "step": 6071
+    },
+    {
+      "epoch": 1.047528681100664,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.317339733188324e-06,
+      "loss": 1.4507,
+      "step": 6072
+    },
+    {
+      "epoch": 1.0477011989993963,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.314624229311285e-06,
+      "loss": 1.4125,
+      "step": 6073
+    },
+    {
+      "epoch": 1.0478737168981282,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.311908776211167e-06,
+      "loss": 1.4712,
+      "step": 6074
+    },
+    {
+      "epoch": 1.0480462347968602,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.309193374089141e-06,
+      "loss": 1.4621,
+      "step": 6075
+    },
+    {
+      "epoch": 1.0482187526955922,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.306478023146393e-06,
+      "loss": 1.5664,
+      "step": 6076
+    },
+    {
+      "epoch": 1.0483912705943241,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.303762723584082e-06,
+      "loss": 1.4725,
+      "step": 6077
+    },
+    {
+      "epoch": 1.048563788493056,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.301047475603382e-06,
+      "loss": 1.3968,
+      "step": 6078
+    },
+    {
+      "epoch": 1.048736306391788,
+      "grad_norm": 0.75390625,
+      "learning_rate": 9.298332279405447e-06,
+      "loss": 1.4459,
+      "step": 6079
+    },
+    {
+      "epoch": 1.0489088242905202,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.295617135191445e-06,
+      "loss": 1.353,
+      "step": 6080
+    },
+    {
+      "epoch": 1.0490813421892522,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.292902043162526e-06,
+      "loss": 1.428,
+      "step": 6081
+    },
+    {
+      "epoch": 1.0492538600879842,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.290187003519841e-06,
+      "loss": 1.3785,
+      "step": 6082
+    },
+    {
+      "epoch": 1.0494263779867161,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.28747201646454e-06,
+      "loss": 1.4798,
+      "step": 6083
+    },
+    {
+      "epoch": 1.049598895885448,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.284757082197763e-06,
+      "loss": 1.3638,
+      "step": 6084
+    },
+    {
+      "epoch": 1.04977141378418,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.282042200920652e-06,
+      "loss": 1.3908,
+      "step": 6085
+    },
+    {
+      "epoch": 1.049943931682912,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.279327372834335e-06,
+      "loss": 1.3872,
+      "step": 6086
+    },
+    {
+      "epoch": 1.0501164495816442,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.276612598139956e-06,
+      "loss": 1.5071,
+      "step": 6087
+    },
+    {
+      "epoch": 1.0502889674803761,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.27389787703863e-06,
+      "loss": 1.454,
+      "step": 6088
+    },
+    {
+      "epoch": 1.050461485379108,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.27118320973149e-06,
+      "loss": 1.4289,
+      "step": 6089
+    },
+    {
+      "epoch": 1.05063400327784,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.268468596419651e-06,
+      "loss": 1.473,
+      "step": 6090
+    },
+    {
+      "epoch": 1.050806521176572,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.265754037304232e-06,
+      "loss": 1.4485,
+      "step": 6091
+    },
+    {
+      "epoch": 1.050979039075304,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.263039532586336e-06,
+      "loss": 1.4458,
+      "step": 6092
+    },
+    {
+      "epoch": 1.051151556974036,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.26032508246708e-06,
+      "loss": 1.3861,
+      "step": 6093
+    },
+    {
+      "epoch": 1.0513240748727681,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.257610687147562e-06,
+      "loss": 1.4952,
+      "step": 6094
+    },
+    {
+      "epoch": 1.0514965927715,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.25489634682888e-06,
+      "loss": 1.419,
+      "step": 6095
+    },
+    {
+      "epoch": 1.051669110670232,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.252182061712137e-06,
+      "loss": 1.3722,
+      "step": 6096
+    },
+    {
+      "epoch": 1.051841628568964,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.249467831998417e-06,
+      "loss": 1.4752,
+      "step": 6097
+    },
+    {
+      "epoch": 1.052014146467696,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.246753657888814e-06,
+      "loss": 1.4241,
+      "step": 6098
+    },
+    {
+      "epoch": 1.052186664366428,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.244039539584399e-06,
+      "loss": 1.3496,
+      "step": 6099
+    },
+    {
+      "epoch": 1.05235918226516,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.241325477286265e-06,
+      "loss": 1.4379,
+      "step": 6100
+    },
+    {
+      "epoch": 1.05235918226516,
+      "eval_loss": 1.4098589420318604,
+      "eval_runtime": 10.8253,
+      "eval_samples_per_second": 94.593,
+      "eval_steps_per_second": 23.648,
+      "step": 6100
+    },
+    {
+      "epoch": 1.052531700163892,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.238611471195477e-06,
+      "loss": 1.3604,
+      "step": 6101
+    },
+    {
+      "epoch": 1.052704218062624,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.235897521513111e-06,
+      "loss": 1.401,
+      "step": 6102
+    },
+    {
+      "epoch": 1.052876735961356,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.233183628440227e-06,
+      "loss": 1.4434,
+      "step": 6103
+    },
+    {
+      "epoch": 1.053049253860088,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.230469792177894e-06,
+      "loss": 1.3778,
+      "step": 6104
+    },
+    {
+      "epoch": 1.05322177175882,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.227756012927166e-06,
+      "loss": 1.3925,
+      "step": 6105
+    },
+    {
+      "epoch": 1.0533942896575519,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.2250422908891e-06,
+      "loss": 1.4522,
+      "step": 6106
+    },
+    {
+      "epoch": 1.053566807556284,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.222328626264738e-06,
+      "loss": 1.3521,
+      "step": 6107
+    },
+    {
+      "epoch": 1.053739325455016,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.219615019255136e-06,
+      "loss": 1.4829,
+      "step": 6108
+    },
+    {
+      "epoch": 1.053911843353748,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.216901470061326e-06,
+      "loss": 1.3929,
+      "step": 6109
+    },
+    {
+      "epoch": 1.05408436125248,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.21418797888435e-06,
+      "loss": 1.4325,
+      "step": 6110
+    },
+    {
+      "epoch": 1.054256879151212,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.211474545925237e-06,
+      "loss": 1.4691,
+      "step": 6111
+    },
+    {
+      "epoch": 1.0544293970499439,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.208761171385017e-06,
+      "loss": 1.4809,
+      "step": 6112
+    },
+    {
+      "epoch": 1.0546019149486758,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.206047855464715e-06,
+      "loss": 1.4147,
+      "step": 6113
+    },
+    {
+      "epoch": 1.054774432847408,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.203334598365345e-06,
+      "loss": 1.3983,
+      "step": 6114
+    },
+    {
+      "epoch": 1.05494695074614,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.200621400287929e-06,
+      "loss": 1.4555,
+      "step": 6115
+    },
+    {
+      "epoch": 1.055119468644872,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.19790826143347e-06,
+      "loss": 1.4068,
+      "step": 6116
+    },
+    {
+      "epoch": 1.055291986543604,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.195195182002987e-06,
+      "loss": 1.381,
+      "step": 6117
+    },
+    {
+      "epoch": 1.0554645044423359,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.192482162197466e-06,
+      "loss": 1.4942,
+      "step": 6118
+    },
+    {
+      "epoch": 1.0556370223410678,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.189769202217918e-06,
+      "loss": 1.4388,
+      "step": 6119
+    },
+    {
+      "epoch": 1.0558095402397998,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.187056302265324e-06,
+      "loss": 1.5254,
+      "step": 6120
+    },
+    {
+      "epoch": 1.055982058138532,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.184343462540683e-06,
+      "loss": 1.3021,
+      "step": 6121
+    },
+    {
+      "epoch": 1.056154576037264,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.181630683244972e-06,
+      "loss": 1.3728,
+      "step": 6122
+    },
+    {
+      "epoch": 1.0563270939359959,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.178917964579176e-06,
+      "loss": 1.3719,
+      "step": 6123
+    },
+    {
+      "epoch": 1.0564996118347278,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.176205306744265e-06,
+      "loss": 1.4278,
+      "step": 6124
+    },
+    {
+      "epoch": 1.0566721297334598,
+      "grad_norm": 0.62109375,
+      "learning_rate": 9.173492709941215e-06,
+      "loss": 1.459,
+      "step": 6125
+    },
+    {
+      "epoch": 1.0568446476321918,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.170780174370988e-06,
+      "loss": 1.4969,
+      "step": 6126
+    },
+    {
+      "epoch": 1.0570171655309237,
+      "grad_norm": 0.625,
+      "learning_rate": 9.168067700234542e-06,
+      "loss": 1.4081,
+      "step": 6127
+    },
+    {
+      "epoch": 1.057189683429656,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.165355287732846e-06,
+      "loss": 1.3269,
+      "step": 6128
+    },
+    {
+      "epoch": 1.0573622013283879,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.162642937066843e-06,
+      "loss": 1.3879,
+      "step": 6129
+    },
+    {
+      "epoch": 1.0575347192271198,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.159930648437484e-06,
+      "loss": 1.5195,
+      "step": 6130
+    },
+    {
+      "epoch": 1.0577072371258518,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.157218422045708e-06,
+      "loss": 1.5369,
+      "step": 6131
+    },
+    {
+      "epoch": 1.0578797550245838,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.154506258092462e-06,
+      "loss": 1.4377,
+      "step": 6132
+    },
+    {
+      "epoch": 1.0580522729233157,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.151794156778673e-06,
+      "loss": 1.4826,
+      "step": 6133
+    },
+    {
+      "epoch": 1.058224790822048,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.149082118305274e-06,
+      "loss": 1.4615,
+      "step": 6134
+    },
+    {
+      "epoch": 1.0583973087207799,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.146370142873185e-06,
+      "loss": 1.5071,
+      "step": 6135
+    },
+    {
+      "epoch": 1.0585698266195118,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.143658230683335e-06,
+      "loss": 1.412,
+      "step": 6136
+    },
+    {
+      "epoch": 1.0587423445182438,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.140946381936629e-06,
+      "loss": 1.322,
+      "step": 6137
+    },
+    {
+      "epoch": 1.0589148624169757,
+      "grad_norm": 0.625,
+      "learning_rate": 9.138234596833987e-06,
+      "loss": 1.4451,
+      "step": 6138
+    },
+    {
+      "epoch": 1.0590873803157077,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.135522875576305e-06,
+      "loss": 1.3518,
+      "step": 6139
+    },
+    {
+      "epoch": 1.0592598982144397,
+      "grad_norm": 0.74609375,
+      "learning_rate": 9.132811218364494e-06,
+      "loss": 1.46,
+      "step": 6140
+    },
+    {
+      "epoch": 1.0594324161131718,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.130099625399446e-06,
+      "loss": 1.4321,
+      "step": 6141
+    },
+    {
+      "epoch": 1.0596049340119038,
+      "grad_norm": 0.63671875,
+      "learning_rate": 9.127388096882054e-06,
+      "loss": 1.4989,
+      "step": 6142
+    },
+    {
+      "epoch": 1.0597774519106358,
+      "grad_norm": 0.796875,
+      "learning_rate": 9.1246766330132e-06,
+      "loss": 1.4134,
+      "step": 6143
+    },
+    {
+      "epoch": 1.0599499698093677,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.121965233993773e-06,
+      "loss": 1.3734,
+      "step": 6144
+    },
+    {
+      "epoch": 1.0601224877080997,
+      "grad_norm": 0.671875,
+      "learning_rate": 9.119253900024649e-06,
+      "loss": 1.4286,
+      "step": 6145
+    },
+    {
+      "epoch": 1.0602950056068317,
+      "grad_norm": 0.9609375,
+      "learning_rate": 9.116542631306695e-06,
+      "loss": 1.5147,
+      "step": 6146
+    },
+    {
+      "epoch": 1.0604675235055636,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.113831428040789e-06,
+      "loss": 1.4119,
+      "step": 6147
+    },
+    {
+      "epoch": 1.0606400414042958,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.11112029042778e-06,
+      "loss": 1.4222,
+      "step": 6148
+    },
+    {
+      "epoch": 1.0608125593030278,
+      "grad_norm": 0.65234375,
+      "learning_rate": 9.108409218668542e-06,
+      "loss": 1.3741,
+      "step": 6149
+    },
+    {
+      "epoch": 1.0609850772017597,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.105698212963915e-06,
+      "loss": 1.4079,
+      "step": 6150
+    },
+    {
+      "epoch": 1.0611575951004917,
+      "grad_norm": 0.6953125,
+      "learning_rate": 9.102987273514757e-06,
+      "loss": 1.4237,
+      "step": 6151
+    },
+    {
+      "epoch": 1.0613301129992236,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.100276400521898e-06,
+      "loss": 1.3756,
+      "step": 6152
+    },
+    {
+      "epoch": 1.0615026308979556,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.097565594186194e-06,
+      "loss": 1.4186,
+      "step": 6153
+    },
+    {
+      "epoch": 1.0616751487966876,
+      "grad_norm": 0.72265625,
+      "learning_rate": 9.094854854708464e-06,
+      "loss": 1.4798,
+      "step": 6154
+    },
+    {
+      "epoch": 1.0618476666954197,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.092144182289546e-06,
+      "loss": 1.4217,
+      "step": 6155
+    },
+    {
+      "epoch": 1.0620201845941517,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.089433577130256e-06,
+      "loss": 1.4153,
+      "step": 6156
+    },
+    {
+      "epoch": 1.0621927024928837,
+      "grad_norm": 1.125,
+      "learning_rate": 9.086723039431418e-06,
+      "loss": 1.3932,
+      "step": 6157
+    },
+    {
+      "epoch": 1.0623652203916156,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.084012569393842e-06,
+      "loss": 1.4848,
+      "step": 6158
+    },
+    {
+      "epoch": 1.0625377382903476,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.081302167218339e-06,
+      "loss": 1.4844,
+      "step": 6159
+    },
+    {
+      "epoch": 1.0627102561890795,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.078591833105712e-06,
+      "loss": 1.4191,
+      "step": 6160
+    },
+    {
+      "epoch": 1.0628827740878115,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.075881567256759e-06,
+      "loss": 1.3469,
+      "step": 6161
+    },
+    {
+      "epoch": 1.0630552919865437,
+      "grad_norm": 0.7421875,
+      "learning_rate": 9.073171369872275e-06,
+      "loss": 1.4488,
+      "step": 6162
+    },
+    {
+      "epoch": 1.0632278098852757,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.070461241153044e-06,
+      "loss": 1.3692,
+      "step": 6163
+    },
+    {
+      "epoch": 1.0634003277840076,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.067751181299856e-06,
+      "loss": 1.3982,
+      "step": 6164
+    },
+    {
+      "epoch": 1.0635728456827396,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.065041190513483e-06,
+      "loss": 1.406,
+      "step": 6165
+    },
+    {
+      "epoch": 1.0637453635814715,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.062331268994704e-06,
+      "loss": 1.4365,
+      "step": 6166
+    },
+    {
+      "epoch": 1.0639178814802035,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.059621416944277e-06,
+      "loss": 1.3919,
+      "step": 6167
+    },
+    {
+      "epoch": 1.0640903993789355,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.056911634562975e-06,
+      "loss": 1.2976,
+      "step": 6168
+    },
+    {
+      "epoch": 1.0642629172776676,
+      "grad_norm": 0.625,
+      "learning_rate": 9.054201922051552e-06,
+      "loss": 1.3825,
+      "step": 6169
+    },
+    {
+      "epoch": 1.0644354351763996,
+      "grad_norm": 0.53125,
+      "learning_rate": 9.051492279610763e-06,
+      "loss": 1.3458,
+      "step": 6170
+    },
+    {
+      "epoch": 1.0646079530751316,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.048782707441346e-06,
+      "loss": 1.3666,
+      "step": 6171
+    },
+    {
+      "epoch": 1.0647804709738635,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.046073205744053e-06,
+      "loss": 1.4677,
+      "step": 6172
+    },
+    {
+      "epoch": 1.0649529888725955,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.043363774719618e-06,
+      "loss": 1.4026,
+      "step": 6173
+    },
+    {
+      "epoch": 1.0651255067713274,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.040654414568772e-06,
+      "loss": 1.5799,
+      "step": 6174
+    },
+    {
+      "epoch": 1.0652980246700596,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.037945125492238e-06,
+      "loss": 1.4668,
+      "step": 6175
+    },
+    {
+      "epoch": 1.0654705425687916,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.035235907690739e-06,
+      "loss": 1.4445,
+      "step": 6176
+    },
+    {
+      "epoch": 1.0656430604675236,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.032526761364999e-06,
+      "loss": 1.4619,
+      "step": 6177
+    },
+    {
+      "epoch": 1.0658155783662555,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.029817686715717e-06,
+      "loss": 1.4089,
+      "step": 6178
+    },
+    {
+      "epoch": 1.0659880962649875,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.027108683943605e-06,
+      "loss": 1.4093,
+      "step": 6179
+    },
+    {
+      "epoch": 1.0661606141637194,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.024399753249358e-06,
+      "loss": 1.4857,
+      "step": 6180
+    },
+    {
+      "epoch": 1.0663331320624514,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.021690894833676e-06,
+      "loss": 1.4078,
+      "step": 6181
+    },
+    {
+      "epoch": 1.0665056499611834,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.018982108897243e-06,
+      "loss": 1.4882,
+      "step": 6182
+    },
+    {
+      "epoch": 1.0666781678599155,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.016273395640748e-06,
+      "loss": 1.362,
+      "step": 6183
+    },
+    {
+      "epoch": 1.0668506857586475,
+      "grad_norm": 0.80859375,
+      "learning_rate": 9.013564755264862e-06,
+      "loss": 1.411,
+      "step": 6184
+    },
+    {
+      "epoch": 1.0670232036573795,
+      "grad_norm": 0.69140625,
+      "learning_rate": 9.010856187970267e-06,
+      "loss": 1.3623,
+      "step": 6185
+    },
+    {
+      "epoch": 1.0671957215561114,
+      "grad_norm": 0.6875,
+      "learning_rate": 9.008147693957624e-06,
+      "loss": 1.4222,
+      "step": 6186
+    },
+    {
+      "epoch": 1.0673682394548434,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.005439273427597e-06,
+      "loss": 1.3998,
+      "step": 6187
+    },
+    {
+      "epoch": 1.0675407573535753,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.00273092658084e-06,
+      "loss": 1.4059,
+      "step": 6188
+    },
+    {
+      "epoch": 1.0677132752523075,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.000022653618012e-06,
+      "loss": 1.3935,
+      "step": 6189
+    },
+    {
+      "epoch": 1.0678857931510395,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.997314454739752e-06,
+      "loss": 1.2628,
+      "step": 6190
+    },
+    {
+      "epoch": 1.0680583110497714,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.9946063301467e-06,
+      "loss": 1.4216,
+      "step": 6191
+    },
+    {
+      "epoch": 1.0682308289485034,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.991898280039498e-06,
+      "loss": 1.4487,
+      "step": 6192
+    },
+    {
+      "epoch": 1.0684033468472354,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.989190304618767e-06,
+      "loss": 1.4705,
+      "step": 6193
+    },
+    {
+      "epoch": 1.0685758647459673,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.986482404085137e-06,
+      "loss": 1.4954,
+      "step": 6194
+    },
+    {
+      "epoch": 1.0687483826446993,
+      "grad_norm": 1.3515625,
+      "learning_rate": 8.983774578639219e-06,
+      "loss": 1.4177,
+      "step": 6195
+    },
+    {
+      "epoch": 1.0689209005434315,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.981066828481635e-06,
+      "loss": 1.4546,
+      "step": 6196
+    },
+    {
+      "epoch": 1.0690934184421634,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.978359153812982e-06,
+      "loss": 1.4029,
+      "step": 6197
+    },
+    {
+      "epoch": 1.0692659363408954,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.975651554833869e-06,
+      "loss": 1.3723,
+      "step": 6198
+    },
+    {
+      "epoch": 1.0694384542396274,
+      "grad_norm": 0.625,
+      "learning_rate": 8.972944031744886e-06,
+      "loss": 1.4171,
+      "step": 6199
+    },
+    {
+      "epoch": 1.0696109721383593,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.970236584746631e-06,
+      "loss": 1.4983,
+      "step": 6200
+    },
+    {
+      "epoch": 1.0696109721383593,
+      "eval_loss": 1.4095999002456665,
+      "eval_runtime": 11.2941,
+      "eval_samples_per_second": 90.667,
+      "eval_steps_per_second": 22.667,
+      "step": 6200
+    },
+    {
+      "epoch": 1.0697834900370913,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.967529214039682e-06,
+      "loss": 1.3528,
+      "step": 6201
+    },
+    {
+      "epoch": 1.0699560079358232,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.964821919824623e-06,
+      "loss": 1.4585,
+      "step": 6202
+    },
+    {
+      "epoch": 1.0701285258345554,
+      "grad_norm": 6.25,
+      "learning_rate": 8.962114702302018e-06,
+      "loss": 1.4366,
+      "step": 6203
+    },
+    {
+      "epoch": 1.0703010437332874,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.959407561672447e-06,
+      "loss": 1.4571,
+      "step": 6204
+    },
+    {
+      "epoch": 1.0704735616320193,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.956700498136461e-06,
+      "loss": 1.4125,
+      "step": 6205
+    },
+    {
+      "epoch": 1.0706460795307513,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.953993511894626e-06,
+      "loss": 1.4351,
+      "step": 6206
+    },
+    {
+      "epoch": 1.0708185974294833,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.951286603147481e-06,
+      "loss": 1.4297,
+      "step": 6207
+    },
+    {
+      "epoch": 1.0709911153282152,
+      "grad_norm": 1.765625,
+      "learning_rate": 8.948579772095578e-06,
+      "loss": 1.4696,
+      "step": 6208
+    },
+    {
+      "epoch": 1.0711636332269472,
+      "grad_norm": 0.625,
+      "learning_rate": 8.94587301893946e-06,
+      "loss": 1.4402,
+      "step": 6209
+    },
+    {
+      "epoch": 1.0713361511256794,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.943166343879652e-06,
+      "loss": 1.4923,
+      "step": 6210
+    },
+    {
+      "epoch": 1.0715086690244113,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.940459747116688e-06,
+      "loss": 1.403,
+      "step": 6211
+    },
+    {
+      "epoch": 1.0716811869231433,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.93775322885108e-06,
+      "loss": 1.4063,
+      "step": 6212
+    },
+    {
+      "epoch": 1.0718537048218753,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.935046789283356e-06,
+      "loss": 1.4808,
+      "step": 6213
+    },
+    {
+      "epoch": 1.0720262227206072,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.932340428614016e-06,
+      "loss": 1.4331,
+      "step": 6214
+    },
+    {
+      "epoch": 1.0721987406193392,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.929634147043573e-06,
+      "loss": 1.4485,
+      "step": 6215
+    },
+    {
+      "epoch": 1.0723712585180714,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.926927944772514e-06,
+      "loss": 1.4356,
+      "step": 6216
+    },
+    {
+      "epoch": 1.0725437764168033,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.924221822001342e-06,
+      "loss": 1.4756,
+      "step": 6217
+    },
+    {
+      "epoch": 1.0727162943155353,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.921515778930538e-06,
+      "loss": 1.4703,
+      "step": 6218
+    },
+    {
+      "epoch": 1.0728888122142672,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.918809815760585e-06,
+      "loss": 1.3884,
+      "step": 6219
+    },
+    {
+      "epoch": 1.0730613301129992,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.916103932691953e-06,
+      "loss": 1.4818,
+      "step": 6220
+    },
+    {
+      "epoch": 1.0732338480117312,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.913398129925118e-06,
+      "loss": 1.5478,
+      "step": 6221
+    },
+    {
+      "epoch": 1.0734063659104631,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.910692407660538e-06,
+      "loss": 1.3961,
+      "step": 6222
+    },
+    {
+      "epoch": 1.0735788838091953,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.907986766098666e-06,
+      "loss": 1.5076,
+      "step": 6223
+    },
+    {
+      "epoch": 1.0737514017079273,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.905281205439965e-06,
+      "loss": 1.4828,
+      "step": 6224
+    },
+    {
+      "epoch": 1.0739239196066592,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.902575725884867e-06,
+      "loss": 1.41,
+      "step": 6225
+    },
+    {
+      "epoch": 1.0740964375053912,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.89987032763382e-06,
+      "loss": 1.431,
+      "step": 6226
+    },
+    {
+      "epoch": 1.0742689554041231,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.89716501088725e-06,
+      "loss": 1.3323,
+      "step": 6227
+    },
+    {
+      "epoch": 1.074441473302855,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.89445977584559e-06,
+      "loss": 1.4044,
+      "step": 6228
+    },
+    {
+      "epoch": 1.074613991201587,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.891754622709254e-06,
+      "loss": 1.4322,
+      "step": 6229
+    },
+    {
+      "epoch": 1.0747865091003193,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.889049551678664e-06,
+      "loss": 1.4238,
+      "step": 6230
+    },
+    {
+      "epoch": 1.0749590269990512,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.886344562954221e-06,
+      "loss": 1.3572,
+      "step": 6231
+    },
+    {
+      "epoch": 1.0751315448977832,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.883639656736334e-06,
+      "loss": 1.4339,
+      "step": 6232
+    },
+    {
+      "epoch": 1.0753040627965151,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.880934833225395e-06,
+      "loss": 1.4185,
+      "step": 6233
+    },
+    {
+      "epoch": 1.075476580695247,
+      "grad_norm": 0.625,
+      "learning_rate": 8.878230092621799e-06,
+      "loss": 1.4326,
+      "step": 6234
+    },
+    {
+      "epoch": 1.075649098593979,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.87552543512592e-06,
+      "loss": 1.4075,
+      "step": 6235
+    },
+    {
+      "epoch": 1.075821616492711,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.872820860938149e-06,
+      "loss": 1.315,
+      "step": 6236
+    },
+    {
+      "epoch": 1.0759941343914432,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.870116370258847e-06,
+      "loss": 1.3736,
+      "step": 6237
+    },
+    {
+      "epoch": 1.0761666522901752,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.86741196328839e-06,
+      "loss": 1.4035,
+      "step": 6238
+    },
+    {
+      "epoch": 1.0763391701889071,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.864707640227127e-06,
+      "loss": 1.3408,
+      "step": 6239
+    },
+    {
+      "epoch": 1.076511688087639,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.862003401275414e-06,
+      "loss": 1.4941,
+      "step": 6240
+    },
+    {
+      "epoch": 1.076684205986371,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.859299246633604e-06,
+      "loss": 1.3956,
+      "step": 6241
+    },
+    {
+      "epoch": 1.076856723885103,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.85659517650203e-06,
+      "loss": 1.329,
+      "step": 6242
+    },
+    {
+      "epoch": 1.0770292417838352,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.853891191081035e-06,
+      "loss": 1.3659,
+      "step": 6243
+    },
+    {
+      "epoch": 1.0772017596825672,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.851187290570934e-06,
+      "loss": 1.4607,
+      "step": 6244
+    },
+    {
+      "epoch": 1.0773742775812991,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.848483475172063e-06,
+      "loss": 1.4711,
+      "step": 6245
+    },
+    {
+      "epoch": 1.077546795480031,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.845779745084728e-06,
+      "loss": 1.4921,
+      "step": 6246
+    },
+    {
+      "epoch": 1.077719313378763,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.843076100509244e-06,
+      "loss": 1.4246,
+      "step": 6247
+    },
+    {
+      "epoch": 1.077891831277495,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.840372541645907e-06,
+      "loss": 1.4466,
+      "step": 6248
+    },
+    {
+      "epoch": 1.078064349176227,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.837669068695023e-06,
+      "loss": 1.4841,
+      "step": 6249
+    },
+    {
+      "epoch": 1.078236867074959,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.834965681856873e-06,
+      "loss": 1.4025,
+      "step": 6250
+    },
+    {
+      "epoch": 1.078409384973691,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.832262381331747e-06,
+      "loss": 1.4657,
+      "step": 6251
+    },
+    {
+      "epoch": 1.078581902872423,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.829559167319917e-06,
+      "loss": 1.3731,
+      "step": 6252
+    },
+    {
+      "epoch": 1.078754420771155,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.826856040021661e-06,
+      "loss": 1.5165,
+      "step": 6253
+    },
+    {
+      "epoch": 1.078926938669887,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.824152999637237e-06,
+      "loss": 1.3957,
+      "step": 6254
+    },
+    {
+      "epoch": 1.079099456568619,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.821450046366909e-06,
+      "loss": 1.4961,
+      "step": 6255
+    },
+    {
+      "epoch": 1.079271974467351,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.81874718041092e-06,
+      "loss": 1.2706,
+      "step": 6256
+    },
+    {
+      "epoch": 1.079444492366083,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.816044401969524e-06,
+      "loss": 1.5226,
+      "step": 6257
+    },
+    {
+      "epoch": 1.079617010264815,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.813341711242959e-06,
+      "loss": 1.4211,
+      "step": 6258
+    },
+    {
+      "epoch": 1.079789528163547,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.810639108431449e-06,
+      "loss": 1.4257,
+      "step": 6259
+    },
+    {
+      "epoch": 1.079962046062279,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.80793659373523e-06,
+      "loss": 1.5166,
+      "step": 6260
+    },
+    {
+      "epoch": 1.080134563961011,
+      "grad_norm": 0.625,
+      "learning_rate": 8.805234167354515e-06,
+      "loss": 1.4059,
+      "step": 6261
+    },
+    {
+      "epoch": 1.080307081859743,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.80253182948952e-06,
+      "loss": 1.4157,
+      "step": 6262
+    },
+    {
+      "epoch": 1.0804795997584749,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.799829580340444e-06,
+      "loss": 1.4555,
+      "step": 6263
+    },
+    {
+      "epoch": 1.080652117657207,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.797127420107496e-06,
+      "loss": 1.4794,
+      "step": 6264
+    },
+    {
+      "epoch": 1.080824635555939,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.794425348990861e-06,
+      "loss": 1.4365,
+      "step": 6265
+    },
+    {
+      "epoch": 1.080997153454671,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.791723367190736e-06,
+      "loss": 1.4629,
+      "step": 6266
+    },
+    {
+      "epoch": 1.081169671353403,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.789021474907286e-06,
+      "loss": 1.4438,
+      "step": 6267
+    },
+    {
+      "epoch": 1.0813421892521349,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.786319672340696e-06,
+      "loss": 1.4537,
+      "step": 6268
+    },
+    {
+      "epoch": 1.0815147071508668,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.783617959691126e-06,
+      "loss": 1.4335,
+      "step": 6269
+    },
+    {
+      "epoch": 1.0816872250495988,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.780916337158739e-06,
+      "loss": 1.4098,
+      "step": 6270
+    },
+    {
+      "epoch": 1.081859742948331,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.778214804943687e-06,
+      "loss": 1.4097,
+      "step": 6271
+    },
+    {
+      "epoch": 1.082032260847063,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.775513363246113e-06,
+      "loss": 1.4887,
+      "step": 6272
+    },
+    {
+      "epoch": 1.082204778745795,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.772812012266165e-06,
+      "loss": 1.3853,
+      "step": 6273
+    },
+    {
+      "epoch": 1.0823772966445269,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.770110752203968e-06,
+      "loss": 1.3535,
+      "step": 6274
+    },
+    {
+      "epoch": 1.0825498145432588,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.767409583259654e-06,
+      "loss": 1.4843,
+      "step": 6275
+    },
+    {
+      "epoch": 1.0827223324419908,
+      "grad_norm": 1.078125,
+      "learning_rate": 8.764708505633334e-06,
+      "loss": 1.4779,
+      "step": 6276
+    },
+    {
+      "epoch": 1.0828948503407227,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.762007519525132e-06,
+      "loss": 1.306,
+      "step": 6277
+    },
+    {
+      "epoch": 1.083067368239455,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.759306625135147e-06,
+      "loss": 1.4475,
+      "step": 6278
+    },
+    {
+      "epoch": 1.083239886138187,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.75660582266348e-06,
+      "loss": 1.4575,
+      "step": 6279
+    },
+    {
+      "epoch": 1.0834124040369189,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.753905112310217e-06,
+      "loss": 1.3896,
+      "step": 6280
+    },
+    {
+      "epoch": 1.0835849219356508,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.751204494275457e-06,
+      "loss": 1.3266,
+      "step": 6281
+    },
+    {
+      "epoch": 1.0837574398343828,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.748503968759267e-06,
+      "loss": 1.4157,
+      "step": 6282
+    },
+    {
+      "epoch": 1.0839299577331147,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.745803535961725e-06,
+      "loss": 1.3011,
+      "step": 6283
+    },
+    {
+      "epoch": 1.084102475631847,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.74310319608289e-06,
+      "loss": 1.3948,
+      "step": 6284
+    },
+    {
+      "epoch": 1.0842749935305789,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.740402949322827e-06,
+      "loss": 1.3987,
+      "step": 6285
+    },
+    {
+      "epoch": 1.0844475114293108,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.737702795881581e-06,
+      "loss": 1.4962,
+      "step": 6286
+    },
+    {
+      "epoch": 1.0846200293280428,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.735002735959203e-06,
+      "loss": 1.4318,
+      "step": 6287
+    },
+    {
+      "epoch": 1.0847925472267748,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.732302769755722e-06,
+      "loss": 1.4156,
+      "step": 6288
+    },
+    {
+      "epoch": 1.0849650651255067,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.729602897471175e-06,
+      "loss": 1.3992,
+      "step": 6289
+    },
+    {
+      "epoch": 1.0851375830242387,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.726903119305583e-06,
+      "loss": 1.3411,
+      "step": 6290
+    },
+    {
+      "epoch": 1.0853101009229706,
+      "grad_norm": 0.78515625,
+      "learning_rate": 8.72420343545896e-06,
+      "loss": 1.4526,
+      "step": 6291
+    },
+    {
+      "epoch": 1.0854826188217028,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.721503846131322e-06,
+      "loss": 1.4865,
+      "step": 6292
+    },
+    {
+      "epoch": 1.0856551367204348,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.718804351522666e-06,
+      "loss": 1.3999,
+      "step": 6293
+    },
+    {
+      "epoch": 1.0858276546191667,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.716104951832992e-06,
+      "loss": 1.3276,
+      "step": 6294
+    },
+    {
+      "epoch": 1.0860001725178987,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.713405647262279e-06,
+      "loss": 1.3087,
+      "step": 6295
+    },
+    {
+      "epoch": 1.0861726904166307,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.71070643801052e-06,
+      "loss": 1.3814,
+      "step": 6296
+    },
+    {
+      "epoch": 1.0863452083153626,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.708007324277678e-06,
+      "loss": 1.4415,
+      "step": 6297
+    },
+    {
+      "epoch": 1.0865177262140948,
+      "grad_norm": 0.7578125,
+      "learning_rate": 8.705308306263732e-06,
+      "loss": 1.2513,
+      "step": 6298
+    },
+    {
+      "epoch": 1.0866902441128268,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.702609384168634e-06,
+      "loss": 1.4319,
+      "step": 6299
+    },
+    {
+      "epoch": 1.0868627620115587,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.69991055819234e-06,
+      "loss": 1.4037,
+      "step": 6300
+    },
+    {
+      "epoch": 1.0868627620115587,
+      "eval_loss": 1.4092991352081299,
+      "eval_runtime": 10.9216,
+      "eval_samples_per_second": 93.759,
+      "eval_steps_per_second": 23.44,
+      "step": 6300
+    },
+    {
+      "epoch": 1.0870352799102907,
+      "grad_norm": 0.7421875,
+      "learning_rate": 8.697211828534793e-06,
+      "loss": 1.4708,
+      "step": 6301
+    },
+    {
+      "epoch": 1.0872077978090227,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.694513195395937e-06,
+      "loss": 1.2764,
+      "step": 6302
+    },
+    {
+      "epoch": 1.0873803157077546,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.691814658975699e-06,
+      "loss": 1.4155,
+      "step": 6303
+    },
+    {
+      "epoch": 1.0875528336064866,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.689116219474004e-06,
+      "loss": 1.3576,
+      "step": 6304
+    },
+    {
+      "epoch": 1.0877253515052188,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.686417877090772e-06,
+      "loss": 1.4314,
+      "step": 6305
+    },
+    {
+      "epoch": 1.0878978694039507,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.68371963202591e-06,
+      "loss": 1.4714,
+      "step": 6306
+    },
+    {
+      "epoch": 1.0880703873026827,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.681021484479327e-06,
+      "loss": 1.465,
+      "step": 6307
+    },
+    {
+      "epoch": 1.0882429052014146,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.678323434650906e-06,
+      "loss": 1.4521,
+      "step": 6308
+    },
+    {
+      "epoch": 1.0884154231001466,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.675625482740549e-06,
+      "loss": 1.5038,
+      "step": 6309
+    },
+    {
+      "epoch": 1.0885879409988786,
+      "grad_norm": 0.69921875,
+      "learning_rate": 8.672927628948128e-06,
+      "loss": 1.4076,
+      "step": 6310
+    },
+    {
+      "epoch": 1.0887604588976105,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.670229873473524e-06,
+      "loss": 1.5556,
+      "step": 6311
+    },
+    {
+      "epoch": 1.0889329767963427,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.667532216516594e-06,
+      "loss": 1.4564,
+      "step": 6312
+    },
+    {
+      "epoch": 1.0891054946950747,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.664834658277208e-06,
+      "loss": 1.4006,
+      "step": 6313
+    },
+    {
+      "epoch": 1.0892780125938066,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.662137198955211e-06,
+      "loss": 1.4791,
+      "step": 6314
+    },
+    {
+      "epoch": 1.0894505304925386,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.659439838750451e-06,
+      "loss": 1.4249,
+      "step": 6315
+    },
+    {
+      "epoch": 1.0896230483912706,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.656742577862761e-06,
+      "loss": 1.4844,
+      "step": 6316
+    },
+    {
+      "epoch": 1.0897955662900025,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.654045416491975e-06,
+      "loss": 1.3683,
+      "step": 6317
+    },
+    {
+      "epoch": 1.0899680841887345,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.651348354837917e-06,
+      "loss": 1.4811,
+      "step": 6318
+    },
+    {
+      "epoch": 1.0901406020874667,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.6486513931004e-06,
+      "loss": 1.4244,
+      "step": 6319
+    },
+    {
+      "epoch": 1.0903131199861986,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.645954531479226e-06,
+      "loss": 1.3503,
+      "step": 6320
+    },
+    {
+      "epoch": 1.0904856378849306,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.643257770174202e-06,
+      "loss": 1.4309,
+      "step": 6321
+    },
+    {
+      "epoch": 1.0906581557836625,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.640561109385125e-06,
+      "loss": 1.4698,
+      "step": 6322
+    },
+    {
+      "epoch": 1.0908306736823945,
+      "grad_norm": 0.52734375,
+      "learning_rate": 8.637864549311768e-06,
+      "loss": 1.299,
+      "step": 6323
+    },
+    {
+      "epoch": 1.0910031915811265,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.635168090153922e-06,
+      "loss": 1.3882,
+      "step": 6324
+    },
+    {
+      "epoch": 1.0911757094798586,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.632471732111349e-06,
+      "loss": 1.348,
+      "step": 6325
+    },
+    {
+      "epoch": 1.0913482273785906,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.629775475383816e-06,
+      "loss": 1.4612,
+      "step": 6326
+    },
+    {
+      "epoch": 1.0915207452773226,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.627079320171076e-06,
+      "loss": 1.3347,
+      "step": 6327
+    },
+    {
+      "epoch": 1.0916932631760545,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.624383266672882e-06,
+      "loss": 1.432,
+      "step": 6328
+    },
+    {
+      "epoch": 1.0918657810747865,
+      "grad_norm": 0.625,
+      "learning_rate": 8.621687315088965e-06,
+      "loss": 1.4275,
+      "step": 6329
+    },
+    {
+      "epoch": 1.0920382989735185,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.618991465619068e-06,
+      "loss": 1.4922,
+      "step": 6330
+    },
+    {
+      "epoch": 1.0922108168722504,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.616295718462913e-06,
+      "loss": 1.3674,
+      "step": 6331
+    },
+    {
+      "epoch": 1.0923833347709824,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.613600073820216e-06,
+      "loss": 1.5326,
+      "step": 6332
+    },
+    {
+      "epoch": 1.0925558526697146,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.610904531890685e-06,
+      "loss": 1.432,
+      "step": 6333
+    },
+    {
+      "epoch": 1.0927283705684465,
+      "grad_norm": 0.6875,
+      "learning_rate": 8.60820909287403e-06,
+      "loss": 1.434,
+      "step": 6334
+    },
+    {
+      "epoch": 1.0929008884671785,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.60551375696994e-06,
+      "loss": 1.4066,
+      "step": 6335
+    },
+    {
+      "epoch": 1.0930734063659104,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.602818524378098e-06,
+      "loss": 1.4648,
+      "step": 6336
+    },
+    {
+      "epoch": 1.0932459242646424,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.600123395298198e-06,
+      "loss": 1.3308,
+      "step": 6337
+    },
+    {
+      "epoch": 1.0934184421633744,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.597428369929902e-06,
+      "loss": 1.36,
+      "step": 6338
+    },
+    {
+      "epoch": 1.0935909600621065,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.594733448472876e-06,
+      "loss": 1.4629,
+      "step": 6339
+    },
+    {
+      "epoch": 1.0937634779608385,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.592038631126774e-06,
+      "loss": 1.4091,
+      "step": 6340
+    },
+    {
+      "epoch": 1.0939359958595705,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.589343918091251e-06,
+      "loss": 1.3806,
+      "step": 6341
+    },
+    {
+      "epoch": 1.0941085137583024,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.586649309565942e-06,
+      "loss": 1.4089,
+      "step": 6342
+    },
+    {
+      "epoch": 1.0942810316570344,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.583954805750488e-06,
+      "loss": 1.5163,
+      "step": 6343
+    },
+    {
+      "epoch": 1.0944535495557663,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.581260406844503e-06,
+      "loss": 1.3662,
+      "step": 6344
+    },
+    {
+      "epoch": 1.0946260674544983,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.578566113047616e-06,
+      "loss": 1.4788,
+      "step": 6345
+    },
+    {
+      "epoch": 1.0947985853532305,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.575871924559431e-06,
+      "loss": 1.4255,
+      "step": 6346
+    },
+    {
+      "epoch": 1.0949711032519625,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.573177841579556e-06,
+      "loss": 1.3274,
+      "step": 6347
+    },
+    {
+      "epoch": 1.0951436211506944,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.570483864307575e-06,
+      "loss": 1.3774,
+      "step": 6348
+    },
+    {
+      "epoch": 1.0953161390494264,
+      "grad_norm": 0.796875,
+      "learning_rate": 8.567789992943088e-06,
+      "loss": 1.5031,
+      "step": 6349
+    },
+    {
+      "epoch": 1.0954886569481583,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.565096227685663e-06,
+      "loss": 1.5199,
+      "step": 6350
+    },
+    {
+      "epoch": 1.0956611748468903,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.562402568734879e-06,
+      "loss": 1.4346,
+      "step": 6351
+    },
+    {
+      "epoch": 1.0958336927456223,
+      "grad_norm": 0.703125,
+      "learning_rate": 8.559709016290288e-06,
+      "loss": 1.3992,
+      "step": 6352
+    },
+    {
+      "epoch": 1.0960062106443544,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.557015570551455e-06,
+      "loss": 1.3908,
+      "step": 6353
+    },
+    {
+      "epoch": 1.0961787285430864,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.55432223171793e-06,
+      "loss": 1.3654,
+      "step": 6354
+    },
+    {
+      "epoch": 1.0963512464418184,
+      "grad_norm": 0.625,
+      "learning_rate": 8.551628999989242e-06,
+      "loss": 1.4114,
+      "step": 6355
+    },
+    {
+      "epoch": 1.0965237643405503,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.548935875564931e-06,
+      "loss": 1.4118,
+      "step": 6356
+    },
+    {
+      "epoch": 1.0966962822392823,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.546242858644513e-06,
+      "loss": 1.4611,
+      "step": 6357
+    },
+    {
+      "epoch": 1.0968688001380142,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.543549949427512e-06,
+      "loss": 1.4758,
+      "step": 6358
+    },
+    {
+      "epoch": 1.0970413180367462,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.540857148113429e-06,
+      "loss": 1.3938,
+      "step": 6359
+    },
+    {
+      "epoch": 1.0972138359354784,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.538164454901766e-06,
+      "loss": 1.4961,
+      "step": 6360
+    },
+    {
+      "epoch": 1.0973863538342103,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.535471869992011e-06,
+      "loss": 1.4775,
+      "step": 6361
+    },
+    {
+      "epoch": 1.0975588717329423,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.532779393583656e-06,
+      "loss": 1.4067,
+      "step": 6362
+    },
+    {
+      "epoch": 1.0977313896316743,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.530087025876168e-06,
+      "loss": 1.3677,
+      "step": 6363
+    },
+    {
+      "epoch": 1.0979039075304062,
+      "grad_norm": 0.546875,
+      "learning_rate": 8.52739476706902e-06,
+      "loss": 1.2977,
+      "step": 6364
+    },
+    {
+      "epoch": 1.0980764254291382,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.524702617361665e-06,
+      "loss": 1.4171,
+      "step": 6365
+    },
+    {
+      "epoch": 1.0982489433278704,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.522010576953561e-06,
+      "loss": 1.3564,
+      "step": 6366
+    },
+    {
+      "epoch": 1.0984214612266023,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.519318646044147e-06,
+      "loss": 1.4582,
+      "step": 6367
+    },
+    {
+      "epoch": 1.0985939791253343,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.516626824832858e-06,
+      "loss": 1.4448,
+      "step": 6368
+    },
+    {
+      "epoch": 1.0987664970240663,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.513935113519126e-06,
+      "loss": 1.3989,
+      "step": 6369
+    },
+    {
+      "epoch": 1.0989390149227982,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.511243512302362e-06,
+      "loss": 1.4482,
+      "step": 6370
+    },
+    {
+      "epoch": 1.0991115328215302,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.508552021381987e-06,
+      "loss": 1.345,
+      "step": 6371
+    },
+    {
+      "epoch": 1.0992840507202621,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.50586064095739e-06,
+      "loss": 1.3918,
+      "step": 6372
+    },
+    {
+      "epoch": 1.0994565686189943,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.503169371227978e-06,
+      "loss": 1.5311,
+      "step": 6373
+    },
+    {
+      "epoch": 1.0996290865177263,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.50047821239313e-06,
+      "loss": 1.4589,
+      "step": 6374
+    },
+    {
+      "epoch": 1.0998016044164582,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.497787164652227e-06,
+      "loss": 1.3397,
+      "step": 6375
+    },
+    {
+      "epoch": 1.0999741223151902,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.495096228204632e-06,
+      "loss": 1.4847,
+      "step": 6376
+    },
+    {
+      "epoch": 1.1001466402139222,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.492405403249717e-06,
+      "loss": 1.4609,
+      "step": 6377
+    },
+    {
+      "epoch": 1.1003191581126541,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.489714689986826e-06,
+      "loss": 1.4853,
+      "step": 6378
+    },
+    {
+      "epoch": 1.100491676011386,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.48702408861531e-06,
+      "loss": 1.5215,
+      "step": 6379
+    },
+    {
+      "epoch": 1.1006641939101183,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.4843335993345e-06,
+      "loss": 1.3049,
+      "step": 6380
+    },
+    {
+      "epoch": 1.1008367118088502,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.48164322234373e-06,
+      "loss": 1.464,
+      "step": 6381
+    },
+    {
+      "epoch": 1.1010092297075822,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.478952957842317e-06,
+      "loss": 1.4325,
+      "step": 6382
+    },
+    {
+      "epoch": 1.1011817476063142,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.476262806029573e-06,
+      "loss": 1.3597,
+      "step": 6383
+    },
+    {
+      "epoch": 1.1013542655050461,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.473572767104799e-06,
+      "loss": 1.3886,
+      "step": 6384
+    },
+    {
+      "epoch": 1.101526783403778,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.47088284126729e-06,
+      "loss": 1.3507,
+      "step": 6385
+    },
+    {
+      "epoch": 1.10169930130251,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.46819302871634e-06,
+      "loss": 1.541,
+      "step": 6386
+    },
+    {
+      "epoch": 1.1018718192012422,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.465503329651219e-06,
+      "loss": 1.405,
+      "step": 6387
+    },
+    {
+      "epoch": 1.1020443370999742,
+      "grad_norm": 0.74609375,
+      "learning_rate": 8.462813744271202e-06,
+      "loss": 1.519,
+      "step": 6388
+    },
+    {
+      "epoch": 1.1022168549987061,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.460124272775542e-06,
+      "loss": 1.3921,
+      "step": 6389
+    },
+    {
+      "epoch": 1.102389372897438,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.457434915363501e-06,
+      "loss": 1.3868,
+      "step": 6390
+    },
+    {
+      "epoch": 1.10256189079617,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.454745672234321e-06,
+      "loss": 1.4336,
+      "step": 6391
+    },
+    {
+      "epoch": 1.102734408694902,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.452056543587236e-06,
+      "loss": 1.3515,
+      "step": 6392
+    },
+    {
+      "epoch": 1.1029069265936342,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.44936752962147e-06,
+      "loss": 1.4774,
+      "step": 6393
+    },
+    {
+      "epoch": 1.1030794444923662,
+      "grad_norm": 0.671875,
+      "learning_rate": 8.446678630536252e-06,
+      "loss": 1.4492,
+      "step": 6394
+    },
+    {
+      "epoch": 1.1032519623910981,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.443989846530784e-06,
+      "loss": 1.4031,
+      "step": 6395
+    },
+    {
+      "epoch": 1.10342448028983,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.441301177804273e-06,
+      "loss": 1.3862,
+      "step": 6396
+    },
+    {
+      "epoch": 1.103596998188562,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.438612624555905e-06,
+      "loss": 1.3648,
+      "step": 6397
+    },
+    {
+      "epoch": 1.103769516087294,
+      "grad_norm": 0.73046875,
+      "learning_rate": 8.435924186984875e-06,
+      "loss": 1.4533,
+      "step": 6398
+    },
+    {
+      "epoch": 1.103942033986026,
+      "grad_norm": 0.80859375,
+      "learning_rate": 8.433235865290351e-06,
+      "loss": 1.4442,
+      "step": 6399
+    },
+    {
+      "epoch": 1.104114551884758,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.430547659671503e-06,
+      "loss": 1.4276,
+      "step": 6400
+    },
+    {
+      "epoch": 1.104114551884758,
+      "eval_loss": 1.4093081951141357,
+      "eval_runtime": 10.9921,
+      "eval_samples_per_second": 93.158,
+      "eval_steps_per_second": 23.29,
+      "step": 6400
+    },
+    {
+      "epoch": 1.1042870697834901,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.427859570327494e-06,
+      "loss": 1.4678,
+      "step": 6401
+    },
+    {
+      "epoch": 1.104459587682222,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.425171597457469e-06,
+      "loss": 1.4255,
+      "step": 6402
+    },
+    {
+      "epoch": 1.104632105580954,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.422483741260575e-06,
+      "loss": 1.4794,
+      "step": 6403
+    },
+    {
+      "epoch": 1.104804623479686,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.419796001935935e-06,
+      "loss": 1.4145,
+      "step": 6404
+    },
+    {
+      "epoch": 1.104977141378418,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.417108379682688e-06,
+      "loss": 1.4402,
+      "step": 6405
+    },
+    {
+      "epoch": 1.10514965927715,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.41442087469994e-06,
+      "loss": 1.4904,
+      "step": 6406
+    },
+    {
+      "epoch": 1.105322177175882,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.411733487186802e-06,
+      "loss": 1.3821,
+      "step": 6407
+    },
+    {
+      "epoch": 1.105494695074614,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.409046217342367e-06,
+      "loss": 1.4523,
+      "step": 6408
+    },
+    {
+      "epoch": 1.105667212973346,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.406359065365735e-06,
+      "loss": 1.3564,
+      "step": 6409
+    },
+    {
+      "epoch": 1.105839730872078,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.403672031455977e-06,
+      "loss": 1.456,
+      "step": 6410
+    },
+    {
+      "epoch": 1.10601224877081,
+      "grad_norm": 0.625,
+      "learning_rate": 8.40098511581217e-06,
+      "loss": 1.4374,
+      "step": 6411
+    },
+    {
+      "epoch": 1.106184766669542,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.398298318633376e-06,
+      "loss": 1.449,
+      "step": 6412
+    },
+    {
+      "epoch": 1.1063572845682739,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.395611640118653e-06,
+      "loss": 1.3961,
+      "step": 6413
+    },
+    {
+      "epoch": 1.106529802467006,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.39292508046704e-06,
+      "loss": 1.3924,
+      "step": 6414
+    },
+    {
+      "epoch": 1.106702320365738,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.390238639877584e-06,
+      "loss": 1.4491,
+      "step": 6415
+    },
+    {
+      "epoch": 1.10687483826447,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.387552318549304e-06,
+      "loss": 1.4087,
+      "step": 6416
+    },
+    {
+      "epoch": 1.107047356163202,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.384866116681221e-06,
+      "loss": 1.5279,
+      "step": 6417
+    },
+    {
+      "epoch": 1.107219874061934,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.382180034472353e-06,
+      "loss": 1.4544,
+      "step": 6418
+    },
+    {
+      "epoch": 1.1073923919606659,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.379494072121695e-06,
+      "loss": 1.5367,
+      "step": 6419
+    },
+    {
+      "epoch": 1.1075649098593978,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.376808229828242e-06,
+      "loss": 1.4524,
+      "step": 6420
+    },
+    {
+      "epoch": 1.10773742775813,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.374122507790975e-06,
+      "loss": 1.4472,
+      "step": 6421
+    },
+    {
+      "epoch": 1.107909945656862,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.371436906208876e-06,
+      "loss": 1.5322,
+      "step": 6422
+    },
+    {
+      "epoch": 1.108082463555594,
+      "grad_norm": 0.77734375,
+      "learning_rate": 8.368751425280904e-06,
+      "loss": 1.4576,
+      "step": 6423
+    },
+    {
+      "epoch": 1.1082549814543259,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.36606606520602e-06,
+      "loss": 1.4851,
+      "step": 6424
+    },
+    {
+      "epoch": 1.1084274993530578,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.363380826183167e-06,
+      "loss": 1.3564,
+      "step": 6425
+    },
+    {
+      "epoch": 1.1086000172517898,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.360695708411295e-06,
+      "loss": 1.428,
+      "step": 6426
+    },
+    {
+      "epoch": 1.1087725351505218,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.358010712089324e-06,
+      "loss": 1.4501,
+      "step": 6427
+    },
+    {
+      "epoch": 1.108945053049254,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.355325837416182e-06,
+      "loss": 1.363,
+      "step": 6428
+    },
+    {
+      "epoch": 1.109117570947986,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.352641084590772e-06,
+      "loss": 1.4789,
+      "step": 6429
+    },
+    {
+      "epoch": 1.1092900888467179,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.349956453812009e-06,
+      "loss": 1.4532,
+      "step": 6430
+    },
+    {
+      "epoch": 1.1094626067454498,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.34727194527878e-06,
+      "loss": 1.412,
+      "step": 6431
+    },
+    {
+      "epoch": 1.1096351246441818,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.34458755918997e-06,
+      "loss": 1.401,
+      "step": 6432
+    },
+    {
+      "epoch": 1.1098076425429138,
+      "grad_norm": 0.625,
+      "learning_rate": 8.341903295744463e-06,
+      "loss": 1.4473,
+      "step": 6433
+    },
+    {
+      "epoch": 1.109980160441646,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.339219155141115e-06,
+      "loss": 1.3622,
+      "step": 6434
+    },
+    {
+      "epoch": 1.110152678340378,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.336535137578792e-06,
+      "loss": 1.4421,
+      "step": 6435
+    },
+    {
+      "epoch": 1.1103251962391099,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.333851243256337e-06,
+      "loss": 1.4733,
+      "step": 6436
+    },
+    {
+      "epoch": 1.1104977141378418,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.331167472372596e-06,
+      "loss": 1.3693,
+      "step": 6437
+    },
+    {
+      "epoch": 1.1106702320365738,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.328483825126393e-06,
+      "loss": 1.4495,
+      "step": 6438
+    },
+    {
+      "epoch": 1.1108427499353057,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.325800301716555e-06,
+      "loss": 1.4169,
+      "step": 6439
+    },
+    {
+      "epoch": 1.1110152678340377,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.323116902341888e-06,
+      "loss": 1.4493,
+      "step": 6440
+    },
+    {
+      "epoch": 1.1111877857327697,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.320433627201202e-06,
+      "loss": 1.3748,
+      "step": 6441
+    },
+    {
+      "epoch": 1.1113603036315018,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.317750476493282e-06,
+      "loss": 1.5276,
+      "step": 6442
+    },
+    {
+      "epoch": 1.1115328215302338,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.315067450416926e-06,
+      "loss": 1.3542,
+      "step": 6443
+    },
+    {
+      "epoch": 1.1117053394289658,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.312384549170894e-06,
+      "loss": 1.426,
+      "step": 6444
+    },
+    {
+      "epoch": 1.1118778573276977,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.309701772953964e-06,
+      "loss": 1.4683,
+      "step": 6445
+    },
+    {
+      "epoch": 1.1120503752264297,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.307019121964885e-06,
+      "loss": 1.4291,
+      "step": 6446
+    },
+    {
+      "epoch": 1.1122228931251616,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.30433659640241e-06,
+      "loss": 1.4038,
+      "step": 6447
+    },
+    {
+      "epoch": 1.1123954110238938,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.301654196465273e-06,
+      "loss": 1.422,
+      "step": 6448
+    },
+    {
+      "epoch": 1.1125679289226258,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.298971922352203e-06,
+      "loss": 1.3519,
+      "step": 6449
+    },
+    {
+      "epoch": 1.1127404468213578,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.296289774261926e-06,
+      "loss": 1.4286,
+      "step": 6450
+    },
+    {
+      "epoch": 1.1129129647200897,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.293607752393145e-06,
+      "loss": 1.3807,
+      "step": 6451
+    },
+    {
+      "epoch": 1.1130854826188217,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.290925856944567e-06,
+      "loss": 1.3515,
+      "step": 6452
+    },
+    {
+      "epoch": 1.1132580005175536,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.288244088114876e-06,
+      "loss": 1.3681,
+      "step": 6453
+    },
+    {
+      "epoch": 1.1134305184162856,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.285562446102761e-06,
+      "loss": 1.4408,
+      "step": 6454
+    },
+    {
+      "epoch": 1.1136030363150178,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.282880931106893e-06,
+      "loss": 1.4276,
+      "step": 6455
+    },
+    {
+      "epoch": 1.1137755542137497,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.280199543325935e-06,
+      "loss": 1.467,
+      "step": 6456
+    },
+    {
+      "epoch": 1.1139480721124817,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.277518282958536e-06,
+      "loss": 1.4664,
+      "step": 6457
+    },
+    {
+      "epoch": 1.1141205900112137,
+      "grad_norm": 0.6484375,
+      "learning_rate": 8.274837150203352e-06,
+      "loss": 1.4037,
+      "step": 6458
+    },
+    {
+      "epoch": 1.1142931079099456,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.272156145259006e-06,
+      "loss": 1.4159,
+      "step": 6459
+    },
+    {
+      "epoch": 1.1144656258086776,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.269475268324131e-06,
+      "loss": 1.3251,
+      "step": 6460
+    },
+    {
+      "epoch": 1.1146381437074095,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.266794519597339e-06,
+      "loss": 1.4099,
+      "step": 6461
+    },
+    {
+      "epoch": 1.1148106616061417,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.264113899277241e-06,
+      "loss": 1.3554,
+      "step": 6462
+    },
+    {
+      "epoch": 1.1149831795048737,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.26143340756243e-06,
+      "loss": 1.3989,
+      "step": 6463
+    },
+    {
+      "epoch": 1.1151556974036057,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.258753044651499e-06,
+      "loss": 1.4215,
+      "step": 6464
+    },
+    {
+      "epoch": 1.1153282153023376,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.256072810743015e-06,
+      "loss": 1.4646,
+      "step": 6465
+    },
+    {
+      "epoch": 1.1155007332010696,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.253392706035558e-06,
+      "loss": 1.4461,
+      "step": 6466
+    },
+    {
+      "epoch": 1.1156732510998015,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.250712730727685e-06,
+      "loss": 1.4213,
+      "step": 6467
+    },
+    {
+      "epoch": 1.1158457689985335,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.248032885017937e-06,
+      "loss": 1.3849,
+      "step": 6468
+    },
+    {
+      "epoch": 1.1160182868972657,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.245353169104865e-06,
+      "loss": 1.4312,
+      "step": 6469
+    },
+    {
+      "epoch": 1.1161908047959976,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.242673583186991e-06,
+      "loss": 1.3815,
+      "step": 6470
+    },
+    {
+      "epoch": 1.1163633226947296,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.239994127462842e-06,
+      "loss": 1.2654,
+      "step": 6471
+    },
+    {
+      "epoch": 1.1165358405934616,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.237314802130919e-06,
+      "loss": 1.4075,
+      "step": 6472
+    },
+    {
+      "epoch": 1.1167083584921935,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.234635607389733e-06,
+      "loss": 1.3882,
+      "step": 6473
+    },
+    {
+      "epoch": 1.1168808763909255,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.231956543437768e-06,
+      "loss": 1.4488,
+      "step": 6474
+    },
+    {
+      "epoch": 1.1170533942896577,
+      "grad_norm": 0.625,
+      "learning_rate": 8.229277610473516e-06,
+      "loss": 1.4515,
+      "step": 6475
+    },
+    {
+      "epoch": 1.1172259121883896,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.226598808695438e-06,
+      "loss": 1.4751,
+      "step": 6476
+    },
+    {
+      "epoch": 1.1173984300871216,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.223920138302006e-06,
+      "loss": 1.4667,
+      "step": 6477
+    },
+    {
+      "epoch": 1.1175709479858535,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.22124159949166e-06,
+      "loss": 1.4327,
+      "step": 6478
+    },
+    {
+      "epoch": 1.1177434658845855,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.218563192462857e-06,
+      "loss": 1.4853,
+      "step": 6479
+    },
+    {
+      "epoch": 1.1179159837833175,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.21588491741402e-06,
+      "loss": 1.432,
+      "step": 6480
+    },
+    {
+      "epoch": 1.1180885016820494,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.213206774543574e-06,
+      "loss": 1.5029,
+      "step": 6481
+    },
+    {
+      "epoch": 1.1182610195807814,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.21052876404994e-06,
+      "loss": 1.4315,
+      "step": 6482
+    },
+    {
+      "epoch": 1.1184335374795136,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.207850886131512e-06,
+      "loss": 1.3277,
+      "step": 6483
+    },
+    {
+      "epoch": 1.1186060553782455,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.205173140986691e-06,
+      "loss": 1.4484,
+      "step": 6484
+    },
+    {
+      "epoch": 1.1187785732769775,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.202495528813856e-06,
+      "loss": 1.4537,
+      "step": 6485
+    },
+    {
+      "epoch": 1.1189510911757095,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.199818049811387e-06,
+      "loss": 1.4245,
+      "step": 6486
+    },
+    {
+      "epoch": 1.1191236090744414,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.197140704177642e-06,
+      "loss": 1.412,
+      "step": 6487
+    },
+    {
+      "epoch": 1.1192961269731734,
+      "grad_norm": 0.65625,
+      "learning_rate": 8.194463492110982e-06,
+      "loss": 1.4611,
+      "step": 6488
+    },
+    {
+      "epoch": 1.1194686448719056,
+      "grad_norm": 0.625,
+      "learning_rate": 8.191786413809742e-06,
+      "loss": 1.502,
+      "step": 6489
+    },
+    {
+      "epoch": 1.1196411627706375,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.189109469472267e-06,
+      "loss": 1.44,
+      "step": 6490
+    },
+    {
+      "epoch": 1.1198136806693695,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.186432659296876e-06,
+      "loss": 1.4693,
+      "step": 6491
+    },
+    {
+      "epoch": 1.1199861985681014,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.183755983481888e-06,
+      "loss": 1.3884,
+      "step": 6492
+    },
+    {
+      "epoch": 1.1201587164668334,
+      "grad_norm": 0.625,
+      "learning_rate": 8.1810794422256e-06,
+      "loss": 1.4728,
+      "step": 6493
+    },
+    {
+      "epoch": 1.1203312343655654,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.178403035726316e-06,
+      "loss": 1.5149,
+      "step": 6494
+    },
+    {
+      "epoch": 1.1205037522642973,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.175726764182315e-06,
+      "loss": 1.4063,
+      "step": 6495
+    },
+    {
+      "epoch": 1.1206762701630295,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.173050627791877e-06,
+      "loss": 1.4264,
+      "step": 6496
+    },
+    {
+      "epoch": 1.1208487880617615,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.170374626753259e-06,
+      "loss": 1.483,
+      "step": 6497
+    },
+    {
+      "epoch": 1.1210213059604934,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.167698761264723e-06,
+      "loss": 1.4241,
+      "step": 6498
+    },
+    {
+      "epoch": 1.1211938238592254,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.165023031524513e-06,
+      "loss": 1.3938,
+      "step": 6499
+    },
+    {
+      "epoch": 1.1213663417579574,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.162347437730859e-06,
+      "loss": 1.4848,
+      "step": 6500
+    },
+    {
+      "epoch": 1.1213663417579574,
+      "eval_loss": 1.4090793132781982,
+      "eval_runtime": 10.9381,
+      "eval_samples_per_second": 93.618,
+      "eval_steps_per_second": 23.405,
+      "step": 6500
+    },
+    {
+      "epoch": 1.1215388596566893,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.159671980081994e-06,
+      "loss": 1.3506,
+      "step": 6501
+    },
+    {
+      "epoch": 1.1217113775554213,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.156996658776121e-06,
+      "loss": 1.3836,
+      "step": 6502
+    },
+    {
+      "epoch": 1.1218838954541535,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.154321474011457e-06,
+      "loss": 1.368,
+      "step": 6503
+    },
+    {
+      "epoch": 1.1220564133528854,
+      "grad_norm": 0.7109375,
+      "learning_rate": 8.151646425986187e-06,
+      "loss": 1.3849,
+      "step": 6504
+    },
+    {
+      "epoch": 1.1222289312516174,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.148971514898503e-06,
+      "loss": 1.4842,
+      "step": 6505
+    },
+    {
+      "epoch": 1.1224014491503493,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.14629674094657e-06,
+      "loss": 1.5188,
+      "step": 6506
+    },
+    {
+      "epoch": 1.1225739670490813,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.14362210432856e-06,
+      "loss": 1.3948,
+      "step": 6507
+    },
+    {
+      "epoch": 1.1227464849478133,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.140947605242622e-06,
+      "loss": 1.3766,
+      "step": 6508
+    },
+    {
+      "epoch": 1.1229190028465452,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.138273243886902e-06,
+      "loss": 1.4794,
+      "step": 6509
+    },
+    {
+      "epoch": 1.1230915207452774,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.135599020459531e-06,
+      "loss": 1.4689,
+      "step": 6510
+    },
+    {
+      "epoch": 1.1232640386440094,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.132924935158638e-06,
+      "loss": 1.4573,
+      "step": 6511
+    },
+    {
+      "epoch": 1.1234365565427413,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.130250988182328e-06,
+      "loss": 1.4931,
+      "step": 6512
+    },
+    {
+      "epoch": 1.1236090744414733,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.127577179728708e-06,
+      "loss": 1.3708,
+      "step": 6513
+    },
+    {
+      "epoch": 1.1237815923402052,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.124903509995872e-06,
+      "loss": 1.4501,
+      "step": 6514
+    },
+    {
+      "epoch": 1.1239541102389372,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.122229979181899e-06,
+      "loss": 1.3516,
+      "step": 6515
+    },
+    {
+      "epoch": 1.1241266281376694,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.119556587484863e-06,
+      "loss": 1.4109,
+      "step": 6516
+    },
+    {
+      "epoch": 1.1242991460364014,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.116883335102821e-06,
+      "loss": 1.4759,
+      "step": 6517
+    },
+    {
+      "epoch": 1.1244716639351333,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.114210222233832e-06,
+      "loss": 1.3568,
+      "step": 6518
+    },
+    {
+      "epoch": 1.1246441818338653,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.11153724907593e-06,
+      "loss": 1.3548,
+      "step": 6519
+    },
+    {
+      "epoch": 1.1248166997325972,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.108864415827152e-06,
+      "loss": 1.4087,
+      "step": 6520
+    },
+    {
+      "epoch": 1.1249892176313292,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.10619172268551e-06,
+      "loss": 1.4023,
+      "step": 6521
+    },
+    {
+      "epoch": 1.1251617355300612,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.10351916984902e-06,
+      "loss": 1.4742,
+      "step": 6522
+    },
+    {
+      "epoch": 1.1253342534287931,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.10084675751568e-06,
+      "loss": 1.4095,
+      "step": 6523
+    },
+    {
+      "epoch": 1.1255067713275253,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.09817448588348e-06,
+      "loss": 1.4618,
+      "step": 6524
+    },
+    {
+      "epoch": 1.1256792892262573,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.095502355150392e-06,
+      "loss": 1.3808,
+      "step": 6525
+    },
+    {
+      "epoch": 1.1258518071249892,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.092830365514395e-06,
+      "loss": 1.4675,
+      "step": 6526
+    },
+    {
+      "epoch": 1.1260243250237212,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.090158517173438e-06,
+      "loss": 1.4233,
+      "step": 6527
+    },
+    {
+      "epoch": 1.1261968429224531,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.087486810325475e-06,
+      "loss": 1.5255,
+      "step": 6528
+    },
+    {
+      "epoch": 1.126369360821185,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.084815245168434e-06,
+      "loss": 1.4081,
+      "step": 6529
+    },
+    {
+      "epoch": 1.1265418787199173,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.082143821900246e-06,
+      "loss": 1.4038,
+      "step": 6530
+    },
+    {
+      "epoch": 1.1267143966186492,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.079472540718833e-06,
+      "loss": 1.4907,
+      "step": 6531
+    },
+    {
+      "epoch": 1.1268869145173812,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.076801401822088e-06,
+      "loss": 1.413,
+      "step": 6532
+    },
+    {
+      "epoch": 1.1270594324161132,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.074130405407915e-06,
+      "loss": 1.4993,
+      "step": 6533
+    },
+    {
+      "epoch": 1.1272319503148451,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.071459551674193e-06,
+      "loss": 1.4363,
+      "step": 6534
+    },
+    {
+      "epoch": 1.127404468213577,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.0687888408188e-06,
+      "loss": 1.4667,
+      "step": 6535
+    },
+    {
+      "epoch": 1.127576986112309,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.066118273039597e-06,
+      "loss": 1.5419,
+      "step": 6536
+    },
+    {
+      "epoch": 1.1277495040110412,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.063447848534435e-06,
+      "loss": 1.4496,
+      "step": 6537
+    },
+    {
+      "epoch": 1.1279220219097732,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.060777567501154e-06,
+      "loss": 1.5354,
+      "step": 6538
+    },
+    {
+      "epoch": 1.1280945398085052,
+      "grad_norm": 0.65234375,
+      "learning_rate": 8.05810743013759e-06,
+      "loss": 1.5538,
+      "step": 6539
+    },
+    {
+      "epoch": 1.1282670577072371,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.05543743664156e-06,
+      "loss": 1.4138,
+      "step": 6540
+    },
+    {
+      "epoch": 1.128439575605969,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.052767587210878e-06,
+      "loss": 1.3768,
+      "step": 6541
+    },
+    {
+      "epoch": 1.128612093504701,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.050097882043334e-06,
+      "loss": 1.494,
+      "step": 6542
+    },
+    {
+      "epoch": 1.1287846114034332,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.047428321336728e-06,
+      "loss": 1.3528,
+      "step": 6543
+    },
+    {
+      "epoch": 1.1289571293021652,
+      "grad_norm": 0.63671875,
+      "learning_rate": 8.04475890528883e-06,
+      "loss": 1.4149,
+      "step": 6544
+    },
+    {
+      "epoch": 1.1291296472008971,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.042089634097406e-06,
+      "loss": 1.5056,
+      "step": 6545
+    },
+    {
+      "epoch": 1.129302165099629,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.03942050796022e-06,
+      "loss": 1.4004,
+      "step": 6546
+    },
+    {
+      "epoch": 1.129474682998361,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.036751527075011e-06,
+      "loss": 1.476,
+      "step": 6547
+    },
+    {
+      "epoch": 1.129647200897093,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.034082691639519e-06,
+      "loss": 1.4687,
+      "step": 6548
+    },
+    {
+      "epoch": 1.129819718795825,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.031414001851459e-06,
+      "loss": 1.5199,
+      "step": 6549
+    },
+    {
+      "epoch": 1.129992236694557,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.028745457908555e-06,
+      "loss": 1.3649,
+      "step": 6550
+    },
+    {
+      "epoch": 1.1301647545932891,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.026077060008503e-06,
+      "loss": 1.4537,
+      "step": 6551
+    },
+    {
+      "epoch": 1.130337272492021,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.023408808348999e-06,
+      "loss": 1.4067,
+      "step": 6552
+    },
+    {
+      "epoch": 1.130509790390753,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.020740703127715e-06,
+      "loss": 1.4795,
+      "step": 6553
+    },
+    {
+      "epoch": 1.130682308289485,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.01807274454233e-06,
+      "loss": 1.472,
+      "step": 6554
+    },
+    {
+      "epoch": 1.130854826188217,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.015404932790499e-06,
+      "loss": 1.4342,
+      "step": 6555
+    },
+    {
+      "epoch": 1.131027344086949,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.012737268069873e-06,
+      "loss": 1.3848,
+      "step": 6556
+    },
+    {
+      "epoch": 1.1311998619856811,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.010069750578082e-06,
+      "loss": 1.3666,
+      "step": 6557
+    },
+    {
+      "epoch": 1.131372379884413,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.007402380512763e-06,
+      "loss": 1.3574,
+      "step": 6558
+    },
+    {
+      "epoch": 1.131544897783145,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.004735158071524e-06,
+      "loss": 1.5037,
+      "step": 6559
+    },
+    {
+      "epoch": 1.131717415681877,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.002068083451973e-06,
+      "loss": 1.4382,
+      "step": 6560
+    },
+    {
+      "epoch": 1.131889933580609,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.999401156851697e-06,
+      "loss": 1.452,
+      "step": 6561
+    },
+    {
+      "epoch": 1.132062451479341,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.996734378468284e-06,
+      "loss": 1.4871,
+      "step": 6562
+    },
+    {
+      "epoch": 1.1322349693780729,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.99406774849931e-06,
+      "loss": 1.4092,
+      "step": 6563
+    },
+    {
+      "epoch": 1.1324074872768048,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.991401267142329e-06,
+      "loss": 1.3844,
+      "step": 6564
+    },
+    {
+      "epoch": 1.132580005175537,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.988734934594893e-06,
+      "loss": 1.4034,
+      "step": 6565
+    },
+    {
+      "epoch": 1.132752523074269,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.986068751054537e-06,
+      "loss": 1.4734,
+      "step": 6566
+    },
+    {
+      "epoch": 1.132925040973001,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.983402716718796e-06,
+      "loss": 1.3407,
+      "step": 6567
+    },
+    {
+      "epoch": 1.133097558871733,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.980736831785178e-06,
+      "loss": 1.5068,
+      "step": 6568
+    },
+    {
+      "epoch": 1.1332700767704649,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.978071096451198e-06,
+      "loss": 1.4126,
+      "step": 6569
+    },
+    {
+      "epoch": 1.1334425946691968,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.975405510914338e-06,
+      "loss": 1.3754,
+      "step": 6570
+    },
+    {
+      "epoch": 1.133615112567929,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.972740075372094e-06,
+      "loss": 1.4839,
+      "step": 6571
+    },
+    {
+      "epoch": 1.133787630466661,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.970074790021928e-06,
+      "loss": 1.379,
+      "step": 6572
+    },
+    {
+      "epoch": 1.133960148365393,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.967409655061308e-06,
+      "loss": 1.4018,
+      "step": 6573
+    },
+    {
+      "epoch": 1.134132666264125,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.964744670687676e-06,
+      "loss": 1.3949,
+      "step": 6574
+    },
+    {
+      "epoch": 1.1343051841628569,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.962079837098481e-06,
+      "loss": 1.4509,
+      "step": 6575
+    },
+    {
+      "epoch": 1.1344777020615888,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.959415154491142e-06,
+      "loss": 1.3816,
+      "step": 6576
+    },
+    {
+      "epoch": 1.1346502199603208,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.956750623063077e-06,
+      "loss": 1.4659,
+      "step": 6577
+    },
+    {
+      "epoch": 1.134822737859053,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.954086243011698e-06,
+      "loss": 1.3871,
+      "step": 6578
+    },
+    {
+      "epoch": 1.134995255757785,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.95142201453439e-06,
+      "loss": 1.354,
+      "step": 6579
+    },
+    {
+      "epoch": 1.1351677736565169,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.948757937828542e-06,
+      "loss": 1.405,
+      "step": 6580
+    },
+    {
+      "epoch": 1.1353402915552488,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.946094013091518e-06,
+      "loss": 1.4215,
+      "step": 6581
+    },
+    {
+      "epoch": 1.1355128094539808,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.943430240520689e-06,
+      "loss": 1.3913,
+      "step": 6582
+    },
+    {
+      "epoch": 1.1356853273527128,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.940766620313394e-06,
+      "loss": 1.394,
+      "step": 6583
+    },
+    {
+      "epoch": 1.135857845251445,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.938103152666976e-06,
+      "loss": 1.374,
+      "step": 6584
+    },
+    {
+      "epoch": 1.136030363150177,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.935439837778757e-06,
+      "loss": 1.3971,
+      "step": 6585
+    },
+    {
+      "epoch": 1.1362028810489089,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.932776675846058e-06,
+      "loss": 1.4629,
+      "step": 6586
+    },
+    {
+      "epoch": 1.1363753989476408,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.930113667066177e-06,
+      "loss": 1.4402,
+      "step": 6587
+    },
+    {
+      "epoch": 1.1365479168463728,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.927450811636413e-06,
+      "loss": 1.3988,
+      "step": 6588
+    },
+    {
+      "epoch": 1.1367204347451048,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.924788109754036e-06,
+      "loss": 1.355,
+      "step": 6589
+    },
+    {
+      "epoch": 1.1368929526438367,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.922125561616329e-06,
+      "loss": 1.397,
+      "step": 6590
+    },
+    {
+      "epoch": 1.1370654705425687,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.919463167420538e-06,
+      "loss": 1.5036,
+      "step": 6591
+    },
+    {
+      "epoch": 1.1372379884413009,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.91680092736392e-06,
+      "loss": 1.4029,
+      "step": 6592
+    },
+    {
+      "epoch": 1.1374105063400328,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.914138841643702e-06,
+      "loss": 1.4439,
+      "step": 6593
+    },
+    {
+      "epoch": 1.1375830242387648,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.91147691045711e-06,
+      "loss": 1.4961,
+      "step": 6594
+    },
+    {
+      "epoch": 1.1377555421374967,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.908815134001363e-06,
+      "loss": 1.3272,
+      "step": 6595
+    },
+    {
+      "epoch": 1.1379280600362287,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.906153512473656e-06,
+      "loss": 1.4737,
+      "step": 6596
+    },
+    {
+      "epoch": 1.1381005779349607,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.903492046071182e-06,
+      "loss": 1.3569,
+      "step": 6597
+    },
+    {
+      "epoch": 1.1382730958336928,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.900830734991111e-06,
+      "loss": 1.4397,
+      "step": 6598
+    },
+    {
+      "epoch": 1.1384456137324248,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.89816957943062e-06,
+      "loss": 1.3377,
+      "step": 6599
+    },
+    {
+      "epoch": 1.1386181316311568,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.895508579586857e-06,
+      "loss": 1.5259,
+      "step": 6600
+    },
+    {
+      "epoch": 1.1386181316311568,
+      "eval_loss": 1.408766746520996,
+      "eval_runtime": 10.7693,
+      "eval_samples_per_second": 95.085,
+      "eval_steps_per_second": 23.771,
+      "step": 6600
+    },
+    {
+      "epoch": 1.1387906495298887,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.892847735656972e-06,
+      "loss": 1.528,
+      "step": 6601
+    },
+    {
+      "epoch": 1.1389631674286207,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.890187047838087e-06,
+      "loss": 1.4519,
+      "step": 6602
+    },
+    {
+      "epoch": 1.1391356853273527,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.887526516327334e-06,
+      "loss": 1.4995,
+      "step": 6603
+    },
+    {
+      "epoch": 1.1393082032260846,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.884866141321811e-06,
+      "loss": 1.4422,
+      "step": 6604
+    },
+    {
+      "epoch": 1.1394807211248166,
+      "grad_norm": 0.54296875,
+      "learning_rate": 7.882205923018624e-06,
+      "loss": 1.4153,
+      "step": 6605
+    },
+    {
+      "epoch": 1.1396532390235488,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.879545861614851e-06,
+      "loss": 1.4228,
+      "step": 6606
+    },
+    {
+      "epoch": 1.1398257569222807,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.876885957307573e-06,
+      "loss": 1.3649,
+      "step": 6607
+    },
+    {
+      "epoch": 1.1399982748210127,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.874226210293847e-06,
+      "loss": 1.4664,
+      "step": 6608
+    },
+    {
+      "epoch": 1.1401707927197446,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.871566620770726e-06,
+      "loss": 1.4722,
+      "step": 6609
+    },
+    {
+      "epoch": 1.1403433106184766,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.86890718893525e-06,
+      "loss": 1.3472,
+      "step": 6610
+    },
+    {
+      "epoch": 1.1405158285172088,
+      "grad_norm": 0.8515625,
+      "learning_rate": 7.866247914984444e-06,
+      "loss": 1.3521,
+      "step": 6611
+    },
+    {
+      "epoch": 1.1406883464159407,
+      "grad_norm": 0.82421875,
+      "learning_rate": 7.863588799115327e-06,
+      "loss": 1.3674,
+      "step": 6612
+    },
+    {
+      "epoch": 1.1408608643146727,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.860929841524898e-06,
+      "loss": 1.3509,
+      "step": 6613
+    },
+    {
+      "epoch": 1.1410333822134047,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.858271042410153e-06,
+      "loss": 1.4391,
+      "step": 6614
+    },
+    {
+      "epoch": 1.1412059001121366,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.855612401968072e-06,
+      "loss": 1.5101,
+      "step": 6615
+    },
+    {
+      "epoch": 1.1413784180108686,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.852953920395623e-06,
+      "loss": 1.4194,
+      "step": 6616
+    },
+    {
+      "epoch": 1.1415509359096006,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.85029559788976e-06,
+      "loss": 1.458,
+      "step": 6617
+    },
+    {
+      "epoch": 1.1417234538083325,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.847637434647436e-06,
+      "loss": 1.4617,
+      "step": 6618
+    },
+    {
+      "epoch": 1.1418959717070647,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.844979430865575e-06,
+      "loss": 1.3958,
+      "step": 6619
+    },
+    {
+      "epoch": 1.1420684896057967,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.842321586741107e-06,
+      "loss": 1.516,
+      "step": 6620
+    },
+    {
+      "epoch": 1.1422410075045286,
+      "grad_norm": 0.86328125,
+      "learning_rate": 7.839663902470933e-06,
+      "loss": 1.4452,
+      "step": 6621
+    },
+    {
+      "epoch": 1.1424135254032606,
+      "grad_norm": 4.40625,
+      "learning_rate": 7.837006378251959e-06,
+      "loss": 1.3733,
+      "step": 6622
+    },
+    {
+      "epoch": 1.1425860433019925,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.834349014281065e-06,
+      "loss": 1.4408,
+      "step": 6623
+    },
+    {
+      "epoch": 1.1427585612007245,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.83169181075513e-06,
+      "loss": 1.4705,
+      "step": 6624
+    },
+    {
+      "epoch": 1.1429310790994567,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.82903476787101e-06,
+      "loss": 1.3875,
+      "step": 6625
+    },
+    {
+      "epoch": 1.1431035969981886,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.826377885825561e-06,
+      "loss": 1.5407,
+      "step": 6626
+    },
+    {
+      "epoch": 1.1432761148969206,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.823721164815624e-06,
+      "loss": 1.4636,
+      "step": 6627
+    },
+    {
+      "epoch": 1.1434486327956526,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.821064605038016e-06,
+      "loss": 1.4006,
+      "step": 6628
+    },
+    {
+      "epoch": 1.1436211506943845,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.818408206689561e-06,
+      "loss": 1.4363,
+      "step": 6629
+    },
+    {
+      "epoch": 1.1437936685931165,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.815751969967052e-06,
+      "loss": 1.3846,
+      "step": 6630
+    },
+    {
+      "epoch": 1.1439661864918484,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.813095895067289e-06,
+      "loss": 1.4288,
+      "step": 6631
+    },
+    {
+      "epoch": 1.1441387043905804,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.810439982187045e-06,
+      "loss": 1.3967,
+      "step": 6632
+    },
+    {
+      "epoch": 1.1443112222893126,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.80778423152309e-06,
+      "loss": 1.3882,
+      "step": 6633
+    },
+    {
+      "epoch": 1.1444837401880446,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.805128643272171e-06,
+      "loss": 1.4812,
+      "step": 6634
+    },
+    {
+      "epoch": 1.1446562580867765,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.802473217631043e-06,
+      "loss": 1.4707,
+      "step": 6635
+    },
+    {
+      "epoch": 1.1448287759855085,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.799817954796427e-06,
+      "loss": 1.4223,
+      "step": 6636
+    },
+    {
+      "epoch": 1.1450012938842404,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.797162854965046e-06,
+      "loss": 1.401,
+      "step": 6637
+    },
+    {
+      "epoch": 1.1451738117829724,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.7945079183336e-06,
+      "loss": 1.3815,
+      "step": 6638
+    },
+    {
+      "epoch": 1.1453463296817046,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.791853145098792e-06,
+      "loss": 1.4514,
+      "step": 6639
+    },
+    {
+      "epoch": 1.1455188475804365,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.7891985354573e-06,
+      "loss": 1.4598,
+      "step": 6640
+    },
+    {
+      "epoch": 1.1456913654791685,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.78654408960579e-06,
+      "loss": 1.3204,
+      "step": 6641
+    },
+    {
+      "epoch": 1.1458638833779005,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.78388980774093e-06,
+      "loss": 1.5014,
+      "step": 6642
+    },
+    {
+      "epoch": 1.1460364012766324,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.781235690059356e-06,
+      "loss": 1.3385,
+      "step": 6643
+    },
+    {
+      "epoch": 1.1462089191753644,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.77858173675771e-06,
+      "loss": 1.4659,
+      "step": 6644
+    },
+    {
+      "epoch": 1.1463814370740963,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.775927948032602e-06,
+      "loss": 1.3328,
+      "step": 6645
+    },
+    {
+      "epoch": 1.1465539549728285,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.773274324080655e-06,
+      "loss": 1.4962,
+      "step": 6646
+    },
+    {
+      "epoch": 1.1467264728715605,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.770620865098455e-06,
+      "loss": 1.3894,
+      "step": 6647
+    },
+    {
+      "epoch": 1.1468989907702924,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.767967571282595e-06,
+      "loss": 1.4642,
+      "step": 6648
+    },
+    {
+      "epoch": 1.1470715086690244,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.76531444282964e-06,
+      "loss": 1.4378,
+      "step": 6649
+    },
+    {
+      "epoch": 1.1472440265677564,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.762661479936157e-06,
+      "loss": 1.4204,
+      "step": 6650
+    },
+    {
+      "epoch": 1.1474165444664883,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.760008682798687e-06,
+      "loss": 1.5035,
+      "step": 6651
+    },
+    {
+      "epoch": 1.1475890623652205,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.757356051613774e-06,
+      "loss": 1.4768,
+      "step": 6652
+    },
+    {
+      "epoch": 1.1477615802639525,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.754703586577935e-06,
+      "loss": 1.4607,
+      "step": 6653
+    },
+    {
+      "epoch": 1.1479340981626844,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.752051287887685e-06,
+      "loss": 1.3546,
+      "step": 6654
+    },
+    {
+      "epoch": 1.1481066160614164,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.74939915573952e-06,
+      "loss": 1.3577,
+      "step": 6655
+    },
+    {
+      "epoch": 1.1482791339601484,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.74674719032993e-06,
+      "loss": 1.402,
+      "step": 6656
+    },
+    {
+      "epoch": 1.1484516518588803,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.744095391855386e-06,
+      "loss": 1.4069,
+      "step": 6657
+    },
+    {
+      "epoch": 1.1486241697576123,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.741443760512348e-06,
+      "loss": 1.4762,
+      "step": 6658
+    },
+    {
+      "epoch": 1.1487966876563442,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.738792296497272e-06,
+      "loss": 1.416,
+      "step": 6659
+    },
+    {
+      "epoch": 1.1489692055550764,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.736141000006589e-06,
+      "loss": 1.4218,
+      "step": 6660
+    },
+    {
+      "epoch": 1.1491417234538084,
+      "grad_norm": 0.734375,
+      "learning_rate": 7.73348987123673e-06,
+      "loss": 1.4855,
+      "step": 6661
+    },
+    {
+      "epoch": 1.1493142413525403,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.730838910384098e-06,
+      "loss": 1.4101,
+      "step": 6662
+    },
+    {
+      "epoch": 1.1494867592512723,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.728188117645103e-06,
+      "loss": 1.3949,
+      "step": 6663
+    },
+    {
+      "epoch": 1.1496592771500043,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.725537493216125e-06,
+      "loss": 1.3943,
+      "step": 6664
+    },
+    {
+      "epoch": 1.1498317950487362,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.72288703729354e-06,
+      "loss": 1.3493,
+      "step": 6665
+    },
+    {
+      "epoch": 1.1500043129474684,
+      "grad_norm": 1.0546875,
+      "learning_rate": 7.72023675007371e-06,
+      "loss": 1.3391,
+      "step": 6666
+    },
+    {
+      "epoch": 1.1501768308462004,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.71758663175299e-06,
+      "loss": 1.3756,
+      "step": 6667
+    },
+    {
+      "epoch": 1.1503493487449323,
+      "grad_norm": 0.625,
+      "learning_rate": 7.714936682527712e-06,
+      "loss": 1.4111,
+      "step": 6668
+    },
+    {
+      "epoch": 1.1505218666436643,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.712286902594205e-06,
+      "loss": 1.3056,
+      "step": 6669
+    },
+    {
+      "epoch": 1.1506943845423963,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.709637292148771e-06,
+      "loss": 1.4415,
+      "step": 6670
+    },
+    {
+      "epoch": 1.1508669024411282,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.706987851387724e-06,
+      "loss": 1.4628,
+      "step": 6671
+    },
+    {
+      "epoch": 1.1510394203398602,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.704338580507341e-06,
+      "loss": 1.4112,
+      "step": 6672
+    },
+    {
+      "epoch": 1.1512119382385921,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.701689479703899e-06,
+      "loss": 1.422,
+      "step": 6673
+    },
+    {
+      "epoch": 1.1513844561373243,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.699040549173664e-06,
+      "loss": 1.4545,
+      "step": 6674
+    },
+    {
+      "epoch": 1.1515569740360563,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.69639178911288e-06,
+      "loss": 1.3825,
+      "step": 6675
+    },
+    {
+      "epoch": 1.1517294919347882,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.693743199717789e-06,
+      "loss": 1.4206,
+      "step": 6676
+    },
+    {
+      "epoch": 1.1519020098335202,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.691094781184608e-06,
+      "loss": 1.4968,
+      "step": 6677
+    },
+    {
+      "epoch": 1.1520745277322522,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.688446533709556e-06,
+      "loss": 1.4569,
+      "step": 6678
+    },
+    {
+      "epoch": 1.1522470456309841,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.685798457488824e-06,
+      "loss": 1.435,
+      "step": 6679
+    },
+    {
+      "epoch": 1.1524195635297163,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.683150552718608e-06,
+      "loss": 1.4661,
+      "step": 6680
+    },
+    {
+      "epoch": 1.1525920814284483,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.680502819595067e-06,
+      "loss": 1.3847,
+      "step": 6681
+    },
+    {
+      "epoch": 1.1527645993271802,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.677855258314378e-06,
+      "loss": 1.5205,
+      "step": 6682
+    },
+    {
+      "epoch": 1.1529371172259122,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.675207869072675e-06,
+      "loss": 1.4652,
+      "step": 6683
+    },
+    {
+      "epoch": 1.1531096351246442,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.672560652066104e-06,
+      "loss": 1.3736,
+      "step": 6684
+    },
+    {
+      "epoch": 1.153282153023376,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.66991360749078e-06,
+      "loss": 1.4111,
+      "step": 6685
+    },
+    {
+      "epoch": 1.153454670922108,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.667266735542816e-06,
+      "loss": 1.4514,
+      "step": 6686
+    },
+    {
+      "epoch": 1.1536271888208403,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.664620036418304e-06,
+      "loss": 1.3545,
+      "step": 6687
+    },
+    {
+      "epoch": 1.1537997067195722,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.661973510313336e-06,
+      "loss": 1.2812,
+      "step": 6688
+    },
+    {
+      "epoch": 1.1539722246183042,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.659327157423977e-06,
+      "loss": 1.4696,
+      "step": 6689
+    },
+    {
+      "epoch": 1.1541447425170361,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.656680977946286e-06,
+      "loss": 1.4847,
+      "step": 6690
+    },
+    {
+      "epoch": 1.154317260415768,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.654034972076314e-06,
+      "loss": 1.3304,
+      "step": 6691
+    },
+    {
+      "epoch": 1.1544897783145,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.651389140010087e-06,
+      "loss": 1.3784,
+      "step": 6692
+    },
+    {
+      "epoch": 1.1546622962132322,
+      "grad_norm": 0.78125,
+      "learning_rate": 7.648743481943628e-06,
+      "loss": 1.4342,
+      "step": 6693
+    },
+    {
+      "epoch": 1.1548348141119642,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.646097998072941e-06,
+      "loss": 1.4824,
+      "step": 6694
+    },
+    {
+      "epoch": 1.1550073320106962,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.643452688594026e-06,
+      "loss": 1.3911,
+      "step": 6695
+    },
+    {
+      "epoch": 1.1551798499094281,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.640807553702858e-06,
+      "loss": 1.3827,
+      "step": 6696
+    },
+    {
+      "epoch": 1.15535236780816,
+      "grad_norm": 0.71484375,
+      "learning_rate": 7.63816259359541e-06,
+      "loss": 1.3959,
+      "step": 6697
+    },
+    {
+      "epoch": 1.155524885706892,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.63551780846763e-06,
+      "loss": 1.4016,
+      "step": 6698
+    },
+    {
+      "epoch": 1.155697403605624,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.632873198515468e-06,
+      "loss": 1.4195,
+      "step": 6699
+    },
+    {
+      "epoch": 1.155869921504356,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.630228763934848e-06,
+      "loss": 1.396,
+      "step": 6700
+    },
+    {
+      "epoch": 1.155869921504356,
+      "eval_loss": 1.408650517463684,
+      "eval_runtime": 10.8332,
+      "eval_samples_per_second": 94.525,
+      "eval_steps_per_second": 23.631,
+      "step": 6700
+    },
+    {
+      "epoch": 1.1560424394030882,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.6275845049216914e-06,
+      "loss": 1.4485,
+      "step": 6701
+    },
+    {
+      "epoch": 1.1562149573018201,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.624940421671893e-06,
+      "loss": 1.4597,
+      "step": 6702
+    },
+    {
+      "epoch": 1.156387475200552,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.622296514381353e-06,
+      "loss": 1.4152,
+      "step": 6703
+    },
+    {
+      "epoch": 1.156559993099284,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.619652783245941e-06,
+      "loss": 1.4197,
+      "step": 6704
+    },
+    {
+      "epoch": 1.156732510998016,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.617009228461527e-06,
+      "loss": 1.4941,
+      "step": 6705
+    },
+    {
+      "epoch": 1.156905028896748,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.6143658502239546e-06,
+      "loss": 1.4696,
+      "step": 6706
+    },
+    {
+      "epoch": 1.1570775467954801,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.611722648729065e-06,
+      "loss": 1.3503,
+      "step": 6707
+    },
+    {
+      "epoch": 1.157250064694212,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.609079624172692e-06,
+      "loss": 1.4071,
+      "step": 6708
+    },
+    {
+      "epoch": 1.157422582592944,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.606436776750632e-06,
+      "loss": 1.3535,
+      "step": 6709
+    },
+    {
+      "epoch": 1.157595100491676,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.603794106658696e-06,
+      "loss": 1.515,
+      "step": 6710
+    },
+    {
+      "epoch": 1.157767618390408,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.601151614092661e-06,
+      "loss": 1.4618,
+      "step": 6711
+    },
+    {
+      "epoch": 1.15794013628914,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.598509299248307e-06,
+      "loss": 1.4592,
+      "step": 6712
+    },
+    {
+      "epoch": 1.158112654187872,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.595867162321388e-06,
+      "loss": 1.4808,
+      "step": 6713
+    },
+    {
+      "epoch": 1.1582851720866039,
+      "grad_norm": 0.8125,
+      "learning_rate": 7.593225203507652e-06,
+      "loss": 1.3933,
+      "step": 6714
+    },
+    {
+      "epoch": 1.158457689985336,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.590583423002828e-06,
+      "loss": 1.5316,
+      "step": 6715
+    },
+    {
+      "epoch": 1.158630207884068,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.5879418210026425e-06,
+      "loss": 1.4468,
+      "step": 6716
+    },
+    {
+      "epoch": 1.1588027257828,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.585300397702795e-06,
+      "loss": 1.4647,
+      "step": 6717
+    },
+    {
+      "epoch": 1.158975243681532,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.5826591532989855e-06,
+      "loss": 1.4042,
+      "step": 6718
+    },
+    {
+      "epoch": 1.159147761580264,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.580018087986886e-06,
+      "loss": 1.4184,
+      "step": 6719
+    },
+    {
+      "epoch": 1.1593202794789959,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.57737720196217e-06,
+      "loss": 1.5034,
+      "step": 6720
+    },
+    {
+      "epoch": 1.159492797377728,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.574736495420487e-06,
+      "loss": 1.4559,
+      "step": 6721
+    },
+    {
+      "epoch": 1.15966531527646,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.572095968557476e-06,
+      "loss": 1.4681,
+      "step": 6722
+    },
+    {
+      "epoch": 1.159837833175192,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.56945562156877e-06,
+      "loss": 1.4418,
+      "step": 6723
+    },
+    {
+      "epoch": 1.160010351073924,
+      "grad_norm": 0.625,
+      "learning_rate": 7.566815454649976e-06,
+      "loss": 1.3457,
+      "step": 6724
+    },
+    {
+      "epoch": 1.1601828689726559,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.5641754679967e-06,
+      "loss": 1.3257,
+      "step": 6725
+    },
+    {
+      "epoch": 1.1603553868713878,
+      "grad_norm": 0.68359375,
+      "learning_rate": 7.561535661804519e-06,
+      "loss": 1.4105,
+      "step": 6726
+    },
+    {
+      "epoch": 1.1605279047701198,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.558896036269017e-06,
+      "loss": 1.4474,
+      "step": 6727
+    },
+    {
+      "epoch": 1.160700422668852,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.556256591585747e-06,
+      "loss": 1.4649,
+      "step": 6728
+    },
+    {
+      "epoch": 1.160872940567584,
+      "grad_norm": 0.625,
+      "learning_rate": 7.5536173279502615e-06,
+      "loss": 1.423,
+      "step": 6729
+    },
+    {
+      "epoch": 1.161045458466316,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.550978245558084e-06,
+      "loss": 1.4107,
+      "step": 6730
+    },
+    {
+      "epoch": 1.1612179763650479,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.548339344604745e-06,
+      "loss": 1.4978,
+      "step": 6731
+    },
+    {
+      "epoch": 1.1613904942637798,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.5457006252857445e-06,
+      "loss": 1.5156,
+      "step": 6732
+    },
+    {
+      "epoch": 1.1615630121625118,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.543062087796579e-06,
+      "loss": 1.3659,
+      "step": 6733
+    },
+    {
+      "epoch": 1.161735530061244,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.540423732332721e-06,
+      "loss": 1.4551,
+      "step": 6734
+    },
+    {
+      "epoch": 1.161908047959976,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.537785559089646e-06,
+      "loss": 1.3231,
+      "step": 6735
+    },
+    {
+      "epoch": 1.162080565858708,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.535147568262799e-06,
+      "loss": 1.4627,
+      "step": 6736
+    },
+    {
+      "epoch": 1.1622530837574399,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.532509760047621e-06,
+      "loss": 1.4135,
+      "step": 6737
+    },
+    {
+      "epoch": 1.1624256016561718,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.529872134639537e-06,
+      "loss": 1.3578,
+      "step": 6738
+    },
+    {
+      "epoch": 1.1625981195549038,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.527234692233957e-06,
+      "loss": 1.3597,
+      "step": 6739
+    },
+    {
+      "epoch": 1.1627706374536357,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.524597433026286e-06,
+      "loss": 1.3543,
+      "step": 6740
+    },
+    {
+      "epoch": 1.1629431553523677,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.521960357211904e-06,
+      "loss": 1.4293,
+      "step": 6741
+    },
+    {
+      "epoch": 1.1631156732510999,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.519323464986182e-06,
+      "loss": 1.4101,
+      "step": 6742
+    },
+    {
+      "epoch": 1.1632881911498318,
+      "grad_norm": 1.1640625,
+      "learning_rate": 7.516686756544475e-06,
+      "loss": 1.3559,
+      "step": 6743
+    },
+    {
+      "epoch": 1.1634607090485638,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.514050232082133e-06,
+      "loss": 1.5265,
+      "step": 6744
+    },
+    {
+      "epoch": 1.1636332269472958,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.511413891794482e-06,
+      "loss": 1.4018,
+      "step": 6745
+    },
+    {
+      "epoch": 1.1638057448460277,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.508777735876839e-06,
+      "loss": 1.4592,
+      "step": 6746
+    },
+    {
+      "epoch": 1.1639782627447597,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.5061417645245034e-06,
+      "loss": 1.3904,
+      "step": 6747
+    },
+    {
+      "epoch": 1.1641507806434919,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.503505977932775e-06,
+      "loss": 1.4482,
+      "step": 6748
+    },
+    {
+      "epoch": 1.1643232985422238,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.500870376296918e-06,
+      "loss": 1.4214,
+      "step": 6749
+    },
+    {
+      "epoch": 1.1644958164409558,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.4982349598122e-06,
+      "loss": 1.3156,
+      "step": 6750
+    },
+    {
+      "epoch": 1.1646683343396877,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.495599728673867e-06,
+      "loss": 1.3671,
+      "step": 6751
+    },
+    {
+      "epoch": 1.1648408522384197,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.492964683077156e-06,
+      "loss": 1.4576,
+      "step": 6752
+    },
+    {
+      "epoch": 1.1650133701371517,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.490329823217286e-06,
+      "loss": 1.321,
+      "step": 6753
+    },
+    {
+      "epoch": 1.1651858880358836,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.48769514928946e-06,
+      "loss": 1.3814,
+      "step": 6754
+    },
+    {
+      "epoch": 1.1653584059346156,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.485060661488879e-06,
+      "loss": 1.4575,
+      "step": 6755
+    },
+    {
+      "epoch": 1.1655309238333478,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.482426360010717e-06,
+      "loss": 1.442,
+      "step": 6756
+    },
+    {
+      "epoch": 1.1657034417320797,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.479792245050142e-06,
+      "loss": 1.4058,
+      "step": 6757
+    },
+    {
+      "epoch": 1.1658759596308117,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.477158316802302e-06,
+      "loss": 1.4523,
+      "step": 6758
+    },
+    {
+      "epoch": 1.1660484775295437,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.474524575462341e-06,
+      "loss": 1.4831,
+      "step": 6759
+    },
+    {
+      "epoch": 1.1662209954282756,
+      "grad_norm": 0.671875,
+      "learning_rate": 7.471891021225376e-06,
+      "loss": 1.5114,
+      "step": 6760
+    },
+    {
+      "epoch": 1.1663935133270076,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.469257654286524e-06,
+      "loss": 1.4559,
+      "step": 6761
+    },
+    {
+      "epoch": 1.1665660312257398,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.466624474840872e-06,
+      "loss": 1.3233,
+      "step": 6762
+    },
+    {
+      "epoch": 1.1667385491244717,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.463991483083512e-06,
+      "loss": 1.432,
+      "step": 6763
+    },
+    {
+      "epoch": 1.1669110670232037,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.461358679209509e-06,
+      "loss": 1.3795,
+      "step": 6764
+    },
+    {
+      "epoch": 1.1670835849219356,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.458726063413918e-06,
+      "loss": 1.2887,
+      "step": 6765
+    },
+    {
+      "epoch": 1.1672561028206676,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.4560936358917745e-06,
+      "loss": 1.4637,
+      "step": 6766
+    },
+    {
+      "epoch": 1.1674286207193996,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.4534613968381146e-06,
+      "loss": 1.4384,
+      "step": 6767
+    },
+    {
+      "epoch": 1.1676011386181315,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.450829346447941e-06,
+      "loss": 1.4334,
+      "step": 6768
+    },
+    {
+      "epoch": 1.1677736565168637,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.448197484916264e-06,
+      "loss": 1.4752,
+      "step": 6769
+    },
+    {
+      "epoch": 1.1679461744155957,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.4455658124380545e-06,
+      "loss": 1.3971,
+      "step": 6770
+    },
+    {
+      "epoch": 1.1681186923143276,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.442934329208291e-06,
+      "loss": 1.457,
+      "step": 6771
+    },
+    {
+      "epoch": 1.1682912102130596,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.440303035421934e-06,
+      "loss": 1.3178,
+      "step": 6772
+    },
+    {
+      "epoch": 1.1684637281117916,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.43767193127392e-06,
+      "loss": 1.4777,
+      "step": 6773
+    },
+    {
+      "epoch": 1.1686362460105235,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.4350410169591815e-06,
+      "loss": 1.3621,
+      "step": 6774
+    },
+    {
+      "epoch": 1.1688087639092557,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.432410292672627e-06,
+      "loss": 1.3563,
+      "step": 6775
+    },
+    {
+      "epoch": 1.1689812818079877,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.429779758609165e-06,
+      "loss": 1.3639,
+      "step": 6776
+    },
+    {
+      "epoch": 1.1691537997067196,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.427149414963676e-06,
+      "loss": 1.4728,
+      "step": 6777
+    },
+    {
+      "epoch": 1.1693263176054516,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.424519261931036e-06,
+      "loss": 1.426,
+      "step": 6778
+    },
+    {
+      "epoch": 1.1694988355041835,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.421889299706098e-06,
+      "loss": 1.4206,
+      "step": 6779
+    },
+    {
+      "epoch": 1.1696713534029155,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.419259528483713e-06,
+      "loss": 1.4275,
+      "step": 6780
+    },
+    {
+      "epoch": 1.1698438713016475,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.416629948458705e-06,
+      "loss": 1.5017,
+      "step": 6781
+    },
+    {
+      "epoch": 1.1700163892003794,
+      "grad_norm": 0.74609375,
+      "learning_rate": 7.414000559825893e-06,
+      "loss": 1.3737,
+      "step": 6782
+    },
+    {
+      "epoch": 1.1701889070991116,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.411371362780076e-06,
+      "loss": 1.4951,
+      "step": 6783
+    },
+    {
+      "epoch": 1.1703614249978436,
+      "grad_norm": 0.66796875,
+      "learning_rate": 7.408742357516046e-06,
+      "loss": 1.4681,
+      "step": 6784
+    },
+    {
+      "epoch": 1.1705339428965755,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.406113544228571e-06,
+      "loss": 1.4076,
+      "step": 6785
+    },
+    {
+      "epoch": 1.1707064607953075,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.40348492311241e-06,
+      "loss": 1.5682,
+      "step": 6786
+    },
+    {
+      "epoch": 1.1708789786940395,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.400856494362314e-06,
+      "loss": 1.4475,
+      "step": 6787
+    },
+    {
+      "epoch": 1.1710514965927714,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.398228258173006e-06,
+      "loss": 1.3377,
+      "step": 6788
+    },
+    {
+      "epoch": 1.1712240144915036,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.395600214739209e-06,
+      "loss": 1.4687,
+      "step": 6789
+    },
+    {
+      "epoch": 1.1713965323902356,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.392972364255615e-06,
+      "loss": 1.4127,
+      "step": 6790
+    },
+    {
+      "epoch": 1.1715690502889675,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.390344706916923e-06,
+      "loss": 1.4401,
+      "step": 6791
+    },
+    {
+      "epoch": 1.1717415681876995,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.3877172429178e-06,
+      "loss": 1.4409,
+      "step": 6792
+    },
+    {
+      "epoch": 1.1719140860864314,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.385089972452907e-06,
+      "loss": 1.3877,
+      "step": 6793
+    },
+    {
+      "epoch": 1.1720866039851634,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.382462895716882e-06,
+      "loss": 1.3832,
+      "step": 6794
+    },
+    {
+      "epoch": 1.1722591218838954,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.379836012904367e-06,
+      "loss": 1.3738,
+      "step": 6795
+    },
+    {
+      "epoch": 1.1724316397826275,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.377209324209968e-06,
+      "loss": 1.4374,
+      "step": 6796
+    },
+    {
+      "epoch": 1.1726041576813595,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.374582829828294e-06,
+      "loss": 1.3218,
+      "step": 6797
+    },
+    {
+      "epoch": 1.1727766755800915,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.3719565299539235e-06,
+      "loss": 1.457,
+      "step": 6798
+    },
+    {
+      "epoch": 1.1729491934788234,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.369330424781438e-06,
+      "loss": 1.4775,
+      "step": 6799
+    },
+    {
+      "epoch": 1.1731217113775554,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.3667045145053875e-06,
+      "loss": 1.3732,
+      "step": 6800
+    },
+    {
+      "epoch": 1.1731217113775554,
+      "eval_loss": 1.4084804058074951,
+      "eval_runtime": 10.8179,
+      "eval_samples_per_second": 94.658,
+      "eval_steps_per_second": 23.665,
+      "step": 6800
+    },
+    {
+      "epoch": 1.1732942292762873,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.364078799320324e-06,
+      "loss": 1.3311,
+      "step": 6801
+    },
+    {
+      "epoch": 1.1734667471750195,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.3614532794207714e-06,
+      "loss": 1.3944,
+      "step": 6802
+    },
+    {
+      "epoch": 1.1736392650737515,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.358827955001244e-06,
+      "loss": 1.3655,
+      "step": 6803
+    },
+    {
+      "epoch": 1.1738117829724835,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.35620282625625e-06,
+      "loss": 1.3973,
+      "step": 6804
+    },
+    {
+      "epoch": 1.1739843008712154,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.353577893380266e-06,
+      "loss": 1.4846,
+      "step": 6805
+    },
+    {
+      "epoch": 1.1741568187699474,
+      "grad_norm": 0.95703125,
+      "learning_rate": 7.350953156567771e-06,
+      "loss": 1.422,
+      "step": 6806
+    },
+    {
+      "epoch": 1.1743293366686793,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.348328616013213e-06,
+      "loss": 1.3805,
+      "step": 6807
+    },
+    {
+      "epoch": 1.1745018545674113,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.345704271911043e-06,
+      "loss": 1.427,
+      "step": 6808
+    },
+    {
+      "epoch": 1.1746743724661433,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.343080124455684e-06,
+      "loss": 1.4001,
+      "step": 6809
+    },
+    {
+      "epoch": 1.1748468903648754,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.340456173841552e-06,
+      "loss": 1.468,
+      "step": 6810
+    },
+    {
+      "epoch": 1.1750194082636074,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.337832420263042e-06,
+      "loss": 1.4489,
+      "step": 6811
+    },
+    {
+      "epoch": 1.1751919261623394,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.3352088639145425e-06,
+      "loss": 1.4517,
+      "step": 6812
+    },
+    {
+      "epoch": 1.1753644440610713,
+      "grad_norm": 0.6640625,
+      "learning_rate": 7.332585504990419e-06,
+      "loss": 1.3657,
+      "step": 6813
+    },
+    {
+      "epoch": 1.1755369619598033,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.329962343685031e-06,
+      "loss": 1.336,
+      "step": 6814
+    },
+    {
+      "epoch": 1.1757094798585352,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.327339380192712e-06,
+      "loss": 1.3534,
+      "step": 6815
+    },
+    {
+      "epoch": 1.1758819977572674,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.324716614707794e-06,
+      "loss": 1.5315,
+      "step": 6816
+    },
+    {
+      "epoch": 1.1760545156559994,
+      "grad_norm": 0.69921875,
+      "learning_rate": 7.322094047424584e-06,
+      "loss": 1.4136,
+      "step": 6817
+    },
+    {
+      "epoch": 1.1762270335547313,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.319471678537376e-06,
+      "loss": 1.4698,
+      "step": 6818
+    },
+    {
+      "epoch": 1.1763995514534633,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.31684950824046e-06,
+      "loss": 1.3532,
+      "step": 6819
+    },
+    {
+      "epoch": 1.1765720693521953,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.314227536728096e-06,
+      "loss": 1.4208,
+      "step": 6820
+    },
+    {
+      "epoch": 1.1767445872509272,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.311605764194538e-06,
+      "loss": 1.4014,
+      "step": 6821
+    },
+    {
+      "epoch": 1.1769171051496592,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.308984190834019e-06,
+      "loss": 1.4367,
+      "step": 6822
+    },
+    {
+      "epoch": 1.1770896230483912,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.306362816840771e-06,
+      "loss": 1.4118,
+      "step": 6823
+    },
+    {
+      "epoch": 1.1772621409471233,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.3037416424089925e-06,
+      "loss": 1.4064,
+      "step": 6824
+    },
+    {
+      "epoch": 1.1774346588458553,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.301120667732884e-06,
+      "loss": 1.4058,
+      "step": 6825
+    },
+    {
+      "epoch": 1.1776071767445873,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.2984998930066145e-06,
+      "loss": 1.5118,
+      "step": 6826
+    },
+    {
+      "epoch": 1.1777796946433192,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.295879318424356e-06,
+      "loss": 1.559,
+      "step": 6827
+    },
+    {
+      "epoch": 1.1779522125420512,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.2932589441802506e-06,
+      "loss": 1.4733,
+      "step": 6828
+    },
+    {
+      "epoch": 1.1781247304407831,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.290638770468439e-06,
+      "loss": 1.412,
+      "step": 6829
+    },
+    {
+      "epoch": 1.1782972483395153,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.288018797483034e-06,
+      "loss": 1.3695,
+      "step": 6830
+    },
+    {
+      "epoch": 1.1784697662382473,
+      "grad_norm": 0.98828125,
+      "learning_rate": 7.285399025418144e-06,
+      "loss": 1.4087,
+      "step": 6831
+    },
+    {
+      "epoch": 1.1786422841369792,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.282779454467851e-06,
+      "loss": 1.4296,
+      "step": 6832
+    },
+    {
+      "epoch": 1.1788148020357112,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.280160084826239e-06,
+      "loss": 1.4193,
+      "step": 6833
+    },
+    {
+      "epoch": 1.1789873199344432,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.2775409166873604e-06,
+      "loss": 1.4459,
+      "step": 6834
+    },
+    {
+      "epoch": 1.1791598378331751,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.274921950245258e-06,
+      "loss": 1.4489,
+      "step": 6835
+    },
+    {
+      "epoch": 1.179332355731907,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.27230318569397e-06,
+      "loss": 1.3256,
+      "step": 6836
+    },
+    {
+      "epoch": 1.1795048736306393,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.269684623227502e-06,
+      "loss": 1.3541,
+      "step": 6837
+    },
+    {
+      "epoch": 1.1796773915293712,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.267066263039862e-06,
+      "loss": 1.4544,
+      "step": 6838
+    },
+    {
+      "epoch": 1.1798499094281032,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.2644481053250215e-06,
+      "loss": 1.3666,
+      "step": 6839
+    },
+    {
+      "epoch": 1.1800224273268352,
+      "grad_norm": 0.625,
+      "learning_rate": 7.261830150276964e-06,
+      "loss": 1.3835,
+      "step": 6840
+    },
+    {
+      "epoch": 1.1801949452255671,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.259212398089636e-06,
+      "loss": 1.4254,
+      "step": 6841
+    },
+    {
+      "epoch": 1.180367463124299,
+      "grad_norm": 0.8359375,
+      "learning_rate": 7.25659484895698e-06,
+      "loss": 1.4535,
+      "step": 6842
+    },
+    {
+      "epoch": 1.1805399810230313,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.253977503072916e-06,
+      "loss": 1.3889,
+      "step": 6843
+    },
+    {
+      "epoch": 1.1807124989217632,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.251360360631359e-06,
+      "loss": 1.4511,
+      "step": 6844
+    },
+    {
+      "epoch": 1.1808850168204952,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.2487434218262e-06,
+      "loss": 1.4949,
+      "step": 6845
+    },
+    {
+      "epoch": 1.1810575347192271,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.24612668685132e-06,
+      "loss": 1.4106,
+      "step": 6846
+    },
+    {
+      "epoch": 1.181230052617959,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.24351015590058e-06,
+      "loss": 1.4674,
+      "step": 6847
+    },
+    {
+      "epoch": 1.181402570516691,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.240893829167834e-06,
+      "loss": 1.4853,
+      "step": 6848
+    },
+    {
+      "epoch": 1.181575088415423,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.23827770684691e-06,
+      "loss": 1.4398,
+      "step": 6849
+    },
+    {
+      "epoch": 1.181747606314155,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.2356617891316275e-06,
+      "loss": 1.4522,
+      "step": 6850
+    },
+    {
+      "epoch": 1.1819201242128872,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.233046076215798e-06,
+      "loss": 1.5239,
+      "step": 6851
+    },
+    {
+      "epoch": 1.1820926421116191,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.230430568293199e-06,
+      "loss": 1.5447,
+      "step": 6852
+    },
+    {
+      "epoch": 1.182265160010351,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.227815265557611e-06,
+      "loss": 1.4014,
+      "step": 6853
+    },
+    {
+      "epoch": 1.182437677909083,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.225200168202789e-06,
+      "loss": 1.4451,
+      "step": 6854
+    },
+    {
+      "epoch": 1.182610195807815,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.222585276422477e-06,
+      "loss": 1.4846,
+      "step": 6855
+    },
+    {
+      "epoch": 1.182782713706547,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.219970590410399e-06,
+      "loss": 1.4114,
+      "step": 6856
+    },
+    {
+      "epoch": 1.1829552316052792,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.217356110360275e-06,
+      "loss": 1.4652,
+      "step": 6857
+    },
+    {
+      "epoch": 1.1831277495040111,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.214741836465793e-06,
+      "loss": 1.4706,
+      "step": 6858
+    },
+    {
+      "epoch": 1.183300267402743,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.21212776892064e-06,
+      "loss": 1.4584,
+      "step": 6859
+    },
+    {
+      "epoch": 1.183472785301475,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.209513907918479e-06,
+      "loss": 1.4025,
+      "step": 6860
+    },
+    {
+      "epoch": 1.183645303200207,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.206900253652969e-06,
+      "loss": 1.3999,
+      "step": 6861
+    },
+    {
+      "epoch": 1.183817821098939,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.204286806317736e-06,
+      "loss": 1.3659,
+      "step": 6862
+    },
+    {
+      "epoch": 1.183990338997671,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.2016735661064086e-06,
+      "loss": 1.4903,
+      "step": 6863
+    },
+    {
+      "epoch": 1.1841628568964029,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.1990605332125825e-06,
+      "loss": 1.4975,
+      "step": 6864
+    },
+    {
+      "epoch": 1.184335374795135,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.1964477078298574e-06,
+      "loss": 1.3395,
+      "step": 6865
+    },
+    {
+      "epoch": 1.184507892693867,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.193835090151803e-06,
+      "loss": 1.3884,
+      "step": 6866
+    },
+    {
+      "epoch": 1.184680410592599,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.191222680371975e-06,
+      "loss": 1.3982,
+      "step": 6867
+    },
+    {
+      "epoch": 1.184852928491331,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.188610478683926e-06,
+      "loss": 1.3685,
+      "step": 6868
+    },
+    {
+      "epoch": 1.185025446390063,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.1859984852811775e-06,
+      "loss": 1.4035,
+      "step": 6869
+    },
+    {
+      "epoch": 1.1851979642887949,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.183386700357245e-06,
+      "loss": 1.3266,
+      "step": 6870
+    },
+    {
+      "epoch": 1.185370482187527,
+      "grad_norm": 0.69140625,
+      "learning_rate": 7.1807751241056215e-06,
+      "loss": 1.4277,
+      "step": 6871
+    },
+    {
+      "epoch": 1.185543000086259,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.178163756719795e-06,
+      "loss": 1.3906,
+      "step": 6872
+    },
+    {
+      "epoch": 1.185715517984991,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.175552598393227e-06,
+      "loss": 1.4967,
+      "step": 6873
+    },
+    {
+      "epoch": 1.185888035883723,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.172941649319374e-06,
+      "loss": 1.3458,
+      "step": 6874
+    },
+    {
+      "epoch": 1.186060553782455,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.1703309096916625e-06,
+      "loss": 1.4214,
+      "step": 6875
+    },
+    {
+      "epoch": 1.1862330716811869,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.167720379703522e-06,
+      "loss": 1.3223,
+      "step": 6876
+    },
+    {
+      "epoch": 1.1864055895799188,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.165110059548353e-06,
+      "loss": 1.4065,
+      "step": 6877
+    },
+    {
+      "epoch": 1.186578107478651,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.162499949419543e-06,
+      "loss": 1.4116,
+      "step": 6878
+    },
+    {
+      "epoch": 1.186750625377383,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.159890049510463e-06,
+      "loss": 1.431,
+      "step": 6879
+    },
+    {
+      "epoch": 1.186923143276115,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.157280360014478e-06,
+      "loss": 1.3818,
+      "step": 6880
+    },
+    {
+      "epoch": 1.1870956611748469,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.154670881124925e-06,
+      "loss": 1.4078,
+      "step": 6881
+    },
+    {
+      "epoch": 1.1872681790735788,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.152061613035128e-06,
+      "loss": 1.3175,
+      "step": 6882
+    },
+    {
+      "epoch": 1.1874406969723108,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.149452555938407e-06,
+      "loss": 1.4258,
+      "step": 6883
+    },
+    {
+      "epoch": 1.187613214871043,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.146843710028049e-06,
+      "loss": 1.3981,
+      "step": 6884
+    },
+    {
+      "epoch": 1.187785732769775,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.144235075497339e-06,
+      "loss": 1.4337,
+      "step": 6885
+    },
+    {
+      "epoch": 1.187958250668507,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.141626652539533e-06,
+      "loss": 1.4078,
+      "step": 6886
+    },
+    {
+      "epoch": 1.1881307685672389,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.139018441347889e-06,
+      "loss": 1.3799,
+      "step": 6887
+    },
+    {
+      "epoch": 1.1883032864659708,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.136410442115631e-06,
+      "loss": 1.4632,
+      "step": 6888
+    },
+    {
+      "epoch": 1.1884758043647028,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.133802655035984e-06,
+      "loss": 1.4898,
+      "step": 6889
+    },
+    {
+      "epoch": 1.1886483222634348,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.131195080302144e-06,
+      "loss": 1.4195,
+      "step": 6890
+    },
+    {
+      "epoch": 1.1888208401621667,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.128587718107298e-06,
+      "loss": 1.4984,
+      "step": 6891
+    },
+    {
+      "epoch": 1.188993358060899,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.125980568644612e-06,
+      "loss": 1.4518,
+      "step": 6892
+    },
+    {
+      "epoch": 1.1891658759596309,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.123373632107247e-06,
+      "loss": 1.4977,
+      "step": 6893
+    },
+    {
+      "epoch": 1.1893383938583628,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.1207669086883366e-06,
+      "loss": 1.3991,
+      "step": 6894
+    },
+    {
+      "epoch": 1.1895109117570948,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.118160398581004e-06,
+      "loss": 1.3576,
+      "step": 6895
+    },
+    {
+      "epoch": 1.1896834296558267,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.115554101978354e-06,
+      "loss": 1.4036,
+      "step": 6896
+    },
+    {
+      "epoch": 1.1898559475545587,
+      "grad_norm": 0.6953125,
+      "learning_rate": 7.112948019073481e-06,
+      "loss": 1.3978,
+      "step": 6897
+    },
+    {
+      "epoch": 1.1900284654532909,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.110342150059457e-06,
+      "loss": 1.4511,
+      "step": 6898
+    },
+    {
+      "epoch": 1.1902009833520228,
+      "grad_norm": 0.65625,
+      "learning_rate": 7.107736495129338e-06,
+      "loss": 1.4181,
+      "step": 6899
+    },
+    {
+      "epoch": 1.1903735012507548,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.1051310544761775e-06,
+      "loss": 1.3462,
+      "step": 6900
+    },
+    {
+      "epoch": 1.1903735012507548,
+      "eval_loss": 1.408332109451294,
+      "eval_runtime": 10.8573,
+      "eval_samples_per_second": 94.315,
+      "eval_steps_per_second": 23.579,
+      "step": 6900
+    },
+    {
+      "epoch": 1.1905460191494868,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.102525828292993e-06,
+      "loss": 1.5048,
+      "step": 6901
+    },
+    {
+      "epoch": 1.1907185370482187,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.099920816772803e-06,
+      "loss": 1.4587,
+      "step": 6902
+    },
+    {
+      "epoch": 1.1908910549469507,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.097316020108594e-06,
+      "loss": 1.4227,
+      "step": 6903
+    },
+    {
+      "epoch": 1.1910635728456826,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.094711438493355e-06,
+      "loss": 1.349,
+      "step": 6904
+    },
+    {
+      "epoch": 1.1912360907444146,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.0921070721200445e-06,
+      "loss": 1.4222,
+      "step": 6905
+    },
+    {
+      "epoch": 1.1914086086431468,
+      "grad_norm": 0.89453125,
+      "learning_rate": 7.089502921181613e-06,
+      "loss": 1.3994,
+      "step": 6906
+    },
+    {
+      "epoch": 1.1915811265418788,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.086898985870987e-06,
+      "loss": 1.4528,
+      "step": 6907
+    },
+    {
+      "epoch": 1.1917536444406107,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.084295266381089e-06,
+      "loss": 1.4027,
+      "step": 6908
+    },
+    {
+      "epoch": 1.1919261623393427,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.081691762904814e-06,
+      "loss": 1.4115,
+      "step": 6909
+    },
+    {
+      "epoch": 1.1920986802380746,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.079088475635051e-06,
+      "loss": 1.4856,
+      "step": 6910
+    },
+    {
+      "epoch": 1.1922711981368066,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.076485404764659e-06,
+      "loss": 1.4102,
+      "step": 6911
+    },
+    {
+      "epoch": 1.1924437160355388,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.073882550486501e-06,
+      "loss": 1.3498,
+      "step": 6912
+    },
+    {
+      "epoch": 1.1926162339342707,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.071279912993403e-06,
+      "loss": 1.4475,
+      "step": 6913
+    },
+    {
+      "epoch": 1.1927887518330027,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.068677492478191e-06,
+      "loss": 1.4666,
+      "step": 6914
+    },
+    {
+      "epoch": 1.1929612697317347,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.066075289133662e-06,
+      "loss": 1.4479,
+      "step": 6915
+    },
+    {
+      "epoch": 1.1931337876304666,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.063473303152608e-06,
+      "loss": 1.3826,
+      "step": 6916
+    },
+    {
+      "epoch": 1.1933063055291986,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.0608715347278045e-06,
+      "loss": 1.4623,
+      "step": 6917
+    },
+    {
+      "epoch": 1.1934788234279305,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.0582699840519996e-06,
+      "loss": 1.3564,
+      "step": 6918
+    },
+    {
+      "epoch": 1.1936513413266627,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.055668651317937e-06,
+      "loss": 1.3768,
+      "step": 6919
+    },
+    {
+      "epoch": 1.1938238592253947,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.053067536718334e-06,
+      "loss": 1.2849,
+      "step": 6920
+    },
+    {
+      "epoch": 1.1939963771241267,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.050466640445906e-06,
+      "loss": 1.3967,
+      "step": 6921
+    },
+    {
+      "epoch": 1.1941688950228586,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.0478659626933345e-06,
+      "loss": 1.3141,
+      "step": 6922
+    },
+    {
+      "epoch": 1.1943414129215906,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.045265503653302e-06,
+      "loss": 1.4541,
+      "step": 6923
+    },
+    {
+      "epoch": 1.1945139308203225,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.042665263518458e-06,
+      "loss": 1.481,
+      "step": 6924
+    },
+    {
+      "epoch": 1.1946864487190547,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.040065242481455e-06,
+      "loss": 1.3615,
+      "step": 6925
+    },
+    {
+      "epoch": 1.1948589666177867,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.0374654407349095e-06,
+      "loss": 1.3896,
+      "step": 6926
+    },
+    {
+      "epoch": 1.1950314845165186,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.034865858471438e-06,
+      "loss": 1.4479,
+      "step": 6927
+    },
+    {
+      "epoch": 1.1952040024152506,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.032266495883627e-06,
+      "loss": 1.473,
+      "step": 6928
+    },
+    {
+      "epoch": 1.1953765203139826,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.029667353164061e-06,
+      "loss": 1.4585,
+      "step": 6929
+    },
+    {
+      "epoch": 1.1955490382127145,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.027068430505295e-06,
+      "loss": 1.3895,
+      "step": 6930
+    },
+    {
+      "epoch": 1.1957215561114465,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.024469728099873e-06,
+      "loss": 1.4092,
+      "step": 6931
+    },
+    {
+      "epoch": 1.1958940740101784,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.02187124614033e-06,
+      "loss": 1.4369,
+      "step": 6932
+    },
+    {
+      "epoch": 1.1960665919089106,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.019272984819171e-06,
+      "loss": 1.4359,
+      "step": 6933
+    },
+    {
+      "epoch": 1.1962391098076426,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.016674944328896e-06,
+      "loss": 1.3549,
+      "step": 6934
+    },
+    {
+      "epoch": 1.1964116277063745,
+      "grad_norm": 0.625,
+      "learning_rate": 7.014077124861978e-06,
+      "loss": 1.447,
+      "step": 6935
+    },
+    {
+      "epoch": 1.1965841456051065,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.011479526610887e-06,
+      "loss": 1.4539,
+      "step": 6936
+    },
+    {
+      "epoch": 1.1967566635038385,
+      "grad_norm": 0.84375,
+      "learning_rate": 7.0088821497680635e-06,
+      "loss": 1.348,
+      "step": 6937
+    },
+    {
+      "epoch": 1.1969291814025704,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.006284994525943e-06,
+      "loss": 1.3846,
+      "step": 6938
+    },
+    {
+      "epoch": 1.1971016993013026,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.003688061076929e-06,
+      "loss": 1.404,
+      "step": 6939
+    },
+    {
+      "epoch": 1.1972742172000346,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.001091349613433e-06,
+      "loss": 1.4452,
+      "step": 6940
+    },
+    {
+      "epoch": 1.1974467350987665,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.9984948603278225e-06,
+      "loss": 1.4511,
+      "step": 6941
+    },
+    {
+      "epoch": 1.1976192529974985,
+      "grad_norm": 0.6953125,
+      "learning_rate": 6.995898593412471e-06,
+      "loss": 1.4435,
+      "step": 6942
+    },
+    {
+      "epoch": 1.1977917708962305,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.993302549059717e-06,
+      "loss": 1.425,
+      "step": 6943
+    },
+    {
+      "epoch": 1.1979642887949624,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.9907067274619025e-06,
+      "loss": 1.427,
+      "step": 6944
+    },
+    {
+      "epoch": 1.1981368066936944,
+      "grad_norm": 0.70703125,
+      "learning_rate": 6.988111128811334e-06,
+      "loss": 1.4501,
+      "step": 6945
+    },
+    {
+      "epoch": 1.1983093245924266,
+      "grad_norm": 0.8046875,
+      "learning_rate": 6.985515753300314e-06,
+      "loss": 1.4704,
+      "step": 6946
+    },
+    {
+      "epoch": 1.1984818424911585,
+      "grad_norm": 0.625,
+      "learning_rate": 6.982920601121117e-06,
+      "loss": 1.3816,
+      "step": 6947
+    },
+    {
+      "epoch": 1.1986543603898905,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.980325672466015e-06,
+      "loss": 1.4104,
+      "step": 6948
+    },
+    {
+      "epoch": 1.1988268782886224,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.977730967527259e-06,
+      "loss": 1.4798,
+      "step": 6949
+    },
+    {
+      "epoch": 1.1989993961873544,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.975136486497074e-06,
+      "loss": 1.4175,
+      "step": 6950
+    },
+    {
+      "epoch": 1.1991719140860864,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.972542229567682e-06,
+      "loss": 1.3979,
+      "step": 6951
+    },
+    {
+      "epoch": 1.1993444319848185,
+      "grad_norm": 0.625,
+      "learning_rate": 6.969948196931272e-06,
+      "loss": 1.3222,
+      "step": 6952
+    },
+    {
+      "epoch": 1.1995169498835505,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.967354388780037e-06,
+      "loss": 1.3931,
+      "step": 6953
+    },
+    {
+      "epoch": 1.1996894677822825,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.964760805306137e-06,
+      "loss": 1.4464,
+      "step": 6954
+    },
+    {
+      "epoch": 1.1998619856810144,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.962167446701722e-06,
+      "loss": 1.4952,
+      "step": 6955
+    },
+    {
+      "epoch": 1.2000345035797464,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.95957431315892e-06,
+      "loss": 1.4631,
+      "step": 6956
+    },
+    {
+      "epoch": 1.2002070214784784,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.956981404869855e-06,
+      "loss": 1.5089,
+      "step": 6957
+    },
+    {
+      "epoch": 1.2003795393772103,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.954388722026618e-06,
+      "loss": 1.4603,
+      "step": 6958
+    },
+    {
+      "epoch": 1.2005520572759423,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.951796264821298e-06,
+      "loss": 1.4433,
+      "step": 6959
+    },
+    {
+      "epoch": 1.2007245751746745,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.949204033445951e-06,
+      "loss": 1.4816,
+      "step": 6960
+    },
+    {
+      "epoch": 1.2008970930734064,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.946612028092636e-06,
+      "loss": 1.4718,
+      "step": 6961
+    },
+    {
+      "epoch": 1.2010696109721384,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.944020248953379e-06,
+      "loss": 1.4309,
+      "step": 6962
+    },
+    {
+      "epoch": 1.2012421288708703,
+      "grad_norm": 0.6953125,
+      "learning_rate": 6.941428696220195e-06,
+      "loss": 1.5134,
+      "step": 6963
+    },
+    {
+      "epoch": 1.2014146467696023,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.938837370085087e-06,
+      "loss": 1.3666,
+      "step": 6964
+    },
+    {
+      "epoch": 1.2015871646683343,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.9362462707400325e-06,
+      "loss": 1.4257,
+      "step": 6965
+    },
+    {
+      "epoch": 1.2017596825670664,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.9336553983769995e-06,
+      "loss": 1.43,
+      "step": 6966
+    },
+    {
+      "epoch": 1.2019322004657984,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.931064753187929e-06,
+      "loss": 1.4498,
+      "step": 6967
+    },
+    {
+      "epoch": 1.2021047183645304,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.928474335364761e-06,
+      "loss": 1.3838,
+      "step": 6968
+    },
+    {
+      "epoch": 1.2022772362632623,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.925884145099405e-06,
+      "loss": 1.3933,
+      "step": 6969
+    },
+    {
+      "epoch": 1.2024497541619943,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.923294182583762e-06,
+      "loss": 1.4558,
+      "step": 6970
+    },
+    {
+      "epoch": 1.2026222720607262,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.920704448009705e-06,
+      "loss": 1.4645,
+      "step": 6971
+    },
+    {
+      "epoch": 1.2027947899594582,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.918114941569108e-06,
+      "loss": 1.4875,
+      "step": 6972
+    },
+    {
+      "epoch": 1.2029673078581902,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.915525663453808e-06,
+      "loss": 1.4333,
+      "step": 6973
+    },
+    {
+      "epoch": 1.2031398257569224,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.912936613855643e-06,
+      "loss": 1.4254,
+      "step": 6974
+    },
+    {
+      "epoch": 1.2033123436556543,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.910347792966418e-06,
+      "loss": 1.4324,
+      "step": 6975
+    },
+    {
+      "epoch": 1.2034848615543863,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.907759200977939e-06,
+      "loss": 1.5757,
+      "step": 6976
+    },
+    {
+      "epoch": 1.2036573794531182,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.905170838081976e-06,
+      "loss": 1.3948,
+      "step": 6977
+    },
+    {
+      "epoch": 1.2038298973518502,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.902582704470298e-06,
+      "loss": 1.5119,
+      "step": 6978
+    },
+    {
+      "epoch": 1.2040024152505822,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.899994800334644e-06,
+      "loss": 1.4551,
+      "step": 6979
+    },
+    {
+      "epoch": 1.2041749331493143,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.897407125866743e-06,
+      "loss": 1.364,
+      "step": 6980
+    },
+    {
+      "epoch": 1.2043474510480463,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.894819681258312e-06,
+      "loss": 1.2945,
+      "step": 6981
+    },
+    {
+      "epoch": 1.2045199689467783,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.89223246670104e-06,
+      "loss": 1.428,
+      "step": 6982
+    },
+    {
+      "epoch": 1.2046924868455102,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.889645482386607e-06,
+      "loss": 1.4469,
+      "step": 6983
+    },
+    {
+      "epoch": 1.2048650047442422,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.887058728506666e-06,
+      "loss": 1.3184,
+      "step": 6984
+    },
+    {
+      "epoch": 1.2050375226429741,
+      "grad_norm": 0.625,
+      "learning_rate": 6.8844722052528704e-06,
+      "loss": 1.4146,
+      "step": 6985
+    },
+    {
+      "epoch": 1.205210040541706,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.881885912816837e-06,
+      "loss": 1.4958,
+      "step": 6986
+    },
+    {
+      "epoch": 1.2053825584404383,
+      "grad_norm": 0.625,
+      "learning_rate": 6.87929985139018e-06,
+      "loss": 1.3745,
+      "step": 6987
+    },
+    {
+      "epoch": 1.2055550763391703,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.876714021164486e-06,
+      "loss": 1.3671,
+      "step": 6988
+    },
+    {
+      "epoch": 1.2057275942379022,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.874128422331336e-06,
+      "loss": 1.4298,
+      "step": 6989
+    },
+    {
+      "epoch": 1.2059001121366342,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.871543055082283e-06,
+      "loss": 1.4688,
+      "step": 6990
+    },
+    {
+      "epoch": 1.2060726300353661,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.8689579196088694e-06,
+      "loss": 1.3904,
+      "step": 6991
+    },
+    {
+      "epoch": 1.206245147934098,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.8663730161026125e-06,
+      "loss": 1.4428,
+      "step": 6992
+    },
+    {
+      "epoch": 1.2064176658328303,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.863788344755026e-06,
+      "loss": 1.4108,
+      "step": 6993
+    },
+    {
+      "epoch": 1.2065901837315622,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.861203905757593e-06,
+      "loss": 1.3966,
+      "step": 6994
+    },
+    {
+      "epoch": 1.2067627016302942,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.858619699301785e-06,
+      "loss": 1.4039,
+      "step": 6995
+    },
+    {
+      "epoch": 1.2069352195290262,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.856035725579062e-06,
+      "loss": 1.5806,
+      "step": 6996
+    },
+    {
+      "epoch": 1.2071077374277581,
+      "grad_norm": 0.70703125,
+      "learning_rate": 6.853451984780854e-06,
+      "loss": 1.4333,
+      "step": 6997
+    },
+    {
+      "epoch": 1.20728025532649,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.850868477098587e-06,
+      "loss": 1.3247,
+      "step": 6998
+    },
+    {
+      "epoch": 1.207452773225222,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.848285202723655e-06,
+      "loss": 1.4973,
+      "step": 6999
+    },
+    {
+      "epoch": 1.207625291123954,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.8457021618474514e-06,
+      "loss": 1.4564,
+      "step": 7000
+    },
+    {
+      "epoch": 1.207625291123954,
+      "eval_loss": 1.4081740379333496,
+      "eval_runtime": 11.0377,
+      "eval_samples_per_second": 92.773,
+      "eval_steps_per_second": 23.193,
+      "step": 7000
+    },
+    {
+      "epoch": 1.2077978090226862,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.843119354661341e-06,
+      "loss": 1.4374,
+      "step": 7001
+    },
+    {
+      "epoch": 1.2079703269214181,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.840536781356674e-06,
+      "loss": 1.3482,
+      "step": 7002
+    },
+    {
+      "epoch": 1.20814284482015,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.837954442124779e-06,
+      "loss": 1.5448,
+      "step": 7003
+    },
+    {
+      "epoch": 1.208315362718882,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.835372337156981e-06,
+      "loss": 1.5004,
+      "step": 7004
+    },
+    {
+      "epoch": 1.208487880617614,
+      "grad_norm": 0.78125,
+      "learning_rate": 6.83279046664457e-06,
+      "loss": 1.4206,
+      "step": 7005
+    },
+    {
+      "epoch": 1.208660398516346,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.830208830778837e-06,
+      "loss": 1.4897,
+      "step": 7006
+    },
+    {
+      "epoch": 1.2088329164150782,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.8276274297510336e-06,
+      "loss": 1.4595,
+      "step": 7007
+    },
+    {
+      "epoch": 1.2090054343138101,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.825046263752415e-06,
+      "loss": 1.4146,
+      "step": 7008
+    },
+    {
+      "epoch": 1.209177952212542,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.822465332974204e-06,
+      "loss": 1.386,
+      "step": 7009
+    },
+    {
+      "epoch": 1.209350470111274,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.819884637607619e-06,
+      "loss": 1.3649,
+      "step": 7010
+    },
+    {
+      "epoch": 1.209522988010006,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.817304177843848e-06,
+      "loss": 1.4369,
+      "step": 7011
+    },
+    {
+      "epoch": 1.209695505908738,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.8147239538740695e-06,
+      "loss": 1.3845,
+      "step": 7012
+    },
+    {
+      "epoch": 1.20986802380747,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.812143965889446e-06,
+      "loss": 1.4777,
+      "step": 7013
+    },
+    {
+      "epoch": 1.210040541706202,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.809564214081114e-06,
+      "loss": 1.5328,
+      "step": 7014
+    },
+    {
+      "epoch": 1.210213059604934,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.806984698640202e-06,
+      "loss": 1.4277,
+      "step": 7015
+    },
+    {
+      "epoch": 1.210385577503666,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.8044054197578115e-06,
+      "loss": 1.4582,
+      "step": 7016
+    },
+    {
+      "epoch": 1.210558095402398,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.801826377625036e-06,
+      "loss": 1.4857,
+      "step": 7017
+    },
+    {
+      "epoch": 1.21073061330113,
+      "grad_norm": 0.67578125,
+      "learning_rate": 6.799247572432945e-06,
+      "loss": 1.5168,
+      "step": 7018
+    },
+    {
+      "epoch": 1.210903131199862,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.796669004372596e-06,
+      "loss": 1.4525,
+      "step": 7019
+    },
+    {
+      "epoch": 1.2110756490985939,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.794090673635017e-06,
+      "loss": 1.4734,
+      "step": 7020
+    },
+    {
+      "epoch": 1.211248166997326,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.791512580411237e-06,
+      "loss": 1.3496,
+      "step": 7021
+    },
+    {
+      "epoch": 1.211420684896058,
+      "grad_norm": 0.625,
+      "learning_rate": 6.788934724892251e-06,
+      "loss": 1.4423,
+      "step": 7022
+    },
+    {
+      "epoch": 1.21159320279479,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.786357107269045e-06,
+      "loss": 1.3485,
+      "step": 7023
+    },
+    {
+      "epoch": 1.211765720693522,
+      "grad_norm": 0.76171875,
+      "learning_rate": 6.78377972773258e-06,
+      "loss": 1.4124,
+      "step": 7024
+    },
+    {
+      "epoch": 1.211938238592254,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.781202586473814e-06,
+      "loss": 1.4283,
+      "step": 7025
+    },
+    {
+      "epoch": 1.2121107564909859,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.778625683683671e-06,
+      "loss": 1.4206,
+      "step": 7026
+    },
+    {
+      "epoch": 1.2122832743897178,
+      "grad_norm": 0.703125,
+      "learning_rate": 6.776049019553062e-06,
+      "loss": 1.6036,
+      "step": 7027
+    },
+    {
+      "epoch": 1.21245579228845,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.773472594272889e-06,
+      "loss": 1.4514,
+      "step": 7028
+    },
+    {
+      "epoch": 1.212628310187182,
+      "grad_norm": 0.71484375,
+      "learning_rate": 6.7708964080340265e-06,
+      "loss": 1.4119,
+      "step": 7029
+    },
+    {
+      "epoch": 1.212800828085914,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.768320461027336e-06,
+      "loss": 1.4975,
+      "step": 7030
+    },
+    {
+      "epoch": 1.212973345984646,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.765744753443654e-06,
+      "loss": 1.3481,
+      "step": 7031
+    },
+    {
+      "epoch": 1.2131458638833779,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.763169285473813e-06,
+      "loss": 1.4414,
+      "step": 7032
+    },
+    {
+      "epoch": 1.2133183817821098,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.760594057308614e-06,
+      "loss": 1.3994,
+      "step": 7033
+    },
+    {
+      "epoch": 1.213490899680842,
+      "grad_norm": 0.80078125,
+      "learning_rate": 6.758019069138851e-06,
+      "loss": 1.3516,
+      "step": 7034
+    },
+    {
+      "epoch": 1.213663417579574,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.7554443211552864e-06,
+      "loss": 1.4604,
+      "step": 7035
+    },
+    {
+      "epoch": 1.213835935478306,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.752869813548684e-06,
+      "loss": 1.4742,
+      "step": 7036
+    },
+    {
+      "epoch": 1.2140084533770379,
+      "grad_norm": 0.625,
+      "learning_rate": 6.750295546509771e-06,
+      "loss": 1.4507,
+      "step": 7037
+    },
+    {
+      "epoch": 1.2141809712757698,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.747721520229273e-06,
+      "loss": 1.4602,
+      "step": 7038
+    },
+    {
+      "epoch": 1.2143534891745018,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.7451477348978835e-06,
+      "loss": 1.3104,
+      "step": 7039
+    },
+    {
+      "epoch": 1.2145260070732338,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.74257419070629e-06,
+      "loss": 1.416,
+      "step": 7040
+    },
+    {
+      "epoch": 1.2146985249719657,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.740000887845149e-06,
+      "loss": 1.4529,
+      "step": 7041
+    },
+    {
+      "epoch": 1.214871042870698,
+      "grad_norm": 0.890625,
+      "learning_rate": 6.737427826505116e-06,
+      "loss": 1.3673,
+      "step": 7042
+    },
+    {
+      "epoch": 1.2150435607694299,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.734855006876814e-06,
+      "loss": 1.4521,
+      "step": 7043
+    },
+    {
+      "epoch": 1.2152160786681618,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.732282429150852e-06,
+      "loss": 1.3779,
+      "step": 7044
+    },
+    {
+      "epoch": 1.2153885965668938,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.729710093517829e-06,
+      "loss": 1.4222,
+      "step": 7045
+    },
+    {
+      "epoch": 1.2155611144656258,
+      "grad_norm": 0.84765625,
+      "learning_rate": 6.727138000168314e-06,
+      "loss": 1.4788,
+      "step": 7046
+    },
+    {
+      "epoch": 1.2157336323643577,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.7245661492928674e-06,
+      "loss": 1.4366,
+      "step": 7047
+    },
+    {
+      "epoch": 1.21590615026309,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.721994541082022e-06,
+      "loss": 1.4501,
+      "step": 7048
+    },
+    {
+      "epoch": 1.2160786681618219,
+      "grad_norm": 0.7734375,
+      "learning_rate": 6.719423175726308e-06,
+      "loss": 1.3713,
+      "step": 7049
+    },
+    {
+      "epoch": 1.2162511860605538,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.716852053416221e-06,
+      "loss": 1.47,
+      "step": 7050
+    },
+    {
+      "epoch": 1.2164237039592858,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.7142811743422495e-06,
+      "loss": 1.4462,
+      "step": 7051
+    },
+    {
+      "epoch": 1.2165962218580177,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.711710538694855e-06,
+      "loss": 1.4658,
+      "step": 7052
+    },
+    {
+      "epoch": 1.2167687397567497,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.709140146664494e-06,
+      "loss": 1.4135,
+      "step": 7053
+    },
+    {
+      "epoch": 1.2169412576554817,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.706569998441591e-06,
+      "loss": 1.4711,
+      "step": 7054
+    },
+    {
+      "epoch": 1.2171137755542136,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.704000094216563e-06,
+      "loss": 1.4631,
+      "step": 7055
+    },
+    {
+      "epoch": 1.2172862934529458,
+      "grad_norm": 0.625,
+      "learning_rate": 6.701430434179799e-06,
+      "loss": 1.452,
+      "step": 7056
+    },
+    {
+      "epoch": 1.2174588113516778,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.698861018521681e-06,
+      "loss": 1.5188,
+      "step": 7057
+    },
+    {
+      "epoch": 1.2176313292504097,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.696291847432565e-06,
+      "loss": 1.4292,
+      "step": 7058
+    },
+    {
+      "epoch": 1.2178038471491417,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.693722921102788e-06,
+      "loss": 1.4083,
+      "step": 7059
+    },
+    {
+      "epoch": 1.2179763650478737,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.691154239722681e-06,
+      "loss": 1.4582,
+      "step": 7060
+    },
+    {
+      "epoch": 1.2181488829466056,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.688585803482539e-06,
+      "loss": 1.53,
+      "step": 7061
+    },
+    {
+      "epoch": 1.2183214008453378,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.686017612572655e-06,
+      "loss": 1.432,
+      "step": 7062
+    },
+    {
+      "epoch": 1.2184939187440698,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.683449667183288e-06,
+      "loss": 1.3916,
+      "step": 7063
+    },
+    {
+      "epoch": 1.2186664366428017,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.680881967504698e-06,
+      "loss": 1.4147,
+      "step": 7064
+    },
+    {
+      "epoch": 1.2188389545415337,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.678314513727105e-06,
+      "loss": 1.4874,
+      "step": 7065
+    },
+    {
+      "epoch": 1.2190114724402656,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.675747306040732e-06,
+      "loss": 1.4433,
+      "step": 7066
+    },
+    {
+      "epoch": 1.2191839903389976,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.673180344635767e-06,
+      "loss": 1.4913,
+      "step": 7067
+    },
+    {
+      "epoch": 1.2193565082377296,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.670613629702391e-06,
+      "loss": 1.4379,
+      "step": 7068
+    },
+    {
+      "epoch": 1.2195290261364617,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.6680471614307576e-06,
+      "loss": 1.4167,
+      "step": 7069
+    },
+    {
+      "epoch": 1.2197015440351937,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.665480940011012e-06,
+      "loss": 1.5605,
+      "step": 7070
+    },
+    {
+      "epoch": 1.2198740619339257,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.662914965633272e-06,
+      "loss": 1.4221,
+      "step": 7071
+    },
+    {
+      "epoch": 1.2200465798326576,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.660349238487644e-06,
+      "loss": 1.5069,
+      "step": 7072
+    },
+    {
+      "epoch": 1.2202190977313896,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.657783758764208e-06,
+      "loss": 1.4822,
+      "step": 7073
+    },
+    {
+      "epoch": 1.2203916156301216,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.655218526653038e-06,
+      "loss": 1.3407,
+      "step": 7074
+    },
+    {
+      "epoch": 1.2205641335288537,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.6526535423441775e-06,
+      "loss": 1.4482,
+      "step": 7075
+    },
+    {
+      "epoch": 1.2207366514275857,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.650088806027655e-06,
+      "loss": 1.3931,
+      "step": 7076
+    },
+    {
+      "epoch": 1.2209091693263177,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.647524317893489e-06,
+      "loss": 1.4404,
+      "step": 7077
+    },
+    {
+      "epoch": 1.2210816872250496,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.644960078131667e-06,
+      "loss": 1.3433,
+      "step": 7078
+    },
+    {
+      "epoch": 1.2212542051237816,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.642396086932168e-06,
+      "loss": 1.4951,
+      "step": 7079
+    },
+    {
+      "epoch": 1.2214267230225135,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.639832344484942e-06,
+      "loss": 1.4083,
+      "step": 7080
+    },
+    {
+      "epoch": 1.2215992409212455,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.637268850979934e-06,
+      "loss": 1.3391,
+      "step": 7081
+    },
+    {
+      "epoch": 1.2217717588199775,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.6347056066070605e-06,
+      "loss": 1.5557,
+      "step": 7082
+    },
+    {
+      "epoch": 1.2219442767187096,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.632142611556225e-06,
+      "loss": 1.4448,
+      "step": 7083
+    },
+    {
+      "epoch": 1.2221167946174416,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.629579866017303e-06,
+      "loss": 1.4612,
+      "step": 7084
+    },
+    {
+      "epoch": 1.2222893125161736,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.6270173701801685e-06,
+      "loss": 1.4145,
+      "step": 7085
+    },
+    {
+      "epoch": 1.2224618304149055,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.62445512423466e-06,
+      "loss": 1.4511,
+      "step": 7086
+    },
+    {
+      "epoch": 1.2226343483136375,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.621893128370609e-06,
+      "loss": 1.3963,
+      "step": 7087
+    },
+    {
+      "epoch": 1.2228068662123694,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.61933138277782e-06,
+      "loss": 1.4633,
+      "step": 7088
+    },
+    {
+      "epoch": 1.2229793841111016,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.616769887646088e-06,
+      "loss": 1.4498,
+      "step": 7089
+    },
+    {
+      "epoch": 1.2231519020098336,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.614208643165181e-06,
+      "loss": 1.3724,
+      "step": 7090
+    },
+    {
+      "epoch": 1.2233244199085656,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.61164764952485e-06,
+      "loss": 1.4823,
+      "step": 7091
+    },
+    {
+      "epoch": 1.2234969378072975,
+      "grad_norm": 0.703125,
+      "learning_rate": 6.609086906914839e-06,
+      "loss": 1.3061,
+      "step": 7092
+    },
+    {
+      "epoch": 1.2236694557060295,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.606526415524852e-06,
+      "loss": 1.4084,
+      "step": 7093
+    },
+    {
+      "epoch": 1.2238419736047614,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.603966175544595e-06,
+      "loss": 1.5003,
+      "step": 7094
+    },
+    {
+      "epoch": 1.2240144915034934,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.601406187163741e-06,
+      "loss": 1.4785,
+      "step": 7095
+    },
+    {
+      "epoch": 1.2241870094022254,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.598846450571956e-06,
+      "loss": 1.3297,
+      "step": 7096
+    },
+    {
+      "epoch": 1.2243595273009575,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.596286965958872e-06,
+      "loss": 1.4479,
+      "step": 7097
+    },
+    {
+      "epoch": 1.2245320451996895,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.593727733514119e-06,
+      "loss": 1.4002,
+      "step": 7098
+    },
+    {
+      "epoch": 1.2247045630984215,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.5911687534272995e-06,
+      "loss": 1.4138,
+      "step": 7099
+    },
+    {
+      "epoch": 1.2248770809971534,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.588610025887999e-06,
+      "loss": 1.3873,
+      "step": 7100
+    },
+    {
+      "epoch": 1.2248770809971534,
+      "eval_loss": 1.408066987991333,
+      "eval_runtime": 10.9497,
+      "eval_samples_per_second": 93.518,
+      "eval_steps_per_second": 23.38,
+      "step": 7100
+    },
+    {
+      "epoch": 1.2250495988958854,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.5860515510857795e-06,
+      "loss": 1.4459,
+      "step": 7101
+    },
+    {
+      "epoch": 1.2252221167946176,
+      "grad_norm": 0.8515625,
+      "learning_rate": 6.583493329210197e-06,
+      "loss": 1.3754,
+      "step": 7102
+    },
+    {
+      "epoch": 1.2253946346933495,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.5809353604507735e-06,
+      "loss": 1.4302,
+      "step": 7103
+    },
+    {
+      "epoch": 1.2255671525920815,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.578377644997022e-06,
+      "loss": 1.4781,
+      "step": 7104
+    },
+    {
+      "epoch": 1.2257396704908134,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.575820183038433e-06,
+      "loss": 1.4126,
+      "step": 7105
+    },
+    {
+      "epoch": 1.2259121883895454,
+      "grad_norm": 0.70703125,
+      "learning_rate": 6.573262974764483e-06,
+      "loss": 1.3968,
+      "step": 7106
+    },
+    {
+      "epoch": 1.2260847062882774,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.570706020364619e-06,
+      "loss": 1.4346,
+      "step": 7107
+    },
+    {
+      "epoch": 1.2262572241870093,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.568149320028281e-06,
+      "loss": 1.4194,
+      "step": 7108
+    },
+    {
+      "epoch": 1.2264297420857413,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.5655928739448874e-06,
+      "loss": 1.465,
+      "step": 7109
+    },
+    {
+      "epoch": 1.2266022599844735,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.56303668230383e-06,
+      "loss": 1.3937,
+      "step": 7110
+    },
+    {
+      "epoch": 1.2267747778832054,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.560480745294493e-06,
+      "loss": 1.4778,
+      "step": 7111
+    },
+    {
+      "epoch": 1.2269472957819374,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.557925063106229e-06,
+      "loss": 1.3817,
+      "step": 7112
+    },
+    {
+      "epoch": 1.2271198136806694,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.5553696359283905e-06,
+      "loss": 1.4118,
+      "step": 7113
+    },
+    {
+      "epoch": 1.2272923315794013,
+      "grad_norm": 0.625,
+      "learning_rate": 6.552814463950288e-06,
+      "loss": 1.3556,
+      "step": 7114
+    },
+    {
+      "epoch": 1.2274648494781333,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.5502595473612304e-06,
+      "loss": 1.5199,
+      "step": 7115
+    },
+    {
+      "epoch": 1.2276373673768655,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.5477048863504965e-06,
+      "loss": 1.5357,
+      "step": 7116
+    },
+    {
+      "epoch": 1.2278098852755974,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.5451504811073604e-06,
+      "loss": 1.3962,
+      "step": 7117
+    },
+    {
+      "epoch": 1.2279824031743294,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.542596331821061e-06,
+      "loss": 1.4034,
+      "step": 7118
+    },
+    {
+      "epoch": 1.2281549210730613,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.540042438680832e-06,
+      "loss": 1.3123,
+      "step": 7119
+    },
+    {
+      "epoch": 1.2283274389717933,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.537488801875872e-06,
+      "loss": 1.3925,
+      "step": 7120
+    },
+    {
+      "epoch": 1.2284999568705253,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.534935421595381e-06,
+      "loss": 1.3863,
+      "step": 7121
+    },
+    {
+      "epoch": 1.2286724747692572,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.5323822980285224e-06,
+      "loss": 1.4513,
+      "step": 7122
+    },
+    {
+      "epoch": 1.2288449926679892,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.5298294313644515e-06,
+      "loss": 1.4978,
+      "step": 7123
+    },
+    {
+      "epoch": 1.2290175105667214,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.527276821792297e-06,
+      "loss": 1.3822,
+      "step": 7124
+    },
+    {
+      "epoch": 1.2291900284654533,
+      "grad_norm": 0.67578125,
+      "learning_rate": 6.524724469501172e-06,
+      "loss": 1.363,
+      "step": 7125
+    },
+    {
+      "epoch": 1.2293625463641853,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.522172374680177e-06,
+      "loss": 1.4761,
+      "step": 7126
+    },
+    {
+      "epoch": 1.2295350642629173,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.5196205375183804e-06,
+      "loss": 1.419,
+      "step": 7127
+    },
+    {
+      "epoch": 1.2297075821616492,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.517068958204844e-06,
+      "loss": 1.3345,
+      "step": 7128
+    },
+    {
+      "epoch": 1.2298801000603812,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.514517636928598e-06,
+      "loss": 1.4148,
+      "step": 7129
+    },
+    {
+      "epoch": 1.2300526179591134,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.511966573878667e-06,
+      "loss": 1.4199,
+      "step": 7130
+    },
+    {
+      "epoch": 1.2302251358578453,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.509415769244044e-06,
+      "loss": 1.3635,
+      "step": 7131
+    },
+    {
+      "epoch": 1.2303976537565773,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.506865223213714e-06,
+      "loss": 1.3639,
+      "step": 7132
+    },
+    {
+      "epoch": 1.2305701716553092,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.50431493597663e-06,
+      "loss": 1.359,
+      "step": 7133
+    },
+    {
+      "epoch": 1.2307426895540412,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.5017649077217415e-06,
+      "loss": 1.4499,
+      "step": 7134
+    },
+    {
+      "epoch": 1.2309152074527732,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.499215138637964e-06,
+      "loss": 1.4122,
+      "step": 7135
+    },
+    {
+      "epoch": 1.2310877253515051,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.496665628914205e-06,
+      "loss": 1.5291,
+      "step": 7136
+    },
+    {
+      "epoch": 1.2312602432502373,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.494116378739344e-06,
+      "loss": 1.3532,
+      "step": 7137
+    },
+    {
+      "epoch": 1.2314327611489693,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.491567388302249e-06,
+      "loss": 1.4024,
+      "step": 7138
+    },
+    {
+      "epoch": 1.2316052790477012,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.489018657791763e-06,
+      "loss": 1.3638,
+      "step": 7139
+    },
+    {
+      "epoch": 1.2317777969464332,
+      "grad_norm": 3.328125,
+      "learning_rate": 6.48647018739671e-06,
+      "loss": 1.5342,
+      "step": 7140
+    },
+    {
+      "epoch": 1.2319503148451652,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.483921977305903e-06,
+      "loss": 1.4922,
+      "step": 7141
+    },
+    {
+      "epoch": 1.2321228327438971,
+      "grad_norm": 0.66015625,
+      "learning_rate": 6.481374027708123e-06,
+      "loss": 1.4724,
+      "step": 7142
+    },
+    {
+      "epoch": 1.2322953506426293,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.478826338792144e-06,
+      "loss": 1.4799,
+      "step": 7143
+    },
+    {
+      "epoch": 1.2324678685413613,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.476278910746705e-06,
+      "loss": 1.4143,
+      "step": 7144
+    },
+    {
+      "epoch": 1.2326403864400932,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.4737317437605475e-06,
+      "loss": 1.3686,
+      "step": 7145
+    },
+    {
+      "epoch": 1.2328129043388252,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.471184838022372e-06,
+      "loss": 1.406,
+      "step": 7146
+    },
+    {
+      "epoch": 1.2329854222375571,
+      "grad_norm": 0.70703125,
+      "learning_rate": 6.468638193720875e-06,
+      "loss": 1.4347,
+      "step": 7147
+    },
+    {
+      "epoch": 1.233157940136289,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.4660918110447215e-06,
+      "loss": 1.3784,
+      "step": 7148
+    },
+    {
+      "epoch": 1.233330458035021,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.463545690182573e-06,
+      "loss": 1.4489,
+      "step": 7149
+    },
+    {
+      "epoch": 1.233502975933753,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.460999831323054e-06,
+      "loss": 1.3989,
+      "step": 7150
+    },
+    {
+      "epoch": 1.2336754938324852,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.458454234654781e-06,
+      "loss": 1.4118,
+      "step": 7151
+    },
+    {
+      "epoch": 1.2338480117312172,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.4559089003663434e-06,
+      "loss": 1.4362,
+      "step": 7152
+    },
+    {
+      "epoch": 1.2340205296299491,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.453363828646323e-06,
+      "loss": 1.4336,
+      "step": 7153
+    },
+    {
+      "epoch": 1.234193047528681,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.450819019683267e-06,
+      "loss": 1.501,
+      "step": 7154
+    },
+    {
+      "epoch": 1.234365565427413,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.448274473665718e-06,
+      "loss": 1.3423,
+      "step": 7155
+    },
+    {
+      "epoch": 1.234538083326145,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.445730190782187e-06,
+      "loss": 1.4127,
+      "step": 7156
+    },
+    {
+      "epoch": 1.2347106012248772,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.443186171221167e-06,
+      "loss": 1.4934,
+      "step": 7157
+    },
+    {
+      "epoch": 1.2348831191236092,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.4406424151711456e-06,
+      "loss": 1.4181,
+      "step": 7158
+    },
+    {
+      "epoch": 1.2350556370223411,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.438098922820573e-06,
+      "loss": 1.44,
+      "step": 7159
+    },
+    {
+      "epoch": 1.235228154921073,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.435555694357888e-06,
+      "loss": 1.3861,
+      "step": 7160
+    },
+    {
+      "epoch": 1.235400672819805,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.433012729971506e-06,
+      "loss": 1.4417,
+      "step": 7161
+    },
+    {
+      "epoch": 1.235573190718537,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.430470029849832e-06,
+      "loss": 1.5808,
+      "step": 7162
+    },
+    {
+      "epoch": 1.235745708617269,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.4279275941812395e-06,
+      "loss": 1.3912,
+      "step": 7163
+    },
+    {
+      "epoch": 1.235918226516001,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.425385423154091e-06,
+      "loss": 1.3512,
+      "step": 7164
+    },
+    {
+      "epoch": 1.236090744414733,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.422843516956724e-06,
+      "loss": 1.3951,
+      "step": 7165
+    },
+    {
+      "epoch": 1.236263262313465,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.420301875777464e-06,
+      "loss": 1.4184,
+      "step": 7166
+    },
+    {
+      "epoch": 1.236435780212197,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.417760499804604e-06,
+      "loss": 1.4245,
+      "step": 7167
+    },
+    {
+      "epoch": 1.236608298110929,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.415219389226432e-06,
+      "loss": 1.4479,
+      "step": 7168
+    },
+    {
+      "epoch": 1.236780816009661,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.412678544231203e-06,
+      "loss": 1.4777,
+      "step": 7169
+    },
+    {
+      "epoch": 1.236953333908393,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.410137965007166e-06,
+      "loss": 1.3913,
+      "step": 7170
+    },
+    {
+      "epoch": 1.237125851807125,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.4075976517425365e-06,
+      "loss": 1.4419,
+      "step": 7171
+    },
+    {
+      "epoch": 1.237298369705857,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.4050576046255176e-06,
+      "loss": 1.4678,
+      "step": 7172
+    },
+    {
+      "epoch": 1.237470887604589,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.402517823844299e-06,
+      "loss": 1.4013,
+      "step": 7173
+    },
+    {
+      "epoch": 1.237643405503321,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.399978309587034e-06,
+      "loss": 1.4265,
+      "step": 7174
+    },
+    {
+      "epoch": 1.237815923402053,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.397439062041873e-06,
+      "loss": 1.3889,
+      "step": 7175
+    },
+    {
+      "epoch": 1.237988441300785,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.394900081396931e-06,
+      "loss": 1.4773,
+      "step": 7176
+    },
+    {
+      "epoch": 1.2381609591995169,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.392361367840322e-06,
+      "loss": 1.4759,
+      "step": 7177
+    },
+    {
+      "epoch": 1.238333477098249,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.3898229215601215e-06,
+      "loss": 1.3627,
+      "step": 7178
+    },
+    {
+      "epoch": 1.238505994996981,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.3872847427443985e-06,
+      "loss": 1.5092,
+      "step": 7179
+    },
+    {
+      "epoch": 1.238678512895713,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.384746831581191e-06,
+      "loss": 1.4332,
+      "step": 7180
+    },
+    {
+      "epoch": 1.238851030794445,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.382209188258533e-06,
+      "loss": 1.3706,
+      "step": 7181
+    },
+    {
+      "epoch": 1.2390235486931769,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.379671812964416e-06,
+      "loss": 1.4427,
+      "step": 7182
+    },
+    {
+      "epoch": 1.2391960665919088,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.3771347058868404e-06,
+      "loss": 1.4604,
+      "step": 7183
+    },
+    {
+      "epoch": 1.239368584490641,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.374597867213756e-06,
+      "loss": 1.4729,
+      "step": 7184
+    },
+    {
+      "epoch": 1.239541102389373,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.372061297133119e-06,
+      "loss": 1.3559,
+      "step": 7185
+    },
+    {
+      "epoch": 1.239713620288105,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.369524995832844e-06,
+      "loss": 1.4233,
+      "step": 7186
+    },
+    {
+      "epoch": 1.239886138186837,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.366988963500846e-06,
+      "loss": 1.452,
+      "step": 7187
+    },
+    {
+      "epoch": 1.2400586560855689,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.364453200325005e-06,
+      "loss": 1.4131,
+      "step": 7188
+    },
+    {
+      "epoch": 1.2402311739843008,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.361917706493184e-06,
+      "loss": 1.4352,
+      "step": 7189
+    },
+    {
+      "epoch": 1.2404036918830328,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.359382482193234e-06,
+      "loss": 1.4955,
+      "step": 7190
+    },
+    {
+      "epoch": 1.2405762097817647,
+      "grad_norm": 0.546875,
+      "learning_rate": 6.356847527612976e-06,
+      "loss": 1.3062,
+      "step": 7191
+    },
+    {
+      "epoch": 1.240748727680497,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.354312842940219e-06,
+      "loss": 1.4652,
+      "step": 7192
+    },
+    {
+      "epoch": 1.240921245579229,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.351778428362742e-06,
+      "loss": 1.4002,
+      "step": 7193
+    },
+    {
+      "epoch": 1.2410937634779609,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.349244284068318e-06,
+      "loss": 1.3301,
+      "step": 7194
+    },
+    {
+      "epoch": 1.2412662813766928,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.346710410244685e-06,
+      "loss": 1.4002,
+      "step": 7195
+    },
+    {
+      "epoch": 1.2414387992754248,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.344176807079576e-06,
+      "loss": 1.3778,
+      "step": 7196
+    },
+    {
+      "epoch": 1.2416113171741567,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.341643474760686e-06,
+      "loss": 1.4565,
+      "step": 7197
+    },
+    {
+      "epoch": 1.241783835072889,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.339110413475711e-06,
+      "loss": 1.3695,
+      "step": 7198
+    },
+    {
+      "epoch": 1.2419563529716209,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.336577623412308e-06,
+      "loss": 1.2846,
+      "step": 7199
+    },
+    {
+      "epoch": 1.2421288708703528,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.3340451047581275e-06,
+      "loss": 1.4077,
+      "step": 7200
+    },
+    {
+      "epoch": 1.2421288708703528,
+      "eval_loss": 1.4079400300979614,
+      "eval_runtime": 10.842,
+      "eval_samples_per_second": 94.447,
+      "eval_steps_per_second": 23.612,
+      "step": 7200
+    },
+    {
+      "epoch": 1.2423013887690848,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.3315128577007874e-06,
+      "loss": 1.3817,
+      "step": 7201
+    },
+    {
+      "epoch": 1.2424739066678168,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.3289808824279e-06,
+      "loss": 1.3751,
+      "step": 7202
+    },
+    {
+      "epoch": 1.2426464245665487,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.3264491791270455e-06,
+      "loss": 1.3145,
+      "step": 7203
+    },
+    {
+      "epoch": 1.2428189424652807,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.323917747985786e-06,
+      "loss": 1.4272,
+      "step": 7204
+    },
+    {
+      "epoch": 1.2429914603640126,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.321386589191674e-06,
+      "loss": 1.3329,
+      "step": 7205
+    },
+    {
+      "epoch": 1.2431639782627448,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.318855702932225e-06,
+      "loss": 1.4386,
+      "step": 7206
+    },
+    {
+      "epoch": 1.2433364961614768,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.31632508939495e-06,
+      "loss": 1.5471,
+      "step": 7207
+    },
+    {
+      "epoch": 1.2435090140602088,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.313794748767324e-06,
+      "loss": 1.3682,
+      "step": 7208
+    },
+    {
+      "epoch": 1.2436815319589407,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.31126468123682e-06,
+      "loss": 1.3712,
+      "step": 7209
+    },
+    {
+      "epoch": 1.2438540498576727,
+      "grad_norm": 0.8203125,
+      "learning_rate": 6.308734886990875e-06,
+      "loss": 1.3938,
+      "step": 7210
+    },
+    {
+      "epoch": 1.2440265677564046,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.306205366216915e-06,
+      "loss": 1.4611,
+      "step": 7211
+    },
+    {
+      "epoch": 1.2441990856551368,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.3036761191023374e-06,
+      "loss": 1.4358,
+      "step": 7212
+    },
+    {
+      "epoch": 1.2443716035538688,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.301147145834534e-06,
+      "loss": 1.4512,
+      "step": 7213
+    },
+    {
+      "epoch": 1.2445441214526007,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.298618446600856e-06,
+      "loss": 1.4666,
+      "step": 7214
+    },
+    {
+      "epoch": 1.2447166393513327,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.2960900215886556e-06,
+      "loss": 1.3919,
+      "step": 7215
+    },
+    {
+      "epoch": 1.2448891572500647,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.293561870985248e-06,
+      "loss": 1.3503,
+      "step": 7216
+    },
+    {
+      "epoch": 1.2450616751487966,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.291033994977935e-06,
+      "loss": 1.4322,
+      "step": 7217
+    },
+    {
+      "epoch": 1.2452341930475286,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.288506393753997e-06,
+      "loss": 1.3567,
+      "step": 7218
+    },
+    {
+      "epoch": 1.2454067109462608,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.285979067500699e-06,
+      "loss": 1.4231,
+      "step": 7219
+    },
+    {
+      "epoch": 1.2455792288449927,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.283452016405276e-06,
+      "loss": 1.3525,
+      "step": 7220
+    },
+    {
+      "epoch": 1.2457517467437247,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.280925240654948e-06,
+      "loss": 1.4268,
+      "step": 7221
+    },
+    {
+      "epoch": 1.2459242646424566,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.27839874043692e-06,
+      "loss": 1.438,
+      "step": 7222
+    },
+    {
+      "epoch": 1.2460967825411886,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.275872515938365e-06,
+      "loss": 1.4832,
+      "step": 7223
+    },
+    {
+      "epoch": 1.2462693004399206,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.2733465673464456e-06,
+      "loss": 1.49,
+      "step": 7224
+    },
+    {
+      "epoch": 1.2464418183386528,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.270820894848293e-06,
+      "loss": 1.3996,
+      "step": 7225
+    },
+    {
+      "epoch": 1.2466143362373847,
+      "grad_norm": 1.046875,
+      "learning_rate": 6.268295498631034e-06,
+      "loss": 1.3661,
+      "step": 7226
+    },
+    {
+      "epoch": 1.2467868541361167,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.265770378881759e-06,
+      "loss": 1.3727,
+      "step": 7227
+    },
+    {
+      "epoch": 1.2469593720348486,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.263245535787548e-06,
+      "loss": 1.4667,
+      "step": 7228
+    },
+    {
+      "epoch": 1.2471318899335806,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.260720969535453e-06,
+      "loss": 1.439,
+      "step": 7229
+    },
+    {
+      "epoch": 1.2473044078323126,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.258196680312517e-06,
+      "loss": 1.341,
+      "step": 7230
+    },
+    {
+      "epoch": 1.2474769257310445,
+      "grad_norm": 0.69921875,
+      "learning_rate": 6.255672668305748e-06,
+      "loss": 1.4412,
+      "step": 7231
+    },
+    {
+      "epoch": 1.2476494436297765,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.253148933702147e-06,
+      "loss": 1.3463,
+      "step": 7232
+    },
+    {
+      "epoch": 1.2478219615285087,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.250625476688679e-06,
+      "loss": 1.3639,
+      "step": 7233
+    },
+    {
+      "epoch": 1.2479944794272406,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.248102297452307e-06,
+      "loss": 1.4123,
+      "step": 7234
+    },
+    {
+      "epoch": 1.2481669973259726,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.245579396179957e-06,
+      "loss": 1.5242,
+      "step": 7235
+    },
+    {
+      "epoch": 1.2483395152247045,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.243056773058542e-06,
+      "loss": 1.541,
+      "step": 7236
+    },
+    {
+      "epoch": 1.2485120331234365,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.240534428274961e-06,
+      "loss": 1.5029,
+      "step": 7237
+    },
+    {
+      "epoch": 1.2486845510221685,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.238012362016077e-06,
+      "loss": 1.4879,
+      "step": 7238
+    },
+    {
+      "epoch": 1.2488570689209006,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.235490574468745e-06,
+      "loss": 1.4711,
+      "step": 7239
+    },
+    {
+      "epoch": 1.2490295868196326,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.232969065819791e-06,
+      "loss": 1.4517,
+      "step": 7240
+    },
+    {
+      "epoch": 1.2492021047183646,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.230447836256028e-06,
+      "loss": 1.4584,
+      "step": 7241
+    },
+    {
+      "epoch": 1.2493746226170965,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.2279268859642396e-06,
+      "loss": 1.4252,
+      "step": 7242
+    },
+    {
+      "epoch": 1.2495471405158285,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.2254062151312e-06,
+      "loss": 1.3844,
+      "step": 7243
+    },
+    {
+      "epoch": 1.2497196584145605,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.222885823943651e-06,
+      "loss": 1.4895,
+      "step": 7244
+    },
+    {
+      "epoch": 1.2498921763132924,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.220365712588322e-06,
+      "loss": 1.4183,
+      "step": 7245
+    },
+    {
+      "epoch": 1.2500646942120244,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.217845881251913e-06,
+      "loss": 1.3925,
+      "step": 7246
+    },
+    {
+      "epoch": 1.2502372121107566,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.215326330121119e-06,
+      "loss": 1.4546,
+      "step": 7247
+    },
+    {
+      "epoch": 1.2504097300094885,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.212807059382595e-06,
+      "loss": 1.3991,
+      "step": 7248
+    },
+    {
+      "epoch": 1.2505822479082205,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.210288069222989e-06,
+      "loss": 1.4733,
+      "step": 7249
+    },
+    {
+      "epoch": 1.2507547658069524,
+      "grad_norm": 0.8671875,
+      "learning_rate": 6.207769359828919e-06,
+      "loss": 1.4307,
+      "step": 7250
+    },
+    {
+      "epoch": 1.2509272837056844,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.205250931386992e-06,
+      "loss": 1.4177,
+      "step": 7251
+    },
+    {
+      "epoch": 1.2510998016044166,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.202732784083787e-06,
+      "loss": 1.3977,
+      "step": 7252
+    },
+    {
+      "epoch": 1.2512723195031485,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.20021491810586e-06,
+      "loss": 1.4586,
+      "step": 7253
+    },
+    {
+      "epoch": 1.2514448374018805,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.197697333639759e-06,
+      "loss": 1.5194,
+      "step": 7254
+    },
+    {
+      "epoch": 1.2516173553006125,
+      "grad_norm": 0.79296875,
+      "learning_rate": 6.195180030871995e-06,
+      "loss": 1.3996,
+      "step": 7255
+    },
+    {
+      "epoch": 1.2517898731993444,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.19266300998907e-06,
+      "loss": 1.3985,
+      "step": 7256
+    },
+    {
+      "epoch": 1.2519623910980764,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.190146271177455e-06,
+      "loss": 1.3829,
+      "step": 7257
+    },
+    {
+      "epoch": 1.2521349089968083,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.187629814623613e-06,
+      "loss": 1.4638,
+      "step": 7258
+    },
+    {
+      "epoch": 1.2523074268955403,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.1851136405139735e-06,
+      "loss": 1.4106,
+      "step": 7259
+    },
+    {
+      "epoch": 1.2524799447942723,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.182597749034953e-06,
+      "loss": 1.3867,
+      "step": 7260
+    },
+    {
+      "epoch": 1.2526524626930045,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.1800821403729405e-06,
+      "loss": 1.3823,
+      "step": 7261
+    },
+    {
+      "epoch": 1.2528249805917364,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.177566814714316e-06,
+      "loss": 1.4487,
+      "step": 7262
+    },
+    {
+      "epoch": 1.2529974984904684,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.175051772245421e-06,
+      "loss": 1.4645,
+      "step": 7263
+    },
+    {
+      "epoch": 1.2531700163892003,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.172537013152593e-06,
+      "loss": 1.4401,
+      "step": 7264
+    },
+    {
+      "epoch": 1.2533425342879323,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.170022537622135e-06,
+      "loss": 1.5324,
+      "step": 7265
+    },
+    {
+      "epoch": 1.2535150521866645,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.1675083458403405e-06,
+      "loss": 1.3808,
+      "step": 7266
+    },
+    {
+      "epoch": 1.2536875700853964,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.164994437993474e-06,
+      "loss": 1.4153,
+      "step": 7267
+    },
+    {
+      "epoch": 1.2538600879841284,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.162480814267779e-06,
+      "loss": 1.492,
+      "step": 7268
+    },
+    {
+      "epoch": 1.2540326058828604,
+      "grad_norm": 0.625,
+      "learning_rate": 6.1599674748494875e-06,
+      "loss": 1.3307,
+      "step": 7269
+    },
+    {
+      "epoch": 1.2542051237815923,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.1574544199247955e-06,
+      "loss": 1.4724,
+      "step": 7270
+    },
+    {
+      "epoch": 1.2543776416803243,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.154941649679894e-06,
+      "loss": 1.4853,
+      "step": 7271
+    },
+    {
+      "epoch": 1.2545501595790562,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.152429164300935e-06,
+      "loss": 1.3948,
+      "step": 7272
+    },
+    {
+      "epoch": 1.2547226774777882,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.149916963974068e-06,
+      "loss": 1.4597,
+      "step": 7273
+    },
+    {
+      "epoch": 1.2548951953765204,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.1474050488854055e-06,
+      "loss": 1.3844,
+      "step": 7274
+    },
+    {
+      "epoch": 1.2550677132752524,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.144893419221052e-06,
+      "loss": 1.4039,
+      "step": 7275
+    },
+    {
+      "epoch": 1.2552402311739843,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.142382075167082e-06,
+      "loss": 1.4968,
+      "step": 7276
+    },
+    {
+      "epoch": 1.2554127490727163,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.139871016909551e-06,
+      "loss": 1.5144,
+      "step": 7277
+    },
+    {
+      "epoch": 1.2555852669714482,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.1373602446344904e-06,
+      "loss": 1.404,
+      "step": 7278
+    },
+    {
+      "epoch": 1.2557577848701804,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.134849758527923e-06,
+      "loss": 1.4736,
+      "step": 7279
+    },
+    {
+      "epoch": 1.2559303027689124,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.1323395587758325e-06,
+      "loss": 1.3478,
+      "step": 7280
+    },
+    {
+      "epoch": 1.2561028206676443,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.129829645564197e-06,
+      "loss": 1.3936,
+      "step": 7281
+    },
+    {
+      "epoch": 1.2562753385663763,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.127320019078959e-06,
+      "loss": 1.4741,
+      "step": 7282
+    },
+    {
+      "epoch": 1.2564478564651083,
+      "grad_norm": 0.625,
+      "learning_rate": 6.124810679506056e-06,
+      "loss": 1.4305,
+      "step": 7283
+    },
+    {
+      "epoch": 1.2566203743638402,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.122301627031388e-06,
+      "loss": 1.5184,
+      "step": 7284
+    },
+    {
+      "epoch": 1.2567928922625722,
+      "grad_norm": 0.625,
+      "learning_rate": 6.119792861840843e-06,
+      "loss": 1.4044,
+      "step": 7285
+    },
+    {
+      "epoch": 1.2569654101613041,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.117284384120292e-06,
+      "loss": 1.422,
+      "step": 7286
+    },
+    {
+      "epoch": 1.257137928060036,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.1147761940555714e-06,
+      "loss": 1.4325,
+      "step": 7287
+    },
+    {
+      "epoch": 1.2573104459587683,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.112268291832509e-06,
+      "loss": 1.387,
+      "step": 7288
+    },
+    {
+      "epoch": 1.2574829638575002,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.1097606776369e-06,
+      "loss": 1.4262,
+      "step": 7289
+    },
+    {
+      "epoch": 1.2576554817562322,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.107253351654529e-06,
+      "loss": 1.3348,
+      "step": 7290
+    },
+    {
+      "epoch": 1.2578279996549642,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.104746314071153e-06,
+      "loss": 1.3996,
+      "step": 7291
+    },
+    {
+      "epoch": 1.2580005175536961,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.1022395650725095e-06,
+      "loss": 1.3755,
+      "step": 7292
+    },
+    {
+      "epoch": 1.2581730354524283,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.09973310484431e-06,
+      "loss": 1.4054,
+      "step": 7293
+    },
+    {
+      "epoch": 1.2583455533511603,
+      "grad_norm": 0.69921875,
+      "learning_rate": 6.0972269335722555e-06,
+      "loss": 1.4913,
+      "step": 7294
+    },
+    {
+      "epoch": 1.2585180712498922,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.094721051442013e-06,
+      "loss": 1.5199,
+      "step": 7295
+    },
+    {
+      "epoch": 1.2586905891486242,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.092215458639239e-06,
+      "loss": 1.4474,
+      "step": 7296
+    },
+    {
+      "epoch": 1.2588631070473562,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.089710155349558e-06,
+      "loss": 1.3976,
+      "step": 7297
+    },
+    {
+      "epoch": 1.2590356249460881,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.087205141758583e-06,
+      "loss": 1.4169,
+      "step": 7298
+    },
+    {
+      "epoch": 1.25920814284482,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.0847004180518985e-06,
+      "loss": 1.4104,
+      "step": 7299
+    },
+    {
+      "epoch": 1.259380660743552,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.082195984415069e-06,
+      "loss": 1.3966,
+      "step": 7300
+    },
+    {
+      "epoch": 1.259380660743552,
+      "eval_loss": 1.4078376293182373,
+      "eval_runtime": 10.8355,
+      "eval_samples_per_second": 94.504,
+      "eval_steps_per_second": 23.626,
+      "step": 7300
+    },
+    {
+      "epoch": 1.259553178642284,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.079691841033643e-06,
+      "loss": 1.5522,
+      "step": 7301
+    },
+    {
+      "epoch": 1.2597256965410162,
+      "grad_norm": 0.703125,
+      "learning_rate": 6.077187988093138e-06,
+      "loss": 1.414,
+      "step": 7302
+    },
+    {
+      "epoch": 1.2598982144397481,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.074684425779063e-06,
+      "loss": 1.4165,
+      "step": 7303
+    },
+    {
+      "epoch": 1.26007073233848,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.072181154276888e-06,
+      "loss": 1.3574,
+      "step": 7304
+    },
+    {
+      "epoch": 1.260243250237212,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.069678173772079e-06,
+      "loss": 1.3784,
+      "step": 7305
+    },
+    {
+      "epoch": 1.260415768135944,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.067175484450063e-06,
+      "loss": 1.4636,
+      "step": 7306
+    },
+    {
+      "epoch": 1.2605882860346762,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.064673086496267e-06,
+      "loss": 1.3886,
+      "step": 7307
+    },
+    {
+      "epoch": 1.2607608039334082,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.062170980096073e-06,
+      "loss": 1.376,
+      "step": 7308
+    },
+    {
+      "epoch": 1.2609333218321401,
+      "grad_norm": 0.625,
+      "learning_rate": 6.05966916543486e-06,
+      "loss": 1.3851,
+      "step": 7309
+    },
+    {
+      "epoch": 1.261105839730872,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.057167642697973e-06,
+      "loss": 1.483,
+      "step": 7310
+    },
+    {
+      "epoch": 1.261278357629604,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.054666412070746e-06,
+      "loss": 1.4069,
+      "step": 7311
+    },
+    {
+      "epoch": 1.261450875528336,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.0521654737384804e-06,
+      "loss": 1.3993,
+      "step": 7312
+    },
+    {
+      "epoch": 1.261623393427068,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.049664827886468e-06,
+      "loss": 1.4738,
+      "step": 7313
+    },
+    {
+      "epoch": 1.2617959113258,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.047164474699962e-06,
+      "loss": 1.5021,
+      "step": 7314
+    },
+    {
+      "epoch": 1.2619684292245321,
+      "grad_norm": 0.625,
+      "learning_rate": 6.044664414364214e-06,
+      "loss": 1.4437,
+      "step": 7315
+    },
+    {
+      "epoch": 1.262140947123264,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.0421646470644394e-06,
+      "loss": 1.4966,
+      "step": 7316
+    },
+    {
+      "epoch": 1.262313465021996,
+      "grad_norm": 0.6484375,
+      "learning_rate": 6.039665172985834e-06,
+      "loss": 1.4586,
+      "step": 7317
+    },
+    {
+      "epoch": 1.262485982920728,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.0371659923135825e-06,
+      "loss": 1.4142,
+      "step": 7318
+    },
+    {
+      "epoch": 1.26265850081946,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.0346671052328345e-06,
+      "loss": 1.5371,
+      "step": 7319
+    },
+    {
+      "epoch": 1.2628310187181921,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.0321685119287245e-06,
+      "loss": 1.4882,
+      "step": 7320
+    },
+    {
+      "epoch": 1.263003536616924,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.0296702125863586e-06,
+      "loss": 1.4118,
+      "step": 7321
+    },
+    {
+      "epoch": 1.263176054515656,
+      "grad_norm": 0.6953125,
+      "learning_rate": 6.027172207390836e-06,
+      "loss": 1.4162,
+      "step": 7322
+    },
+    {
+      "epoch": 1.263348572414388,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.024674496527219e-06,
+      "loss": 1.3656,
+      "step": 7323
+    },
+    {
+      "epoch": 1.26352109031312,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.022177080180553e-06,
+      "loss": 1.4969,
+      "step": 7324
+    },
+    {
+      "epoch": 1.263693608211852,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.019679958535862e-06,
+      "loss": 1.378,
+      "step": 7325
+    },
+    {
+      "epoch": 1.263866126110584,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.017183131778154e-06,
+      "loss": 1.4315,
+      "step": 7326
+    },
+    {
+      "epoch": 1.2640386440093159,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.0146866000924035e-06,
+      "loss": 1.4713,
+      "step": 7327
+    },
+    {
+      "epoch": 1.2642111619080478,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.012190363663571e-06,
+      "loss": 1.3467,
+      "step": 7328
+    },
+    {
+      "epoch": 1.26438367980678,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.009694422676591e-06,
+      "loss": 1.4483,
+      "step": 7329
+    },
+    {
+      "epoch": 1.264556197705512,
+      "grad_norm": 0.7890625,
+      "learning_rate": 6.007198777316385e-06,
+      "loss": 1.4766,
+      "step": 7330
+    },
+    {
+      "epoch": 1.264728715604244,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.004703427767837e-06,
+      "loss": 1.3881,
+      "step": 7331
+    },
+    {
+      "epoch": 1.264901233502976,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.002208374215829e-06,
+      "loss": 1.4425,
+      "step": 7332
+    },
+    {
+      "epoch": 1.2650737514017079,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.999713616845197e-06,
+      "loss": 1.4362,
+      "step": 7333
+    },
+    {
+      "epoch": 1.26524626930044,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.997219155840777e-06,
+      "loss": 1.5301,
+      "step": 7334
+    },
+    {
+      "epoch": 1.265418787199172,
+      "grad_norm": 0.625,
+      "learning_rate": 5.994724991387375e-06,
+      "loss": 1.4292,
+      "step": 7335
+    },
+    {
+      "epoch": 1.265591305097904,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.992231123669771e-06,
+      "loss": 1.4772,
+      "step": 7336
+    },
+    {
+      "epoch": 1.265763822996636,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.989737552872729e-06,
+      "loss": 1.4154,
+      "step": 7337
+    },
+    {
+      "epoch": 1.2659363408953679,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.9872442791809815e-06,
+      "loss": 1.4353,
+      "step": 7338
+    },
+    {
+      "epoch": 1.2661088587940998,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.984751302779255e-06,
+      "loss": 1.4214,
+      "step": 7339
+    },
+    {
+      "epoch": 1.2662813766928318,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.982258623852239e-06,
+      "loss": 1.4753,
+      "step": 7340
+    },
+    {
+      "epoch": 1.2664538945915638,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.979766242584608e-06,
+      "loss": 1.5378,
+      "step": 7341
+    },
+    {
+      "epoch": 1.266626412490296,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.977274159161012e-06,
+      "loss": 1.3289,
+      "step": 7342
+    },
+    {
+      "epoch": 1.266798930389028,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.974782373766084e-06,
+      "loss": 1.405,
+      "step": 7343
+    },
+    {
+      "epoch": 1.2669714482877599,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.972290886584426e-06,
+      "loss": 1.4091,
+      "step": 7344
+    },
+    {
+      "epoch": 1.2671439661864918,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.969799697800629e-06,
+      "loss": 1.3101,
+      "step": 7345
+    },
+    {
+      "epoch": 1.2673164840852238,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.967308807599248e-06,
+      "loss": 1.472,
+      "step": 7346
+    },
+    {
+      "epoch": 1.2674890019839558,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.964818216164832e-06,
+      "loss": 1.5133,
+      "step": 7347
+    },
+    {
+      "epoch": 1.267661519882688,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.962327923681892e-06,
+      "loss": 1.3993,
+      "step": 7348
+    },
+    {
+      "epoch": 1.26783403778142,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.959837930334926e-06,
+      "loss": 1.4841,
+      "step": 7349
+    },
+    {
+      "epoch": 1.2680065556801519,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.9573482363084155e-06,
+      "loss": 1.4165,
+      "step": 7350
+    },
+    {
+      "epoch": 1.2681790735788838,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.954858841786806e-06,
+      "loss": 1.4749,
+      "step": 7351
+    },
+    {
+      "epoch": 1.2683515914776158,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.952369746954529e-06,
+      "loss": 1.3882,
+      "step": 7352
+    },
+    {
+      "epoch": 1.2685241093763477,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.9498809519959875e-06,
+      "loss": 1.5357,
+      "step": 7353
+    },
+    {
+      "epoch": 1.2686966272750797,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.9473924570955776e-06,
+      "loss": 1.3692,
+      "step": 7354
+    },
+    {
+      "epoch": 1.2688691451738117,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.944904262437653e-06,
+      "loss": 1.3353,
+      "step": 7355
+    },
+    {
+      "epoch": 1.2690416630725438,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.94241636820656e-06,
+      "loss": 1.4578,
+      "step": 7356
+    },
+    {
+      "epoch": 1.2692141809712758,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.939928774586612e-06,
+      "loss": 1.4813,
+      "step": 7357
+    },
+    {
+      "epoch": 1.2693866988700078,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.937441481762112e-06,
+      "loss": 1.464,
+      "step": 7358
+    },
+    {
+      "epoch": 1.2695592167687397,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.934954489917329e-06,
+      "loss": 1.4479,
+      "step": 7359
+    },
+    {
+      "epoch": 1.2697317346674717,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.93246779923652e-06,
+      "loss": 1.3553,
+      "step": 7360
+    },
+    {
+      "epoch": 1.2699042525662039,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.929981409903907e-06,
+      "loss": 1.4249,
+      "step": 7361
+    },
+    {
+      "epoch": 1.2700767704649358,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.927495322103707e-06,
+      "loss": 1.4009,
+      "step": 7362
+    },
+    {
+      "epoch": 1.2702492883636678,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.925009536020094e-06,
+      "loss": 1.4347,
+      "step": 7363
+    },
+    {
+      "epoch": 1.2704218062623998,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.922524051837241e-06,
+      "loss": 1.4666,
+      "step": 7364
+    },
+    {
+      "epoch": 1.2705943241611317,
+      "grad_norm": 0.6875,
+      "learning_rate": 5.9200388697392805e-06,
+      "loss": 1.4344,
+      "step": 7365
+    },
+    {
+      "epoch": 1.2707668420598637,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.9175539899103315e-06,
+      "loss": 1.351,
+      "step": 7366
+    },
+    {
+      "epoch": 1.2709393599585956,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.915069412534498e-06,
+      "loss": 1.347,
+      "step": 7367
+    },
+    {
+      "epoch": 1.2711118778573276,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.912585137795841e-06,
+      "loss": 1.4984,
+      "step": 7368
+    },
+    {
+      "epoch": 1.2712843957560596,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.910101165878419e-06,
+      "loss": 1.4683,
+      "step": 7369
+    },
+    {
+      "epoch": 1.2714569136547917,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.907617496966254e-06,
+      "loss": 1.3853,
+      "step": 7370
+    },
+    {
+      "epoch": 1.2716294315535237,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.90513413124336e-06,
+      "loss": 1.4295,
+      "step": 7371
+    },
+    {
+      "epoch": 1.2718019494522557,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.902651068893712e-06,
+      "loss": 1.5122,
+      "step": 7372
+    },
+    {
+      "epoch": 1.2719744673509876,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.9001683101012775e-06,
+      "loss": 1.3092,
+      "step": 7373
+    },
+    {
+      "epoch": 1.2721469852497196,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.897685855049986e-06,
+      "loss": 1.4631,
+      "step": 7374
+    },
+    {
+      "epoch": 1.2723195031484518,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.895203703923764e-06,
+      "loss": 1.4884,
+      "step": 7375
+    },
+    {
+      "epoch": 1.2724920210471837,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.892721856906498e-06,
+      "loss": 1.3907,
+      "step": 7376
+    },
+    {
+      "epoch": 1.2726645389459157,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.890240314182061e-06,
+      "loss": 1.4388,
+      "step": 7377
+    },
+    {
+      "epoch": 1.2728370568446477,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.887759075934297e-06,
+      "loss": 1.4239,
+      "step": 7378
+    },
+    {
+      "epoch": 1.2730095747433796,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.8852781423470395e-06,
+      "loss": 1.407,
+      "step": 7379
+    },
+    {
+      "epoch": 1.2731820926421116,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.882797513604085e-06,
+      "loss": 1.4078,
+      "step": 7380
+    },
+    {
+      "epoch": 1.2733546105408435,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.880317189889213e-06,
+      "loss": 1.528,
+      "step": 7381
+    },
+    {
+      "epoch": 1.2735271284395755,
+      "grad_norm": 0.625,
+      "learning_rate": 5.87783717138619e-06,
+      "loss": 1.4042,
+      "step": 7382
+    },
+    {
+      "epoch": 1.2736996463383077,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.8753574582787435e-06,
+      "loss": 1.325,
+      "step": 7383
+    },
+    {
+      "epoch": 1.2738721642370396,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.87287805075059e-06,
+      "loss": 1.5028,
+      "step": 7384
+    },
+    {
+      "epoch": 1.2740446821357716,
+      "grad_norm": 0.6953125,
+      "learning_rate": 5.870398948985414e-06,
+      "loss": 1.4158,
+      "step": 7385
+    },
+    {
+      "epoch": 1.2742172000345036,
+      "grad_norm": 0.6796875,
+      "learning_rate": 5.867920153166892e-06,
+      "loss": 1.4555,
+      "step": 7386
+    },
+    {
+      "epoch": 1.2743897179332355,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.865441663478661e-06,
+      "loss": 1.4297,
+      "step": 7387
+    },
+    {
+      "epoch": 1.2745622358319677,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.862963480104347e-06,
+      "loss": 1.4885,
+      "step": 7388
+    },
+    {
+      "epoch": 1.2747347537306997,
+      "grad_norm": 0.76171875,
+      "learning_rate": 5.860485603227544e-06,
+      "loss": 1.4481,
+      "step": 7389
+    },
+    {
+      "epoch": 1.2749072716294316,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.858008033031836e-06,
+      "loss": 1.4486,
+      "step": 7390
+    },
+    {
+      "epoch": 1.2750797895281636,
+      "grad_norm": 0.9453125,
+      "learning_rate": 5.855530769700769e-06,
+      "loss": 1.4396,
+      "step": 7391
+    },
+    {
+      "epoch": 1.2752523074268955,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.853053813417883e-06,
+      "loss": 1.5185,
+      "step": 7392
+    },
+    {
+      "epoch": 1.2754248253256275,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.85057716436668e-06,
+      "loss": 1.3693,
+      "step": 7393
+    },
+    {
+      "epoch": 1.2755973432243595,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.848100822730649e-06,
+      "loss": 1.494,
+      "step": 7394
+    },
+    {
+      "epoch": 1.2757698611230914,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.8456247886932475e-06,
+      "loss": 1.4265,
+      "step": 7395
+    },
+    {
+      "epoch": 1.2759423790218234,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.843149062437923e-06,
+      "loss": 1.4153,
+      "step": 7396
+    },
+    {
+      "epoch": 1.2761148969205556,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.840673644148087e-06,
+      "loss": 1.4869,
+      "step": 7397
+    },
+    {
+      "epoch": 1.2762874148192875,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.838198534007138e-06,
+      "loss": 1.4606,
+      "step": 7398
+    },
+    {
+      "epoch": 1.2764599327180195,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.835723732198444e-06,
+      "loss": 1.4313,
+      "step": 7399
+    },
+    {
+      "epoch": 1.2766324506167515,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.833249238905357e-06,
+      "loss": 1.446,
+      "step": 7400
+    },
+    {
+      "epoch": 1.2766324506167515,
+      "eval_loss": 1.4078235626220703,
+      "eval_runtime": 10.7712,
+      "eval_samples_per_second": 95.069,
+      "eval_steps_per_second": 23.767,
+      "step": 7400
+    },
+    {
+      "epoch": 1.2768049685154834,
+      "grad_norm": 0.74609375,
+      "learning_rate": 5.8307750543111996e-06,
+      "loss": 1.41,
+      "step": 7401
+    },
+    {
+      "epoch": 1.2769774864142156,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.828301178599277e-06,
+      "loss": 1.4281,
+      "step": 7402
+    },
+    {
+      "epoch": 1.2771500043129476,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.825827611952874e-06,
+      "loss": 1.5065,
+      "step": 7403
+    },
+    {
+      "epoch": 1.2773225222116795,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.823354354555234e-06,
+      "loss": 1.3859,
+      "step": 7404
+    },
+    {
+      "epoch": 1.2774950401104115,
+      "grad_norm": 0.875,
+      "learning_rate": 5.820881406589609e-06,
+      "loss": 1.4492,
+      "step": 7405
+    },
+    {
+      "epoch": 1.2776675580091434,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.818408768239197e-06,
+      "loss": 1.4723,
+      "step": 7406
+    },
+    {
+      "epoch": 1.2778400759078754,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.815936439687192e-06,
+      "loss": 1.3366,
+      "step": 7407
+    },
+    {
+      "epoch": 1.2780125938066074,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.813464421116759e-06,
+      "loss": 1.389,
+      "step": 7408
+    },
+    {
+      "epoch": 1.2781851117053393,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.810992712711039e-06,
+      "loss": 1.4805,
+      "step": 7409
+    },
+    {
+      "epoch": 1.2783576296040713,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.808521314653155e-06,
+      "loss": 1.4856,
+      "step": 7410
+    },
+    {
+      "epoch": 1.2785301475028035,
+      "grad_norm": 0.625,
+      "learning_rate": 5.806050227126203e-06,
+      "loss": 1.4922,
+      "step": 7411
+    },
+    {
+      "epoch": 1.2787026654015354,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.803579450313249e-06,
+      "loss": 1.4169,
+      "step": 7412
+    },
+    {
+      "epoch": 1.2788751833002674,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.801108984397355e-06,
+      "loss": 1.4809,
+      "step": 7413
+    },
+    {
+      "epoch": 1.2790477011989994,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.798638829561545e-06,
+      "loss": 1.4217,
+      "step": 7414
+    },
+    {
+      "epoch": 1.2792202190977313,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.796168985988814e-06,
+      "loss": 1.3787,
+      "step": 7415
+    },
+    {
+      "epoch": 1.2793927369964635,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.793699453862161e-06,
+      "loss": 1.5047,
+      "step": 7416
+    },
+    {
+      "epoch": 1.2795652548951955,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.791230233364529e-06,
+      "loss": 1.4522,
+      "step": 7417
+    },
+    {
+      "epoch": 1.2797377727939274,
+      "grad_norm": 0.80859375,
+      "learning_rate": 5.788761324678859e-06,
+      "loss": 1.3559,
+      "step": 7418
+    },
+    {
+      "epoch": 1.2799102906926594,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.7862927279880635e-06,
+      "loss": 1.4391,
+      "step": 7419
+    },
+    {
+      "epoch": 1.2800828085913913,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.78382444347503e-06,
+      "loss": 1.4629,
+      "step": 7420
+    },
+    {
+      "epoch": 1.2802553264901233,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.781356471322628e-06,
+      "loss": 1.4712,
+      "step": 7421
+    },
+    {
+      "epoch": 1.2804278443888553,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.7788888117136964e-06,
+      "loss": 1.4271,
+      "step": 7422
+    },
+    {
+      "epoch": 1.2806003622875872,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.7764214648310564e-06,
+      "loss": 1.4672,
+      "step": 7423
+    },
+    {
+      "epoch": 1.2807728801863194,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.773954430857509e-06,
+      "loss": 1.3926,
+      "step": 7424
+    },
+    {
+      "epoch": 1.2809453980850514,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.771487709975814e-06,
+      "loss": 1.3844,
+      "step": 7425
+    },
+    {
+      "epoch": 1.2811179159837833,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.769021302368739e-06,
+      "loss": 1.4494,
+      "step": 7426
+    },
+    {
+      "epoch": 1.2812904338825153,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.766555208218998e-06,
+      "loss": 1.4814,
+      "step": 7427
+    },
+    {
+      "epoch": 1.2814629517812473,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.7640894277093e-06,
+      "loss": 1.5364,
+      "step": 7428
+    },
+    {
+      "epoch": 1.2816354696799794,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.761623961022323e-06,
+      "loss": 1.4008,
+      "step": 7429
+    },
+    {
+      "epoch": 1.2818079875787114,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.759158808340726e-06,
+      "loss": 1.3162,
+      "step": 7430
+    },
+    {
+      "epoch": 1.2819805054774434,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.756693969847142e-06,
+      "loss": 1.3761,
+      "step": 7431
+    },
+    {
+      "epoch": 1.2821530233761753,
+      "grad_norm": 0.984375,
+      "learning_rate": 5.754229445724184e-06,
+      "loss": 1.4603,
+      "step": 7432
+    },
+    {
+      "epoch": 1.2823255412749073,
+      "grad_norm": 1.3046875,
+      "learning_rate": 5.751765236154436e-06,
+      "loss": 1.4425,
+      "step": 7433
+    },
+    {
+      "epoch": 1.2824980591736392,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.749301341320464e-06,
+      "loss": 1.4418,
+      "step": 7434
+    },
+    {
+      "epoch": 1.2826705770723712,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.746837761404811e-06,
+      "loss": 1.4066,
+      "step": 7435
+    },
+    {
+      "epoch": 1.2828430949711032,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.744374496589985e-06,
+      "loss": 1.3874,
+      "step": 7436
+    },
+    {
+      "epoch": 1.2830156128698351,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.741911547058494e-06,
+      "loss": 1.4895,
+      "step": 7437
+    },
+    {
+      "epoch": 1.2831881307685673,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.739448912992798e-06,
+      "loss": 1.4584,
+      "step": 7438
+    },
+    {
+      "epoch": 1.2833606486672993,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.736986594575347e-06,
+      "loss": 1.3845,
+      "step": 7439
+    },
+    {
+      "epoch": 1.2835331665660312,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.734524591988566e-06,
+      "loss": 1.3687,
+      "step": 7440
+    },
+    {
+      "epoch": 1.2837056844647632,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.732062905414855e-06,
+      "loss": 1.4917,
+      "step": 7441
+    },
+    {
+      "epoch": 1.2838782023634951,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.7296015350365905e-06,
+      "loss": 1.4639,
+      "step": 7442
+    },
+    {
+      "epoch": 1.2840507202622273,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.727140481036133e-06,
+      "loss": 1.4227,
+      "step": 7443
+    },
+    {
+      "epoch": 1.2842232381609593,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.7246797435957965e-06,
+      "loss": 1.4035,
+      "step": 7444
+    },
+    {
+      "epoch": 1.2843957560596913,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.722219322897904e-06,
+      "loss": 1.3356,
+      "step": 7445
+    },
+    {
+      "epoch": 1.2845682739584232,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.719759219124735e-06,
+      "loss": 1.4335,
+      "step": 7446
+    },
+    {
+      "epoch": 1.2847407918571552,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.71729943245854e-06,
+      "loss": 1.4359,
+      "step": 7447
+    },
+    {
+      "epoch": 1.2849133097558871,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.714839963081571e-06,
+      "loss": 1.383,
+      "step": 7448
+    },
+    {
+      "epoch": 1.285085827654619,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.712380811176024e-06,
+      "loss": 1.4616,
+      "step": 7449
+    },
+    {
+      "epoch": 1.285258345553351,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.709921976924106e-06,
+      "loss": 1.4082,
+      "step": 7450
+    },
+    {
+      "epoch": 1.285430863452083,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.707463460507967e-06,
+      "loss": 1.5528,
+      "step": 7451
+    },
+    {
+      "epoch": 1.2856033813508152,
+      "grad_norm": 0.6953125,
+      "learning_rate": 5.705005262109759e-06,
+      "loss": 1.4262,
+      "step": 7452
+    },
+    {
+      "epoch": 1.2857758992495472,
+      "grad_norm": 0.859375,
+      "learning_rate": 5.702547381911595e-06,
+      "loss": 1.4013,
+      "step": 7453
+    },
+    {
+      "epoch": 1.2859484171482791,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.700089820095573e-06,
+      "loss": 1.3427,
+      "step": 7454
+    },
+    {
+      "epoch": 1.286120935047011,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.697632576843766e-06,
+      "loss": 1.3756,
+      "step": 7455
+    },
+    {
+      "epoch": 1.286293452945743,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.695175652338223e-06,
+      "loss": 1.4234,
+      "step": 7456
+    },
+    {
+      "epoch": 1.2864659708444752,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.692719046760957e-06,
+      "loss": 1.4469,
+      "step": 7457
+    },
+    {
+      "epoch": 1.2866384887432072,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.690262760293986e-06,
+      "loss": 1.4501,
+      "step": 7458
+    },
+    {
+      "epoch": 1.2868110066419391,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.687806793119273e-06,
+      "loss": 1.3342,
+      "step": 7459
+    },
+    {
+      "epoch": 1.286983524540671,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.685351145418778e-06,
+      "loss": 1.427,
+      "step": 7460
+    },
+    {
+      "epoch": 1.287156042439403,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.682895817374429e-06,
+      "loss": 1.4885,
+      "step": 7461
+    },
+    {
+      "epoch": 1.287328560338135,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.680440809168131e-06,
+      "loss": 1.4662,
+      "step": 7462
+    },
+    {
+      "epoch": 1.287501078236867,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.677986120981769e-06,
+      "loss": 1.4228,
+      "step": 7463
+    },
+    {
+      "epoch": 1.287673596135599,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.675531752997201e-06,
+      "loss": 1.4616,
+      "step": 7464
+    },
+    {
+      "epoch": 1.2878461140343311,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.67307770539626e-06,
+      "loss": 1.3393,
+      "step": 7465
+    },
+    {
+      "epoch": 1.288018631933063,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.670623978360759e-06,
+      "loss": 1.374,
+      "step": 7466
+    },
+    {
+      "epoch": 1.288191149831795,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.668170572072489e-06,
+      "loss": 1.4372,
+      "step": 7467
+    },
+    {
+      "epoch": 1.288363667730527,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.6657174867132e-06,
+      "loss": 1.4121,
+      "step": 7468
+    },
+    {
+      "epoch": 1.288536185629259,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.663264722464651e-06,
+      "loss": 1.4619,
+      "step": 7469
+    },
+    {
+      "epoch": 1.2887087035279912,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.6608122795085444e-06,
+      "loss": 1.4514,
+      "step": 7470
+    },
+    {
+      "epoch": 1.2888812214267231,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.658360158026577e-06,
+      "loss": 1.3991,
+      "step": 7471
+    },
+    {
+      "epoch": 1.289053739325455,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.655908358200415e-06,
+      "loss": 1.4462,
+      "step": 7472
+    },
+    {
+      "epoch": 1.289226257224187,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.653456880211707e-06,
+      "loss": 1.4311,
+      "step": 7473
+    },
+    {
+      "epoch": 1.289398775122919,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.651005724242072e-06,
+      "loss": 1.4442,
+      "step": 7474
+    },
+    {
+      "epoch": 1.289571293021651,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.648554890473108e-06,
+      "loss": 1.5117,
+      "step": 7475
+    },
+    {
+      "epoch": 1.289743810920383,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.646104379086381e-06,
+      "loss": 1.4792,
+      "step": 7476
+    },
+    {
+      "epoch": 1.2899163288191149,
+      "grad_norm": 0.625,
+      "learning_rate": 5.64365419026345e-06,
+      "loss": 1.4261,
+      "step": 7477
+    },
+    {
+      "epoch": 1.2900888467178468,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.64120432418584e-06,
+      "loss": 1.3697,
+      "step": 7478
+    },
+    {
+      "epoch": 1.290261364616579,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.63875478103504e-06,
+      "loss": 1.379,
+      "step": 7479
+    },
+    {
+      "epoch": 1.290433882515311,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.636305560992545e-06,
+      "loss": 1.3918,
+      "step": 7480
+    },
+    {
+      "epoch": 1.290606400414043,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.6338566642397915e-06,
+      "loss": 1.4982,
+      "step": 7481
+    },
+    {
+      "epoch": 1.290778918312775,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.631408090958225e-06,
+      "loss": 1.4698,
+      "step": 7482
+    },
+    {
+      "epoch": 1.2909514362115069,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.62895984132924e-06,
+      "loss": 1.4574,
+      "step": 7483
+    },
+    {
+      "epoch": 1.291123954110239,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.626511915534221e-06,
+      "loss": 1.3892,
+      "step": 7484
+    },
+    {
+      "epoch": 1.291296472008971,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.624064313754525e-06,
+      "loss": 1.4437,
+      "step": 7485
+    },
+    {
+      "epoch": 1.291468989907703,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.621617036171487e-06,
+      "loss": 1.3831,
+      "step": 7486
+    },
+    {
+      "epoch": 1.291641507806435,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.619170082966419e-06,
+      "loss": 1.3423,
+      "step": 7487
+    },
+    {
+      "epoch": 1.291814025705167,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.616723454320605e-06,
+      "loss": 1.4161,
+      "step": 7488
+    },
+    {
+      "epoch": 1.2919865436038989,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.6142771504152995e-06,
+      "loss": 1.4779,
+      "step": 7489
+    },
+    {
+      "epoch": 1.2921590615026308,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.611831171431752e-06,
+      "loss": 1.4647,
+      "step": 7490
+    },
+    {
+      "epoch": 1.2923315794013628,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.609385517551167e-06,
+      "loss": 1.3995,
+      "step": 7491
+    },
+    {
+      "epoch": 1.292504097300095,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.606940188954737e-06,
+      "loss": 1.5,
+      "step": 7492
+    },
+    {
+      "epoch": 1.292676615198827,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.604495185823626e-06,
+      "loss": 1.4466,
+      "step": 7493
+    },
+    {
+      "epoch": 1.292849133097559,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.602050508338976e-06,
+      "loss": 1.5505,
+      "step": 7494
+    },
+    {
+      "epoch": 1.2930216509962908,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.599606156681905e-06,
+      "loss": 1.3898,
+      "step": 7495
+    },
+    {
+      "epoch": 1.2931941688950228,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.597162131033505e-06,
+      "loss": 1.4221,
+      "step": 7496
+    },
+    {
+      "epoch": 1.2933666867937548,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.594718431574844e-06,
+      "loss": 1.3453,
+      "step": 7497
+    },
+    {
+      "epoch": 1.293539204692487,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.592275058486967e-06,
+      "loss": 1.4244,
+      "step": 7498
+    },
+    {
+      "epoch": 1.293711722591219,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.589832011950897e-06,
+      "loss": 1.3701,
+      "step": 7499
+    },
+    {
+      "epoch": 1.2938842404899509,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.5873892921476215e-06,
+      "loss": 1.4186,
+      "step": 7500
+    },
+    {
+      "epoch": 1.2938842404899509,
+      "eval_loss": 1.4076613187789917,
+      "eval_runtime": 10.8861,
+      "eval_samples_per_second": 94.065,
+      "eval_steps_per_second": 23.516,
+      "step": 7500
+    },
+    {
+      "epoch": 1.2940567583886828,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.584946899258125e-06,
+      "loss": 1.3333,
+      "step": 7501
+    },
+    {
+      "epoch": 1.2942292762874148,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.582504833463347e-06,
+      "loss": 1.4381,
+      "step": 7502
+    },
+    {
+      "epoch": 1.2944017941861468,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.58006309494421e-06,
+      "loss": 1.4592,
+      "step": 7503
+    },
+    {
+      "epoch": 1.2945743120848787,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.577621683881618e-06,
+      "loss": 1.4971,
+      "step": 7504
+    },
+    {
+      "epoch": 1.2947468299836107,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.5751806004564435e-06,
+      "loss": 1.396,
+      "step": 7505
+    },
+    {
+      "epoch": 1.2949193478823429,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.572739844849537e-06,
+      "loss": 1.4507,
+      "step": 7506
+    },
+    {
+      "epoch": 1.2950918657810748,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.57029941724173e-06,
+      "loss": 1.3708,
+      "step": 7507
+    },
+    {
+      "epoch": 1.2952643836798068,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.5678593178138125e-06,
+      "loss": 1.4396,
+      "step": 7508
+    },
+    {
+      "epoch": 1.2954369015785387,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.565419546746574e-06,
+      "loss": 1.4847,
+      "step": 7509
+    },
+    {
+      "epoch": 1.2956094194772707,
+      "grad_norm": 0.6796875,
+      "learning_rate": 5.562980104220763e-06,
+      "loss": 1.3488,
+      "step": 7510
+    },
+    {
+      "epoch": 1.295781937376003,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.560540990417111e-06,
+      "loss": 1.3707,
+      "step": 7511
+    },
+    {
+      "epoch": 1.2959544552747349,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.558102205516325e-06,
+      "loss": 1.4186,
+      "step": 7512
+    },
+    {
+      "epoch": 1.2961269731734668,
+      "grad_norm": 0.625,
+      "learning_rate": 5.555663749699074e-06,
+      "loss": 1.3705,
+      "step": 7513
+    },
+    {
+      "epoch": 1.2962994910721988,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.55322562314603e-06,
+      "loss": 1.5207,
+      "step": 7514
+    },
+    {
+      "epoch": 1.2964720089709307,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.5507878260378115e-06,
+      "loss": 1.4564,
+      "step": 7515
+    },
+    {
+      "epoch": 1.2966445268696627,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.5483503585550326e-06,
+      "loss": 1.4188,
+      "step": 7516
+    },
+    {
+      "epoch": 1.2968170447683947,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.545913220878272e-06,
+      "loss": 1.3734,
+      "step": 7517
+    },
+    {
+      "epoch": 1.2969895626671266,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.5434764131880915e-06,
+      "loss": 1.3631,
+      "step": 7518
+    },
+    {
+      "epoch": 1.2971620805658586,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.541039935665025e-06,
+      "loss": 1.4003,
+      "step": 7519
+    },
+    {
+      "epoch": 1.2973345984645908,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.538603788489584e-06,
+      "loss": 1.454,
+      "step": 7520
+    },
+    {
+      "epoch": 1.2975071163633227,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.5361679718422426e-06,
+      "loss": 1.3602,
+      "step": 7521
+    },
+    {
+      "epoch": 1.2976796342620547,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.533732485903477e-06,
+      "loss": 1.4205,
+      "step": 7522
+    },
+    {
+      "epoch": 1.2978521521607866,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.531297330853711e-06,
+      "loss": 1.425,
+      "step": 7523
+    },
+    {
+      "epoch": 1.2980246700595186,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.528862506873361e-06,
+      "loss": 1.419,
+      "step": 7524
+    },
+    {
+      "epoch": 1.2981971879582508,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.526428014142814e-06,
+      "loss": 1.5736,
+      "step": 7525
+    },
+    {
+      "epoch": 1.2983697058569827,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.523993852842431e-06,
+      "loss": 1.3834,
+      "step": 7526
+    },
+    {
+      "epoch": 1.2985422237557147,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.521560023152552e-06,
+      "loss": 1.4649,
+      "step": 7527
+    },
+    {
+      "epoch": 1.2987147416544467,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.519126525253486e-06,
+      "loss": 1.465,
+      "step": 7528
+    },
+    {
+      "epoch": 1.2988872595531786,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.516693359325528e-06,
+      "loss": 1.4627,
+      "step": 7529
+    },
+    {
+      "epoch": 1.2990597774519106,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.514260525548938e-06,
+      "loss": 1.3427,
+      "step": 7530
+    },
+    {
+      "epoch": 1.2992322953506426,
+      "grad_norm": 1.4375,
+      "learning_rate": 5.51182802410396e-06,
+      "loss": 1.3953,
+      "step": 7531
+    },
+    {
+      "epoch": 1.2994048132493745,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.509395855170798e-06,
+      "loss": 1.345,
+      "step": 7532
+    },
+    {
+      "epoch": 1.2995773311481067,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.506964018929657e-06,
+      "loss": 1.3181,
+      "step": 7533
+    },
+    {
+      "epoch": 1.2997498490468387,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.5045325155606925e-06,
+      "loss": 1.5654,
+      "step": 7534
+    },
+    {
+      "epoch": 1.2999223669455706,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.502101345244047e-06,
+      "loss": 1.441,
+      "step": 7535
+    },
+    {
+      "epoch": 1.3000948848443026,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.499670508159838e-06,
+      "loss": 1.4018,
+      "step": 7536
+    },
+    {
+      "epoch": 1.3002674027430345,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.497240004488158e-06,
+      "loss": 1.3829,
+      "step": 7537
+    },
+    {
+      "epoch": 1.3004399206417665,
+      "grad_norm": 1.0,
+      "learning_rate": 5.494809834409071e-06,
+      "loss": 1.4306,
+      "step": 7538
+    },
+    {
+      "epoch": 1.3006124385404987,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.492379998102627e-06,
+      "loss": 1.4568,
+      "step": 7539
+    },
+    {
+      "epoch": 1.3007849564392306,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.48995049574883e-06,
+      "loss": 1.5109,
+      "step": 7540
+    },
+    {
+      "epoch": 1.3009574743379626,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.4875213275276875e-06,
+      "loss": 1.3706,
+      "step": 7541
+    },
+    {
+      "epoch": 1.3011299922366946,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.485092493619153e-06,
+      "loss": 1.3853,
+      "step": 7542
+    },
+    {
+      "epoch": 1.3013025101354265,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.482663994203179e-06,
+      "loss": 1.426,
+      "step": 7543
+    },
+    {
+      "epoch": 1.3014750280341585,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.480235829459688e-06,
+      "loss": 1.4036,
+      "step": 7544
+    },
+    {
+      "epoch": 1.3016475459328904,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.477807999568558e-06,
+      "loss": 1.4661,
+      "step": 7545
+    },
+    {
+      "epoch": 1.3018200638316224,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.475380504709678e-06,
+      "loss": 1.3434,
+      "step": 7546
+    },
+    {
+      "epoch": 1.3019925817303546,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.472953345062875e-06,
+      "loss": 1.3539,
+      "step": 7547
+    },
+    {
+      "epoch": 1.3021650996290866,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.470526520807975e-06,
+      "loss": 1.4634,
+      "step": 7548
+    },
+    {
+      "epoch": 1.3023376175278185,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.4681000321247725e-06,
+      "loss": 1.3594,
+      "step": 7549
+    },
+    {
+      "epoch": 1.3025101354265505,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.465673879193035e-06,
+      "loss": 1.5056,
+      "step": 7550
+    },
+    {
+      "epoch": 1.3026826533252824,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.46324806219251e-06,
+      "loss": 1.3833,
+      "step": 7551
+    },
+    {
+      "epoch": 1.3028551712240146,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.460822581302918e-06,
+      "loss": 1.3744,
+      "step": 7552
+    },
+    {
+      "epoch": 1.3030276891227466,
+      "grad_norm": 0.67578125,
+      "learning_rate": 5.458397436703944e-06,
+      "loss": 1.4649,
+      "step": 7553
+    },
+    {
+      "epoch": 1.3032002070214785,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.455972628575272e-06,
+      "loss": 1.4986,
+      "step": 7554
+    },
+    {
+      "epoch": 1.3033727249202105,
+      "grad_norm": 0.67578125,
+      "learning_rate": 5.453548157096538e-06,
+      "loss": 1.5157,
+      "step": 7555
+    },
+    {
+      "epoch": 1.3035452428189425,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.4511240224473625e-06,
+      "loss": 1.4927,
+      "step": 7556
+    },
+    {
+      "epoch": 1.3037177607176744,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.448700224807342e-06,
+      "loss": 1.4667,
+      "step": 7557
+    },
+    {
+      "epoch": 1.3038902786164064,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.446276764356048e-06,
+      "loss": 1.4254,
+      "step": 7558
+    },
+    {
+      "epoch": 1.3040627965151383,
+      "grad_norm": 1.484375,
+      "learning_rate": 5.443853641273024e-06,
+      "loss": 1.338,
+      "step": 7559
+    },
+    {
+      "epoch": 1.3042353144138703,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.441430855737789e-06,
+      "loss": 1.3658,
+      "step": 7560
+    },
+    {
+      "epoch": 1.3044078323126025,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.43900840792984e-06,
+      "loss": 1.4133,
+      "step": 7561
+    },
+    {
+      "epoch": 1.3045803502113344,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.436586298028647e-06,
+      "loss": 1.4209,
+      "step": 7562
+    },
+    {
+      "epoch": 1.3047528681100664,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.434164526213659e-06,
+      "loss": 1.4423,
+      "step": 7563
+    },
+    {
+      "epoch": 1.3049253860087984,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.431743092664283e-06,
+      "loss": 1.3585,
+      "step": 7564
+    },
+    {
+      "epoch": 1.3050979039075303,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.429321997559931e-06,
+      "loss": 1.4644,
+      "step": 7565
+    },
+    {
+      "epoch": 1.3052704218062625,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.42690124107996e-06,
+      "loss": 1.4258,
+      "step": 7566
+    },
+    {
+      "epoch": 1.3054429397049945,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.4244808234037195e-06,
+      "loss": 1.3962,
+      "step": 7567
+    },
+    {
+      "epoch": 1.3056154576037264,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.422060744710527e-06,
+      "loss": 1.3899,
+      "step": 7568
+    },
+    {
+      "epoch": 1.3057879755024584,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.419641005179681e-06,
+      "loss": 1.474,
+      "step": 7569
+    },
+    {
+      "epoch": 1.3059604934011904,
+      "grad_norm": 0.81640625,
+      "learning_rate": 5.417221604990448e-06,
+      "loss": 1.5104,
+      "step": 7570
+    },
+    {
+      "epoch": 1.3061330112999223,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.414802544322072e-06,
+      "loss": 1.5238,
+      "step": 7571
+    },
+    {
+      "epoch": 1.3063055291986543,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.412383823353774e-06,
+      "loss": 1.3963,
+      "step": 7572
+    },
+    {
+      "epoch": 1.3064780470973862,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.409965442264751e-06,
+      "loss": 1.3876,
+      "step": 7573
+    },
+    {
+      "epoch": 1.3066505649961184,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.407547401234161e-06,
+      "loss": 1.4604,
+      "step": 7574
+    },
+    {
+      "epoch": 1.3068230828948504,
+      "grad_norm": 0.6875,
+      "learning_rate": 5.405129700441157e-06,
+      "loss": 1.5196,
+      "step": 7575
+    },
+    {
+      "epoch": 1.3069956007935823,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.402712340064859e-06,
+      "loss": 1.3634,
+      "step": 7576
+    },
+    {
+      "epoch": 1.3071681186923143,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.400295320284348e-06,
+      "loss": 1.5047,
+      "step": 7577
+    },
+    {
+      "epoch": 1.3073406365910463,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.397878641278709e-06,
+      "loss": 1.3541,
+      "step": 7578
+    },
+    {
+      "epoch": 1.3075131544897785,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.3954623032269705e-06,
+      "loss": 1.4108,
+      "step": 7579
+    },
+    {
+      "epoch": 1.3076856723885104,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.3930463063081564e-06,
+      "loss": 1.4113,
+      "step": 7580
+    },
+    {
+      "epoch": 1.3078581902872424,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.390630650701257e-06,
+      "loss": 1.4792,
+      "step": 7581
+    },
+    {
+      "epoch": 1.3080307081859743,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.388215336585239e-06,
+      "loss": 1.4458,
+      "step": 7582
+    },
+    {
+      "epoch": 1.3082032260847063,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.385800364139044e-06,
+      "loss": 1.4197,
+      "step": 7583
+    },
+    {
+      "epoch": 1.3083757439834383,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.383385733541594e-06,
+      "loss": 1.3762,
+      "step": 7584
+    },
+    {
+      "epoch": 1.3085482618821702,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.380971444971766e-06,
+      "loss": 1.4279,
+      "step": 7585
+    },
+    {
+      "epoch": 1.3087207797809022,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.3785574986084435e-06,
+      "loss": 1.4792,
+      "step": 7586
+    },
+    {
+      "epoch": 1.3088932976796341,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.376143894630454e-06,
+      "loss": 1.4466,
+      "step": 7587
+    },
+    {
+      "epoch": 1.3090658155783663,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.373730633216614e-06,
+      "loss": 1.4329,
+      "step": 7588
+    },
+    {
+      "epoch": 1.3092383334770983,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.3713177145457165e-06,
+      "loss": 1.4848,
+      "step": 7589
+    },
+    {
+      "epoch": 1.3094108513758302,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.368905138796523e-06,
+      "loss": 1.4444,
+      "step": 7590
+    },
+    {
+      "epoch": 1.3095833692745622,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.366492906147775e-06,
+      "loss": 1.466,
+      "step": 7591
+    },
+    {
+      "epoch": 1.3097558871732942,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.364081016778182e-06,
+      "loss": 1.406,
+      "step": 7592
+    },
+    {
+      "epoch": 1.3099284050720263,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.361669470866435e-06,
+      "loss": 1.3943,
+      "step": 7593
+    },
+    {
+      "epoch": 1.3101009229707583,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.359258268591195e-06,
+      "loss": 1.4474,
+      "step": 7594
+    },
+    {
+      "epoch": 1.3102734408694903,
+      "grad_norm": 0.625,
+      "learning_rate": 5.356847410131103e-06,
+      "loss": 1.4458,
+      "step": 7595
+    },
+    {
+      "epoch": 1.3104459587682222,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.354436895664759e-06,
+      "loss": 1.4278,
+      "step": 7596
+    },
+    {
+      "epoch": 1.3106184766669542,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.352026725370763e-06,
+      "loss": 1.4711,
+      "step": 7597
+    },
+    {
+      "epoch": 1.3107909945656862,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.3496168994276635e-06,
+      "loss": 1.423,
+      "step": 7598
+    },
+    {
+      "epoch": 1.3109635124644181,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.347207418014006e-06,
+      "loss": 1.4974,
+      "step": 7599
+    },
+    {
+      "epoch": 1.31113603036315,
+      "grad_norm": 0.75,
+      "learning_rate": 5.344798281308295e-06,
+      "loss": 1.3676,
+      "step": 7600
+    },
+    {
+      "epoch": 1.31113603036315,
+      "eval_loss": 1.407556176185608,
+      "eval_runtime": 10.9219,
+      "eval_samples_per_second": 93.756,
+      "eval_steps_per_second": 23.439,
+      "step": 7600
+    },
+    {
+      "epoch": 1.311308548261882,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.34238948948901e-06,
+      "loss": 1.477,
+      "step": 7601
+    },
+    {
+      "epoch": 1.3114810661606142,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.339981042734617e-06,
+      "loss": 1.4956,
+      "step": 7602
+    },
+    {
+      "epoch": 1.3116535840593462,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.337572941223544e-06,
+      "loss": 1.4456,
+      "step": 7603
+    },
+    {
+      "epoch": 1.3118261019580781,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.3351651851342e-06,
+      "loss": 1.5027,
+      "step": 7604
+    },
+    {
+      "epoch": 1.31199861985681,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.33275777464497e-06,
+      "loss": 1.4509,
+      "step": 7605
+    },
+    {
+      "epoch": 1.312171137755542,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.3303507099342e-06,
+      "loss": 1.4749,
+      "step": 7606
+    },
+    {
+      "epoch": 1.3123436556542742,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.3279439911802286e-06,
+      "loss": 1.4949,
+      "step": 7607
+    },
+    {
+      "epoch": 1.3125161735530062,
+      "grad_norm": 0.83203125,
+      "learning_rate": 5.325537618561364e-06,
+      "loss": 1.5091,
+      "step": 7608
+    },
+    {
+      "epoch": 1.3126886914517382,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.323131592255871e-06,
+      "loss": 1.5434,
+      "step": 7609
+    },
+    {
+      "epoch": 1.3128612093504701,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.3207259124420205e-06,
+      "loss": 1.4441,
+      "step": 7610
+    },
+    {
+      "epoch": 1.313033727249202,
+      "grad_norm": 0.73828125,
+      "learning_rate": 5.318320579298028e-06,
+      "loss": 1.4234,
+      "step": 7611
+    },
+    {
+      "epoch": 1.313206245147934,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.3159155930021e-06,
+      "loss": 1.4529,
+      "step": 7612
+    },
+    {
+      "epoch": 1.313378763046666,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.313510953732411e-06,
+      "loss": 1.4523,
+      "step": 7613
+    },
+    {
+      "epoch": 1.313551280945398,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.311106661667115e-06,
+      "loss": 1.4488,
+      "step": 7614
+    },
+    {
+      "epoch": 1.3137237988441302,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.308702716984333e-06,
+      "loss": 1.3764,
+      "step": 7615
+    },
+    {
+      "epoch": 1.3138963167428621,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.306299119862171e-06,
+      "loss": 1.3542,
+      "step": 7616
+    },
+    {
+      "epoch": 1.314068834641594,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.30389587047869e-06,
+      "loss": 1.4034,
+      "step": 7617
+    },
+    {
+      "epoch": 1.314241352540326,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.301492969011954e-06,
+      "loss": 1.4876,
+      "step": 7618
+    },
+    {
+      "epoch": 1.314413870439058,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.299090415639973e-06,
+      "loss": 1.5078,
+      "step": 7619
+    },
+    {
+      "epoch": 1.3145863883377902,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.296688210540746e-06,
+      "loss": 1.4435,
+      "step": 7620
+    },
+    {
+      "epoch": 1.3147589062365221,
+      "grad_norm": 0.9375,
+      "learning_rate": 5.294286353892243e-06,
+      "loss": 1.5348,
+      "step": 7621
+    },
+    {
+      "epoch": 1.314931424135254,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.29188484587241e-06,
+      "loss": 1.3537,
+      "step": 7622
+    },
+    {
+      "epoch": 1.315103942033986,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.2894836866591655e-06,
+      "loss": 1.4277,
+      "step": 7623
+    },
+    {
+      "epoch": 1.315276459932718,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.287082876430403e-06,
+      "loss": 1.3934,
+      "step": 7624
+    },
+    {
+      "epoch": 1.31544897783145,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.284682415363988e-06,
+      "loss": 1.4736,
+      "step": 7625
+    },
+    {
+      "epoch": 1.315621495730182,
+      "grad_norm": 0.66015625,
+      "learning_rate": 5.2822823036377625e-06,
+      "loss": 1.3568,
+      "step": 7626
+    },
+    {
+      "epoch": 1.315794013628914,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.279882541429544e-06,
+      "loss": 1.36,
+      "step": 7627
+    },
+    {
+      "epoch": 1.3159665315276459,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.2774831289171136e-06,
+      "loss": 1.4064,
+      "step": 7628
+    },
+    {
+      "epoch": 1.316139049426378,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.275084066278248e-06,
+      "loss": 1.4406,
+      "step": 7629
+    },
+    {
+      "epoch": 1.31631156732511,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.27268535369067e-06,
+      "loss": 1.4174,
+      "step": 7630
+    },
+    {
+      "epoch": 1.316484085223842,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.270286991332106e-06,
+      "loss": 1.2914,
+      "step": 7631
+    },
+    {
+      "epoch": 1.316656603122574,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.267888979380229e-06,
+      "loss": 1.3725,
+      "step": 7632
+    },
+    {
+      "epoch": 1.316829121021306,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.265491318012705e-06,
+      "loss": 1.392,
+      "step": 7633
+    },
+    {
+      "epoch": 1.317001638920038,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.263094007407168e-06,
+      "loss": 1.4334,
+      "step": 7634
+    },
+    {
+      "epoch": 1.31717415681877,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.2606970477412236e-06,
+      "loss": 1.4093,
+      "step": 7635
+    },
+    {
+      "epoch": 1.317346674717502,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.258300439192454e-06,
+      "loss": 1.4355,
+      "step": 7636
+    },
+    {
+      "epoch": 1.317519192616234,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.255904181938419e-06,
+      "loss": 1.4449,
+      "step": 7637
+    },
+    {
+      "epoch": 1.317691710514966,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.253508276156638e-06,
+      "loss": 1.447,
+      "step": 7638
+    },
+    {
+      "epoch": 1.3178642284136979,
+      "grad_norm": 0.91796875,
+      "learning_rate": 5.251112722024625e-06,
+      "loss": 1.4952,
+      "step": 7639
+    },
+    {
+      "epoch": 1.3180367463124298,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.248717519719857e-06,
+      "loss": 1.4281,
+      "step": 7640
+    },
+    {
+      "epoch": 1.3182092642111618,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.246322669419775e-06,
+      "loss": 1.3986,
+      "step": 7641
+    },
+    {
+      "epoch": 1.3183817821098938,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.2439281713018196e-06,
+      "loss": 1.4189,
+      "step": 7642
+    },
+    {
+      "epoch": 1.318554300008626,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.24153402554338e-06,
+      "loss": 1.4394,
+      "step": 7643
+    },
+    {
+      "epoch": 1.318726817907358,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.239140232321831e-06,
+      "loss": 1.4017,
+      "step": 7644
+    },
+    {
+      "epoch": 1.3188993358060899,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.236746791814522e-06,
+      "loss": 1.4656,
+      "step": 7645
+    },
+    {
+      "epoch": 1.3190718537048218,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.2343537041987715e-06,
+      "loss": 1.4927,
+      "step": 7646
+    },
+    {
+      "epoch": 1.3192443716035538,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.231960969651876e-06,
+      "loss": 1.4623,
+      "step": 7647
+    },
+    {
+      "epoch": 1.319416889502286,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.2295685883511086e-06,
+      "loss": 1.5245,
+      "step": 7648
+    },
+    {
+      "epoch": 1.319589407401018,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.227176560473698e-06,
+      "loss": 1.3987,
+      "step": 7649
+    },
+    {
+      "epoch": 1.31976192529975,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.224784886196878e-06,
+      "loss": 1.4556,
+      "step": 7650
+    },
+    {
+      "epoch": 1.3199344431984819,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.222393565697828e-06,
+      "loss": 1.4211,
+      "step": 7651
+    },
+    {
+      "epoch": 1.3201069610972138,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.2200025991537126e-06,
+      "loss": 1.4429,
+      "step": 7652
+    },
+    {
+      "epoch": 1.3202794789959458,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.217611986741673e-06,
+      "loss": 1.4242,
+      "step": 7653
+    },
+    {
+      "epoch": 1.3204519968946777,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.215221728638815e-06,
+      "loss": 1.4632,
+      "step": 7654
+    },
+    {
+      "epoch": 1.3206245147934097,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.2128318250222355e-06,
+      "loss": 1.4538,
+      "step": 7655
+    },
+    {
+      "epoch": 1.3207970326921419,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.210442276068981e-06,
+      "loss": 1.4823,
+      "step": 7656
+    },
+    {
+      "epoch": 1.3209695505908738,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.208053081956091e-06,
+      "loss": 1.3459,
+      "step": 7657
+    },
+    {
+      "epoch": 1.3211420684896058,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.205664242860568e-06,
+      "loss": 1.402,
+      "step": 7658
+    },
+    {
+      "epoch": 1.3213145863883378,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.203275758959396e-06,
+      "loss": 1.376,
+      "step": 7659
+    },
+    {
+      "epoch": 1.3214871042870697,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.200887630429528e-06,
+      "loss": 1.3909,
+      "step": 7660
+    },
+    {
+      "epoch": 1.321659622185802,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.198499857447894e-06,
+      "loss": 1.4473,
+      "step": 7661
+    },
+    {
+      "epoch": 1.3218321400845339,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.196112440191383e-06,
+      "loss": 1.3559,
+      "step": 7662
+    },
+    {
+      "epoch": 1.3220046579832658,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.193725378836886e-06,
+      "loss": 1.4613,
+      "step": 7663
+    },
+    {
+      "epoch": 1.3221771758819978,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.19133867356124e-06,
+      "loss": 1.3656,
+      "step": 7664
+    },
+    {
+      "epoch": 1.3223496937807298,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.188952324541272e-06,
+      "loss": 1.4035,
+      "step": 7665
+    },
+    {
+      "epoch": 1.3225222116794617,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.1865663319537764e-06,
+      "loss": 1.4547,
+      "step": 7666
+    },
+    {
+      "epoch": 1.3226947295781937,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.184180695975522e-06,
+      "loss": 1.4893,
+      "step": 7667
+    },
+    {
+      "epoch": 1.3228672474769256,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.181795416783253e-06,
+      "loss": 1.4425,
+      "step": 7668
+    },
+    {
+      "epoch": 1.3230397653756576,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.1794104945536886e-06,
+      "loss": 1.4272,
+      "step": 7669
+    },
+    {
+      "epoch": 1.3232122832743898,
+      "grad_norm": 2.5,
+      "learning_rate": 5.1770259294635075e-06,
+      "loss": 1.4514,
+      "step": 7670
+    },
+    {
+      "epoch": 1.3233848011731217,
+      "grad_norm": 0.7890625,
+      "learning_rate": 5.1746417216893845e-06,
+      "loss": 1.4506,
+      "step": 7671
+    },
+    {
+      "epoch": 1.3235573190718537,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.172257871407957e-06,
+      "loss": 1.4556,
+      "step": 7672
+    },
+    {
+      "epoch": 1.3237298369705857,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.169874378795824e-06,
+      "loss": 1.3554,
+      "step": 7673
+    },
+    {
+      "epoch": 1.3239023548693176,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.167491244029584e-06,
+      "loss": 1.4715,
+      "step": 7674
+    },
+    {
+      "epoch": 1.3240748727680498,
+      "grad_norm": 3.78125,
+      "learning_rate": 5.165108467285784e-06,
+      "loss": 1.4466,
+      "step": 7675
+    },
+    {
+      "epoch": 1.3242473906667818,
+      "grad_norm": 0.70703125,
+      "learning_rate": 5.16272604874096e-06,
+      "loss": 1.4296,
+      "step": 7676
+    },
+    {
+      "epoch": 1.3244199085655137,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.160343988571614e-06,
+      "loss": 1.4277,
+      "step": 7677
+    },
+    {
+      "epoch": 1.3245924264642457,
+      "grad_norm": 0.625,
+      "learning_rate": 5.157962286954224e-06,
+      "loss": 1.5272,
+      "step": 7678
+    },
+    {
+      "epoch": 1.3247649443629776,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.155580944065244e-06,
+      "loss": 1.5264,
+      "step": 7679
+    },
+    {
+      "epoch": 1.3249374622617096,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.153199960081099e-06,
+      "loss": 1.3906,
+      "step": 7680
+    },
+    {
+      "epoch": 1.3251099801604416,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.150819335178179e-06,
+      "loss": 1.4098,
+      "step": 7681
+    },
+    {
+      "epoch": 1.3252824980591735,
+      "grad_norm": 0.73828125,
+      "learning_rate": 5.148439069532868e-06,
+      "loss": 1.3736,
+      "step": 7682
+    },
+    {
+      "epoch": 1.3254550159579057,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.1460591633215015e-06,
+      "loss": 1.36,
+      "step": 7683
+    },
+    {
+      "epoch": 1.3256275338566377,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.143679616720401e-06,
+      "loss": 1.4018,
+      "step": 7684
+    },
+    {
+      "epoch": 1.3258000517553696,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.141300429905858e-06,
+      "loss": 1.4803,
+      "step": 7685
+    },
+    {
+      "epoch": 1.3259725696541016,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.1389216030541345e-06,
+      "loss": 1.3943,
+      "step": 7686
+    },
+    {
+      "epoch": 1.3261450875528336,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.1365431363414784e-06,
+      "loss": 1.3702,
+      "step": 7687
+    },
+    {
+      "epoch": 1.3263176054515655,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.134165029944094e-06,
+      "loss": 1.5292,
+      "step": 7688
+    },
+    {
+      "epoch": 1.3264901233502977,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.1317872840381645e-06,
+      "loss": 1.4318,
+      "step": 7689
+    },
+    {
+      "epoch": 1.3266626412490297,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.129409898799852e-06,
+      "loss": 1.4077,
+      "step": 7690
+    },
+    {
+      "epoch": 1.3268351591477616,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.1270328744052864e-06,
+      "loss": 1.4297,
+      "step": 7691
+    },
+    {
+      "epoch": 1.3270076770464936,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.124656211030574e-06,
+      "loss": 1.4418,
+      "step": 7692
+    },
+    {
+      "epoch": 1.3271801949452255,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.122279908851796e-06,
+      "loss": 1.4186,
+      "step": 7693
+    },
+    {
+      "epoch": 1.3273527128439575,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.119903968044992e-06,
+      "loss": 1.4164,
+      "step": 7694
+    },
+    {
+      "epoch": 1.3275252307426895,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.117528388786201e-06,
+      "loss": 1.4222,
+      "step": 7695
+    },
+    {
+      "epoch": 1.3276977486414214,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.1151531712514115e-06,
+      "loss": 1.3717,
+      "step": 7696
+    },
+    {
+      "epoch": 1.3278702665401536,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.112778315616596e-06,
+      "loss": 1.4239,
+      "step": 7697
+    },
+    {
+      "epoch": 1.3280427844388856,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.1104038220577e-06,
+      "loss": 1.4251,
+      "step": 7698
+    },
+    {
+      "epoch": 1.3282153023376175,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.108029690750641e-06,
+      "loss": 1.3776,
+      "step": 7699
+    },
+    {
+      "epoch": 1.3283878202363495,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.105655921871309e-06,
+      "loss": 1.4722,
+      "step": 7700
+    },
+    {
+      "epoch": 1.3283878202363495,
+      "eval_loss": 1.4075722694396973,
+      "eval_runtime": 11.0299,
+      "eval_samples_per_second": 92.839,
+      "eval_steps_per_second": 23.21,
+      "step": 7700
+    },
+    {
+      "epoch": 1.3285603381350815,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.10328251559557e-06,
+      "loss": 1.4283,
+      "step": 7701
+    },
+    {
+      "epoch": 1.3287328560338136,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.100909472099251e-06,
+      "loss": 1.4533,
+      "step": 7702
+    },
+    {
+      "epoch": 1.3289053739325456,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.098536791558175e-06,
+      "loss": 1.2669,
+      "step": 7703
+    },
+    {
+      "epoch": 1.3290778918312776,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.096164474148122e-06,
+      "loss": 1.3087,
+      "step": 7704
+    },
+    {
+      "epoch": 1.3292504097300095,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.093792520044837e-06,
+      "loss": 1.3846,
+      "step": 7705
+    },
+    {
+      "epoch": 1.3294229276287415,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.091420929424065e-06,
+      "loss": 1.4604,
+      "step": 7706
+    },
+    {
+      "epoch": 1.3295954455274734,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.089049702461497e-06,
+      "loss": 1.4529,
+      "step": 7707
+    },
+    {
+      "epoch": 1.3297679634262054,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.086678839332813e-06,
+      "loss": 1.4119,
+      "step": 7708
+    },
+    {
+      "epoch": 1.3299404813249374,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.084308340213661e-06,
+      "loss": 1.4156,
+      "step": 7709
+    },
+    {
+      "epoch": 1.3301129992236693,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.08193820527966e-06,
+      "loss": 1.5028,
+      "step": 7710
+    },
+    {
+      "epoch": 1.3302855171224015,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.079568434706408e-06,
+      "loss": 1.4327,
+      "step": 7711
+    },
+    {
+      "epoch": 1.3304580350211335,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.0771990286694725e-06,
+      "loss": 1.5038,
+      "step": 7712
+    },
+    {
+      "epoch": 1.3306305529198654,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.0748299873443855e-06,
+      "loss": 1.4443,
+      "step": 7713
+    },
+    {
+      "epoch": 1.3308030708185974,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.072461310906675e-06,
+      "loss": 1.4156,
+      "step": 7714
+    },
+    {
+      "epoch": 1.3309755887173293,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.07009299953181e-06,
+      "loss": 1.4279,
+      "step": 7715
+    },
+    {
+      "epoch": 1.3311481066160615,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.0677250533952695e-06,
+      "loss": 1.5014,
+      "step": 7716
+    },
+    {
+      "epoch": 1.3313206245147935,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.065357472672469e-06,
+      "loss": 1.4707,
+      "step": 7717
+    },
+    {
+      "epoch": 1.3314931424135255,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.0629902575388165e-06,
+      "loss": 1.4062,
+      "step": 7718
+    },
+    {
+      "epoch": 1.3316656603122574,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.060623408169703e-06,
+      "loss": 1.4613,
+      "step": 7719
+    },
+    {
+      "epoch": 1.3318381782109894,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.058256924740463e-06,
+      "loss": 1.4527,
+      "step": 7720
+    },
+    {
+      "epoch": 1.3320106961097213,
+      "grad_norm": 0.70703125,
+      "learning_rate": 5.0558908074264315e-06,
+      "loss": 1.5059,
+      "step": 7721
+    },
+    {
+      "epoch": 1.3321832140084533,
+      "grad_norm": 0.74609375,
+      "learning_rate": 5.053525056402898e-06,
+      "loss": 1.4908,
+      "step": 7722
+    },
+    {
+      "epoch": 1.3323557319071853,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.051159671845136e-06,
+      "loss": 1.529,
+      "step": 7723
+    },
+    {
+      "epoch": 1.3325282498059174,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.048794653928389e-06,
+      "loss": 1.4887,
+      "step": 7724
+    },
+    {
+      "epoch": 1.3327007677046494,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.046430002827874e-06,
+      "loss": 1.3807,
+      "step": 7725
+    },
+    {
+      "epoch": 1.3328732856033814,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.044065718718766e-06,
+      "loss": 1.3835,
+      "step": 7726
+    },
+    {
+      "epoch": 1.3330458035021133,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.041701801776244e-06,
+      "loss": 1.3587,
+      "step": 7727
+    },
+    {
+      "epoch": 1.3332183214008453,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.039338252175431e-06,
+      "loss": 1.4684,
+      "step": 7728
+    },
+    {
+      "epoch": 1.3333908392995775,
+      "grad_norm": 0.71484375,
+      "learning_rate": 5.0369750700914345e-06,
+      "loss": 1.3529,
+      "step": 7729
+    },
+    {
+      "epoch": 1.3335633571983094,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.034612255699336e-06,
+      "loss": 1.448,
+      "step": 7730
+    },
+    {
+      "epoch": 1.3337358750970414,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.032249809174187e-06,
+      "loss": 1.361,
+      "step": 7731
+    },
+    {
+      "epoch": 1.3339083929957734,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.029887730691011e-06,
+      "loss": 1.3377,
+      "step": 7732
+    },
+    {
+      "epoch": 1.3340809108945053,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.027526020424811e-06,
+      "loss": 1.4446,
+      "step": 7733
+    },
+    {
+      "epoch": 1.3342534287932373,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.025164678550545e-06,
+      "loss": 1.3311,
+      "step": 7734
+    },
+    {
+      "epoch": 1.3344259466919692,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.0228037052431685e-06,
+      "loss": 1.4285,
+      "step": 7735
+    },
+    {
+      "epoch": 1.3345984645907012,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.020443100677595e-06,
+      "loss": 1.3957,
+      "step": 7736
+    },
+    {
+      "epoch": 1.3347709824894332,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.018082865028704e-06,
+      "loss": 1.4813,
+      "step": 7737
+    },
+    {
+      "epoch": 1.3349435003881653,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.01572299847137e-06,
+      "loss": 1.3646,
+      "step": 7738
+    },
+    {
+      "epoch": 1.3351160182868973,
+      "grad_norm": 0.625,
+      "learning_rate": 5.013363501180415e-06,
+      "loss": 1.5413,
+      "step": 7739
+    },
+    {
+      "epoch": 1.3352885361856293,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.01100437333065e-06,
+      "loss": 1.4785,
+      "step": 7740
+    },
+    {
+      "epoch": 1.3354610540843612,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.008645615096855e-06,
+      "loss": 1.4102,
+      "step": 7741
+    },
+    {
+      "epoch": 1.3356335719830932,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.006287226653779e-06,
+      "loss": 1.4611,
+      "step": 7742
+    },
+    {
+      "epoch": 1.3358060898818254,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.003929208176148e-06,
+      "loss": 1.4491,
+      "step": 7743
+    },
+    {
+      "epoch": 1.3359786077805573,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.00157155983866e-06,
+      "loss": 1.5095,
+      "step": 7744
+    },
+    {
+      "epoch": 1.3361511256792893,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.999214281815977e-06,
+      "loss": 1.496,
+      "step": 7745
+    },
+    {
+      "epoch": 1.3363236435780212,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.996857374282754e-06,
+      "loss": 1.4236,
+      "step": 7746
+    },
+    {
+      "epoch": 1.3364961614767532,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.99450083741359e-06,
+      "loss": 1.2766,
+      "step": 7747
+    },
+    {
+      "epoch": 1.3366686793754852,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.992144671383087e-06,
+      "loss": 1.3265,
+      "step": 7748
+    },
+    {
+      "epoch": 1.3368411972742171,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.989788876365793e-06,
+      "loss": 1.4192,
+      "step": 7749
+    },
+    {
+      "epoch": 1.337013715172949,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.987433452536244e-06,
+      "loss": 1.476,
+      "step": 7750
+    },
+    {
+      "epoch": 1.337186233071681,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.985078400068947e-06,
+      "loss": 1.4746,
+      "step": 7751
+    },
+    {
+      "epoch": 1.3373587509704132,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.982723719138375e-06,
+      "loss": 1.4289,
+      "step": 7752
+    },
+    {
+      "epoch": 1.3375312688691452,
+      "grad_norm": 1.375,
+      "learning_rate": 4.980369409918979e-06,
+      "loss": 1.3759,
+      "step": 7753
+    },
+    {
+      "epoch": 1.3377037867678772,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.978015472585183e-06,
+      "loss": 1.4064,
+      "step": 7754
+    },
+    {
+      "epoch": 1.3378763046666091,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.975661907311377e-06,
+      "loss": 1.4212,
+      "step": 7755
+    },
+    {
+      "epoch": 1.338048822565341,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.973308714271933e-06,
+      "loss": 1.4666,
+      "step": 7756
+    },
+    {
+      "epoch": 1.3382213404640733,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.97095589364119e-06,
+      "loss": 1.4525,
+      "step": 7757
+    },
+    {
+      "epoch": 1.3383938583628052,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.96860344559345e-06,
+      "loss": 1.4663,
+      "step": 7758
+    },
+    {
+      "epoch": 1.3385663762615372,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.966251370303011e-06,
+      "loss": 1.3687,
+      "step": 7759
+    },
+    {
+      "epoch": 1.3387388941602691,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.963899667944121e-06,
+      "loss": 1.415,
+      "step": 7760
+    },
+    {
+      "epoch": 1.338911412059001,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.961548338691009e-06,
+      "loss": 1.45,
+      "step": 7761
+    },
+    {
+      "epoch": 1.339083929957733,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.959197382717878e-06,
+      "loss": 1.4356,
+      "step": 7762
+    },
+    {
+      "epoch": 1.339256447856465,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.956846800198902e-06,
+      "loss": 1.4391,
+      "step": 7763
+    },
+    {
+      "epoch": 1.339428965755197,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.954496591308227e-06,
+      "loss": 1.4465,
+      "step": 7764
+    },
+    {
+      "epoch": 1.3396014836539292,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.952146756219972e-06,
+      "loss": 1.3614,
+      "step": 7765
+    },
+    {
+      "epoch": 1.3397740015526611,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.949797295108218e-06,
+      "loss": 1.3784,
+      "step": 7766
+    },
+    {
+      "epoch": 1.339946519451393,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.947448208147041e-06,
+      "loss": 1.424,
+      "step": 7767
+    },
+    {
+      "epoch": 1.340119037350125,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.9450994955104736e-06,
+      "loss": 1.4743,
+      "step": 7768
+    },
+    {
+      "epoch": 1.340291555248857,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.9427511573725125e-06,
+      "loss": 1.3808,
+      "step": 7769
+    },
+    {
+      "epoch": 1.3404640731475892,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.940403193907153e-06,
+      "loss": 1.3319,
+      "step": 7770
+    },
+    {
+      "epoch": 1.3406365910463212,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.938055605288334e-06,
+      "loss": 1.5072,
+      "step": 7771
+    },
+    {
+      "epoch": 1.3408091089450531,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.935708391689985e-06,
+      "loss": 1.4005,
+      "step": 7772
+    },
+    {
+      "epoch": 1.340981626843785,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.9333615532860005e-06,
+      "loss": 1.4497,
+      "step": 7773
+    },
+    {
+      "epoch": 1.341154144742517,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.931015090250251e-06,
+      "loss": 1.4624,
+      "step": 7774
+    },
+    {
+      "epoch": 1.341326662641249,
+      "grad_norm": 1.140625,
+      "learning_rate": 4.928669002756576e-06,
+      "loss": 1.4006,
+      "step": 7775
+    },
+    {
+      "epoch": 1.341499180539981,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.926323290978787e-06,
+      "loss": 1.5053,
+      "step": 7776
+    },
+    {
+      "epoch": 1.341671698438713,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.923977955090672e-06,
+      "loss": 1.515,
+      "step": 7777
+    },
+    {
+      "epoch": 1.3418442163374449,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.9216329952659895e-06,
+      "loss": 1.385,
+      "step": 7778
+    },
+    {
+      "epoch": 1.342016734236177,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.919288411678459e-06,
+      "loss": 1.4394,
+      "step": 7779
+    },
+    {
+      "epoch": 1.342189252134909,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.916944204501796e-06,
+      "loss": 1.4153,
+      "step": 7780
+    },
+    {
+      "epoch": 1.342361770033641,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.914600373909662e-06,
+      "loss": 1.356,
+      "step": 7781
+    },
+    {
+      "epoch": 1.342534287932373,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.912256920075708e-06,
+      "loss": 1.4713,
+      "step": 7782
+    },
+    {
+      "epoch": 1.342706805831105,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.909913843173552e-06,
+      "loss": 1.3787,
+      "step": 7783
+    },
+    {
+      "epoch": 1.342879323729837,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.907571143376782e-06,
+      "loss": 1.4151,
+      "step": 7784
+    },
+    {
+      "epoch": 1.343051841628569,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.905228820858959e-06,
+      "loss": 1.5189,
+      "step": 7785
+    },
+    {
+      "epoch": 1.343224359527301,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.902886875793621e-06,
+      "loss": 1.3454,
+      "step": 7786
+    },
+    {
+      "epoch": 1.343396877426033,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.900545308354271e-06,
+      "loss": 1.4785,
+      "step": 7787
+    },
+    {
+      "epoch": 1.343569395324765,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.898204118714387e-06,
+      "loss": 1.4196,
+      "step": 7788
+    },
+    {
+      "epoch": 1.343741913223497,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.895863307047423e-06,
+      "loss": 1.4437,
+      "step": 7789
+    },
+    {
+      "epoch": 1.3439144311222289,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.89352287352679e-06,
+      "loss": 1.4041,
+      "step": 7790
+    },
+    {
+      "epoch": 1.3440869490209608,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.891182818325897e-06,
+      "loss": 1.4244,
+      "step": 7791
+    },
+    {
+      "epoch": 1.3442594669196928,
+      "grad_norm": 1.3359375,
+      "learning_rate": 4.888843141618098e-06,
+      "loss": 1.4494,
+      "step": 7792
+    },
+    {
+      "epoch": 1.344431984818425,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.886503843576736e-06,
+      "loss": 1.4727,
+      "step": 7793
+    },
+    {
+      "epoch": 1.344604502717157,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.884164924375119e-06,
+      "loss": 1.3916,
+      "step": 7794
+    },
+    {
+      "epoch": 1.3447770206158889,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.881826384186529e-06,
+      "loss": 1.3834,
+      "step": 7795
+    },
+    {
+      "epoch": 1.3449495385146208,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.879488223184221e-06,
+      "loss": 1.3876,
+      "step": 7796
+    },
+    {
+      "epoch": 1.3451220564133528,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.877150441541424e-06,
+      "loss": 1.3679,
+      "step": 7797
+    },
+    {
+      "epoch": 1.345294574312085,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.8748130394313234e-06,
+      "loss": 1.4242,
+      "step": 7798
+    },
+    {
+      "epoch": 1.345467092210817,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.8724760170271e-06,
+      "loss": 1.4362,
+      "step": 7799
+    },
+    {
+      "epoch": 1.345639610109549,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.870139374501895e-06,
+      "loss": 1.3885,
+      "step": 7800
+    },
+    {
+      "epoch": 1.345639610109549,
+      "eval_loss": 1.407551646232605,
+      "eval_runtime": 10.8526,
+      "eval_samples_per_second": 94.355,
+      "eval_steps_per_second": 23.589,
+      "step": 7800
+    },
+    {
+      "epoch": 1.3458121280082809,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.8678031120288115e-06,
+      "loss": 1.452,
+      "step": 7801
+    },
+    {
+      "epoch": 1.3459846459070128,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.865467229780948e-06,
+      "loss": 1.4555,
+      "step": 7802
+    },
+    {
+      "epoch": 1.3461571638057448,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.863131727931347e-06,
+      "loss": 1.6427,
+      "step": 7803
+    },
+    {
+      "epoch": 1.3463296817044768,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.860796606653051e-06,
+      "loss": 1.3934,
+      "step": 7804
+    },
+    {
+      "epoch": 1.3465021996032087,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.858461866119051e-06,
+      "loss": 1.5244,
+      "step": 7805
+    },
+    {
+      "epoch": 1.346674717501941,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.856127506502321e-06,
+      "loss": 1.4147,
+      "step": 7806
+    },
+    {
+      "epoch": 1.3468472354006729,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.853793527975806e-06,
+      "loss": 1.5024,
+      "step": 7807
+    },
+    {
+      "epoch": 1.3470197532994048,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.85145993071242e-06,
+      "loss": 1.4646,
+      "step": 7808
+    },
+    {
+      "epoch": 1.3471922711981368,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.849126714885053e-06,
+      "loss": 1.4857,
+      "step": 7809
+    },
+    {
+      "epoch": 1.3473647890968687,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.846793880666567e-06,
+      "loss": 1.3657,
+      "step": 7810
+    },
+    {
+      "epoch": 1.347537306995601,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.844461428229782e-06,
+      "loss": 1.3712,
+      "step": 7811
+    },
+    {
+      "epoch": 1.3477098248943329,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.8421293577475145e-06,
+      "loss": 1.385,
+      "step": 7812
+    },
+    {
+      "epoch": 1.3478823427930648,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.839797669392528e-06,
+      "loss": 1.4094,
+      "step": 7813
+    },
+    {
+      "epoch": 1.3480548606917968,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.837466363337573e-06,
+      "loss": 1.4854,
+      "step": 7814
+    },
+    {
+      "epoch": 1.3482273785905288,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.835135439755367e-06,
+      "loss": 1.3394,
+      "step": 7815
+    },
+    {
+      "epoch": 1.3483998964892607,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.832804898818599e-06,
+      "loss": 1.4,
+      "step": 7816
+    },
+    {
+      "epoch": 1.3485724143879927,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.8304747406999304e-06,
+      "loss": 1.2918,
+      "step": 7817
+    },
+    {
+      "epoch": 1.3487449322867247,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.828144965571994e-06,
+      "loss": 1.2981,
+      "step": 7818
+    },
+    {
+      "epoch": 1.3489174501854566,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.825815573607393e-06,
+      "loss": 1.4247,
+      "step": 7819
+    },
+    {
+      "epoch": 1.3490899680841888,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.823486564978705e-06,
+      "loss": 1.5022,
+      "step": 7820
+    },
+    {
+      "epoch": 1.3492624859829208,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.821157939858479e-06,
+      "loss": 1.3706,
+      "step": 7821
+    },
+    {
+      "epoch": 1.3494350038816527,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.818829698419225e-06,
+      "loss": 1.4452,
+      "step": 7822
+    },
+    {
+      "epoch": 1.3496075217803847,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.816501840833448e-06,
+      "loss": 1.4214,
+      "step": 7823
+    },
+    {
+      "epoch": 1.3497800396791166,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.814174367273599e-06,
+      "loss": 1.4343,
+      "step": 7824
+    },
+    {
+      "epoch": 1.3499525575778488,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.811847277912115e-06,
+      "loss": 1.4916,
+      "step": 7825
+    },
+    {
+      "epoch": 1.3501250754765808,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.8095205729214015e-06,
+      "loss": 1.5595,
+      "step": 7826
+    },
+    {
+      "epoch": 1.3502975933753127,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.8071942524738355e-06,
+      "loss": 1.4548,
+      "step": 7827
+    },
+    {
+      "epoch": 1.3504701112740447,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.8048683167417664e-06,
+      "loss": 1.3287,
+      "step": 7828
+    },
+    {
+      "epoch": 1.3506426291727767,
+      "grad_norm": 0.625,
+      "learning_rate": 4.802542765897516e-06,
+      "loss": 1.4488,
+      "step": 7829
+    },
+    {
+      "epoch": 1.3508151470715086,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.800217600113366e-06,
+      "loss": 1.4453,
+      "step": 7830
+    },
+    {
+      "epoch": 1.3509876649702406,
+      "grad_norm": 0.625,
+      "learning_rate": 4.797892819561589e-06,
+      "loss": 1.4369,
+      "step": 7831
+    },
+    {
+      "epoch": 1.3511601828689725,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.795568424414421e-06,
+      "loss": 1.3552,
+      "step": 7832
+    },
+    {
+      "epoch": 1.3513327007677047,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.793244414844054e-06,
+      "loss": 1.3656,
+      "step": 7833
+    },
+    {
+      "epoch": 1.3515052186664367,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.790920791022682e-06,
+      "loss": 1.4695,
+      "step": 7834
+    },
+    {
+      "epoch": 1.3516777365651687,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.788597553122438e-06,
+      "loss": 1.405,
+      "step": 7835
+    },
+    {
+      "epoch": 1.3518502544639006,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.786274701315458e-06,
+      "loss": 1.3898,
+      "step": 7836
+    },
+    {
+      "epoch": 1.3520227723626326,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.7839522357738196e-06,
+      "loss": 1.5197,
+      "step": 7837
+    },
+    {
+      "epoch": 1.3521952902613645,
+      "grad_norm": 1.46875,
+      "learning_rate": 4.781630156669592e-06,
+      "loss": 1.4992,
+      "step": 7838
+    },
+    {
+      "epoch": 1.3523678081600967,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.7793084641748085e-06,
+      "loss": 1.5419,
+      "step": 7839
+    },
+    {
+      "epoch": 1.3525403260588287,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.776987158461475e-06,
+      "loss": 1.365,
+      "step": 7840
+    },
+    {
+      "epoch": 1.3527128439575606,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.774666239701566e-06,
+      "loss": 1.4914,
+      "step": 7841
+    },
+    {
+      "epoch": 1.3528853618562926,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.772345708067035e-06,
+      "loss": 1.4306,
+      "step": 7842
+    },
+    {
+      "epoch": 1.3530578797550246,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.770025563729792e-06,
+      "loss": 1.4419,
+      "step": 7843
+    },
+    {
+      "epoch": 1.3532303976537565,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.767705806861741e-06,
+      "loss": 1.4563,
+      "step": 7844
+    },
+    {
+      "epoch": 1.3534029155524885,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.765386437634732e-06,
+      "loss": 1.3103,
+      "step": 7845
+    },
+    {
+      "epoch": 1.3535754334512204,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.763067456220604e-06,
+      "loss": 1.4048,
+      "step": 7846
+    },
+    {
+      "epoch": 1.3537479513499526,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.760748862791159e-06,
+      "loss": 1.3718,
+      "step": 7847
+    },
+    {
+      "epoch": 1.3539204692486846,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.758430657518176e-06,
+      "loss": 1.4345,
+      "step": 7848
+    },
+    {
+      "epoch": 1.3540929871474165,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.7561128405734e-06,
+      "loss": 1.3932,
+      "step": 7849
+    },
+    {
+      "epoch": 1.3542655050461485,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.753795412128552e-06,
+      "loss": 1.4566,
+      "step": 7850
+    },
+    {
+      "epoch": 1.3544380229448805,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.751478372355317e-06,
+      "loss": 1.3935,
+      "step": 7851
+    },
+    {
+      "epoch": 1.3546105408436127,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.749161721425359e-06,
+      "loss": 1.4629,
+      "step": 7852
+    },
+    {
+      "epoch": 1.3547830587423446,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.746845459510314e-06,
+      "loss": 1.4346,
+      "step": 7853
+    },
+    {
+      "epoch": 1.3549555766410766,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.744529586781773e-06,
+      "loss": 1.3455,
+      "step": 7854
+    },
+    {
+      "epoch": 1.3551280945398085,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.742214103411325e-06,
+      "loss": 1.5296,
+      "step": 7855
+    },
+    {
+      "epoch": 1.3553006124385405,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.739899009570506e-06,
+      "loss": 1.4611,
+      "step": 7856
+    },
+    {
+      "epoch": 1.3554731303372725,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.7375843054308335e-06,
+      "loss": 1.4042,
+      "step": 7857
+    },
+    {
+      "epoch": 1.3556456482360044,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.735269991163798e-06,
+      "loss": 1.4087,
+      "step": 7858
+    },
+    {
+      "epoch": 1.3558181661347364,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.732956066940856e-06,
+      "loss": 1.3889,
+      "step": 7859
+    },
+    {
+      "epoch": 1.3559906840334683,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.7306425329334386e-06,
+      "loss": 1.4786,
+      "step": 7860
+    },
+    {
+      "epoch": 1.3561632019322005,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.7283293893129515e-06,
+      "loss": 1.4151,
+      "step": 7861
+    },
+    {
+      "epoch": 1.3563357198309325,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.726016636250753e-06,
+      "loss": 1.4943,
+      "step": 7862
+    },
+    {
+      "epoch": 1.3565082377296644,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.7237042739182006e-06,
+      "loss": 1.4195,
+      "step": 7863
+    },
+    {
+      "epoch": 1.3566807556283964,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.721392302486602e-06,
+      "loss": 1.3886,
+      "step": 7864
+    },
+    {
+      "epoch": 1.3568532735271284,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.719080722127246e-06,
+      "loss": 1.5363,
+      "step": 7865
+    },
+    {
+      "epoch": 1.3570257914258606,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.716769533011389e-06,
+      "loss": 1.4005,
+      "step": 7866
+    },
+    {
+      "epoch": 1.3571983093245925,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.714458735310249e-06,
+      "loss": 1.4047,
+      "step": 7867
+    },
+    {
+      "epoch": 1.3573708272233245,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.71214832919504e-06,
+      "loss": 1.3229,
+      "step": 7868
+    },
+    {
+      "epoch": 1.3575433451220564,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.709838314836918e-06,
+      "loss": 1.4652,
+      "step": 7869
+    },
+    {
+      "epoch": 1.3577158630207884,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.707528692407027e-06,
+      "loss": 1.4012,
+      "step": 7870
+    },
+    {
+      "epoch": 1.3578883809195204,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.705219462076481e-06,
+      "loss": 1.4204,
+      "step": 7871
+    },
+    {
+      "epoch": 1.3580608988182523,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.70291062401636e-06,
+      "loss": 1.4563,
+      "step": 7872
+    },
+    {
+      "epoch": 1.3582334167169843,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.700602178397719e-06,
+      "loss": 1.468,
+      "step": 7873
+    },
+    {
+      "epoch": 1.3584059346157165,
+      "grad_norm": 1.390625,
+      "learning_rate": 4.698294125391583e-06,
+      "loss": 1.3792,
+      "step": 7874
+    },
+    {
+      "epoch": 1.3585784525144484,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.695986465168937e-06,
+      "loss": 1.4291,
+      "step": 7875
+    },
+    {
+      "epoch": 1.3587509704131804,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.693679197900763e-06,
+      "loss": 1.3912,
+      "step": 7876
+    },
+    {
+      "epoch": 1.3589234883119123,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.691372323757985e-06,
+      "loss": 1.4333,
+      "step": 7877
+    },
+    {
+      "epoch": 1.3590960062106443,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.689065842911517e-06,
+      "loss": 1.3318,
+      "step": 7878
+    },
+    {
+      "epoch": 1.3592685241093765,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.686759755532234e-06,
+      "loss": 1.4298,
+      "step": 7879
+    },
+    {
+      "epoch": 1.3594410420081084,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.684454061790987e-06,
+      "loss": 1.5446,
+      "step": 7880
+    },
+    {
+      "epoch": 1.3596135599068404,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.6821487618585956e-06,
+      "loss": 1.4827,
+      "step": 7881
+    },
+    {
+      "epoch": 1.3597860778055724,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.679843855905853e-06,
+      "loss": 1.3706,
+      "step": 7882
+    },
+    {
+      "epoch": 1.3599585957043043,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.6775393441035185e-06,
+      "loss": 1.4048,
+      "step": 7883
+    },
+    {
+      "epoch": 1.3601311136030363,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.6752352266223255e-06,
+      "loss": 1.3972,
+      "step": 7884
+    },
+    {
+      "epoch": 1.3603036315017683,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.672931503632981e-06,
+      "loss": 1.4176,
+      "step": 7885
+    },
+    {
+      "epoch": 1.3604761494005002,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.67062817530615e-06,
+      "loss": 1.4636,
+      "step": 7886
+    },
+    {
+      "epoch": 1.3606486672992322,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.6683252418124895e-06,
+      "loss": 1.3784,
+      "step": 7887
+    },
+    {
+      "epoch": 1.3608211851979644,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.666022703322605e-06,
+      "loss": 1.4577,
+      "step": 7888
+    },
+    {
+      "epoch": 1.3609937030966963,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.663720560007087e-06,
+      "loss": 1.4598,
+      "step": 7889
+    },
+    {
+      "epoch": 1.3611662209954283,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.661418812036492e-06,
+      "loss": 1.3935,
+      "step": 7890
+    },
+    {
+      "epoch": 1.3613387388941602,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.659117459581351e-06,
+      "loss": 1.3132,
+      "step": 7891
+    },
+    {
+      "epoch": 1.3615112567928922,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.656816502812157e-06,
+      "loss": 1.498,
+      "step": 7892
+    },
+    {
+      "epoch": 1.3616837746916244,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.6545159418993866e-06,
+      "loss": 1.4276,
+      "step": 7893
+    },
+    {
+      "epoch": 1.3618562925903563,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.652215777013469e-06,
+      "loss": 1.3338,
+      "step": 7894
+    },
+    {
+      "epoch": 1.3620288104890883,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.649916008324824e-06,
+      "loss": 1.4057,
+      "step": 7895
+    },
+    {
+      "epoch": 1.3622013283878203,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.64761663600383e-06,
+      "loss": 1.4477,
+      "step": 7896
+    },
+    {
+      "epoch": 1.3623738462865522,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.645317660220838e-06,
+      "loss": 1.4506,
+      "step": 7897
+    },
+    {
+      "epoch": 1.3625463641852842,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.643019081146177e-06,
+      "loss": 1.4392,
+      "step": 7898
+    },
+    {
+      "epoch": 1.3627188820840161,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.640720898950126e-06,
+      "loss": 1.2814,
+      "step": 7899
+    },
+    {
+      "epoch": 1.362891399982748,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.638423113802964e-06,
+      "loss": 1.4289,
+      "step": 7900
+    },
+    {
+      "epoch": 1.362891399982748,
+      "eval_loss": 1.4074419736862183,
+      "eval_runtime": 10.7675,
+      "eval_samples_per_second": 95.101,
+      "eval_steps_per_second": 23.775,
+      "step": 7900
+    },
+    {
+      "epoch": 1.36306391788148,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.636125725874916e-06,
+      "loss": 1.4032,
+      "step": 7901
+    },
+    {
+      "epoch": 1.3632364357802123,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.6338287353361875e-06,
+      "loss": 1.4314,
+      "step": 7902
+    },
+    {
+      "epoch": 1.3634089536789442,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.631532142356957e-06,
+      "loss": 1.409,
+      "step": 7903
+    },
+    {
+      "epoch": 1.3635814715776762,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.629235947107369e-06,
+      "loss": 1.4498,
+      "step": 7904
+    },
+    {
+      "epoch": 1.3637539894764081,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.626940149757541e-06,
+      "loss": 1.3947,
+      "step": 7905
+    },
+    {
+      "epoch": 1.36392650737514,
+      "grad_norm": 0.625,
+      "learning_rate": 4.62464475047756e-06,
+      "loss": 1.4714,
+      "step": 7906
+    },
+    {
+      "epoch": 1.3640990252738723,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.622349749437478e-06,
+      "loss": 1.3314,
+      "step": 7907
+    },
+    {
+      "epoch": 1.3642715431726042,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.620055146807334e-06,
+      "loss": 1.3417,
+      "step": 7908
+    },
+    {
+      "epoch": 1.3644440610713362,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.617760942757117e-06,
+      "loss": 1.5331,
+      "step": 7909
+    },
+    {
+      "epoch": 1.3646165789700682,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.615467137456798e-06,
+      "loss": 1.3129,
+      "step": 7910
+    },
+    {
+      "epoch": 1.3647890968688001,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.613173731076319e-06,
+      "loss": 1.5621,
+      "step": 7911
+    },
+    {
+      "epoch": 1.364961614767532,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.610880723785588e-06,
+      "loss": 1.4238,
+      "step": 7912
+    },
+    {
+      "epoch": 1.365134132666264,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.608588115754486e-06,
+      "loss": 1.4943,
+      "step": 7913
+    },
+    {
+      "epoch": 1.365306650564996,
+      "grad_norm": 0.625,
+      "learning_rate": 4.606295907152862e-06,
+      "loss": 1.4239,
+      "step": 7914
+    },
+    {
+      "epoch": 1.3654791684637282,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.6040040981505395e-06,
+      "loss": 1.4424,
+      "step": 7915
+    },
+    {
+      "epoch": 1.3656516863624601,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.601712688917309e-06,
+      "loss": 1.3706,
+      "step": 7916
+    },
+    {
+      "epoch": 1.365824204261192,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.599421679622936e-06,
+      "loss": 1.4369,
+      "step": 7917
+    },
+    {
+      "epoch": 1.365996722159924,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.597131070437143e-06,
+      "loss": 1.3608,
+      "step": 7918
+    },
+    {
+      "epoch": 1.366169240058656,
+      "grad_norm": 1.0,
+      "learning_rate": 4.594840861529646e-06,
+      "loss": 1.4133,
+      "step": 7919
+    },
+    {
+      "epoch": 1.3663417579573882,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.5925510530701065e-06,
+      "loss": 1.4296,
+      "step": 7920
+    },
+    {
+      "epoch": 1.3665142758561202,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.590261645228173e-06,
+      "loss": 1.421,
+      "step": 7921
+    },
+    {
+      "epoch": 1.3666867937548521,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.587972638173459e-06,
+      "loss": 1.4531,
+      "step": 7922
+    },
+    {
+      "epoch": 1.366859311653584,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.5856840320755465e-06,
+      "loss": 1.3976,
+      "step": 7923
+    },
+    {
+      "epoch": 1.367031829552316,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.583395827103992e-06,
+      "loss": 1.4801,
+      "step": 7924
+    },
+    {
+      "epoch": 1.367204347451048,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.581108023428319e-06,
+      "loss": 1.4647,
+      "step": 7925
+    },
+    {
+      "epoch": 1.36737686534978,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.578820621218023e-06,
+      "loss": 1.4242,
+      "step": 7926
+    },
+    {
+      "epoch": 1.367549383248512,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.576533620642568e-06,
+      "loss": 1.4149,
+      "step": 7927
+    },
+    {
+      "epoch": 1.367721901147244,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.57424702187139e-06,
+      "loss": 1.3514,
+      "step": 7928
+    },
+    {
+      "epoch": 1.367894419045976,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.5719608250738936e-06,
+      "loss": 1.4513,
+      "step": 7929
+    },
+    {
+      "epoch": 1.368066936944708,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.569675030419459e-06,
+      "loss": 1.4672,
+      "step": 7930
+    },
+    {
+      "epoch": 1.36823945484344,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.567389638077421e-06,
+      "loss": 1.3883,
+      "step": 7931
+    },
+    {
+      "epoch": 1.368411972742172,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.565104648217111e-06,
+      "loss": 1.4488,
+      "step": 7932
+    },
+    {
+      "epoch": 1.368584490640904,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.562820061007803e-06,
+      "loss": 1.3655,
+      "step": 7933
+    },
+    {
+      "epoch": 1.368757008539636,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.560535876618759e-06,
+      "loss": 1.3948,
+      "step": 7934
+    },
+    {
+      "epoch": 1.368929526438368,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.558252095219204e-06,
+      "loss": 1.5201,
+      "step": 7935
+    },
+    {
+      "epoch": 1.3691020443371,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.5559687169783354e-06,
+      "loss": 1.4503,
+      "step": 7936
+    },
+    {
+      "epoch": 1.369274562235832,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.55368574206532e-06,
+      "loss": 1.388,
+      "step": 7937
+    },
+    {
+      "epoch": 1.369447080134564,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.551403170649299e-06,
+      "loss": 1.4712,
+      "step": 7938
+    },
+    {
+      "epoch": 1.369619598033296,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.5491210028993685e-06,
+      "loss": 1.3087,
+      "step": 7939
+    },
+    {
+      "epoch": 1.3697921159320279,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.5468392389846195e-06,
+      "loss": 1.488,
+      "step": 7940
+    },
+    {
+      "epoch": 1.3699646338307598,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.544557879074088e-06,
+      "loss": 1.4399,
+      "step": 7941
+    },
+    {
+      "epoch": 1.3701371517294918,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.542276923336798e-06,
+      "loss": 1.4177,
+      "step": 7942
+    },
+    {
+      "epoch": 1.370309669628224,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.539996371941734e-06,
+      "loss": 1.3468,
+      "step": 7943
+    },
+    {
+      "epoch": 1.370482187526956,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.5377162250578545e-06,
+      "loss": 1.3186,
+      "step": 7944
+    },
+    {
+      "epoch": 1.370654705425688,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.535436482854087e-06,
+      "loss": 1.4187,
+      "step": 7945
+    },
+    {
+      "epoch": 1.3708272233244199,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.533157145499328e-06,
+      "loss": 1.3719,
+      "step": 7946
+    },
+    {
+      "epoch": 1.3709997412231518,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.530878213162447e-06,
+      "loss": 1.4092,
+      "step": 7947
+    },
+    {
+      "epoch": 1.371172259121884,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.528599686012281e-06,
+      "loss": 1.4266,
+      "step": 7948
+    },
+    {
+      "epoch": 1.371344777020616,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.526321564217641e-06,
+      "loss": 1.4805,
+      "step": 7949
+    },
+    {
+      "epoch": 1.371517294919348,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.5240438479472926e-06,
+      "loss": 1.4013,
+      "step": 7950
+    },
+    {
+      "epoch": 1.37168981281808,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.521766537369998e-06,
+      "loss": 1.4912,
+      "step": 7951
+    },
+    {
+      "epoch": 1.3718623307168119,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.519489632654461e-06,
+      "loss": 1.3681,
+      "step": 7952
+    },
+    {
+      "epoch": 1.3720348486155438,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.517213133969385e-06,
+      "loss": 1.516,
+      "step": 7953
+    },
+    {
+      "epoch": 1.3722073665142758,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.5149370414834125e-06,
+      "loss": 1.4078,
+      "step": 7954
+    },
+    {
+      "epoch": 1.3723798844130077,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.512661355365177e-06,
+      "loss": 1.3906,
+      "step": 7955
+    },
+    {
+      "epoch": 1.37255240231174,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.510386075783274e-06,
+      "loss": 1.4647,
+      "step": 7956
+    },
+    {
+      "epoch": 1.3727249202104719,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.508111202906271e-06,
+      "loss": 1.4007,
+      "step": 7957
+    },
+    {
+      "epoch": 1.3728974381092038,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.5058367369027054e-06,
+      "loss": 1.4573,
+      "step": 7958
+    },
+    {
+      "epoch": 1.3730699560079358,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.503562677941088e-06,
+      "loss": 1.4147,
+      "step": 7959
+    },
+    {
+      "epoch": 1.3732424739066678,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.501289026189882e-06,
+      "loss": 1.3457,
+      "step": 7960
+    },
+    {
+      "epoch": 1.3734149918054,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.499015781817547e-06,
+      "loss": 1.4553,
+      "step": 7961
+    },
+    {
+      "epoch": 1.373587509704132,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.496742944992499e-06,
+      "loss": 1.3613,
+      "step": 7962
+    },
+    {
+      "epoch": 1.3737600276028639,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.494470515883111e-06,
+      "loss": 1.4739,
+      "step": 7963
+    },
+    {
+      "epoch": 1.3739325455015958,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.492198494657755e-06,
+      "loss": 1.428,
+      "step": 7964
+    },
+    {
+      "epoch": 1.3741050634003278,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.4899268814847455e-06,
+      "loss": 1.4915,
+      "step": 7965
+    },
+    {
+      "epoch": 1.3742775812990597,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.4876556765323805e-06,
+      "loss": 1.4217,
+      "step": 7966
+    },
+    {
+      "epoch": 1.3744500991977917,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.485384879968926e-06,
+      "loss": 1.4559,
+      "step": 7967
+    },
+    {
+      "epoch": 1.3746226170965237,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.483114491962617e-06,
+      "loss": 1.4177,
+      "step": 7968
+    },
+    {
+      "epoch": 1.3747951349952556,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.480844512681657e-06,
+      "loss": 1.4573,
+      "step": 7969
+    },
+    {
+      "epoch": 1.3749676528939878,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.478574942294225e-06,
+      "loss": 1.4633,
+      "step": 7970
+    },
+    {
+      "epoch": 1.3751401707927198,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.476305780968452e-06,
+      "loss": 1.3839,
+      "step": 7971
+    },
+    {
+      "epoch": 1.3753126886914517,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.474037028872468e-06,
+      "loss": 1.4457,
+      "step": 7972
+    },
+    {
+      "epoch": 1.3754852065901837,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.471768686174347e-06,
+      "loss": 1.3865,
+      "step": 7973
+    },
+    {
+      "epoch": 1.3756577244889157,
+      "grad_norm": 1.2578125,
+      "learning_rate": 4.469500753042142e-06,
+      "loss": 1.3655,
+      "step": 7974
+    },
+    {
+      "epoch": 1.3758302423876478,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.467233229643878e-06,
+      "loss": 1.4127,
+      "step": 7975
+    },
+    {
+      "epoch": 1.3760027602863798,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.464966116147546e-06,
+      "loss": 1.4156,
+      "step": 7976
+    },
+    {
+      "epoch": 1.3761752781851118,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.46269941272111e-06,
+      "loss": 1.4055,
+      "step": 7977
+    },
+    {
+      "epoch": 1.3763477960838437,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.460433119532499e-06,
+      "loss": 1.4684,
+      "step": 7978
+    },
+    {
+      "epoch": 1.3765203139825757,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.458167236749616e-06,
+      "loss": 1.415,
+      "step": 7979
+    },
+    {
+      "epoch": 1.3766928318813076,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.45590176454033e-06,
+      "loss": 1.5098,
+      "step": 7980
+    },
+    {
+      "epoch": 1.3768653497800396,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.453636703072487e-06,
+      "loss": 1.5133,
+      "step": 7981
+    },
+    {
+      "epoch": 1.3770378676787716,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.451372052513884e-06,
+      "loss": 1.394,
+      "step": 7982
+    },
+    {
+      "epoch": 1.3772103855775037,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.4491078130323174e-06,
+      "loss": 1.394,
+      "step": 7983
+    },
+    {
+      "epoch": 1.3773829034762357,
+      "grad_norm": 0.625,
+      "learning_rate": 4.446843984795519e-06,
+      "loss": 1.4961,
+      "step": 7984
+    },
+    {
+      "epoch": 1.3775554213749677,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.444580567971224e-06,
+      "loss": 1.5189,
+      "step": 7985
+    },
+    {
+      "epoch": 1.3777279392736996,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.442317562727109e-06,
+      "loss": 1.4859,
+      "step": 7986
+    },
+    {
+      "epoch": 1.3779004571724316,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.440054969230834e-06,
+      "loss": 1.4891,
+      "step": 7987
+    },
+    {
+      "epoch": 1.3780729750711636,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.437792787650028e-06,
+      "loss": 1.3835,
+      "step": 7988
+    },
+    {
+      "epoch": 1.3782454929698957,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.435531018152286e-06,
+      "loss": 1.3922,
+      "step": 7989
+    },
+    {
+      "epoch": 1.3784180108686277,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.433269660905175e-06,
+      "loss": 1.4601,
+      "step": 7990
+    },
+    {
+      "epoch": 1.3785905287673597,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.431008716076232e-06,
+      "loss": 1.4917,
+      "step": 7991
+    },
+    {
+      "epoch": 1.3787630466660916,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.428748183832955e-06,
+      "loss": 1.5076,
+      "step": 7992
+    },
+    {
+      "epoch": 1.3789355645648236,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.426488064342826e-06,
+      "loss": 1.498,
+      "step": 7993
+    },
+    {
+      "epoch": 1.3791080824635555,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.4242283577732905e-06,
+      "loss": 1.3959,
+      "step": 7994
+    },
+    {
+      "epoch": 1.3792806003622875,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.421969064291749e-06,
+      "loss": 1.4679,
+      "step": 7995
+    },
+    {
+      "epoch": 1.3794531182610195,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.4197101840656e-06,
+      "loss": 1.4727,
+      "step": 7996
+    },
+    {
+      "epoch": 1.3796256361597516,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.417451717262184e-06,
+      "loss": 1.467,
+      "step": 7997
+    },
+    {
+      "epoch": 1.3797981540584836,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.415193664048827e-06,
+      "loss": 1.4182,
+      "step": 7998
+    },
+    {
+      "epoch": 1.3799706719572156,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.412936024592818e-06,
+      "loss": 1.5055,
+      "step": 7999
+    },
+    {
+      "epoch": 1.3801431898559475,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.410678799061417e-06,
+      "loss": 1.4059,
+      "step": 8000
+    },
+    {
+      "epoch": 1.3801431898559475,
+      "eval_loss": 1.407374382019043,
+      "eval_runtime": 11.0415,
+      "eval_samples_per_second": 92.741,
+      "eval_steps_per_second": 23.185,
+      "step": 8000
+    },
+    {
+      "epoch": 1.3803157077546795,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.408421987621856e-06,
+      "loss": 1.4668,
+      "step": 8001
+    },
+    {
+      "epoch": 1.3804882256534117,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.406165590441335e-06,
+      "loss": 1.4717,
+      "step": 8002
+    },
+    {
+      "epoch": 1.3806607435521436,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.403909607687012e-06,
+      "loss": 1.4689,
+      "step": 8003
+    },
+    {
+      "epoch": 1.3808332614508756,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.4016540395260375e-06,
+      "loss": 1.4595,
+      "step": 8004
+    },
+    {
+      "epoch": 1.3810057793496076,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.399398886125508e-06,
+      "loss": 1.4481,
+      "step": 8005
+    },
+    {
+      "epoch": 1.3811782972483395,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.397144147652506e-06,
+      "loss": 1.4332,
+      "step": 8006
+    },
+    {
+      "epoch": 1.3813508151470715,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.3948898242740715e-06,
+      "loss": 1.3817,
+      "step": 8007
+    },
+    {
+      "epoch": 1.3815233330458034,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.392635916157221e-06,
+      "loss": 1.3656,
+      "step": 8008
+    },
+    {
+      "epoch": 1.3816958509445354,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.390382423468938e-06,
+      "loss": 1.2959,
+      "step": 8009
+    },
+    {
+      "epoch": 1.3818683688432674,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.388129346376177e-06,
+      "loss": 1.4901,
+      "step": 8010
+    },
+    {
+      "epoch": 1.3820408867419995,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.385876685045858e-06,
+      "loss": 1.3806,
+      "step": 8011
+    },
+    {
+      "epoch": 1.3822134046407315,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.3836244396448725e-06,
+      "loss": 1.3983,
+      "step": 8012
+    },
+    {
+      "epoch": 1.3823859225394635,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.3813726103400825e-06,
+      "loss": 1.4383,
+      "step": 8013
+    },
+    {
+      "epoch": 1.3825584404381954,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.379121197298315e-06,
+      "loss": 1.3774,
+      "step": 8014
+    },
+    {
+      "epoch": 1.3827309583369274,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.3768702006863735e-06,
+      "loss": 1.4498,
+      "step": 8015
+    },
+    {
+      "epoch": 1.3829034762356596,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.374619620671015e-06,
+      "loss": 1.3819,
+      "step": 8016
+    },
+    {
+      "epoch": 1.3830759941343915,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.372369457418994e-06,
+      "loss": 1.5088,
+      "step": 8017
+    },
+    {
+      "epoch": 1.3832485120331235,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.370119711097e-06,
+      "loss": 1.3558,
+      "step": 8018
+    },
+    {
+      "epoch": 1.3834210299318555,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.3678703818717165e-06,
+      "loss": 1.3744,
+      "step": 8019
+    },
+    {
+      "epoch": 1.3835935478305874,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.365621469909785e-06,
+      "loss": 1.4178,
+      "step": 8020
+    },
+    {
+      "epoch": 1.3837660657293194,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.363372975377821e-06,
+      "loss": 1.4425,
+      "step": 8021
+    },
+    {
+      "epoch": 1.3839385836280513,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.361124898442406e-06,
+      "loss": 1.4155,
+      "step": 8022
+    },
+    {
+      "epoch": 1.3841111015267833,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.358877239270096e-06,
+      "loss": 1.5466,
+      "step": 8023
+    },
+    {
+      "epoch": 1.3842836194255155,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.3566299980273995e-06,
+      "loss": 1.4899,
+      "step": 8024
+    },
+    {
+      "epoch": 1.3844561373242474,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.354383174880819e-06,
+      "loss": 1.3355,
+      "step": 8025
+    },
+    {
+      "epoch": 1.3846286552229794,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.3521367699968105e-06,
+      "loss": 1.3952,
+      "step": 8026
+    },
+    {
+      "epoch": 1.3848011731217114,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.349890783541793e-06,
+      "loss": 1.3639,
+      "step": 8027
+    },
+    {
+      "epoch": 1.3849736910204433,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.3476452156821765e-06,
+      "loss": 1.3987,
+      "step": 8028
+    },
+    {
+      "epoch": 1.3851462089191753,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.345400066584318e-06,
+      "loss": 1.4278,
+      "step": 8029
+    },
+    {
+      "epoch": 1.3853187268179075,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.343155336414553e-06,
+      "loss": 1.4388,
+      "step": 8030
+    },
+    {
+      "epoch": 1.3854912447166394,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.340911025339186e-06,
+      "loss": 1.4773,
+      "step": 8031
+    },
+    {
+      "epoch": 1.3856637626153714,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.33866713352449e-06,
+      "loss": 1.4371,
+      "step": 8032
+    },
+    {
+      "epoch": 1.3858362805141033,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.336423661136708e-06,
+      "loss": 1.4436,
+      "step": 8033
+    },
+    {
+      "epoch": 1.3860087984128353,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.3341806083420504e-06,
+      "loss": 1.4122,
+      "step": 8034
+    },
+    {
+      "epoch": 1.3861813163115673,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.331937975306687e-06,
+      "loss": 1.3701,
+      "step": 8035
+    },
+    {
+      "epoch": 1.3863538342102992,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.329695762196783e-06,
+      "loss": 1.4333,
+      "step": 8036
+    },
+    {
+      "epoch": 1.3865263521090312,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.3274539691784434e-06,
+      "loss": 1.448,
+      "step": 8037
+    },
+    {
+      "epoch": 1.3866988700077634,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.325212596417756e-06,
+      "loss": 1.4208,
+      "step": 8038
+    },
+    {
+      "epoch": 1.3868713879064953,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.322971644080778e-06,
+      "loss": 1.3671,
+      "step": 8039
+    },
+    {
+      "epoch": 1.3870439058052273,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.320731112333528e-06,
+      "loss": 1.4953,
+      "step": 8040
+    },
+    {
+      "epoch": 1.3872164237039593,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.318491001342011e-06,
+      "loss": 1.4839,
+      "step": 8041
+    },
+    {
+      "epoch": 1.3873889416026912,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.316251311272177e-06,
+      "loss": 1.5123,
+      "step": 8042
+    },
+    {
+      "epoch": 1.3875614595014234,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.314012042289959e-06,
+      "loss": 1.5034,
+      "step": 8043
+    },
+    {
+      "epoch": 1.3877339774001554,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.311773194561256e-06,
+      "loss": 1.367,
+      "step": 8044
+    },
+    {
+      "epoch": 1.3879064952988873,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.309534768251937e-06,
+      "loss": 1.5059,
+      "step": 8045
+    },
+    {
+      "epoch": 1.3880790131976193,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.307296763527838e-06,
+      "loss": 1.4364,
+      "step": 8046
+    },
+    {
+      "epoch": 1.3882515310963512,
+      "grad_norm": 0.625,
+      "learning_rate": 4.3050591805547696e-06,
+      "loss": 1.4842,
+      "step": 8047
+    },
+    {
+      "epoch": 1.3884240489950832,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.302822019498492e-06,
+      "loss": 1.4707,
+      "step": 8048
+    },
+    {
+      "epoch": 1.3885965668938152,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.300585280524764e-06,
+      "loss": 1.3719,
+      "step": 8049
+    },
+    {
+      "epoch": 1.3887690847925471,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.2983489637992855e-06,
+      "loss": 1.4976,
+      "step": 8050
+    },
+    {
+      "epoch": 1.388941602691279,
+      "grad_norm": 0.625,
+      "learning_rate": 4.296113069487743e-06,
+      "loss": 1.3982,
+      "step": 8051
+    },
+    {
+      "epoch": 1.3891141205900113,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.293877597755783e-06,
+      "loss": 1.4678,
+      "step": 8052
+    },
+    {
+      "epoch": 1.3892866384887432,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.291642548769024e-06,
+      "loss": 1.4979,
+      "step": 8053
+    },
+    {
+      "epoch": 1.3894591563874752,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.289407922693053e-06,
+      "loss": 1.4575,
+      "step": 8054
+    },
+    {
+      "epoch": 1.3896316742862072,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.287173719693427e-06,
+      "loss": 1.4439,
+      "step": 8055
+    },
+    {
+      "epoch": 1.3898041921849391,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.284939939935662e-06,
+      "loss": 1.4507,
+      "step": 8056
+    },
+    {
+      "epoch": 1.3899767100836713,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.282706583585258e-06,
+      "loss": 1.4287,
+      "step": 8057
+    },
+    {
+      "epoch": 1.3901492279824033,
+      "grad_norm": 0.625,
+      "learning_rate": 4.280473650807676e-06,
+      "loss": 1.5049,
+      "step": 8058
+    },
+    {
+      "epoch": 1.3903217458811352,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.278241141768338e-06,
+      "loss": 1.3884,
+      "step": 8059
+    },
+    {
+      "epoch": 1.3904942637798672,
+      "grad_norm": 0.625,
+      "learning_rate": 4.276009056632653e-06,
+      "loss": 1.4438,
+      "step": 8060
+    },
+    {
+      "epoch": 1.3906667816785991,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.27377739556598e-06,
+      "loss": 1.3712,
+      "step": 8061
+    },
+    {
+      "epoch": 1.390839299577331,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.271546158733656e-06,
+      "loss": 1.3682,
+      "step": 8062
+    },
+    {
+      "epoch": 1.391011817476063,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.269315346300985e-06,
+      "loss": 1.4151,
+      "step": 8063
+    },
+    {
+      "epoch": 1.391184335374795,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.2670849584332405e-06,
+      "loss": 1.5293,
+      "step": 8064
+    },
+    {
+      "epoch": 1.3913568532735272,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.264854995295664e-06,
+      "loss": 1.4793,
+      "step": 8065
+    },
+    {
+      "epoch": 1.3915293711722592,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.262625457053467e-06,
+      "loss": 1.5049,
+      "step": 8066
+    },
+    {
+      "epoch": 1.3917018890709911,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.260396343871819e-06,
+      "loss": 1.4196,
+      "step": 8067
+    },
+    {
+      "epoch": 1.391874406969723,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.258167655915878e-06,
+      "loss": 1.4206,
+      "step": 8068
+    },
+    {
+      "epoch": 1.392046924868455,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.255939393350751e-06,
+      "loss": 1.3965,
+      "step": 8069
+    },
+    {
+      "epoch": 1.3922194427671872,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.253711556341524e-06,
+      "loss": 1.4317,
+      "step": 8070
+    },
+    {
+      "epoch": 1.3923919606659192,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.25148414505325e-06,
+      "loss": 1.3929,
+      "step": 8071
+    },
+    {
+      "epoch": 1.3925644785646512,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.249257159650944e-06,
+      "loss": 1.3943,
+      "step": 8072
+    },
+    {
+      "epoch": 1.3927369964633831,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.2470306002996085e-06,
+      "loss": 1.5479,
+      "step": 8073
+    },
+    {
+      "epoch": 1.392909514362115,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.244804467164189e-06,
+      "loss": 1.4135,
+      "step": 8074
+    },
+    {
+      "epoch": 1.393082032260847,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.242578760409614e-06,
+      "loss": 1.3848,
+      "step": 8075
+    },
+    {
+      "epoch": 1.393254550159579,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.240353480200777e-06,
+      "loss": 1.3422,
+      "step": 8076
+    },
+    {
+      "epoch": 1.393427068058311,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.238128626702545e-06,
+      "loss": 1.4875,
+      "step": 8077
+    },
+    {
+      "epoch": 1.393599585957043,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.2359042000797434e-06,
+      "loss": 1.3744,
+      "step": 8078
+    },
+    {
+      "epoch": 1.393772103855775,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.23368020049718e-06,
+      "loss": 1.393,
+      "step": 8079
+    },
+    {
+      "epoch": 1.393944621754507,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.231456628119609e-06,
+      "loss": 1.4639,
+      "step": 8080
+    },
+    {
+      "epoch": 1.394117139653239,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.229233483111781e-06,
+      "loss": 1.4048,
+      "step": 8081
+    },
+    {
+      "epoch": 1.394289657551971,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.227010765638392e-06,
+      "loss": 1.3948,
+      "step": 8082
+    },
+    {
+      "epoch": 1.394462175450703,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.2247884758641155e-06,
+      "loss": 1.4422,
+      "step": 8083
+    },
+    {
+      "epoch": 1.3946346933494351,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.222566613953594e-06,
+      "loss": 1.5084,
+      "step": 8084
+    },
+    {
+      "epoch": 1.394807211248167,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.220345180071437e-06,
+      "loss": 1.4686,
+      "step": 8085
+    },
+    {
+      "epoch": 1.394979729146899,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.218124174382222e-06,
+      "loss": 1.4242,
+      "step": 8086
+    },
+    {
+      "epoch": 1.395152247045631,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.215903597050499e-06,
+      "loss": 1.4581,
+      "step": 8087
+    },
+    {
+      "epoch": 1.395324764944363,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.213683448240771e-06,
+      "loss": 1.5284,
+      "step": 8088
+    },
+    {
+      "epoch": 1.395497282843095,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.211463728117531e-06,
+      "loss": 1.3873,
+      "step": 8089
+    },
+    {
+      "epoch": 1.395669800741827,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.20924443684523e-06,
+      "loss": 1.3575,
+      "step": 8090
+    },
+    {
+      "epoch": 1.3958423186405589,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.207025574588276e-06,
+      "loss": 1.3796,
+      "step": 8091
+    },
+    {
+      "epoch": 1.3960148365392908,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.204807141511071e-06,
+      "loss": 1.3057,
+      "step": 8092
+    },
+    {
+      "epoch": 1.396187354438023,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.20258913777796e-06,
+      "loss": 1.398,
+      "step": 8093
+    },
+    {
+      "epoch": 1.396359872336755,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.200371563553269e-06,
+      "loss": 1.5357,
+      "step": 8094
+    },
+    {
+      "epoch": 1.396532390235487,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.1981544190012915e-06,
+      "loss": 1.3717,
+      "step": 8095
+    },
+    {
+      "epoch": 1.3967049081342189,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.195937704286285e-06,
+      "loss": 1.4642,
+      "step": 8096
+    },
+    {
+      "epoch": 1.3968774260329508,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.193721419572478e-06,
+      "loss": 1.376,
+      "step": 8097
+    },
+    {
+      "epoch": 1.397049943931683,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.191505565024073e-06,
+      "loss": 1.4358,
+      "step": 8098
+    },
+    {
+      "epoch": 1.397222461830415,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.189290140805221e-06,
+      "loss": 1.4701,
+      "step": 8099
+    },
+    {
+      "epoch": 1.397394979729147,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.18707514708007e-06,
+      "loss": 1.5173,
+      "step": 8100
+    },
+    {
+      "epoch": 1.397394979729147,
+      "eval_loss": 1.4073193073272705,
+      "eval_runtime": 11.2049,
+      "eval_samples_per_second": 91.389,
+      "eval_steps_per_second": 22.847,
+      "step": 8100
+    },
+    {
+      "epoch": 1.397567497627879,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.184860584012704e-06,
+      "loss": 1.38,
+      "step": 8101
+    },
+    {
+      "epoch": 1.3977400155266109,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.182646451767209e-06,
+      "loss": 1.4285,
+      "step": 8102
+    },
+    {
+      "epoch": 1.3979125334253428,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.18043275050761e-06,
+      "loss": 1.4443,
+      "step": 8103
+    },
+    {
+      "epoch": 1.3980850513240748,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.178219480397911e-06,
+      "loss": 1.4382,
+      "step": 8104
+    },
+    {
+      "epoch": 1.3982575692228068,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.1760066416020975e-06,
+      "loss": 1.5393,
+      "step": 8105
+    },
+    {
+      "epoch": 1.398430087121539,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.173794234284096e-06,
+      "loss": 1.3635,
+      "step": 8106
+    },
+    {
+      "epoch": 1.398602605020271,
+      "grad_norm": 0.52734375,
+      "learning_rate": 4.171582258607824e-06,
+      "loss": 1.2441,
+      "step": 8107
+    },
+    {
+      "epoch": 1.3987751229190029,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.169370714737155e-06,
+      "loss": 1.3819,
+      "step": 8108
+    },
+    {
+      "epoch": 1.3989476408177348,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.167159602835934e-06,
+      "loss": 1.419,
+      "step": 8109
+    },
+    {
+      "epoch": 1.3991201587164668,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.164948923067976e-06,
+      "loss": 1.3778,
+      "step": 8110
+    },
+    {
+      "epoch": 1.399292676615199,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.162738675597065e-06,
+      "loss": 1.4298,
+      "step": 8111
+    },
+    {
+      "epoch": 1.399465194513931,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.1605288605869365e-06,
+      "loss": 1.4636,
+      "step": 8112
+    },
+    {
+      "epoch": 1.3996377124126629,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.158319478201325e-06,
+      "loss": 1.3767,
+      "step": 8113
+    },
+    {
+      "epoch": 1.3998102303113948,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.156110528603904e-06,
+      "loss": 1.4037,
+      "step": 8114
+    },
+    {
+      "epoch": 1.3999827482101268,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.153902011958329e-06,
+      "loss": 1.4435,
+      "step": 8115
+    },
+    {
+      "epoch": 1.4001552661088588,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.151693928428221e-06,
+      "loss": 1.4166,
+      "step": 8116
+    },
+    {
+      "epoch": 1.4003277840075907,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.149486278177168e-06,
+      "loss": 1.4646,
+      "step": 8117
+    },
+    {
+      "epoch": 1.4005003019063227,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.147279061368729e-06,
+      "loss": 1.4447,
+      "step": 8118
+    },
+    {
+      "epoch": 1.4006728198050546,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.145072278166428e-06,
+      "loss": 1.3519,
+      "step": 8119
+    },
+    {
+      "epoch": 1.4008453377037868,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.142865928733751e-06,
+      "loss": 1.3877,
+      "step": 8120
+    },
+    {
+      "epoch": 1.4010178556025188,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.1406600132341645e-06,
+      "loss": 1.385,
+      "step": 8121
+    },
+    {
+      "epoch": 1.4011903735012508,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.1384545318311e-06,
+      "loss": 1.4587,
+      "step": 8122
+    },
+    {
+      "epoch": 1.4013628913999827,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.1362494846879405e-06,
+      "loss": 1.4534,
+      "step": 8123
+    },
+    {
+      "epoch": 1.4015354092987147,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.134044871968065e-06,
+      "loss": 1.3575,
+      "step": 8124
+    },
+    {
+      "epoch": 1.4017079271974469,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.131840693834794e-06,
+      "loss": 1.3504,
+      "step": 8125
+    },
+    {
+      "epoch": 1.4018804450961788,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.12963695045143e-06,
+      "loss": 1.3188,
+      "step": 8126
+    },
+    {
+      "epoch": 1.4020529629949108,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.127433641981241e-06,
+      "loss": 1.481,
+      "step": 8127
+    },
+    {
+      "epoch": 1.4022254808936427,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.125230768587461e-06,
+      "loss": 1.3856,
+      "step": 8128
+    },
+    {
+      "epoch": 1.4023979987923747,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.123028330433294e-06,
+      "loss": 1.3378,
+      "step": 8129
+    },
+    {
+      "epoch": 1.4025705166911067,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.120826327681911e-06,
+      "loss": 1.3412,
+      "step": 8130
+    },
+    {
+      "epoch": 1.4027430345898386,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.1186247604964425e-06,
+      "loss": 1.3749,
+      "step": 8131
+    },
+    {
+      "epoch": 1.4029155524885706,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.116423629040007e-06,
+      "loss": 1.4766,
+      "step": 8132
+    },
+    {
+      "epoch": 1.4030880703873025,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.1142229334756645e-06,
+      "loss": 1.4402,
+      "step": 8133
+    },
+    {
+      "epoch": 1.4032605882860347,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.112022673966471e-06,
+      "loss": 1.4126,
+      "step": 8134
+    },
+    {
+      "epoch": 1.4034331061847667,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.109822850675422e-06,
+      "loss": 1.3777,
+      "step": 8135
+    },
+    {
+      "epoch": 1.4036056240834986,
+      "grad_norm": 1.7109375,
+      "learning_rate": 4.107623463765498e-06,
+      "loss": 1.5189,
+      "step": 8136
+    },
+    {
+      "epoch": 1.4037781419822306,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.105424513399652e-06,
+      "loss": 1.3747,
+      "step": 8137
+    },
+    {
+      "epoch": 1.4039506598809626,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.1032259997407844e-06,
+      "loss": 1.4141,
+      "step": 8138
+    },
+    {
+      "epoch": 1.4041231777796948,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.10102792295178e-06,
+      "loss": 1.417,
+      "step": 8139
+    },
+    {
+      "epoch": 1.4042956956784267,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.098830283195485e-06,
+      "loss": 1.4213,
+      "step": 8140
+    },
+    {
+      "epoch": 1.4044682135771587,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.096633080634717e-06,
+      "loss": 1.4074,
+      "step": 8141
+    },
+    {
+      "epoch": 1.4046407314758906,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.094436315432254e-06,
+      "loss": 1.5289,
+      "step": 8142
+    },
+    {
+      "epoch": 1.4048132493746226,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.092239987750852e-06,
+      "loss": 1.4553,
+      "step": 8143
+    },
+    {
+      "epoch": 1.4049857672733546,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.09004409775322e-06,
+      "loss": 1.4543,
+      "step": 8144
+    },
+    {
+      "epoch": 1.4051582851720865,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.0878486456020535e-06,
+      "loss": 1.5197,
+      "step": 8145
+    },
+    {
+      "epoch": 1.4053308030708185,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.085653631459997e-06,
+      "loss": 1.4162,
+      "step": 8146
+    },
+    {
+      "epoch": 1.4055033209695507,
+      "grad_norm": 1.21875,
+      "learning_rate": 4.083459055489675e-06,
+      "loss": 1.4667,
+      "step": 8147
+    },
+    {
+      "epoch": 1.4056758388682826,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.081264917853674e-06,
+      "loss": 1.4256,
+      "step": 8148
+    },
+    {
+      "epoch": 1.4058483567670146,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.0790712187145486e-06,
+      "loss": 1.5337,
+      "step": 8149
+    },
+    {
+      "epoch": 1.4060208746657465,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.076877958234825e-06,
+      "loss": 1.4118,
+      "step": 8150
+    },
+    {
+      "epoch": 1.4061933925644785,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.074685136576993e-06,
+      "loss": 1.384,
+      "step": 8151
+    },
+    {
+      "epoch": 1.4063659104632107,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.0724927539035045e-06,
+      "loss": 1.4003,
+      "step": 8152
+    },
+    {
+      "epoch": 1.4065384283619426,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.070300810376792e-06,
+      "loss": 1.4395,
+      "step": 8153
+    },
+    {
+      "epoch": 1.4067109462606746,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.0681093061592495e-06,
+      "loss": 1.411,
+      "step": 8154
+    },
+    {
+      "epoch": 1.4068834641594066,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.065918241413226e-06,
+      "loss": 1.4444,
+      "step": 8155
+    },
+    {
+      "epoch": 1.4070559820581385,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.063727616301064e-06,
+      "loss": 1.4602,
+      "step": 8156
+    },
+    {
+      "epoch": 1.4072284999568705,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.061537430985049e-06,
+      "loss": 1.4107,
+      "step": 8157
+    },
+    {
+      "epoch": 1.4074010178556025,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.059347685627446e-06,
+      "loss": 1.4984,
+      "step": 8158
+    },
+    {
+      "epoch": 1.4075735357543344,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.057158380390486e-06,
+      "loss": 1.3805,
+      "step": 8159
+    },
+    {
+      "epoch": 1.4077460536530664,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.054969515436366e-06,
+      "loss": 1.3743,
+      "step": 8160
+    },
+    {
+      "epoch": 1.4079185715517986,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.052781090927249e-06,
+      "loss": 1.4984,
+      "step": 8161
+    },
+    {
+      "epoch": 1.4080910894505305,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.050593107025269e-06,
+      "loss": 1.4587,
+      "step": 8162
+    },
+    {
+      "epoch": 1.4082636073492625,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.048405563892527e-06,
+      "loss": 1.4661,
+      "step": 8163
+    },
+    {
+      "epoch": 1.4084361252479944,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.046218461691089e-06,
+      "loss": 1.4686,
+      "step": 8164
+    },
+    {
+      "epoch": 1.4086086431467264,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.044031800582984e-06,
+      "loss": 1.376,
+      "step": 8165
+    },
+    {
+      "epoch": 1.4087811610454586,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.041845580730223e-06,
+      "loss": 1.3829,
+      "step": 8166
+    },
+    {
+      "epoch": 1.4089536789441905,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.039659802294767e-06,
+      "loss": 1.418,
+      "step": 8167
+    },
+    {
+      "epoch": 1.4091261968429225,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.037474465438551e-06,
+      "loss": 1.4028,
+      "step": 8168
+    },
+    {
+      "epoch": 1.4092987147416545,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.035289570323489e-06,
+      "loss": 1.5105,
+      "step": 8169
+    },
+    {
+      "epoch": 1.4094712326403864,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.033105117111441e-06,
+      "loss": 1.4076,
+      "step": 8170
+    },
+    {
+      "epoch": 1.4096437505391184,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.030921105964249e-06,
+      "loss": 1.3904,
+      "step": 8171
+    },
+    {
+      "epoch": 1.4098162684378504,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.028737537043719e-06,
+      "loss": 1.4178,
+      "step": 8172
+    },
+    {
+      "epoch": 1.4099887863365823,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.026554410511622e-06,
+      "loss": 1.4416,
+      "step": 8173
+    },
+    {
+      "epoch": 1.4101613042353145,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.024371726529698e-06,
+      "loss": 1.514,
+      "step": 8174
+    },
+    {
+      "epoch": 1.4103338221340465,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.022189485259656e-06,
+      "loss": 1.5538,
+      "step": 8175
+    },
+    {
+      "epoch": 1.4105063400327784,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.020007686863164e-06,
+      "loss": 1.3949,
+      "step": 8176
+    },
+    {
+      "epoch": 1.4106788579315104,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.017826331501872e-06,
+      "loss": 1.3972,
+      "step": 8177
+    },
+    {
+      "epoch": 1.4108513758302423,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.015645419337381e-06,
+      "loss": 1.3803,
+      "step": 8178
+    },
+    {
+      "epoch": 1.4110238937289743,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.013464950531268e-06,
+      "loss": 1.4745,
+      "step": 8179
+    },
+    {
+      "epoch": 1.4111964116277065,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.011284925245079e-06,
+      "loss": 1.385,
+      "step": 8180
+    },
+    {
+      "epoch": 1.4113689295264384,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.009105343640321e-06,
+      "loss": 1.4544,
+      "step": 8181
+    },
+    {
+      "epoch": 1.4115414474251704,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.006926205878472e-06,
+      "loss": 1.3837,
+      "step": 8182
+    },
+    {
+      "epoch": 1.4117139653239024,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.00474751212098e-06,
+      "loss": 1.4387,
+      "step": 8183
+    },
+    {
+      "epoch": 1.4118864832226343,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.002569262529244e-06,
+      "loss": 1.2922,
+      "step": 8184
+    },
+    {
+      "epoch": 1.4120590011213663,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.000391457264656e-06,
+      "loss": 1.4863,
+      "step": 8185
+    },
+    {
+      "epoch": 1.4122315190200982,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.99821409648856e-06,
+      "loss": 1.4103,
+      "step": 8186
+    },
+    {
+      "epoch": 1.4124040369188302,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.996037180362256e-06,
+      "loss": 1.3828,
+      "step": 8187
+    },
+    {
+      "epoch": 1.4125765548175624,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.99386070904704e-06,
+      "loss": 1.3571,
+      "step": 8188
+    },
+    {
+      "epoch": 1.4127490727162944,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.991684682704143e-06,
+      "loss": 1.5166,
+      "step": 8189
+    },
+    {
+      "epoch": 1.4129215906150263,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.989509101494794e-06,
+      "loss": 1.5186,
+      "step": 8190
+    },
+    {
+      "epoch": 1.4130941085137583,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.987333965580163e-06,
+      "loss": 1.4764,
+      "step": 8191
+    },
+    {
+      "epoch": 1.4132666264124902,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.9851592751213995e-06,
+      "loss": 1.4711,
+      "step": 8192
+    },
+    {
+      "epoch": 1.4134391443112224,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.98298503027962e-06,
+      "loss": 1.4253,
+      "step": 8193
+    },
+    {
+      "epoch": 1.4136116622099544,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.980811231215905e-06,
+      "loss": 1.3507,
+      "step": 8194
+    },
+    {
+      "epoch": 1.4137841801086863,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.978637878091305e-06,
+      "loss": 1.4504,
+      "step": 8195
+    },
+    {
+      "epoch": 1.4139566980074183,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.976464971066837e-06,
+      "loss": 1.4436,
+      "step": 8196
+    },
+    {
+      "epoch": 1.4141292159061503,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.974292510303473e-06,
+      "loss": 1.4446,
+      "step": 8197
+    },
+    {
+      "epoch": 1.4143017338048822,
+      "grad_norm": 0.625,
+      "learning_rate": 3.972120495962178e-06,
+      "loss": 1.4351,
+      "step": 8198
+    },
+    {
+      "epoch": 1.4144742517036142,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.969948928203856e-06,
+      "loss": 1.3947,
+      "step": 8199
+    },
+    {
+      "epoch": 1.4146467696023461,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.967777807189396e-06,
+      "loss": 1.454,
+      "step": 8200
+    },
+    {
+      "epoch": 1.4146467696023461,
+      "eval_loss": 1.4072892665863037,
+      "eval_runtime": 10.8508,
+      "eval_samples_per_second": 94.371,
+      "eval_steps_per_second": 23.593,
+      "step": 8200
+    },
+    {
+      "epoch": 1.414819287501078,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.9656071330796475e-06,
+      "loss": 1.343,
+      "step": 8201
+    },
+    {
+      "epoch": 1.4149918053998103,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.963436906035426e-06,
+      "loss": 1.4482,
+      "step": 8202
+    },
+    {
+      "epoch": 1.4151643232985422,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.961267126217517e-06,
+      "loss": 1.4088,
+      "step": 8203
+    },
+    {
+      "epoch": 1.4153368411972742,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.9590977937866715e-06,
+      "loss": 1.3249,
+      "step": 8204
+    },
+    {
+      "epoch": 1.4155093590960062,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.956928908903607e-06,
+      "loss": 1.5044,
+      "step": 8205
+    },
+    {
+      "epoch": 1.4156818769947381,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.954760471729007e-06,
+      "loss": 1.3926,
+      "step": 8206
+    },
+    {
+      "epoch": 1.4158543948934703,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.952592482423527e-06,
+      "loss": 1.3704,
+      "step": 8207
+    },
+    {
+      "epoch": 1.4160269127922023,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.950424941147776e-06,
+      "loss": 1.446,
+      "step": 8208
+    },
+    {
+      "epoch": 1.4161994306909342,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.948257848062351e-06,
+      "loss": 1.4207,
+      "step": 8209
+    },
+    {
+      "epoch": 1.4163719485896662,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.946091203327794e-06,
+      "loss": 1.3249,
+      "step": 8210
+    },
+    {
+      "epoch": 1.4165444664883982,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.9439250071046274e-06,
+      "loss": 1.4373,
+      "step": 8211
+    },
+    {
+      "epoch": 1.4167169843871301,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.941759259553336e-06,
+      "loss": 1.4268,
+      "step": 8212
+    },
+    {
+      "epoch": 1.416889502285862,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.939593960834374e-06,
+      "loss": 1.3892,
+      "step": 8213
+    },
+    {
+      "epoch": 1.417062020184594,
+      "grad_norm": 0.625,
+      "learning_rate": 3.937429111108157e-06,
+      "loss": 1.3337,
+      "step": 8214
+    },
+    {
+      "epoch": 1.4172345380833262,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.935264710535076e-06,
+      "loss": 1.4283,
+      "step": 8215
+    },
+    {
+      "epoch": 1.4174070559820582,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.9331007592754725e-06,
+      "loss": 1.3613,
+      "step": 8216
+    },
+    {
+      "epoch": 1.4175795738807901,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.9309372574896755e-06,
+      "loss": 1.4294,
+      "step": 8217
+    },
+    {
+      "epoch": 1.417752091779522,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.928774205337972e-06,
+      "loss": 1.5224,
+      "step": 8218
+    },
+    {
+      "epoch": 1.417924609678254,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.926611602980603e-06,
+      "loss": 1.3941,
+      "step": 8219
+    },
+    {
+      "epoch": 1.4180971275769862,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.924449450577801e-06,
+      "loss": 1.4262,
+      "step": 8220
+    },
+    {
+      "epoch": 1.4182696454757182,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.922287748289739e-06,
+      "loss": 1.4873,
+      "step": 8221
+    },
+    {
+      "epoch": 1.4184421633744502,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.920126496276583e-06,
+      "loss": 1.5009,
+      "step": 8222
+    },
+    {
+      "epoch": 1.4186146812731821,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.917965694698441e-06,
+      "loss": 1.3267,
+      "step": 8223
+    },
+    {
+      "epoch": 1.418787199171914,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.9158053437154035e-06,
+      "loss": 1.4316,
+      "step": 8224
+    },
+    {
+      "epoch": 1.418959717070646,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.9136454434875216e-06,
+      "loss": 1.4051,
+      "step": 8225
+    },
+    {
+      "epoch": 1.419132234969378,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.911485994174814e-06,
+      "loss": 1.3459,
+      "step": 8226
+    },
+    {
+      "epoch": 1.41930475286811,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.909326995937267e-06,
+      "loss": 1.4291,
+      "step": 8227
+    },
+    {
+      "epoch": 1.419477270766842,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.907168448934836e-06,
+      "loss": 1.4023,
+      "step": 8228
+    },
+    {
+      "epoch": 1.4196497886655741,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.9050103533274295e-06,
+      "loss": 1.54,
+      "step": 8229
+    },
+    {
+      "epoch": 1.419822306564306,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.902852709274946e-06,
+      "loss": 1.4052,
+      "step": 8230
+    },
+    {
+      "epoch": 1.419994824463038,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.900695516937226e-06,
+      "loss": 1.4389,
+      "step": 8231
+    },
+    {
+      "epoch": 1.42016734236177,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.898538776474095e-06,
+      "loss": 1.4603,
+      "step": 8232
+    },
+    {
+      "epoch": 1.420339860260502,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.896382488045333e-06,
+      "loss": 1.4328,
+      "step": 8233
+    },
+    {
+      "epoch": 1.4205123781592341,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.894226651810693e-06,
+      "loss": 1.3588,
+      "step": 8234
+    },
+    {
+      "epoch": 1.420684896057966,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.892071267929894e-06,
+      "loss": 1.4438,
+      "step": 8235
+    },
+    {
+      "epoch": 1.420857413956698,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.88991633656262e-06,
+      "loss": 1.4963,
+      "step": 8236
+    },
+    {
+      "epoch": 1.42102993185543,
+      "grad_norm": 0.625,
+      "learning_rate": 3.887761857868522e-06,
+      "loss": 1.3939,
+      "step": 8237
+    },
+    {
+      "epoch": 1.421202449754162,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.885607832007214e-06,
+      "loss": 1.3764,
+      "step": 8238
+    },
+    {
+      "epoch": 1.421374967652894,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.883454259138289e-06,
+      "loss": 1.3638,
+      "step": 8239
+    },
+    {
+      "epoch": 1.421547485551626,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.881301139421281e-06,
+      "loss": 1.3644,
+      "step": 8240
+    },
+    {
+      "epoch": 1.4217200034503579,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.879148473015723e-06,
+      "loss": 1.4693,
+      "step": 8241
+    },
+    {
+      "epoch": 1.4218925213490898,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.87699626008109e-06,
+      "loss": 1.4157,
+      "step": 8242
+    },
+    {
+      "epoch": 1.422065039247822,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.87484450077683e-06,
+      "loss": 1.4135,
+      "step": 8243
+    },
+    {
+      "epoch": 1.422237557146554,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.872693195262361e-06,
+      "loss": 1.5038,
+      "step": 8244
+    },
+    {
+      "epoch": 1.422410075045286,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.870542343697067e-06,
+      "loss": 1.4158,
+      "step": 8245
+    },
+    {
+      "epoch": 1.422582592944018,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.868391946240294e-06,
+      "loss": 1.4482,
+      "step": 8246
+    },
+    {
+      "epoch": 1.4227551108427499,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.86624200305136e-06,
+      "loss": 1.3736,
+      "step": 8247
+    },
+    {
+      "epoch": 1.422927628741482,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.864092514289539e-06,
+      "loss": 1.3976,
+      "step": 8248
+    },
+    {
+      "epoch": 1.423100146640214,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.861943480114086e-06,
+      "loss": 1.3573,
+      "step": 8249
+    },
+    {
+      "epoch": 1.423272664538946,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.859794900684212e-06,
+      "loss": 1.4126,
+      "step": 8250
+    },
+    {
+      "epoch": 1.423445182437678,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.857646776159098e-06,
+      "loss": 1.4187,
+      "step": 8251
+    },
+    {
+      "epoch": 1.4236177003364099,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.855499106697893e-06,
+      "loss": 1.4546,
+      "step": 8252
+    },
+    {
+      "epoch": 1.4237902182351418,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.8533518924597e-06,
+      "loss": 1.3757,
+      "step": 8253
+    },
+    {
+      "epoch": 1.4239627361338738,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.851205133603611e-06,
+      "loss": 1.4586,
+      "step": 8254
+    },
+    {
+      "epoch": 1.4241352540326058,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.849058830288663e-06,
+      "loss": 1.4201,
+      "step": 8255
+    },
+    {
+      "epoch": 1.424307771931338,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.846912982673869e-06,
+      "loss": 1.5396,
+      "step": 8256
+    },
+    {
+      "epoch": 1.42448028983007,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.8447675909182095e-06,
+      "loss": 1.4887,
+      "step": 8257
+    },
+    {
+      "epoch": 1.4246528077288019,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.842622655180625e-06,
+      "loss": 1.4912,
+      "step": 8258
+    },
+    {
+      "epoch": 1.4248253256275338,
+      "grad_norm": 0.66796875,
+      "learning_rate": 3.840478175620026e-06,
+      "loss": 1.402,
+      "step": 8259
+    },
+    {
+      "epoch": 1.4249978435262658,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.838334152395295e-06,
+      "loss": 1.4811,
+      "step": 8260
+    },
+    {
+      "epoch": 1.425170361424998,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.836190585665263e-06,
+      "loss": 1.4547,
+      "step": 8261
+    },
+    {
+      "epoch": 1.42534287932373,
+      "grad_norm": 0.640625,
+      "learning_rate": 3.834047475588753e-06,
+      "loss": 1.4213,
+      "step": 8262
+    },
+    {
+      "epoch": 1.425515397222462,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.831904822324527e-06,
+      "loss": 1.5176,
+      "step": 8263
+    },
+    {
+      "epoch": 1.4256879151211939,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.829762626031333e-06,
+      "loss": 1.4017,
+      "step": 8264
+    },
+    {
+      "epoch": 1.4258604330199258,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.827620886867878e-06,
+      "loss": 1.4718,
+      "step": 8265
+    },
+    {
+      "epoch": 1.4260329509186578,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.825479604992835e-06,
+      "loss": 1.3909,
+      "step": 8266
+    },
+    {
+      "epoch": 1.4262054688173897,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.823338780564841e-06,
+      "loss": 1.4747,
+      "step": 8267
+    },
+    {
+      "epoch": 1.4263779867161217,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.821198413742505e-06,
+      "loss": 1.4617,
+      "step": 8268
+    },
+    {
+      "epoch": 1.4265505046148537,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.8190585046843965e-06,
+      "loss": 1.3563,
+      "step": 8269
+    },
+    {
+      "epoch": 1.4267230225135858,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.8169190535490555e-06,
+      "loss": 1.5484,
+      "step": 8270
+    },
+    {
+      "epoch": 1.4268955404123178,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.814780060494987e-06,
+      "loss": 1.4492,
+      "step": 8271
+    },
+    {
+      "epoch": 1.4270680583110498,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.8126415256806527e-06,
+      "loss": 1.3372,
+      "step": 8272
+    },
+    {
+      "epoch": 1.4272405762097817,
+      "grad_norm": 0.671875,
+      "learning_rate": 3.810503449264501e-06,
+      "loss": 1.4254,
+      "step": 8273
+    },
+    {
+      "epoch": 1.4274130941085137,
+      "grad_norm": 0.640625,
+      "learning_rate": 3.808365831404924e-06,
+      "loss": 1.4118,
+      "step": 8274
+    },
+    {
+      "epoch": 1.4275856120072459,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.8062286722602936e-06,
+      "loss": 1.3553,
+      "step": 8275
+    },
+    {
+      "epoch": 1.4277581299059778,
+      "grad_norm": 0.625,
+      "learning_rate": 3.8040919719889435e-06,
+      "loss": 1.4825,
+      "step": 8276
+    },
+    {
+      "epoch": 1.4279306478047098,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.801955730749174e-06,
+      "loss": 1.4204,
+      "step": 8277
+    },
+    {
+      "epoch": 1.4281031657034418,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.7998199486992502e-06,
+      "loss": 1.4068,
+      "step": 8278
+    },
+    {
+      "epoch": 1.4282756836021737,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.7976846259974098e-06,
+      "loss": 1.3543,
+      "step": 8279
+    },
+    {
+      "epoch": 1.4284482015009057,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.795549762801839e-06,
+      "loss": 1.4565,
+      "step": 8280
+    },
+    {
+      "epoch": 1.4286207193996376,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.7934153592707125e-06,
+      "loss": 1.3196,
+      "step": 8281
+    },
+    {
+      "epoch": 1.4287932372983696,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.7912814155621568e-06,
+      "loss": 1.3837,
+      "step": 8282
+    },
+    {
+      "epoch": 1.4289657551971016,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.789147931834267e-06,
+      "loss": 1.4508,
+      "step": 8283
+    },
+    {
+      "epoch": 1.4291382730958337,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.7870149082451104e-06,
+      "loss": 1.4073,
+      "step": 8284
+    },
+    {
+      "epoch": 1.4293107909945657,
+      "grad_norm": 0.66015625,
+      "learning_rate": 3.784882344952702e-06,
+      "loss": 1.5276,
+      "step": 8285
+    },
+    {
+      "epoch": 1.4294833088932977,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.7827502421150497e-06,
+      "loss": 1.3499,
+      "step": 8286
+    },
+    {
+      "epoch": 1.4296558267920296,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.7806185998901034e-06,
+      "loss": 1.4606,
+      "step": 8287
+    },
+    {
+      "epoch": 1.4298283446907616,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.7784874184357923e-06,
+      "loss": 1.3935,
+      "step": 8288
+    },
+    {
+      "epoch": 1.4300008625894938,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.7763566979100076e-06,
+      "loss": 1.4289,
+      "step": 8289
+    },
+    {
+      "epoch": 1.4301733804882257,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.774226438470605e-06,
+      "loss": 1.3783,
+      "step": 8290
+    },
+    {
+      "epoch": 1.4303458983869577,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.7720966402754076e-06,
+      "loss": 1.3995,
+      "step": 8291
+    },
+    {
+      "epoch": 1.4305184162856897,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.7699673034822095e-06,
+      "loss": 1.4908,
+      "step": 8292
+    },
+    {
+      "epoch": 1.4306909341844216,
+      "grad_norm": 0.625,
+      "learning_rate": 3.7678384282487535e-06,
+      "loss": 1.4742,
+      "step": 8293
+    },
+    {
+      "epoch": 1.4308634520831536,
+      "grad_norm": 0.671875,
+      "learning_rate": 3.765710014732774e-06,
+      "loss": 1.398,
+      "step": 8294
+    },
+    {
+      "epoch": 1.4310359699818855,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.7635820630919464e-06,
+      "loss": 1.3722,
+      "step": 8295
+    },
+    {
+      "epoch": 1.4312084878806175,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.761454573483927e-06,
+      "loss": 1.49,
+      "step": 8296
+    },
+    {
+      "epoch": 1.4313810057793497,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.7593275460663323e-06,
+      "loss": 1.4273,
+      "step": 8297
+    },
+    {
+      "epoch": 1.4315535236780816,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.757200980996747e-06,
+      "loss": 1.4282,
+      "step": 8298
+    },
+    {
+      "epoch": 1.4317260415768136,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.7550748784327195e-06,
+      "loss": 1.4377,
+      "step": 8299
+    },
+    {
+      "epoch": 1.4318985594755456,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.752949238531766e-06,
+      "loss": 1.4625,
+      "step": 8300
+    },
+    {
+      "epoch": 1.4318985594755456,
+      "eval_loss": 1.4072835445404053,
+      "eval_runtime": 10.8629,
+      "eval_samples_per_second": 94.266,
+      "eval_steps_per_second": 23.566,
+      "step": 8300
+    },
+    {
+      "epoch": 1.4320710773742775,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.750824061451365e-06,
+      "loss": 1.3686,
+      "step": 8301
+    },
+    {
+      "epoch": 1.4322435952730097,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.7486993473489654e-06,
+      "loss": 1.4269,
+      "step": 8302
+    },
+    {
+      "epoch": 1.4324161131717417,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.7465750963819802e-06,
+      "loss": 1.4262,
+      "step": 8303
+    },
+    {
+      "epoch": 1.4325886310704736,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.7444513087077793e-06,
+      "loss": 1.3747,
+      "step": 8304
+    },
+    {
+      "epoch": 1.4327611489692056,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.742327984483718e-06,
+      "loss": 1.4343,
+      "step": 8305
+    },
+    {
+      "epoch": 1.4329336668679375,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.740205123867097e-06,
+      "loss": 1.3822,
+      "step": 8306
+    },
+    {
+      "epoch": 1.4331061847666695,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.738082727015192e-06,
+      "loss": 1.4472,
+      "step": 8307
+    },
+    {
+      "epoch": 1.4332787026654015,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.735960794085246e-06,
+      "loss": 1.405,
+      "step": 8308
+    },
+    {
+      "epoch": 1.4334512205641334,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.733839325234463e-06,
+      "loss": 1.4689,
+      "step": 8309
+    },
+    {
+      "epoch": 1.4336237384628654,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.731718320620017e-06,
+      "loss": 1.4762,
+      "step": 8310
+    },
+    {
+      "epoch": 1.4337962563615976,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.7295977803990426e-06,
+      "loss": 1.4476,
+      "step": 8311
+    },
+    {
+      "epoch": 1.4339687742603295,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.7274777047286435e-06,
+      "loss": 1.4185,
+      "step": 8312
+    },
+    {
+      "epoch": 1.4341412921590615,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.7253580937658896e-06,
+      "loss": 1.2961,
+      "step": 8313
+    },
+    {
+      "epoch": 1.4343138100577935,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.723238947667813e-06,
+      "loss": 1.4259,
+      "step": 8314
+    },
+    {
+      "epoch": 1.4344863279565254,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.721120266591416e-06,
+      "loss": 1.5421,
+      "step": 8315
+    },
+    {
+      "epoch": 1.4346588458552576,
+      "grad_norm": 0.6796875,
+      "learning_rate": 3.719002050693663e-06,
+      "loss": 1.4687,
+      "step": 8316
+    },
+    {
+      "epoch": 1.4348313637539896,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.716884300131478e-06,
+      "loss": 1.4174,
+      "step": 8317
+    },
+    {
+      "epoch": 1.4350038816527215,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.7147670150617698e-06,
+      "loss": 1.4209,
+      "step": 8318
+    },
+    {
+      "epoch": 1.4351763995514535,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.7126501956413898e-06,
+      "loss": 1.3623,
+      "step": 8319
+    },
+    {
+      "epoch": 1.4353489174501854,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.7105338420271673e-06,
+      "loss": 1.428,
+      "step": 8320
+    },
+    {
+      "epoch": 1.4355214353489174,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.7084179543758968e-06,
+      "loss": 1.4367,
+      "step": 8321
+    },
+    {
+      "epoch": 1.4356939532476494,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.706302532844336e-06,
+      "loss": 1.4826,
+      "step": 8322
+    },
+    {
+      "epoch": 1.4358664711463813,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.7041875775892077e-06,
+      "loss": 1.4315,
+      "step": 8323
+    },
+    {
+      "epoch": 1.4360389890451135,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.7020730887672063e-06,
+      "loss": 1.4028,
+      "step": 8324
+    },
+    {
+      "epoch": 1.4362115069438455,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.699959066534974e-06,
+      "loss": 1.2957,
+      "step": 8325
+    },
+    {
+      "epoch": 1.4363840248425774,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.697845511049146e-06,
+      "loss": 1.3867,
+      "step": 8326
+    },
+    {
+      "epoch": 1.4365565427413094,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.695732422466296e-06,
+      "loss": 1.4988,
+      "step": 8327
+    },
+    {
+      "epoch": 1.4367290606400414,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.69361980094298e-06,
+      "loss": 1.384,
+      "step": 8328
+    },
+    {
+      "epoch": 1.4369015785387733,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.6915076466357123e-06,
+      "loss": 1.3995,
+      "step": 8329
+    },
+    {
+      "epoch": 1.4370740964375055,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.6893959597009766e-06,
+      "loss": 1.3905,
+      "step": 8330
+    },
+    {
+      "epoch": 1.4372466143362375,
+      "grad_norm": 0.640625,
+      "learning_rate": 3.687284740295217e-06,
+      "loss": 1.4135,
+      "step": 8331
+    },
+    {
+      "epoch": 1.4374191322349694,
+      "grad_norm": 0.53125,
+      "learning_rate": 3.6851739885748495e-06,
+      "loss": 1.3253,
+      "step": 8332
+    },
+    {
+      "epoch": 1.4375916501337014,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.683063704696248e-06,
+      "loss": 1.4941,
+      "step": 8333
+    },
+    {
+      "epoch": 1.4377641680324333,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.680953888815758e-06,
+      "loss": 1.4896,
+      "step": 8334
+    },
+    {
+      "epoch": 1.4379366859311653,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.678844541089691e-06,
+      "loss": 1.4125,
+      "step": 8335
+    },
+    {
+      "epoch": 1.4381092038298973,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.6767356616743098e-06,
+      "loss": 1.3701,
+      "step": 8336
+    },
+    {
+      "epoch": 1.4382817217286292,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.674627250725866e-06,
+      "loss": 1.4929,
+      "step": 8337
+    },
+    {
+      "epoch": 1.4384542396273614,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.6725193084005527e-06,
+      "loss": 1.4682,
+      "step": 8338
+    },
+    {
+      "epoch": 1.4386267575260934,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.6704118348545516e-06,
+      "loss": 1.3607,
+      "step": 8339
+    },
+    {
+      "epoch": 1.4387992754248253,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.668304830243987e-06,
+      "loss": 1.4724,
+      "step": 8340
+    },
+    {
+      "epoch": 1.4389717933235573,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.666198294724963e-06,
+      "loss": 1.4779,
+      "step": 8341
+    },
+    {
+      "epoch": 1.4391443112222893,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.664092228453545e-06,
+      "loss": 1.4672,
+      "step": 8342
+    },
+    {
+      "epoch": 1.4393168291210214,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.6619866315857632e-06,
+      "loss": 1.3252,
+      "step": 8343
+    },
+    {
+      "epoch": 1.4394893470197534,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.6598815042776135e-06,
+      "loss": 1.4026,
+      "step": 8344
+    },
+    {
+      "epoch": 1.4396618649184854,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.657776846685057e-06,
+      "loss": 1.5287,
+      "step": 8345
+    },
+    {
+      "epoch": 1.4398343828172173,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.655672658964019e-06,
+      "loss": 1.4182,
+      "step": 8346
+    },
+    {
+      "epoch": 1.4400069007159493,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.6535689412703923e-06,
+      "loss": 1.4253,
+      "step": 8347
+    },
+    {
+      "epoch": 1.4401794186146812,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.651465693760037e-06,
+      "loss": 1.4129,
+      "step": 8348
+    },
+    {
+      "epoch": 1.4403519365134132,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.6493629165887623e-06,
+      "loss": 1.3906,
+      "step": 8349
+    },
+    {
+      "epoch": 1.4405244544121452,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.647260609912371e-06,
+      "loss": 1.5282,
+      "step": 8350
+    },
+    {
+      "epoch": 1.4406969723108771,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.6451587738866035e-06,
+      "loss": 1.4269,
+      "step": 8351
+    },
+    {
+      "epoch": 1.4408694902096093,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.643057408667181e-06,
+      "loss": 1.4873,
+      "step": 8352
+    },
+    {
+      "epoch": 1.4410420081083413,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.6409565144097856e-06,
+      "loss": 1.4226,
+      "step": 8353
+    },
+    {
+      "epoch": 1.4412145260070732,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.6388560912700642e-06,
+      "loss": 1.3989,
+      "step": 8354
+    },
+    {
+      "epoch": 1.4413870439058052,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.636756139403631e-06,
+      "loss": 1.4151,
+      "step": 8355
+    },
+    {
+      "epoch": 1.4415595618045371,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.634656658966066e-06,
+      "loss": 1.4418,
+      "step": 8356
+    },
+    {
+      "epoch": 1.4417320797032693,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.6325576501129002e-06,
+      "loss": 1.3659,
+      "step": 8357
+    },
+    {
+      "epoch": 1.4419045976020013,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.630459112999657e-06,
+      "loss": 1.3226,
+      "step": 8358
+    },
+    {
+      "epoch": 1.4420771155007333,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.6283610477817975e-06,
+      "loss": 1.5103,
+      "step": 8359
+    },
+    {
+      "epoch": 1.4422496333994652,
+      "grad_norm": 0.66796875,
+      "learning_rate": 3.6262634546147635e-06,
+      "loss": 1.4522,
+      "step": 8360
+    },
+    {
+      "epoch": 1.4424221512981972,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.624166333653958e-06,
+      "loss": 1.4022,
+      "step": 8361
+    },
+    {
+      "epoch": 1.4425946691969291,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.622069685054749e-06,
+      "loss": 1.547,
+      "step": 8362
+    },
+    {
+      "epoch": 1.442767187095661,
+      "grad_norm": 0.6640625,
+      "learning_rate": 3.6199735089724676e-06,
+      "loss": 1.4204,
+      "step": 8363
+    },
+    {
+      "epoch": 1.442939704994393,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.617877805562413e-06,
+      "loss": 1.376,
+      "step": 8364
+    },
+    {
+      "epoch": 1.4431122228931252,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.6157825749798493e-06,
+      "loss": 1.4068,
+      "step": 8365
+    },
+    {
+      "epoch": 1.4432847407918572,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.613687817380002e-06,
+      "loss": 1.3867,
+      "step": 8366
+    },
+    {
+      "epoch": 1.4434572586905892,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.611593532918064e-06,
+      "loss": 1.3909,
+      "step": 8367
+    },
+    {
+      "epoch": 1.4436297765893211,
+      "grad_norm": 0.66015625,
+      "learning_rate": 3.6094997217491944e-06,
+      "loss": 1.4072,
+      "step": 8368
+    },
+    {
+      "epoch": 1.443802294488053,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.607406384028519e-06,
+      "loss": 1.4459,
+      "step": 8369
+    },
+    {
+      "epoch": 1.4439748123867853,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.6053135199111144e-06,
+      "loss": 1.398,
+      "step": 8370
+    },
+    {
+      "epoch": 1.4441473302855172,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.603221129552047e-06,
+      "loss": 1.5272,
+      "step": 8371
+    },
+    {
+      "epoch": 1.4443198481842492,
+      "grad_norm": 0.6796875,
+      "learning_rate": 3.6011292131063227e-06,
+      "loss": 1.467,
+      "step": 8372
+    },
+    {
+      "epoch": 1.4444923660829811,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.5990377707289292e-06,
+      "loss": 1.4917,
+      "step": 8373
+    },
+    {
+      "epoch": 1.444664883981713,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.5969468025748135e-06,
+      "loss": 1.4118,
+      "step": 8374
+    },
+    {
+      "epoch": 1.444837401880445,
+      "grad_norm": 0.625,
+      "learning_rate": 3.594856308798885e-06,
+      "loss": 1.3988,
+      "step": 8375
+    },
+    {
+      "epoch": 1.445009919779177,
+      "grad_norm": 0.6640625,
+      "learning_rate": 3.5927662895560235e-06,
+      "loss": 1.562,
+      "step": 8376
+    },
+    {
+      "epoch": 1.445182437677909,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.5906767450010694e-06,
+      "loss": 1.4244,
+      "step": 8377
+    },
+    {
+      "epoch": 1.445354955576641,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.5885876752888295e-06,
+      "loss": 1.3598,
+      "step": 8378
+    },
+    {
+      "epoch": 1.4455274734753731,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.586499080574074e-06,
+      "loss": 1.4699,
+      "step": 8379
+    },
+    {
+      "epoch": 1.445699991374105,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.584410961011544e-06,
+      "loss": 1.5031,
+      "step": 8380
+    },
+    {
+      "epoch": 1.445872509272837,
+      "grad_norm": 0.625,
+      "learning_rate": 3.58232331675593e-06,
+      "loss": 1.3617,
+      "step": 8381
+    },
+    {
+      "epoch": 1.446045027171569,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.580236147961911e-06,
+      "loss": 1.4913,
+      "step": 8382
+    },
+    {
+      "epoch": 1.446217545070301,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.5781494547841066e-06,
+      "loss": 1.4094,
+      "step": 8383
+    },
+    {
+      "epoch": 1.4463900629690332,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.5760632373771163e-06,
+      "loss": 1.3552,
+      "step": 8384
+    },
+    {
+      "epoch": 1.4465625808677651,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.573977495895501e-06,
+      "loss": 1.4264,
+      "step": 8385
+    },
+    {
+      "epoch": 1.446735098766497,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.571892230493783e-06,
+      "loss": 1.4048,
+      "step": 8386
+    },
+    {
+      "epoch": 1.446907616665229,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.569807441326454e-06,
+      "loss": 1.384,
+      "step": 8387
+    },
+    {
+      "epoch": 1.447080134563961,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.567723128547971e-06,
+      "loss": 1.3813,
+      "step": 8388
+    },
+    {
+      "epoch": 1.447252652462693,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.5656392923127424e-06,
+      "loss": 1.3643,
+      "step": 8389
+    },
+    {
+      "epoch": 1.447425170361425,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.5635559327751655e-06,
+      "loss": 1.4991,
+      "step": 8390
+    },
+    {
+      "epoch": 1.447597688260157,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.561473050089579e-06,
+      "loss": 1.4303,
+      "step": 8391
+    },
+    {
+      "epoch": 1.4477702061588889,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.559390644410298e-06,
+      "loss": 1.4665,
+      "step": 8392
+    },
+    {
+      "epoch": 1.447942724057621,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.557308715891601e-06,
+      "loss": 1.4102,
+      "step": 8393
+    },
+    {
+      "epoch": 1.448115241956353,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.555227264687726e-06,
+      "loss": 1.5032,
+      "step": 8394
+    },
+    {
+      "epoch": 1.448287759855085,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.5531462909528925e-06,
+      "loss": 1.4227,
+      "step": 8395
+    },
+    {
+      "epoch": 1.448460277753817,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.5510657948412587e-06,
+      "loss": 1.3451,
+      "step": 8396
+    },
+    {
+      "epoch": 1.4486327956525489,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.548985776506967e-06,
+      "loss": 1.4781,
+      "step": 8397
+    },
+    {
+      "epoch": 1.448805313551281,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.5469062361041152e-06,
+      "loss": 1.4004,
+      "step": 8398
+    },
+    {
+      "epoch": 1.448977831450013,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.5448271737867713e-06,
+      "loss": 1.4395,
+      "step": 8399
+    },
+    {
+      "epoch": 1.449150349348745,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.542748589708963e-06,
+      "loss": 1.4146,
+      "step": 8400
+    },
+    {
+      "epoch": 1.449150349348745,
+      "eval_loss": 1.4072437286376953,
+      "eval_runtime": 10.8573,
+      "eval_samples_per_second": 94.315,
+      "eval_steps_per_second": 23.579,
+      "step": 8400
+    },
+    {
+      "epoch": 1.449322867247477,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.540670484024691e-06,
+      "loss": 1.2855,
+      "step": 8401
+    },
+    {
+      "epoch": 1.449495385146209,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.5385928568879012e-06,
+      "loss": 1.3483,
+      "step": 8402
+    },
+    {
+      "epoch": 1.4496679030449409,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.5365157084525326e-06,
+      "loss": 1.48,
+      "step": 8403
+    },
+    {
+      "epoch": 1.4498404209436728,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.5344390388724625e-06,
+      "loss": 1.388,
+      "step": 8404
+    },
+    {
+      "epoch": 1.4500129388424048,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.532362848301547e-06,
+      "loss": 1.4408,
+      "step": 8405
+    },
+    {
+      "epoch": 1.450185456741137,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.5302871368936043e-06,
+      "loss": 1.3503,
+      "step": 8406
+    },
+    {
+      "epoch": 1.450357974639869,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.5282119048024146e-06,
+      "loss": 1.3585,
+      "step": 8407
+    },
+    {
+      "epoch": 1.450530492538601,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.5261371521817247e-06,
+      "loss": 1.4256,
+      "step": 8408
+    },
+    {
+      "epoch": 1.4507030104373329,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.524062879185247e-06,
+      "loss": 1.333,
+      "step": 8409
+    },
+    {
+      "epoch": 1.4508755283360648,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.5219890859666493e-06,
+      "loss": 1.5221,
+      "step": 8410
+    },
+    {
+      "epoch": 1.451048046234797,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.519915772679581e-06,
+      "loss": 1.431,
+      "step": 8411
+    },
+    {
+      "epoch": 1.451220564133529,
+      "grad_norm": 0.875,
+      "learning_rate": 3.5178429394776436e-06,
+      "loss": 1.5443,
+      "step": 8412
+    },
+    {
+      "epoch": 1.451393082032261,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.5157705865143964e-06,
+      "loss": 1.4704,
+      "step": 8413
+    },
+    {
+      "epoch": 1.4515655999309929,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.5136987139433874e-06,
+      "loss": 1.5263,
+      "step": 8414
+    },
+    {
+      "epoch": 1.4517381178297248,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.511627321918102e-06,
+      "loss": 1.4345,
+      "step": 8415
+    },
+    {
+      "epoch": 1.4519106357284568,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.5095564105920065e-06,
+      "loss": 1.4462,
+      "step": 8416
+    },
+    {
+      "epoch": 1.4520831536271888,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.5074859801185256e-06,
+      "loss": 1.4408,
+      "step": 8417
+    },
+    {
+      "epoch": 1.4522556715259207,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.505416030651051e-06,
+      "loss": 1.3899,
+      "step": 8418
+    },
+    {
+      "epoch": 1.4524281894246527,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.5033465623429362e-06,
+      "loss": 1.4151,
+      "step": 8419
+    },
+    {
+      "epoch": 1.4526007073233849,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.501277575347505e-06,
+      "loss": 1.4596,
+      "step": 8420
+    },
+    {
+      "epoch": 1.4527732252221168,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.4992090698180293e-06,
+      "loss": 1.3831,
+      "step": 8421
+    },
+    {
+      "epoch": 1.4529457431208488,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.4971410459077716e-06,
+      "loss": 1.4109,
+      "step": 8422
+    },
+    {
+      "epoch": 1.4531182610195807,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.4950735037699334e-06,
+      "loss": 1.3633,
+      "step": 8423
+    },
+    {
+      "epoch": 1.4532907789183127,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.493006443557696e-06,
+      "loss": 1.5085,
+      "step": 8424
+    },
+    {
+      "epoch": 1.453463296817045,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.490939865424198e-06,
+      "loss": 1.4088,
+      "step": 8425
+    },
+    {
+      "epoch": 1.4536358147157769,
+      "grad_norm": 0.625,
+      "learning_rate": 3.4888737695225416e-06,
+      "loss": 1.4707,
+      "step": 8426
+    },
+    {
+      "epoch": 1.4538083326145088,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.4868081560058066e-06,
+      "loss": 1.4511,
+      "step": 8427
+    },
+    {
+      "epoch": 1.4539808505132408,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.4847430250270165e-06,
+      "loss": 1.428,
+      "step": 8428
+    },
+    {
+      "epoch": 1.4541533684119727,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.4826783767391727e-06,
+      "loss": 1.5579,
+      "step": 8429
+    },
+    {
+      "epoch": 1.4543258863107047,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.4806142112952356e-06,
+      "loss": 1.5196,
+      "step": 8430
+    },
+    {
+      "epoch": 1.4544984042094367,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.478550528848134e-06,
+      "loss": 1.4335,
+      "step": 8431
+    },
+    {
+      "epoch": 1.4546709221081686,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.4764873295507563e-06,
+      "loss": 1.4063,
+      "step": 8432
+    },
+    {
+      "epoch": 1.4548434400069006,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.4744246135559623e-06,
+      "loss": 1.3783,
+      "step": 8433
+    },
+    {
+      "epoch": 1.4550159579056328,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.4723623810165584e-06,
+      "loss": 1.4814,
+      "step": 8434
+    },
+    {
+      "epoch": 1.4551884758043647,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.470300632085344e-06,
+      "loss": 1.4284,
+      "step": 8435
+    },
+    {
+      "epoch": 1.4553609937030967,
+      "grad_norm": 0.6796875,
+      "learning_rate": 3.4682393669150547e-06,
+      "loss": 1.4607,
+      "step": 8436
+    },
+    {
+      "epoch": 1.4555335116018286,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.466178585658405e-06,
+      "loss": 1.454,
+      "step": 8437
+    },
+    {
+      "epoch": 1.4557060295005606,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.464118288468071e-06,
+      "loss": 1.4989,
+      "step": 8438
+    },
+    {
+      "epoch": 1.4558785473992928,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.462058475496692e-06,
+      "loss": 1.4011,
+      "step": 8439
+    },
+    {
+      "epoch": 1.4560510652980247,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.459999146896873e-06,
+      "loss": 1.4045,
+      "step": 8440
+    },
+    {
+      "epoch": 1.4562235831967567,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.4579403028211835e-06,
+      "loss": 1.3958,
+      "step": 8441
+    },
+    {
+      "epoch": 1.4563961010954887,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.4558819434221456e-06,
+      "loss": 1.4219,
+      "step": 8442
+    },
+    {
+      "epoch": 1.4565686189942206,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.4538240688522684e-06,
+      "loss": 1.3098,
+      "step": 8443
+    },
+    {
+      "epoch": 1.4567411368929526,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.451766679264008e-06,
+      "loss": 1.4552,
+      "step": 8444
+    },
+    {
+      "epoch": 1.4569136547916846,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.449709774809782e-06,
+      "loss": 1.3745,
+      "step": 8445
+    },
+    {
+      "epoch": 1.4570861726904165,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.447653355641989e-06,
+      "loss": 1.4049,
+      "step": 8446
+    },
+    {
+      "epoch": 1.4572586905891487,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.445597421912974e-06,
+      "loss": 1.4325,
+      "step": 8447
+    },
+    {
+      "epoch": 1.4574312084878807,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.4435419737750566e-06,
+      "loss": 1.5556,
+      "step": 8448
+    },
+    {
+      "epoch": 1.4576037263866126,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.4414870113805156e-06,
+      "loss": 1.4107,
+      "step": 8449
+    },
+    {
+      "epoch": 1.4577762442853446,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.439432534881596e-06,
+      "loss": 1.4473,
+      "step": 8450
+    },
+    {
+      "epoch": 1.4579487621840765,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.437378544430506e-06,
+      "loss": 1.4035,
+      "step": 8451
+    },
+    {
+      "epoch": 1.4581212800828087,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.4353250401794238e-06,
+      "loss": 1.4244,
+      "step": 8452
+    },
+    {
+      "epoch": 1.4582937979815407,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.433272022280472e-06,
+      "loss": 1.508,
+      "step": 8453
+    },
+    {
+      "epoch": 1.4584663158802726,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.431219490885768e-06,
+      "loss": 1.399,
+      "step": 8454
+    },
+    {
+      "epoch": 1.4586388337790046,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.429167446147359e-06,
+      "loss": 1.473,
+      "step": 8455
+    },
+    {
+      "epoch": 1.4588113516777366,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.4271158882172904e-06,
+      "loss": 1.3995,
+      "step": 8456
+    },
+    {
+      "epoch": 1.4589838695764685,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.425064817247542e-06,
+      "loss": 1.4109,
+      "step": 8457
+    },
+    {
+      "epoch": 1.4591563874752005,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.4230142333900705e-06,
+      "loss": 1.3566,
+      "step": 8458
+    },
+    {
+      "epoch": 1.4593289053739325,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.420964136796807e-06,
+      "loss": 1.4539,
+      "step": 8459
+    },
+    {
+      "epoch": 1.4595014232726644,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.4189145276196244e-06,
+      "loss": 1.3497,
+      "step": 8460
+    },
+    {
+      "epoch": 1.4596739411713966,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.4168654060103735e-06,
+      "loss": 1.5443,
+      "step": 8461
+    },
+    {
+      "epoch": 1.4598464590701286,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.414816772120867e-06,
+      "loss": 1.4199,
+      "step": 8462
+    },
+    {
+      "epoch": 1.4600189769688605,
+      "grad_norm": 0.66015625,
+      "learning_rate": 3.4127686261028803e-06,
+      "loss": 1.4814,
+      "step": 8463
+    },
+    {
+      "epoch": 1.4601914948675925,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.410720968108153e-06,
+      "loss": 1.5426,
+      "step": 8464
+    },
+    {
+      "epoch": 1.4603640127663244,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.408673798288391e-06,
+      "loss": 1.4995,
+      "step": 8465
+    },
+    {
+      "epoch": 1.4605365306650566,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.4066271167952524e-06,
+      "loss": 1.4793,
+      "step": 8466
+    },
+    {
+      "epoch": 1.4607090485637886,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.4045809237803794e-06,
+      "loss": 1.442,
+      "step": 8467
+    },
+    {
+      "epoch": 1.4608815664625205,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.40253521939536e-06,
+      "loss": 1.5188,
+      "step": 8468
+    },
+    {
+      "epoch": 1.4610540843612525,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.4004900037917545e-06,
+      "loss": 1.4222,
+      "step": 8469
+    },
+    {
+      "epoch": 1.4612266022599845,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.398445277121084e-06,
+      "loss": 1.5028,
+      "step": 8470
+    },
+    {
+      "epoch": 1.4613991201587164,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.3964010395348355e-06,
+      "loss": 1.3889,
+      "step": 8471
+    },
+    {
+      "epoch": 1.4615716380574484,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.3943572911844603e-06,
+      "loss": 1.3595,
+      "step": 8472
+    },
+    {
+      "epoch": 1.4617441559561803,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.3923140322213744e-06,
+      "loss": 1.3079,
+      "step": 8473
+    },
+    {
+      "epoch": 1.4619166738549125,
+      "grad_norm": 0.625,
+      "learning_rate": 3.3902712627969446e-06,
+      "loss": 1.4147,
+      "step": 8474
+    },
+    {
+      "epoch": 1.4620891917536445,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.3882289830625227e-06,
+      "loss": 1.3372,
+      "step": 8475
+    },
+    {
+      "epoch": 1.4622617096523765,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.3861871931694124e-06,
+      "loss": 1.4014,
+      "step": 8476
+    },
+    {
+      "epoch": 1.4624342275511084,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.3841458932688744e-06,
+      "loss": 1.5151,
+      "step": 8477
+    },
+    {
+      "epoch": 1.4626067454498404,
+      "grad_norm": 0.625,
+      "learning_rate": 3.382105083512152e-06,
+      "loss": 1.5069,
+      "step": 8478
+    },
+    {
+      "epoch": 1.4627792633485723,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.3800647640504325e-06,
+      "loss": 1.3539,
+      "step": 8479
+    },
+    {
+      "epoch": 1.4629517812473045,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.378024935034879e-06,
+      "loss": 1.3442,
+      "step": 8480
+    },
+    {
+      "epoch": 1.4631242991460365,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.3759855966166144e-06,
+      "loss": 1.4406,
+      "step": 8481
+    },
+    {
+      "epoch": 1.4632968170447684,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.3739467489467247e-06,
+      "loss": 1.446,
+      "step": 8482
+    },
+    {
+      "epoch": 1.4634693349435004,
+      "grad_norm": 0.6640625,
+      "learning_rate": 3.371908392176262e-06,
+      "loss": 1.5256,
+      "step": 8483
+    },
+    {
+      "epoch": 1.4636418528422324,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.3698705264562426e-06,
+      "loss": 1.4524,
+      "step": 8484
+    },
+    {
+      "epoch": 1.4638143707409643,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.367833151937636e-06,
+      "loss": 1.4549,
+      "step": 8485
+    },
+    {
+      "epoch": 1.4639868886396963,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.365796268771395e-06,
+      "loss": 1.4497,
+      "step": 8486
+    },
+    {
+      "epoch": 1.4641594065384282,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.3637598771084125e-06,
+      "loss": 1.3537,
+      "step": 8487
+    },
+    {
+      "epoch": 1.4643319244371604,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.3617239770995704e-06,
+      "loss": 1.3475,
+      "step": 8488
+    },
+    {
+      "epoch": 1.4645044423358924,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.3596885688956893e-06,
+      "loss": 1.3748,
+      "step": 8489
+    },
+    {
+      "epoch": 1.4646769602346243,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.3576536526475678e-06,
+      "loss": 1.3505,
+      "step": 8490
+    },
+    {
+      "epoch": 1.4648494781333563,
+      "grad_norm": 0.625,
+      "learning_rate": 3.355619228505973e-06,
+      "loss": 1.308,
+      "step": 8491
+    },
+    {
+      "epoch": 1.4650219960320883,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.3535852966216176e-06,
+      "loss": 1.3639,
+      "step": 8492
+    },
+    {
+      "epoch": 1.4651945139308205,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.3515518571451933e-06,
+      "loss": 1.5198,
+      "step": 8493
+    },
+    {
+      "epoch": 1.4653670318295524,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.3495189102273484e-06,
+      "loss": 1.4561,
+      "step": 8494
+    },
+    {
+      "epoch": 1.4655395497282844,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.347486456018697e-06,
+      "loss": 1.3444,
+      "step": 8495
+    },
+    {
+      "epoch": 1.4657120676270163,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.3454544946698153e-06,
+      "loss": 1.4556,
+      "step": 8496
+    },
+    {
+      "epoch": 1.4658845855257483,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.3434230263312483e-06,
+      "loss": 1.4214,
+      "step": 8497
+    },
+    {
+      "epoch": 1.4660571034244803,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.3413920511534883e-06,
+      "loss": 1.4896,
+      "step": 8498
+    },
+    {
+      "epoch": 1.4662296213232122,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.3393615692870175e-06,
+      "loss": 1.3647,
+      "step": 8499
+    },
+    {
+      "epoch": 1.4664021392219442,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.3373315808822547e-06,
+      "loss": 1.4469,
+      "step": 8500
+    },
+    {
+      "epoch": 1.4664021392219442,
+      "eval_loss": 1.4072179794311523,
+      "eval_runtime": 10.9263,
+      "eval_samples_per_second": 93.719,
+      "eval_steps_per_second": 23.43,
+      "step": 8500
+    },
+    {
+      "epoch": 1.4665746571206761,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.3353020860896002e-06,
+      "loss": 1.4388,
+      "step": 8501
+    },
+    {
+      "epoch": 1.4667471750194083,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.3332730850594088e-06,
+      "loss": 1.5377,
+      "step": 8502
+    },
+    {
+      "epoch": 1.4669196929181403,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.3312445779420033e-06,
+      "loss": 1.3911,
+      "step": 8503
+    },
+    {
+      "epoch": 1.4670922108168722,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.3292165648876683e-06,
+      "loss": 1.4262,
+      "step": 8504
+    },
+    {
+      "epoch": 1.4672647287156042,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.3271890460466537e-06,
+      "loss": 1.417,
+      "step": 8505
+    },
+    {
+      "epoch": 1.4674372466143362,
+      "grad_norm": 0.68359375,
+      "learning_rate": 3.3251620215691614e-06,
+      "loss": 1.3963,
+      "step": 8506
+    },
+    {
+      "epoch": 1.4676097645130683,
+      "grad_norm": 0.640625,
+      "learning_rate": 3.323135491605377e-06,
+      "loss": 1.4469,
+      "step": 8507
+    },
+    {
+      "epoch": 1.4677822824118003,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.321109456305438e-06,
+      "loss": 1.5031,
+      "step": 8508
+    },
+    {
+      "epoch": 1.4679548003105323,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.3190839158194344e-06,
+      "loss": 1.3735,
+      "step": 8509
+    },
+    {
+      "epoch": 1.4681273182092642,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.317058870297446e-06,
+      "loss": 1.4582,
+      "step": 8510
+    },
+    {
+      "epoch": 1.4682998361079962,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.3150343198894897e-06,
+      "loss": 1.3811,
+      "step": 8511
+    },
+    {
+      "epoch": 1.4684723540067282,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.3130102647455608e-06,
+      "loss": 1.4364,
+      "step": 8512
+    },
+    {
+      "epoch": 1.4686448719054601,
+      "grad_norm": 0.640625,
+      "learning_rate": 3.310986705015613e-06,
+      "loss": 1.3751,
+      "step": 8513
+    },
+    {
+      "epoch": 1.468817389804192,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.3089636408495662e-06,
+      "loss": 1.504,
+      "step": 8514
+    },
+    {
+      "epoch": 1.4689899077029243,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.3069410723972995e-06,
+      "loss": 1.5245,
+      "step": 8515
+    },
+    {
+      "epoch": 1.4691624256016562,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.3049189998086584e-06,
+      "loss": 1.4545,
+      "step": 8516
+    },
+    {
+      "epoch": 1.4693349435003882,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.302897423233451e-06,
+      "loss": 1.3767,
+      "step": 8517
+    },
+    {
+      "epoch": 1.4695074613991201,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.300876342821451e-06,
+      "loss": 1.3456,
+      "step": 8518
+    },
+    {
+      "epoch": 1.469679979297852,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.2988557587223837e-06,
+      "loss": 1.4042,
+      "step": 8519
+    },
+    {
+      "epoch": 1.469852497196584,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.296835671085957e-06,
+      "loss": 1.4078,
+      "step": 8520
+    },
+    {
+      "epoch": 1.4700250150953162,
+      "grad_norm": 0.6640625,
+      "learning_rate": 3.2948160800618254e-06,
+      "loss": 1.3702,
+      "step": 8521
+    },
+    {
+      "epoch": 1.4701975329940482,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.292796985799611e-06,
+      "loss": 1.4343,
+      "step": 8522
+    },
+    {
+      "epoch": 1.4703700508927802,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.290778388448912e-06,
+      "loss": 1.4191,
+      "step": 8523
+    },
+    {
+      "epoch": 1.4705425687915121,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.288760288159266e-06,
+      "loss": 1.3871,
+      "step": 8524
+    },
+    {
+      "epoch": 1.470715086690244,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.2867426850801932e-06,
+      "loss": 1.4247,
+      "step": 8525
+    },
+    {
+      "epoch": 1.470887604588976,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.2847255793611674e-06,
+      "loss": 1.3644,
+      "step": 8526
+    },
+    {
+      "epoch": 1.471060122487708,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.2827089711516303e-06,
+      "loss": 1.4524,
+      "step": 8527
+    },
+    {
+      "epoch": 1.47123264038644,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.2806928606009836e-06,
+      "loss": 1.4674,
+      "step": 8528
+    },
+    {
+      "epoch": 1.4714051582851722,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.278677247858598e-06,
+      "loss": 1.4114,
+      "step": 8529
+    },
+    {
+      "epoch": 1.4715776761839041,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.2766621330737923e-06,
+      "loss": 1.3431,
+      "step": 8530
+    },
+    {
+      "epoch": 1.471750194082636,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.2746475163958714e-06,
+      "loss": 1.4278,
+      "step": 8531
+    },
+    {
+      "epoch": 1.471922711981368,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.272633397974081e-06,
+      "loss": 1.4383,
+      "step": 8532
+    },
+    {
+      "epoch": 1.4720952298801,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.270619777957642e-06,
+      "loss": 1.4291,
+      "step": 8533
+    },
+    {
+      "epoch": 1.4722677477788322,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.2686066564957387e-06,
+      "loss": 1.3628,
+      "step": 8534
+    },
+    {
+      "epoch": 1.4724402656775641,
+      "grad_norm": 0.640625,
+      "learning_rate": 3.2665940337375125e-06,
+      "loss": 1.4406,
+      "step": 8535
+    },
+    {
+      "epoch": 1.472612783576296,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.264581909832073e-06,
+      "loss": 1.4555,
+      "step": 8536
+    },
+    {
+      "epoch": 1.472785301475028,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.2625702849284947e-06,
+      "loss": 1.4099,
+      "step": 8537
+    },
+    {
+      "epoch": 1.47295781937376,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.2605591591758e-06,
+      "loss": 1.4453,
+      "step": 8538
+    },
+    {
+      "epoch": 1.473130337272492,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.258548532722995e-06,
+      "loss": 1.4416,
+      "step": 8539
+    },
+    {
+      "epoch": 1.473302855171224,
+      "grad_norm": 0.66015625,
+      "learning_rate": 3.256538405719042e-06,
+      "loss": 1.3456,
+      "step": 8540
+    },
+    {
+      "epoch": 1.473475373069956,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.254528778312852e-06,
+      "loss": 1.4016,
+      "step": 8541
+    },
+    {
+      "epoch": 1.4736478909686879,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.2525196506533243e-06,
+      "loss": 1.4421,
+      "step": 8542
+    },
+    {
+      "epoch": 1.47382040886742,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.250511022889293e-06,
+      "loss": 1.4036,
+      "step": 8543
+    },
+    {
+      "epoch": 1.473992926766152,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.248502895169586e-06,
+      "loss": 1.3537,
+      "step": 8544
+    },
+    {
+      "epoch": 1.474165444664884,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.246495267642967e-06,
+      "loss": 1.5216,
+      "step": 8545
+    },
+    {
+      "epoch": 1.474337962563616,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.244488140458175e-06,
+      "loss": 1.4198,
+      "step": 8546
+    },
+    {
+      "epoch": 1.474510480462348,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.2424815137639132e-06,
+      "loss": 1.3578,
+      "step": 8547
+    },
+    {
+      "epoch": 1.47468299836108,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.2404753877088437e-06,
+      "loss": 1.4625,
+      "step": 8548
+    },
+    {
+      "epoch": 1.474855516259812,
+      "grad_norm": 0.66015625,
+      "learning_rate": 3.2384697624415915e-06,
+      "loss": 1.4955,
+      "step": 8549
+    },
+    {
+      "epoch": 1.475028034158544,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.236464638110752e-06,
+      "loss": 1.3116,
+      "step": 8550
+    },
+    {
+      "epoch": 1.475200552057276,
+      "grad_norm": 0.68359375,
+      "learning_rate": 3.2344600148648652e-06,
+      "loss": 1.4343,
+      "step": 8551
+    },
+    {
+      "epoch": 1.475373069956008,
+      "grad_norm": 0.66015625,
+      "learning_rate": 3.232455892852461e-06,
+      "loss": 1.4746,
+      "step": 8552
+    },
+    {
+      "epoch": 1.4755455878547399,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.230452272222007e-06,
+      "loss": 1.3621,
+      "step": 8553
+    },
+    {
+      "epoch": 1.4757181057534718,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.228449153121942e-06,
+      "loss": 1.3907,
+      "step": 8554
+    },
+    {
+      "epoch": 1.4758906236522038,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.2264465357006814e-06,
+      "loss": 1.5396,
+      "step": 8555
+    },
+    {
+      "epoch": 1.476063141550936,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.2244444201065815e-06,
+      "loss": 1.4031,
+      "step": 8556
+    },
+    {
+      "epoch": 1.476235659449668,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.2224428064879743e-06,
+      "loss": 1.3998,
+      "step": 8557
+    },
+    {
+      "epoch": 1.4764081773484,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.2204416949931516e-06,
+      "loss": 1.4665,
+      "step": 8558
+    },
+    {
+      "epoch": 1.4765806952471319,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.2184410857703695e-06,
+      "loss": 1.4619,
+      "step": 8559
+    },
+    {
+      "epoch": 1.4767532131458638,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.2164409789678453e-06,
+      "loss": 1.4198,
+      "step": 8560
+    },
+    {
+      "epoch": 1.476925731044596,
+      "grad_norm": 0.625,
+      "learning_rate": 3.2144413747337622e-06,
+      "loss": 1.4841,
+      "step": 8561
+    },
+    {
+      "epoch": 1.477098248943328,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.212442273216253e-06,
+      "loss": 1.3529,
+      "step": 8562
+    },
+    {
+      "epoch": 1.47727076684206,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.210443674563437e-06,
+      "loss": 1.4495,
+      "step": 8563
+    },
+    {
+      "epoch": 1.477443284740792,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.2084455789233748e-06,
+      "loss": 1.3776,
+      "step": 8564
+    },
+    {
+      "epoch": 1.4776158026395239,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.206447986444099e-06,
+      "loss": 1.4294,
+      "step": 8565
+    },
+    {
+      "epoch": 1.4777883205382558,
+      "grad_norm": 0.625,
+      "learning_rate": 3.2044508972736044e-06,
+      "loss": 1.4277,
+      "step": 8566
+    },
+    {
+      "epoch": 1.4779608384369878,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.202454311559847e-06,
+      "loss": 1.4677,
+      "step": 8567
+    },
+    {
+      "epoch": 1.4781333563357197,
+      "grad_norm": 0.640625,
+      "learning_rate": 3.200458229450749e-06,
+      "loss": 1.3658,
+      "step": 8568
+    },
+    {
+      "epoch": 1.4783058742344517,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.198462651094193e-06,
+      "loss": 1.3626,
+      "step": 8569
+    },
+    {
+      "epoch": 1.4784783921331839,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.1964675766380148e-06,
+      "loss": 1.4493,
+      "step": 8570
+    },
+    {
+      "epoch": 1.4786509100319158,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.194473006230033e-06,
+      "loss": 1.4275,
+      "step": 8571
+    },
+    {
+      "epoch": 1.4788234279306478,
+      "grad_norm": 0.640625,
+      "learning_rate": 3.192478940018018e-06,
+      "loss": 1.4172,
+      "step": 8572
+    },
+    {
+      "epoch": 1.4789959458293798,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.19048537814969e-06,
+      "loss": 1.404,
+      "step": 8573
+    },
+    {
+      "epoch": 1.4791684637281117,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.18849232077276e-06,
+      "loss": 1.4071,
+      "step": 8574
+    },
+    {
+      "epoch": 1.479340981626844,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.186499768034874e-06,
+      "loss": 1.4131,
+      "step": 8575
+    },
+    {
+      "epoch": 1.4795134995255759,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.1845077200836638e-06,
+      "loss": 1.4327,
+      "step": 8576
+    },
+    {
+      "epoch": 1.4796860174243078,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.1825161770667023e-06,
+      "loss": 1.4534,
+      "step": 8577
+    },
+    {
+      "epoch": 1.4798585353230398,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.180525139131542e-06,
+      "loss": 1.432,
+      "step": 8578
+    },
+    {
+      "epoch": 1.4800310532217718,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.1785346064256884e-06,
+      "loss": 1.4916,
+      "step": 8579
+    },
+    {
+      "epoch": 1.4802035711205037,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.176544579096613e-06,
+      "loss": 1.4756,
+      "step": 8580
+    },
+    {
+      "epoch": 1.4803760890192357,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.1745550572917503e-06,
+      "loss": 1.378,
+      "step": 8581
+    },
+    {
+      "epoch": 1.4805486069179676,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.1725660411585e-06,
+      "loss": 1.4055,
+      "step": 8582
+    },
+    {
+      "epoch": 1.4807211248166996,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.170577530844211e-06,
+      "loss": 1.437,
+      "step": 8583
+    },
+    {
+      "epoch": 1.4808936427154318,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.1685895264962173e-06,
+      "loss": 1.4833,
+      "step": 8584
+    },
+    {
+      "epoch": 1.4810661606141637,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.1666020282617925e-06,
+      "loss": 1.4276,
+      "step": 8585
+    },
+    {
+      "epoch": 1.4812386785128957,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.1646150362881835e-06,
+      "loss": 1.4181,
+      "step": 8586
+    },
+    {
+      "epoch": 1.4814111964116277,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.1626285507226072e-06,
+      "loss": 1.4481,
+      "step": 8587
+    },
+    {
+      "epoch": 1.4815837143103596,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.160642571712228e-06,
+      "loss": 1.4209,
+      "step": 8588
+    },
+    {
+      "epoch": 1.4817562322090918,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.158657099404181e-06,
+      "loss": 1.4144,
+      "step": 8589
+    },
+    {
+      "epoch": 1.4819287501078238,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.156672133945563e-06,
+      "loss": 1.4058,
+      "step": 8590
+    },
+    {
+      "epoch": 1.4821012680065557,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.154687675483432e-06,
+      "loss": 1.3381,
+      "step": 8591
+    },
+    {
+      "epoch": 1.4822737859052877,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.1527037241648096e-06,
+      "loss": 1.3612,
+      "step": 8592
+    },
+    {
+      "epoch": 1.4824463038040196,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.150720280136682e-06,
+      "loss": 1.3687,
+      "step": 8593
+    },
+    {
+      "epoch": 1.4826188217027516,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.1487373435459865e-06,
+      "loss": 1.5162,
+      "step": 8594
+    },
+    {
+      "epoch": 1.4827913396014836,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.1467549145396437e-06,
+      "loss": 1.2491,
+      "step": 8595
+    },
+    {
+      "epoch": 1.4829638575002155,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.1447729932645145e-06,
+      "loss": 1.4794,
+      "step": 8596
+    },
+    {
+      "epoch": 1.4831363753989477,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.142791579867435e-06,
+      "loss": 1.3974,
+      "step": 8597
+    },
+    {
+      "epoch": 1.4833088932976797,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.1408106744952018e-06,
+      "loss": 1.3553,
+      "step": 8598
+    },
+    {
+      "epoch": 1.4834814111964116,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.1388302772945713e-06,
+      "loss": 1.5289,
+      "step": 8599
+    },
+    {
+      "epoch": 1.4836539290951436,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.136850388412265e-06,
+      "loss": 1.4299,
+      "step": 8600
+    },
+    {
+      "epoch": 1.4836539290951436,
+      "eval_loss": 1.4071738719940186,
+      "eval_runtime": 10.7878,
+      "eval_samples_per_second": 94.922,
+      "eval_steps_per_second": 23.731,
+      "step": 8600
+    },
+    {
+      "epoch": 1.4838264469938756,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.1348710079949675e-06,
+      "loss": 1.4985,
+      "step": 8601
+    },
+    {
+      "epoch": 1.4839989648926077,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.1328921361893148e-06,
+      "loss": 1.4663,
+      "step": 8602
+    },
+    {
+      "epoch": 1.4841714827913397,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.1309137731419236e-06,
+      "loss": 1.3719,
+      "step": 8603
+    },
+    {
+      "epoch": 1.4843440006900717,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.1289359189993607e-06,
+      "loss": 1.5156,
+      "step": 8604
+    },
+    {
+      "epoch": 1.4845165185888036,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.1269585739081564e-06,
+      "loss": 1.373,
+      "step": 8605
+    },
+    {
+      "epoch": 1.4846890364875356,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.1249817380148097e-06,
+      "loss": 1.4311,
+      "step": 8606
+    },
+    {
+      "epoch": 1.4848615543862675,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.123005411465766e-06,
+      "loss": 1.4612,
+      "step": 8607
+    },
+    {
+      "epoch": 1.4850340722849995,
+      "grad_norm": 0.640625,
+      "learning_rate": 3.1210295944074574e-06,
+      "loss": 1.4437,
+      "step": 8608
+    },
+    {
+      "epoch": 1.4852065901837315,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.119054286986255e-06,
+      "loss": 1.4048,
+      "step": 8609
+    },
+    {
+      "epoch": 1.4853791080824634,
+      "grad_norm": 0.6640625,
+      "learning_rate": 3.1170794893485047e-06,
+      "loss": 1.403,
+      "step": 8610
+    },
+    {
+      "epoch": 1.4855516259811956,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.1151052016405125e-06,
+      "loss": 1.3551,
+      "step": 8611
+    },
+    {
+      "epoch": 1.4857241438799276,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.1131314240085465e-06,
+      "loss": 1.4862,
+      "step": 8612
+    },
+    {
+      "epoch": 1.4858966617786595,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.111158156598836e-06,
+      "loss": 1.3378,
+      "step": 8613
+    },
+    {
+      "epoch": 1.4860691796773915,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.109185399557575e-06,
+      "loss": 1.413,
+      "step": 8614
+    },
+    {
+      "epoch": 1.4862416975761235,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.107213153030909e-06,
+      "loss": 1.4823,
+      "step": 8615
+    },
+    {
+      "epoch": 1.4864142154748556,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.105241417164967e-06,
+      "loss": 1.3963,
+      "step": 8616
+    },
+    {
+      "epoch": 1.4865867333735876,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.1032701921058184e-06,
+      "loss": 1.3688,
+      "step": 8617
+    },
+    {
+      "epoch": 1.4867592512723196,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.1012994779995077e-06,
+      "loss": 1.3553,
+      "step": 8618
+    },
+    {
+      "epoch": 1.4869317691710515,
+      "grad_norm": 0.66796875,
+      "learning_rate": 3.0993292749920355e-06,
+      "loss": 1.4014,
+      "step": 8619
+    },
+    {
+      "epoch": 1.4871042870697835,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.097359583229368e-06,
+      "loss": 1.5023,
+      "step": 8620
+    },
+    {
+      "epoch": 1.4872768049685154,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.095390402857432e-06,
+      "loss": 1.2465,
+      "step": 8621
+    },
+    {
+      "epoch": 1.4874493228672474,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.0934217340221183e-06,
+      "loss": 1.4298,
+      "step": 8622
+    },
+    {
+      "epoch": 1.4876218407659794,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.0914535768692753e-06,
+      "loss": 1.3638,
+      "step": 8623
+    },
+    {
+      "epoch": 1.4877943586647115,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.089485931544719e-06,
+      "loss": 1.4358,
+      "step": 8624
+    },
+    {
+      "epoch": 1.4879668765634435,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.0875187981942266e-06,
+      "loss": 1.3866,
+      "step": 8625
+    },
+    {
+      "epoch": 1.4881393944621755,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.0855521769635266e-06,
+      "loss": 1.4508,
+      "step": 8626
+    },
+    {
+      "epoch": 1.4883119123609074,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.0835860679983308e-06,
+      "loss": 1.3201,
+      "step": 8627
+    },
+    {
+      "epoch": 1.4884844302596394,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.081620471444292e-06,
+      "loss": 1.4419,
+      "step": 8628
+    },
+    {
+      "epoch": 1.4886569481583714,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.079655387447037e-06,
+      "loss": 1.3916,
+      "step": 8629
+    },
+    {
+      "epoch": 1.4888294660571035,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.077690816152151e-06,
+      "loss": 1.3884,
+      "step": 8630
+    },
+    {
+      "epoch": 1.4890019839558355,
+      "grad_norm": 0.66796875,
+      "learning_rate": 3.075726757705182e-06,
+      "loss": 1.4443,
+      "step": 8631
+    },
+    {
+      "epoch": 1.4891745018545675,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.073763212251639e-06,
+      "loss": 1.3651,
+      "step": 8632
+    },
+    {
+      "epoch": 1.4893470197532994,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.071800179936998e-06,
+      "loss": 1.4278,
+      "step": 8633
+    },
+    {
+      "epoch": 1.4895195376520314,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.0698376609066828e-06,
+      "loss": 1.48,
+      "step": 8634
+    },
+    {
+      "epoch": 1.4896920555507633,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.0678756553060984e-06,
+      "loss": 1.3465,
+      "step": 8635
+    },
+    {
+      "epoch": 1.4898645734494953,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.0659141632805987e-06,
+      "loss": 1.3635,
+      "step": 8636
+    },
+    {
+      "epoch": 1.4900370913482273,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.0639531849755044e-06,
+      "loss": 1.3396,
+      "step": 8637
+    },
+    {
+      "epoch": 1.4902096092469594,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.0619927205360998e-06,
+      "loss": 1.4531,
+      "step": 8638
+    },
+    {
+      "epoch": 1.4903821271456914,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.0600327701076193e-06,
+      "loss": 1.4887,
+      "step": 8639
+    },
+    {
+      "epoch": 1.4905546450444234,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.058073333835281e-06,
+      "loss": 1.4059,
+      "step": 8640
+    },
+    {
+      "epoch": 1.4907271629431553,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.0561144118642406e-06,
+      "loss": 1.3808,
+      "step": 8641
+    },
+    {
+      "epoch": 1.4908996808418873,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.0541560043396322e-06,
+      "loss": 1.5228,
+      "step": 8642
+    },
+    {
+      "epoch": 1.4910721987406195,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.052198111406547e-06,
+      "loss": 1.3432,
+      "step": 8643
+    },
+    {
+      "epoch": 1.4912447166393514,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.0502407332100382e-06,
+      "loss": 1.4385,
+      "step": 8644
+    },
+    {
+      "epoch": 1.4914172345380834,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.048283869895119e-06,
+      "loss": 1.4864,
+      "step": 8645
+    },
+    {
+      "epoch": 1.4915897524368154,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.04632752160677e-06,
+      "loss": 1.3705,
+      "step": 8646
+    },
+    {
+      "epoch": 1.4917622703355473,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.044371688489921e-06,
+      "loss": 1.4323,
+      "step": 8647
+    },
+    {
+      "epoch": 1.4919347882342793,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.042416370689485e-06,
+      "loss": 1.4554,
+      "step": 8648
+    },
+    {
+      "epoch": 1.4921073061330112,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.0404615683503136e-06,
+      "loss": 1.4668,
+      "step": 8649
+    },
+    {
+      "epoch": 1.4922798240317432,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.0385072816172344e-06,
+      "loss": 1.3353,
+      "step": 8650
+    },
+    {
+      "epoch": 1.4924523419304752,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.036553510635033e-06,
+      "loss": 1.3784,
+      "step": 8651
+    },
+    {
+      "epoch": 1.4926248598292073,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.0346002555484567e-06,
+      "loss": 1.469,
+      "step": 8652
+    },
+    {
+      "epoch": 1.4927973777279393,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.0326475165022164e-06,
+      "loss": 1.3914,
+      "step": 8653
+    },
+    {
+      "epoch": 1.4929698956266713,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.030695293640982e-06,
+      "loss": 1.3753,
+      "step": 8654
+    },
+    {
+      "epoch": 1.4931424135254032,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.028743587109385e-06,
+      "loss": 1.4428,
+      "step": 8655
+    },
+    {
+      "epoch": 1.4933149314241352,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.026792397052023e-06,
+      "loss": 1.4154,
+      "step": 8656
+    },
+    {
+      "epoch": 1.4934874493228674,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.024841723613453e-06,
+      "loss": 1.4908,
+      "step": 8657
+    },
+    {
+      "epoch": 1.4936599672215993,
+      "grad_norm": 0.66796875,
+      "learning_rate": 3.022891566938185e-06,
+      "loss": 1.2884,
+      "step": 8658
+    },
+    {
+      "epoch": 1.4938324851203313,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.0209419271707118e-06,
+      "loss": 1.4292,
+      "step": 8659
+    },
+    {
+      "epoch": 1.4940050030190632,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.018992804455464e-06,
+      "loss": 1.4774,
+      "step": 8660
+    },
+    {
+      "epoch": 1.4941775209177952,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.017044198936848e-06,
+      "loss": 1.4197,
+      "step": 8661
+    },
+    {
+      "epoch": 1.4943500388165272,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.01509611075923e-06,
+      "loss": 1.3863,
+      "step": 8662
+    },
+    {
+      "epoch": 1.4945225567152591,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.013148540066936e-06,
+      "loss": 1.5052,
+      "step": 8663
+    },
+    {
+      "epoch": 1.494695074613991,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.011201487004254e-06,
+      "loss": 1.3792,
+      "step": 8664
+    },
+    {
+      "epoch": 1.4948675925127233,
+      "grad_norm": 0.671875,
+      "learning_rate": 3.0092549517154336e-06,
+      "loss": 1.4096,
+      "step": 8665
+    },
+    {
+      "epoch": 1.4950401104114552,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.007308934344686e-06,
+      "loss": 1.4694,
+      "step": 8666
+    },
+    {
+      "epoch": 1.4952126283101872,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.005363435036186e-06,
+      "loss": 1.408,
+      "step": 8667
+    },
+    {
+      "epoch": 1.4953851462089192,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.0034184539340663e-06,
+      "loss": 1.4432,
+      "step": 8668
+    },
+    {
+      "epoch": 1.4955576641076511,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.0014739911824244e-06,
+      "loss": 1.4187,
+      "step": 8669
+    },
+    {
+      "epoch": 1.495730182006383,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.999530046925322e-06,
+      "loss": 1.4553,
+      "step": 8670
+    },
+    {
+      "epoch": 1.4959026999051153,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.9975866213067673e-06,
+      "loss": 1.4767,
+      "step": 8671
+    },
+    {
+      "epoch": 1.4960752178038472,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.9956437144707552e-06,
+      "loss": 1.4951,
+      "step": 8672
+    },
+    {
+      "epoch": 1.4962477357025792,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.9937013265612192e-06,
+      "loss": 1.4536,
+      "step": 8673
+    },
+    {
+      "epoch": 1.4964202536013111,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.9917594577220665e-06,
+      "loss": 1.4156,
+      "step": 8674
+    },
+    {
+      "epoch": 1.496592771500043,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.989818108097162e-06,
+      "loss": 1.4029,
+      "step": 8675
+    },
+    {
+      "epoch": 1.496765289398775,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.9878772778303344e-06,
+      "loss": 1.4684,
+      "step": 8676
+    },
+    {
+      "epoch": 1.496937807297507,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.9859369670653715e-06,
+      "loss": 1.4403,
+      "step": 8677
+    },
+    {
+      "epoch": 1.497110325196239,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.983997175946027e-06,
+      "loss": 1.4305,
+      "step": 8678
+    },
+    {
+      "epoch": 1.4972828430949712,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.982057904616004e-06,
+      "loss": 1.4958,
+      "step": 8679
+    },
+    {
+      "epoch": 1.4974553609937031,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.9801191532189876e-06,
+      "loss": 1.4899,
+      "step": 8680
+    },
+    {
+      "epoch": 1.497627878892435,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.9781809218986036e-06,
+      "loss": 1.4251,
+      "step": 8681
+    },
+    {
+      "epoch": 1.497800396791167,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.9762432107984508e-06,
+      "loss": 1.3534,
+      "step": 8682
+    },
+    {
+      "epoch": 1.497972914689899,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.974306020062088e-06,
+      "loss": 1.3928,
+      "step": 8683
+    },
+    {
+      "epoch": 1.4981454325886312,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.972369349833033e-06,
+      "loss": 1.3605,
+      "step": 8684
+    },
+    {
+      "epoch": 1.4983179504873632,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.9704332002547677e-06,
+      "loss": 1.3593,
+      "step": 8685
+    },
+    {
+      "epoch": 1.4984904683860951,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.9684975714707333e-06,
+      "loss": 1.3329,
+      "step": 8686
+    },
+    {
+      "epoch": 1.498662986284827,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.966562463624334e-06,
+      "loss": 1.468,
+      "step": 8687
+    },
+    {
+      "epoch": 1.498835504183559,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.9646278768589345e-06,
+      "loss": 1.4801,
+      "step": 8688
+    },
+    {
+      "epoch": 1.499008022082291,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.962693811317863e-06,
+      "loss": 1.5002,
+      "step": 8689
+    },
+    {
+      "epoch": 1.499180539981023,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.9607602671443993e-06,
+      "loss": 1.3173,
+      "step": 8690
+    },
+    {
+      "epoch": 1.499353057879755,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.9588272444818056e-06,
+      "loss": 1.4184,
+      "step": 8691
+    },
+    {
+      "epoch": 1.4995255757784869,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.9568947434732777e-06,
+      "loss": 1.4755,
+      "step": 8692
+    },
+    {
+      "epoch": 1.499698093677219,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.9549627642620005e-06,
+      "loss": 1.344,
+      "step": 8693
+    },
+    {
+      "epoch": 1.499870611575951,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.9530313069910986e-06,
+      "loss": 1.4186,
+      "step": 8694
+    },
+    {
+      "epoch": 1.500043129474683,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.951100371803669e-06,
+      "loss": 1.3953,
+      "step": 8695
+    },
+    {
+      "epoch": 1.500215647373415,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.949169958842767e-06,
+      "loss": 1.4259,
+      "step": 8696
+    },
+    {
+      "epoch": 1.5003881652721471,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.9472400682514104e-06,
+      "loss": 1.4243,
+      "step": 8697
+    },
+    {
+      "epoch": 1.500560683170879,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.945310700172577e-06,
+      "loss": 1.4715,
+      "step": 8698
+    },
+    {
+      "epoch": 1.500733201069611,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.9433818547492067e-06,
+      "loss": 1.4727,
+      "step": 8699
+    },
+    {
+      "epoch": 1.500905718968343,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.941453532124201e-06,
+      "loss": 1.4355,
+      "step": 8700
+    },
+    {
+      "epoch": 1.500905718968343,
+      "eval_loss": 1.4072085618972778,
+      "eval_runtime": 10.8369,
+      "eval_samples_per_second": 94.492,
+      "eval_steps_per_second": 23.623,
+      "step": 8700
+    },
+    {
+      "epoch": 1.501078236867075,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.9395257324404204e-06,
+      "loss": 1.5278,
+      "step": 8701
+    },
+    {
+      "epoch": 1.501250754765807,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.9375984558406934e-06,
+      "loss": 1.4436,
+      "step": 8702
+    },
+    {
+      "epoch": 1.501423272664539,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.935671702467794e-06,
+      "loss": 1.4489,
+      "step": 8703
+    },
+    {
+      "epoch": 1.5015957905632709,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.933745472464481e-06,
+      "loss": 1.4356,
+      "step": 8704
+    },
+    {
+      "epoch": 1.5017683084620028,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.9318197659734527e-06,
+      "loss": 1.3582,
+      "step": 8705
+    },
+    {
+      "epoch": 1.5019408263607348,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.9298945831373803e-06,
+      "loss": 1.4326,
+      "step": 8706
+    },
+    {
+      "epoch": 1.502113344259467,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.9279699240988936e-06,
+      "loss": 1.4332,
+      "step": 8707
+    },
+    {
+      "epoch": 1.502285862158199,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.9260457890005823e-06,
+      "loss": 1.3855,
+      "step": 8708
+    },
+    {
+      "epoch": 1.5024583800569309,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.924122177984998e-06,
+      "loss": 1.4103,
+      "step": 8709
+    },
+    {
+      "epoch": 1.5026308979556628,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.9221990911946595e-06,
+      "loss": 1.4533,
+      "step": 8710
+    },
+    {
+      "epoch": 1.502803415854395,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.92027652877203e-06,
+      "loss": 1.5407,
+      "step": 8711
+    },
+    {
+      "epoch": 1.502975933753127,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.9183544908595573e-06,
+      "loss": 1.4197,
+      "step": 8712
+    },
+    {
+      "epoch": 1.503148451651859,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.9164329775996293e-06,
+      "loss": 1.4315,
+      "step": 8713
+    },
+    {
+      "epoch": 1.503320969550591,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.9145119891346062e-06,
+      "loss": 1.4136,
+      "step": 8714
+    },
+    {
+      "epoch": 1.5034934874493229,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.912591525606807e-06,
+      "loss": 1.3841,
+      "step": 8715
+    },
+    {
+      "epoch": 1.5036660053480548,
+      "grad_norm": 1.15625,
+      "learning_rate": 2.9106715871585124e-06,
+      "loss": 1.3314,
+      "step": 8716
+    },
+    {
+      "epoch": 1.5038385232467868,
+      "grad_norm": 1.5390625,
+      "learning_rate": 2.9087521739319624e-06,
+      "loss": 1.4136,
+      "step": 8717
+    },
+    {
+      "epoch": 1.5040110411455188,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.9068332860693594e-06,
+      "loss": 1.4267,
+      "step": 8718
+    },
+    {
+      "epoch": 1.5041835590442507,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.904914923712867e-06,
+      "loss": 1.3874,
+      "step": 8719
+    },
+    {
+      "epoch": 1.5043560769429827,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.902997087004609e-06,
+      "loss": 1.415,
+      "step": 8720
+    },
+    {
+      "epoch": 1.5045285948417149,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.901079776086674e-06,
+      "loss": 1.3695,
+      "step": 8721
+    },
+    {
+      "epoch": 1.5047011127404468,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.8991629911011e-06,
+      "loss": 1.4224,
+      "step": 8722
+    },
+    {
+      "epoch": 1.5048736306391788,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.8972467321899045e-06,
+      "loss": 1.4014,
+      "step": 8723
+    },
+    {
+      "epoch": 1.505046148537911,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.895330999495045e-06,
+      "loss": 1.4829,
+      "step": 8724
+    },
+    {
+      "epoch": 1.505218666436643,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.893415793158464e-06,
+      "loss": 1.4623,
+      "step": 8725
+    },
+    {
+      "epoch": 1.5053911843353749,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.891501113322042e-06,
+      "loss": 1.5667,
+      "step": 8726
+    },
+    {
+      "epoch": 1.5055637022341068,
+      "grad_norm": 0.625,
+      "learning_rate": 2.8895869601276326e-06,
+      "loss": 1.453,
+      "step": 8727
+    },
+    {
+      "epoch": 1.5057362201328388,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.8876733337170503e-06,
+      "loss": 1.4474,
+      "step": 8728
+    },
+    {
+      "epoch": 1.5059087380315708,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.8857602342320666e-06,
+      "loss": 1.4552,
+      "step": 8729
+    },
+    {
+      "epoch": 1.5060812559303027,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.883847661814416e-06,
+      "loss": 1.4481,
+      "step": 8730
+    },
+    {
+      "epoch": 1.5062537738290347,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.8819356166057953e-06,
+      "loss": 1.419,
+      "step": 8731
+    },
+    {
+      "epoch": 1.5064262917277667,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.88002409874786e-06,
+      "loss": 1.38,
+      "step": 8732
+    },
+    {
+      "epoch": 1.5065988096264986,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.8781131083822267e-06,
+      "loss": 1.3751,
+      "step": 8733
+    },
+    {
+      "epoch": 1.5067713275252308,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.8762026456504767e-06,
+      "loss": 1.4008,
+      "step": 8734
+    },
+    {
+      "epoch": 1.5069438454239628,
+      "grad_norm": 1.1484375,
+      "learning_rate": 2.874292710694141e-06,
+      "loss": 1.4236,
+      "step": 8735
+    },
+    {
+      "epoch": 1.5071163633226947,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.87238330365473e-06,
+      "loss": 1.3983,
+      "step": 8736
+    },
+    {
+      "epoch": 1.5072888812214267,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.8704744246736972e-06,
+      "loss": 1.4195,
+      "step": 8737
+    },
+    {
+      "epoch": 1.5074613991201589,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.868566073892465e-06,
+      "loss": 1.5079,
+      "step": 8738
+    },
+    {
+      "epoch": 1.5076339170188908,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.8666582514524175e-06,
+      "loss": 1.4868,
+      "step": 8739
+    },
+    {
+      "epoch": 1.5078064349176228,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.8647509574948997e-06,
+      "loss": 1.4152,
+      "step": 8740
+    },
+    {
+      "epoch": 1.5079789528163547,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.8628441921612117e-06,
+      "loss": 1.4388,
+      "step": 8741
+    },
+    {
+      "epoch": 1.5081514707150867,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.8609379555926255e-06,
+      "loss": 1.4466,
+      "step": 8742
+    },
+    {
+      "epoch": 1.5083239886138187,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.8590322479303554e-06,
+      "loss": 1.3899,
+      "step": 8743
+    },
+    {
+      "epoch": 1.5084965065125506,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.8571270693156028e-06,
+      "loss": 1.5198,
+      "step": 8744
+    },
+    {
+      "epoch": 1.5086690244112826,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.855222419889503e-06,
+      "loss": 1.4558,
+      "step": 8745
+    },
+    {
+      "epoch": 1.5088415423100145,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.8533182997931695e-06,
+      "loss": 1.4012,
+      "step": 8746
+    },
+    {
+      "epoch": 1.5090140602087465,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.8514147091676713e-06,
+      "loss": 1.4706,
+      "step": 8747
+    },
+    {
+      "epoch": 1.5091865781074787,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.8495116481540375e-06,
+      "loss": 1.4089,
+      "step": 8748
+    },
+    {
+      "epoch": 1.5093590960062107,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.8476091168932594e-06,
+      "loss": 1.4138,
+      "step": 8749
+    },
+    {
+      "epoch": 1.5095316139049426,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.8457071155262885e-06,
+      "loss": 1.3905,
+      "step": 8750
+    },
+    {
+      "epoch": 1.5097041318036746,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.8438056441940353e-06,
+      "loss": 1.4285,
+      "step": 8751
+    },
+    {
+      "epoch": 1.5098766497024068,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.8419047030373746e-06,
+      "loss": 1.4192,
+      "step": 8752
+    },
+    {
+      "epoch": 1.5100491676011387,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.8400042921971395e-06,
+      "loss": 1.3904,
+      "step": 8753
+    },
+    {
+      "epoch": 1.5102216854998707,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.8381044118141244e-06,
+      "loss": 1.5384,
+      "step": 8754
+    },
+    {
+      "epoch": 1.5103942033986026,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.836205062029086e-06,
+      "loss": 1.3682,
+      "step": 8755
+    },
+    {
+      "epoch": 1.5105667212973346,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.834306242982733e-06,
+      "loss": 1.4479,
+      "step": 8756
+    },
+    {
+      "epoch": 1.5107392391960666,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.8324079548157525e-06,
+      "loss": 1.5293,
+      "step": 8757
+    },
+    {
+      "epoch": 1.5109117570947985,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.830510197668773e-06,
+      "loss": 1.4863,
+      "step": 8758
+    },
+    {
+      "epoch": 1.5110842749935305,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.828612971682395e-06,
+      "loss": 1.3653,
+      "step": 8759
+    },
+    {
+      "epoch": 1.5112567928922624,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.826716276997177e-06,
+      "loss": 1.3778,
+      "step": 8760
+    },
+    {
+      "epoch": 1.5114293107909944,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.824820113753638e-06,
+      "loss": 1.4826,
+      "step": 8761
+    },
+    {
+      "epoch": 1.5116018286897266,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.8229244820922573e-06,
+      "loss": 1.4547,
+      "step": 8762
+    },
+    {
+      "epoch": 1.5117743465884586,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.8210293821534763e-06,
+      "loss": 1.4667,
+      "step": 8763
+    },
+    {
+      "epoch": 1.5119468644871905,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.8191348140776942e-06,
+      "loss": 1.5144,
+      "step": 8764
+    },
+    {
+      "epoch": 1.5121193823859227,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.8172407780052726e-06,
+      "loss": 1.3551,
+      "step": 8765
+    },
+    {
+      "epoch": 1.5122919002846547,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.8153472740765385e-06,
+      "loss": 1.44,
+      "step": 8766
+    },
+    {
+      "epoch": 1.5124644181833866,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.813454302431763e-06,
+      "loss": 1.4299,
+      "step": 8767
+    },
+    {
+      "epoch": 1.5126369360821186,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.8115618632112027e-06,
+      "loss": 1.45,
+      "step": 8768
+    },
+    {
+      "epoch": 1.5128094539808505,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.8096699565550525e-06,
+      "loss": 1.4528,
+      "step": 8769
+    },
+    {
+      "epoch": 1.5129819718795825,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.807778582603479e-06,
+      "loss": 1.4855,
+      "step": 8770
+    },
+    {
+      "epoch": 1.5131544897783145,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.805887741496607e-06,
+      "loss": 1.4791,
+      "step": 8771
+    },
+    {
+      "epoch": 1.5133270076770464,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.803997433374521e-06,
+      "loss": 1.4079,
+      "step": 8772
+    },
+    {
+      "epoch": 1.5134995255757784,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.80210765837727e-06,
+      "loss": 1.428,
+      "step": 8773
+    },
+    {
+      "epoch": 1.5136720434745103,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.80021841664486e-06,
+      "loss": 1.3966,
+      "step": 8774
+    },
+    {
+      "epoch": 1.5138445613732425,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.7983297083172487e-06,
+      "loss": 1.4717,
+      "step": 8775
+    },
+    {
+      "epoch": 1.5140170792719745,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.7964415335343785e-06,
+      "loss": 1.4614,
+      "step": 8776
+    },
+    {
+      "epoch": 1.5141895971707064,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.7945538924361258e-06,
+      "loss": 1.4601,
+      "step": 8777
+    },
+    {
+      "epoch": 1.5143621150694384,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.7926667851623437e-06,
+      "loss": 1.4284,
+      "step": 8778
+    },
+    {
+      "epoch": 1.5145346329681706,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.790780211852838e-06,
+      "loss": 1.459,
+      "step": 8779
+    },
+    {
+      "epoch": 1.5147071508669026,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.7888941726473773e-06,
+      "loss": 1.413,
+      "step": 8780
+    },
+    {
+      "epoch": 1.5148796687656345,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.787008667685699e-06,
+      "loss": 1.3772,
+      "step": 8781
+    },
+    {
+      "epoch": 1.5150521866643665,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.7851236971074848e-06,
+      "loss": 1.5823,
+      "step": 8782
+    },
+    {
+      "epoch": 1.5152247045630984,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.783239261052387e-06,
+      "loss": 1.4732,
+      "step": 8783
+    },
+    {
+      "epoch": 1.5153972224618304,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.7813553596600175e-06,
+      "loss": 1.4055,
+      "step": 8784
+    },
+    {
+      "epoch": 1.5155697403605624,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.7794719930699477e-06,
+      "loss": 1.3886,
+      "step": 8785
+    },
+    {
+      "epoch": 1.5157422582592943,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.7775891614217075e-06,
+      "loss": 1.3907,
+      "step": 8786
+    },
+    {
+      "epoch": 1.5159147761580263,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.7757068648547938e-06,
+      "loss": 1.3861,
+      "step": 8787
+    },
+    {
+      "epoch": 1.5160872940567582,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.7738251035086485e-06,
+      "loss": 1.4257,
+      "step": 8788
+    },
+    {
+      "epoch": 1.5162598119554904,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.771943877522697e-06,
+      "loss": 1.4905,
+      "step": 8789
+    },
+    {
+      "epoch": 1.5164323298542224,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.7700631870363038e-06,
+      "loss": 1.3708,
+      "step": 8790
+    },
+    {
+      "epoch": 1.5166048477529543,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.7681830321888035e-06,
+      "loss": 1.5362,
+      "step": 8791
+    },
+    {
+      "epoch": 1.5167773656516865,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.766303413119491e-06,
+      "loss": 1.3839,
+      "step": 8792
+    },
+    {
+      "epoch": 1.5169498835504185,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.7644243299676197e-06,
+      "loss": 1.405,
+      "step": 8793
+    },
+    {
+      "epoch": 1.5171224014491504,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.7625457828724034e-06,
+      "loss": 1.3364,
+      "step": 8794
+    },
+    {
+      "epoch": 1.5172949193478824,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.760667771973018e-06,
+      "loss": 1.4023,
+      "step": 8795
+    },
+    {
+      "epoch": 1.5174674372466144,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.7587902974085977e-06,
+      "loss": 1.3987,
+      "step": 8796
+    },
+    {
+      "epoch": 1.5176399551453463,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.756913359318237e-06,
+      "loss": 1.4989,
+      "step": 8797
+    },
+    {
+      "epoch": 1.5178124730440783,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.755036957840994e-06,
+      "loss": 1.349,
+      "step": 8798
+    },
+    {
+      "epoch": 1.5179849909428103,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.753161093115877e-06,
+      "loss": 1.3201,
+      "step": 8799
+    },
+    {
+      "epoch": 1.5181575088415422,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.7512857652818716e-06,
+      "loss": 1.517,
+      "step": 8800
+    },
+    {
+      "epoch": 1.5181575088415422,
+      "eval_loss": 1.407160758972168,
+      "eval_runtime": 10.9315,
+      "eval_samples_per_second": 93.674,
+      "eval_steps_per_second": 23.419,
+      "step": 8800
+    },
+    {
+      "epoch": 1.5183300267402742,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.7494109744779062e-06,
+      "loss": 1.3835,
+      "step": 8801
+    },
+    {
+      "epoch": 1.5185025446390061,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.7475367208428793e-06,
+      "loss": 1.5125,
+      "step": 8802
+    },
+    {
+      "epoch": 1.5186750625377383,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.7456630045156473e-06,
+      "loss": 1.4015,
+      "step": 8803
+    },
+    {
+      "epoch": 1.5188475804364703,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.7437898256350284e-06,
+      "loss": 1.3302,
+      "step": 8804
+    },
+    {
+      "epoch": 1.5190200983352022,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.7419171843397975e-06,
+      "loss": 1.4035,
+      "step": 8805
+    },
+    {
+      "epoch": 1.5191926162339344,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.740045080768694e-06,
+      "loss": 1.5455,
+      "step": 8806
+    },
+    {
+      "epoch": 1.5193651341326664,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.7381735150604083e-06,
+      "loss": 1.3257,
+      "step": 8807
+    },
+    {
+      "epoch": 1.5195376520313983,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.7363024873536093e-06,
+      "loss": 1.4662,
+      "step": 8808
+    },
+    {
+      "epoch": 1.5197101699301303,
+      "grad_norm": 0.91796875,
+      "learning_rate": 2.7344319977869037e-06,
+      "loss": 1.4842,
+      "step": 8809
+    },
+    {
+      "epoch": 1.5198826878288623,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.7325620464988733e-06,
+      "loss": 1.4691,
+      "step": 8810
+    },
+    {
+      "epoch": 1.5200552057275942,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.730692633628055e-06,
+      "loss": 1.3634,
+      "step": 8811
+    },
+    {
+      "epoch": 1.5202277236263262,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.728823759312944e-06,
+      "loss": 1.5168,
+      "step": 8812
+    },
+    {
+      "epoch": 1.5204002415250581,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.726955423692008e-06,
+      "loss": 1.4439,
+      "step": 8813
+    },
+    {
+      "epoch": 1.52057275942379,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.7250876269036564e-06,
+      "loss": 1.3487,
+      "step": 8814
+    },
+    {
+      "epoch": 1.520745277322522,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.723220369086267e-06,
+      "loss": 1.4754,
+      "step": 8815
+    },
+    {
+      "epoch": 1.5209177952212543,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.7213536503781813e-06,
+      "loss": 1.4661,
+      "step": 8816
+    },
+    {
+      "epoch": 1.5210903131199862,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.7194874709176967e-06,
+      "loss": 1.4603,
+      "step": 8817
+    },
+    {
+      "epoch": 1.5212628310187182,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.7176218308430703e-06,
+      "loss": 1.4373,
+      "step": 8818
+    },
+    {
+      "epoch": 1.5214353489174501,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.7157567302925235e-06,
+      "loss": 1.4028,
+      "step": 8819
+    },
+    {
+      "epoch": 1.5216078668161823,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.7138921694042264e-06,
+      "loss": 1.3868,
+      "step": 8820
+    },
+    {
+      "epoch": 1.5217803847149143,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.71202814831633e-06,
+      "loss": 1.5207,
+      "step": 8821
+    },
+    {
+      "epoch": 1.5219529026136462,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.7101646671669224e-06,
+      "loss": 1.3747,
+      "step": 8822
+    },
+    {
+      "epoch": 1.5221254205123782,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.708301726094065e-06,
+      "loss": 1.3591,
+      "step": 8823
+    },
+    {
+      "epoch": 1.5222979384111102,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.7064393252357757e-06,
+      "loss": 1.4247,
+      "step": 8824
+    },
+    {
+      "epoch": 1.5224704563098421,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.7045774647300347e-06,
+      "loss": 1.4324,
+      "step": 8825
+    },
+    {
+      "epoch": 1.522642974208574,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.7027161447147786e-06,
+      "loss": 1.4449,
+      "step": 8826
+    },
+    {
+      "epoch": 1.522815492107306,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.700855365327909e-06,
+      "loss": 1.4112,
+      "step": 8827
+    },
+    {
+      "epoch": 1.522988010006038,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.6989951267072744e-06,
+      "loss": 1.3578,
+      "step": 8828
+    },
+    {
+      "epoch": 1.52316052790477,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.697135428990704e-06,
+      "loss": 1.4005,
+      "step": 8829
+    },
+    {
+      "epoch": 1.5233330458035022,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.695276272315973e-06,
+      "loss": 1.4191,
+      "step": 8830
+    },
+    {
+      "epoch": 1.5235055637022341,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.6934176568208126e-06,
+      "loss": 1.4618,
+      "step": 8831
+    },
+    {
+      "epoch": 1.523678081600966,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.691559582642932e-06,
+      "loss": 1.2921,
+      "step": 8832
+    },
+    {
+      "epoch": 1.5238505994996983,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.689702049919979e-06,
+      "loss": 1.4131,
+      "step": 8833
+    },
+    {
+      "epoch": 1.5240231173984302,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.6878450587895754e-06,
+      "loss": 1.3017,
+      "step": 8834
+    },
+    {
+      "epoch": 1.5241956352971622,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.6859886093892983e-06,
+      "loss": 1.5027,
+      "step": 8835
+    },
+    {
+      "epoch": 1.5243681531958941,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.6841327018566842e-06,
+      "loss": 1.4793,
+      "step": 8836
+    },
+    {
+      "epoch": 1.524540671094626,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.6822773363292333e-06,
+      "loss": 1.3827,
+      "step": 8837
+    },
+    {
+      "epoch": 1.524713188993358,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.6804225129444016e-06,
+      "loss": 1.4242,
+      "step": 8838
+    },
+    {
+      "epoch": 1.52488570689209,
+      "grad_norm": 0.625,
+      "learning_rate": 2.6785682318396e-06,
+      "loss": 1.4618,
+      "step": 8839
+    },
+    {
+      "epoch": 1.525058224790822,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.676714493152216e-06,
+      "loss": 1.3629,
+      "step": 8840
+    },
+    {
+      "epoch": 1.525230742689554,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.674861297019574e-06,
+      "loss": 1.3417,
+      "step": 8841
+    },
+    {
+      "epoch": 1.525403260588286,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.6730086435789828e-06,
+      "loss": 1.5157,
+      "step": 8842
+    },
+    {
+      "epoch": 1.5255757784870179,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.671156532967689e-06,
+      "loss": 1.4606,
+      "step": 8843
+    },
+    {
+      "epoch": 1.52574829638575,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.66930496532291e-06,
+      "loss": 1.4213,
+      "step": 8844
+    },
+    {
+      "epoch": 1.525920814284482,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.667453940781829e-06,
+      "loss": 1.4048,
+      "step": 8845
+    },
+    {
+      "epoch": 1.526093332183214,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.665603459481573e-06,
+      "loss": 1.4262,
+      "step": 8846
+    },
+    {
+      "epoch": 1.5262658500819462,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.6637535215592405e-06,
+      "loss": 1.4697,
+      "step": 8847
+    },
+    {
+      "epoch": 1.5264383679806781,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.6619041271518854e-06,
+      "loss": 1.5062,
+      "step": 8848
+    },
+    {
+      "epoch": 1.52661088587941,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.6600552763965238e-06,
+      "loss": 1.349,
+      "step": 8849
+    },
+    {
+      "epoch": 1.526783403778142,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.6582069694301284e-06,
+      "loss": 1.4759,
+      "step": 8850
+    },
+    {
+      "epoch": 1.526955921676874,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.6563592063896383e-06,
+      "loss": 1.413,
+      "step": 8851
+    },
+    {
+      "epoch": 1.527128439575606,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.6545119874119364e-06,
+      "loss": 1.3788,
+      "step": 8852
+    },
+    {
+      "epoch": 1.527300957474338,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.65266531263389e-06,
+      "loss": 1.4434,
+      "step": 8853
+    },
+    {
+      "epoch": 1.5274734753730699,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.6508191821923023e-06,
+      "loss": 1.4395,
+      "step": 8854
+    },
+    {
+      "epoch": 1.5276459932718018,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.6489735962239495e-06,
+      "loss": 1.4694,
+      "step": 8855
+    },
+    {
+      "epoch": 1.5278185111705338,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.6471285548655644e-06,
+      "loss": 1.4755,
+      "step": 8856
+    },
+    {
+      "epoch": 1.527991029069266,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.6452840582538387e-06,
+      "loss": 1.331,
+      "step": 8857
+    },
+    {
+      "epoch": 1.528163546967998,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.6434401065254234e-06,
+      "loss": 1.4054,
+      "step": 8858
+    },
+    {
+      "epoch": 1.52833606486673,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.641596699816935e-06,
+      "loss": 1.4518,
+      "step": 8859
+    },
+    {
+      "epoch": 1.5285085827654619,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.6397538382649334e-06,
+      "loss": 1.4237,
+      "step": 8860
+    },
+    {
+      "epoch": 1.528681100664194,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.6379115220059604e-06,
+      "loss": 1.395,
+      "step": 8861
+    },
+    {
+      "epoch": 1.528853618562926,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.6360697511765053e-06,
+      "loss": 1.429,
+      "step": 8862
+    },
+    {
+      "epoch": 1.529026136461658,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.6342285259130085e-06,
+      "loss": 1.3681,
+      "step": 8863
+    },
+    {
+      "epoch": 1.52919865436039,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.6323878463518915e-06,
+      "loss": 1.4454,
+      "step": 8864
+    },
+    {
+      "epoch": 1.529371172259122,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.6305477126295152e-06,
+      "loss": 1.443,
+      "step": 8865
+    },
+    {
+      "epoch": 1.5295436901578539,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.628708124882212e-06,
+      "loss": 1.5633,
+      "step": 8866
+    },
+    {
+      "epoch": 1.5297162080565858,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.626869083246267e-06,
+      "loss": 1.4025,
+      "step": 8867
+    },
+    {
+      "epoch": 1.5298887259553178,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.625030587857931e-06,
+      "loss": 1.512,
+      "step": 8868
+    },
+    {
+      "epoch": 1.5300612438540497,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.62319263885341e-06,
+      "loss": 1.4404,
+      "step": 8869
+    },
+    {
+      "epoch": 1.5302337617527817,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.6213552363688734e-06,
+      "loss": 1.5006,
+      "step": 8870
+    },
+    {
+      "epoch": 1.5304062796515139,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.619518380540439e-06,
+      "loss": 1.456,
+      "step": 8871
+    },
+    {
+      "epoch": 1.5305787975502458,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.617682071504204e-06,
+      "loss": 1.4089,
+      "step": 8872
+    },
+    {
+      "epoch": 1.5307513154489778,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.6158463093962015e-06,
+      "loss": 1.423,
+      "step": 8873
+    },
+    {
+      "epoch": 1.53092383334771,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.61401109435245e-06,
+      "loss": 1.5205,
+      "step": 8874
+    },
+    {
+      "epoch": 1.531096351246442,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.612176426508902e-06,
+      "loss": 1.408,
+      "step": 8875
+    },
+    {
+      "epoch": 1.531268869145174,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.610342306001484e-06,
+      "loss": 1.4899,
+      "step": 8876
+    },
+    {
+      "epoch": 1.5314413870439059,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.6085087329660864e-06,
+      "loss": 1.4048,
+      "step": 8877
+    },
+    {
+      "epoch": 1.5316139049426378,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.606675707538542e-06,
+      "loss": 1.4059,
+      "step": 8878
+    },
+    {
+      "epoch": 1.5317864228413698,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.6048432298546576e-06,
+      "loss": 1.3969,
+      "step": 8879
+    },
+    {
+      "epoch": 1.5319589407401017,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.6030113000501933e-06,
+      "loss": 1.3853,
+      "step": 8880
+    },
+    {
+      "epoch": 1.5321314586388337,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.601179918260871e-06,
+      "loss": 1.4777,
+      "step": 8881
+    },
+    {
+      "epoch": 1.5323039765375657,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.59934908462237e-06,
+      "loss": 1.321,
+      "step": 8882
+    },
+    {
+      "epoch": 1.5324764944362976,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.5975187992703333e-06,
+      "loss": 1.528,
+      "step": 8883
+    },
+    {
+      "epoch": 1.5326490123350298,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.595689062340351e-06,
+      "loss": 1.4284,
+      "step": 8884
+    },
+    {
+      "epoch": 1.5328215302337618,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.5938598739679934e-06,
+      "loss": 1.3888,
+      "step": 8885
+    },
+    {
+      "epoch": 1.5329940481324937,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.5920312342887687e-06,
+      "loss": 1.4692,
+      "step": 8886
+    },
+    {
+      "epoch": 1.5331665660312257,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.590203143438157e-06,
+      "loss": 1.4418,
+      "step": 8887
+    },
+    {
+      "epoch": 1.5333390839299579,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.588375601551595e-06,
+      "loss": 1.4874,
+      "step": 8888
+    },
+    {
+      "epoch": 1.5335116018286898,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.586548608764479e-06,
+      "loss": 1.4726,
+      "step": 8889
+    },
+    {
+      "epoch": 1.5336841197274218,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.5847221652121634e-06,
+      "loss": 1.3969,
+      "step": 8890
+    },
+    {
+      "epoch": 1.5338566376261538,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.5828962710299655e-06,
+      "loss": 1.3807,
+      "step": 8891
+    },
+    {
+      "epoch": 1.5340291555248857,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.5810709263531496e-06,
+      "loss": 1.4172,
+      "step": 8892
+    },
+    {
+      "epoch": 1.5342016734236177,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.579246131316958e-06,
+      "loss": 1.5168,
+      "step": 8893
+    },
+    {
+      "epoch": 1.5343741913223496,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.5774218860565836e-06,
+      "loss": 1.4052,
+      "step": 8894
+    },
+    {
+      "epoch": 1.5345467092210816,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.5755981907071683e-06,
+      "loss": 1.393,
+      "step": 8895
+    },
+    {
+      "epoch": 1.5347192271198136,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.5737750454038346e-06,
+      "loss": 1.5,
+      "step": 8896
+    },
+    {
+      "epoch": 1.5348917450185455,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.571952450281643e-06,
+      "loss": 1.4573,
+      "step": 8897
+    },
+    {
+      "epoch": 1.5350642629172777,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.570130405475627e-06,
+      "loss": 1.4203,
+      "step": 8898
+    },
+    {
+      "epoch": 1.5352367808160097,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.568308911120775e-06,
+      "loss": 1.3794,
+      "step": 8899
+    },
+    {
+      "epoch": 1.5354092987147416,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.5664879673520337e-06,
+      "loss": 1.4801,
+      "step": 8900
+    },
+    {
+      "epoch": 1.5354092987147416,
+      "eval_loss": 1.4071331024169922,
+      "eval_runtime": 10.9461,
+      "eval_samples_per_second": 93.549,
+      "eval_steps_per_second": 23.387,
+      "step": 8900
+    },
+    {
+      "epoch": 1.5355818166134736,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.5646675743043115e-06,
+      "loss": 1.5236,
+      "step": 8901
+    },
+    {
+      "epoch": 1.5357543345122058,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.562847732112472e-06,
+      "loss": 1.4504,
+      "step": 8902
+    },
+    {
+      "epoch": 1.5359268524109377,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.561028440911343e-06,
+      "loss": 1.3734,
+      "step": 8903
+    },
+    {
+      "epoch": 1.5360993703096697,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.5592097008357107e-06,
+      "loss": 1.4178,
+      "step": 8904
+    },
+    {
+      "epoch": 1.5362718882084017,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.5573915120203097e-06,
+      "loss": 1.447,
+      "step": 8905
+    },
+    {
+      "epoch": 1.5364444061071336,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.555573874599856e-06,
+      "loss": 1.3607,
+      "step": 8906
+    },
+    {
+      "epoch": 1.5366169240058656,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.553756788709001e-06,
+      "loss": 1.4766,
+      "step": 8907
+    },
+    {
+      "epoch": 1.5367894419045975,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.5519402544823656e-06,
+      "loss": 1.4166,
+      "step": 8908
+    },
+    {
+      "epoch": 1.5369619598033295,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.5501242720545404e-06,
+      "loss": 1.4674,
+      "step": 8909
+    },
+    {
+      "epoch": 1.5371344777020615,
+      "grad_norm": 1.2734375,
+      "learning_rate": 2.5483088415600552e-06,
+      "loss": 1.5013,
+      "step": 8910
+    },
+    {
+      "epoch": 1.5373069956007934,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.5464939631334106e-06,
+      "loss": 1.4255,
+      "step": 8911
+    },
+    {
+      "epoch": 1.5374795134995256,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.5446796369090665e-06,
+      "loss": 1.3638,
+      "step": 8912
+    },
+    {
+      "epoch": 1.5376520313982576,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.5428658630214365e-06,
+      "loss": 1.4622,
+      "step": 8913
+    },
+    {
+      "epoch": 1.5378245492969895,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.5410526416048986e-06,
+      "loss": 1.3916,
+      "step": 8914
+    },
+    {
+      "epoch": 1.5379970671957217,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.539239972793789e-06,
+      "loss": 1.4569,
+      "step": 8915
+    },
+    {
+      "epoch": 1.5381695850944537,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.537427856722393e-06,
+      "loss": 1.4265,
+      "step": 8916
+    },
+    {
+      "epoch": 1.5383421029931856,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.5356162935249762e-06,
+      "loss": 1.4431,
+      "step": 8917
+    },
+    {
+      "epoch": 1.5385146208919176,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.5338052833357406e-06,
+      "loss": 1.4271,
+      "step": 8918
+    },
+    {
+      "epoch": 1.5386871387906496,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.5319948262888604e-06,
+      "loss": 1.3865,
+      "step": 8919
+    },
+    {
+      "epoch": 1.5388596566893815,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.5301849225184673e-06,
+      "loss": 1.497,
+      "step": 8920
+    },
+    {
+      "epoch": 1.5390321745881135,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.528375572158647e-06,
+      "loss": 1.3021,
+      "step": 8921
+    },
+    {
+      "epoch": 1.5392046924868454,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.5265667753434498e-06,
+      "loss": 1.3756,
+      "step": 8922
+    },
+    {
+      "epoch": 1.5393772103855774,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.5247585322068847e-06,
+      "loss": 1.4439,
+      "step": 8923
+    },
+    {
+      "epoch": 1.5395497282843094,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.52295084288291e-06,
+      "loss": 1.4724,
+      "step": 8924
+    },
+    {
+      "epoch": 1.5397222461830415,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.521143707505457e-06,
+      "loss": 1.3917,
+      "step": 8925
+    },
+    {
+      "epoch": 1.5398947640817735,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.519337126208412e-06,
+      "loss": 1.3374,
+      "step": 8926
+    },
+    {
+      "epoch": 1.5400672819805055,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.5175310991256085e-06,
+      "loss": 1.4126,
+      "step": 8927
+    },
+    {
+      "epoch": 1.5402397998792374,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.515725626390859e-06,
+      "loss": 1.4605,
+      "step": 8928
+    },
+    {
+      "epoch": 1.5404123177779696,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.5139207081379134e-06,
+      "loss": 1.4168,
+      "step": 8929
+    },
+    {
+      "epoch": 1.5405848356767016,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.5121163445005025e-06,
+      "loss": 1.5422,
+      "step": 8930
+    },
+    {
+      "epoch": 1.5407573535754335,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.510312535612297e-06,
+      "loss": 1.3716,
+      "step": 8931
+    },
+    {
+      "epoch": 1.5409298714741655,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.5085092816069367e-06,
+      "loss": 1.4441,
+      "step": 8932
+    },
+    {
+      "epoch": 1.5411023893728975,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.506706582618017e-06,
+      "loss": 1.5556,
+      "step": 8933
+    },
+    {
+      "epoch": 1.5412749072716294,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.5049044387790943e-06,
+      "loss": 1.3855,
+      "step": 8934
+    },
+    {
+      "epoch": 1.5414474251703614,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.503102850223682e-06,
+      "loss": 1.4681,
+      "step": 8935
+    },
+    {
+      "epoch": 1.5416199430690933,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.5013018170852566e-06,
+      "loss": 1.4861,
+      "step": 8936
+    },
+    {
+      "epoch": 1.5417924609678253,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.499501339497241e-06,
+      "loss": 1.3168,
+      "step": 8937
+    },
+    {
+      "epoch": 1.5419649788665573,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.4977014175930368e-06,
+      "loss": 1.5329,
+      "step": 8938
+    },
+    {
+      "epoch": 1.5421374967652894,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.495902051505986e-06,
+      "loss": 1.4227,
+      "step": 8939
+    },
+    {
+      "epoch": 1.5423100146640214,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.4941032413693955e-06,
+      "loss": 1.4139,
+      "step": 8940
+    },
+    {
+      "epoch": 1.5424825325627534,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.4923049873165415e-06,
+      "loss": 1.3561,
+      "step": 8941
+    },
+    {
+      "epoch": 1.5426550504614855,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.4905072894806414e-06,
+      "loss": 1.3619,
+      "step": 8942
+    },
+    {
+      "epoch": 1.5428275683602175,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.4887101479948826e-06,
+      "loss": 1.438,
+      "step": 8943
+    },
+    {
+      "epoch": 1.5430000862589495,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.486913562992409e-06,
+      "loss": 1.3082,
+      "step": 8944
+    },
+    {
+      "epoch": 1.5431726041576814,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.4851175346063227e-06,
+      "loss": 1.3638,
+      "step": 8945
+    },
+    {
+      "epoch": 1.5433451220564134,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.4833220629696852e-06,
+      "loss": 1.459,
+      "step": 8946
+    },
+    {
+      "epoch": 1.5435176399551453,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.481527148215518e-06,
+      "loss": 1.5179,
+      "step": 8947
+    },
+    {
+      "epoch": 1.5436901578538773,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.479732790476791e-06,
+      "loss": 1.3455,
+      "step": 8948
+    },
+    {
+      "epoch": 1.5438626757526093,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.4779389898864538e-06,
+      "loss": 1.5416,
+      "step": 8949
+    },
+    {
+      "epoch": 1.5440351936513412,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.476145746577394e-06,
+      "loss": 1.5077,
+      "step": 8950
+    },
+    {
+      "epoch": 1.5442077115500732,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.474353060682467e-06,
+      "loss": 1.4062,
+      "step": 8951
+    },
+    {
+      "epoch": 1.5443802294488052,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.472560932334489e-06,
+      "loss": 1.4813,
+      "step": 8952
+    },
+    {
+      "epoch": 1.5445527473475373,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.470769361666231e-06,
+      "loss": 1.407,
+      "step": 8953
+    },
+    {
+      "epoch": 1.5447252652462693,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.4689783488104223e-06,
+      "loss": 1.4856,
+      "step": 8954
+    },
+    {
+      "epoch": 1.5448977831450013,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.4671878938997572e-06,
+      "loss": 1.4617,
+      "step": 8955
+    },
+    {
+      "epoch": 1.5450703010437334,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.465397997066874e-06,
+      "loss": 1.4065,
+      "step": 8956
+    },
+    {
+      "epoch": 1.5452428189424654,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.4636086584443885e-06,
+      "loss": 1.4875,
+      "step": 8957
+    },
+    {
+      "epoch": 1.5454153368411974,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.4618198781648663e-06,
+      "loss": 1.3696,
+      "step": 8958
+    },
+    {
+      "epoch": 1.5455878547399293,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.460031656360822e-06,
+      "loss": 1.3794,
+      "step": 8959
+    },
+    {
+      "epoch": 1.5457603726386613,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.4582439931647507e-06,
+      "loss": 1.4181,
+      "step": 8960
+    },
+    {
+      "epoch": 1.5459328905373932,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.4564568887090814e-06,
+      "loss": 1.54,
+      "step": 8961
+    },
+    {
+      "epoch": 1.5461054084361252,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.4546703431262253e-06,
+      "loss": 1.4814,
+      "step": 8962
+    },
+    {
+      "epoch": 1.5462779263348572,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.452884356548533e-06,
+      "loss": 1.3949,
+      "step": 8963
+    },
+    {
+      "epoch": 1.5464504442335891,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.4510989291083245e-06,
+      "loss": 1.3333,
+      "step": 8964
+    },
+    {
+      "epoch": 1.546622962132321,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.4493140609378753e-06,
+      "loss": 1.3779,
+      "step": 8965
+    },
+    {
+      "epoch": 1.5467954800310533,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.4475297521694187e-06,
+      "loss": 1.4457,
+      "step": 8966
+    },
+    {
+      "epoch": 1.5469679979297852,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.4457460029351476e-06,
+      "loss": 1.4055,
+      "step": 8967
+    },
+    {
+      "epoch": 1.5471405158285172,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.443962813367218e-06,
+      "loss": 1.4068,
+      "step": 8968
+    },
+    {
+      "epoch": 1.5473130337272492,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.4421801835977286e-06,
+      "loss": 1.521,
+      "step": 8969
+    },
+    {
+      "epoch": 1.5474855516259813,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.440398113758761e-06,
+      "loss": 1.4514,
+      "step": 8970
+    },
+    {
+      "epoch": 1.5476580695247133,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.438616603982332e-06,
+      "loss": 1.5065,
+      "step": 8971
+    },
+    {
+      "epoch": 1.5478305874234453,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.436835654400429e-06,
+      "loss": 1.4306,
+      "step": 8972
+    },
+    {
+      "epoch": 1.5480031053221772,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.435055265145003e-06,
+      "loss": 1.4412,
+      "step": 8973
+    },
+    {
+      "epoch": 1.5481756232209092,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.4332754363479483e-06,
+      "loss": 1.4578,
+      "step": 8974
+    },
+    {
+      "epoch": 1.5483481411196411,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.4314961681411276e-06,
+      "loss": 1.4251,
+      "step": 8975
+    },
+    {
+      "epoch": 1.548520659018373,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.429717460656361e-06,
+      "loss": 1.5216,
+      "step": 8976
+    },
+    {
+      "epoch": 1.548693176917105,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.427939314025427e-06,
+      "loss": 1.4381,
+      "step": 8977
+    },
+    {
+      "epoch": 1.548865694815837,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.4261617283800597e-06,
+      "loss": 1.4402,
+      "step": 8978
+    },
+    {
+      "epoch": 1.549038212714569,
+      "grad_norm": 0.625,
+      "learning_rate": 2.42438470385196e-06,
+      "loss": 1.3946,
+      "step": 8979
+    },
+    {
+      "epoch": 1.5492107306133012,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.422608240572768e-06,
+      "loss": 1.4762,
+      "step": 8980
+    },
+    {
+      "epoch": 1.5493832485120331,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.4208323386741107e-06,
+      "loss": 1.3965,
+      "step": 8981
+    },
+    {
+      "epoch": 1.549555766410765,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.419056998287547e-06,
+      "loss": 1.4151,
+      "step": 8982
+    },
+    {
+      "epoch": 1.5497282843094973,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.417282219544609e-06,
+      "loss": 1.538,
+      "step": 8983
+    },
+    {
+      "epoch": 1.5499008022082292,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.415508002576783e-06,
+      "loss": 1.3951,
+      "step": 8984
+    },
+    {
+      "epoch": 1.5500733201069612,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.413734347515514e-06,
+      "loss": 1.3653,
+      "step": 8985
+    },
+    {
+      "epoch": 1.5502458380056932,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.411961254492207e-06,
+      "loss": 1.5114,
+      "step": 8986
+    },
+    {
+      "epoch": 1.5504183559044251,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.4101887236382237e-06,
+      "loss": 1.4353,
+      "step": 8987
+    },
+    {
+      "epoch": 1.550590873803157,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.408416755084878e-06,
+      "loss": 1.3855,
+      "step": 8988
+    },
+    {
+      "epoch": 1.550763391701889,
+      "grad_norm": 0.625,
+      "learning_rate": 2.4066453489634565e-06,
+      "loss": 1.4579,
+      "step": 8989
+    },
+    {
+      "epoch": 1.550935909600621,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.4048745054051924e-06,
+      "loss": 1.3656,
+      "step": 8990
+    },
+    {
+      "epoch": 1.551108427499353,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.403104224541283e-06,
+      "loss": 1.4416,
+      "step": 8991
+    },
+    {
+      "epoch": 1.551280945398085,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.4013345065028816e-06,
+      "loss": 1.3904,
+      "step": 8992
+    },
+    {
+      "epoch": 1.5514534632968169,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.3995653514210936e-06,
+      "loss": 1.4777,
+      "step": 8993
+    },
+    {
+      "epoch": 1.551625981195549,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.3977967594270003e-06,
+      "loss": 1.5117,
+      "step": 8994
+    },
+    {
+      "epoch": 1.551798499094281,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.3960287306516193e-06,
+      "loss": 1.4209,
+      "step": 8995
+    },
+    {
+      "epoch": 1.551971016993013,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.3942612652259436e-06,
+      "loss": 1.4499,
+      "step": 8996
+    },
+    {
+      "epoch": 1.5521435348917452,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.392494363280915e-06,
+      "loss": 1.3548,
+      "step": 8997
+    },
+    {
+      "epoch": 1.5523160527904771,
+      "grad_norm": 0.625,
+      "learning_rate": 2.3907280249474384e-06,
+      "loss": 1.4644,
+      "step": 8998
+    },
+    {
+      "epoch": 1.552488570689209,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.3889622503563734e-06,
+      "loss": 1.52,
+      "step": 8999
+    },
+    {
+      "epoch": 1.552661088587941,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.3871970396385457e-06,
+      "loss": 1.4037,
+      "step": 9000
+    },
+    {
+      "epoch": 1.552661088587941,
+      "eval_loss": 1.4071366786956787,
+      "eval_runtime": 10.8655,
+      "eval_samples_per_second": 94.243,
+      "eval_steps_per_second": 23.561,
+      "step": 9000
+    },
+    {
+      "epoch": 1.552833606486673,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.3854323929247214e-06,
+      "loss": 1.36,
+      "step": 9001
+    },
+    {
+      "epoch": 1.553006124385405,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.3836683103456493e-06,
+      "loss": 1.5039,
+      "step": 9002
+    },
+    {
+      "epoch": 1.553178642284137,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.381904792032015e-06,
+      "loss": 1.3965,
+      "step": 9003
+    },
+    {
+      "epoch": 1.553351160182869,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.3801418381144712e-06,
+      "loss": 1.4204,
+      "step": 9004
+    },
+    {
+      "epoch": 1.5535236780816009,
+      "grad_norm": 8.9375,
+      "learning_rate": 2.3783794487236367e-06,
+      "loss": 1.4602,
+      "step": 9005
+    },
+    {
+      "epoch": 1.5536961959803328,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.3766176239900717e-06,
+      "loss": 1.4365,
+      "step": 9006
+    },
+    {
+      "epoch": 1.553868713879065,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.3748563640443066e-06,
+      "loss": 1.3844,
+      "step": 9007
+    },
+    {
+      "epoch": 1.554041231777797,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.373095669016825e-06,
+      "loss": 1.4673,
+      "step": 9008
+    },
+    {
+      "epoch": 1.554213749676529,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.371335539038073e-06,
+      "loss": 1.3874,
+      "step": 9009
+    },
+    {
+      "epoch": 1.5543862675752609,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.3695759742384495e-06,
+      "loss": 1.485,
+      "step": 9010
+    },
+    {
+      "epoch": 1.554558785473993,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.367816974748317e-06,
+      "loss": 1.4039,
+      "step": 9011
+    },
+    {
+      "epoch": 1.554731303372725,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.3660585406979865e-06,
+      "loss": 1.3879,
+      "step": 9012
+    },
+    {
+      "epoch": 1.554903821271457,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.364300672217744e-06,
+      "loss": 1.3803,
+      "step": 9013
+    },
+    {
+      "epoch": 1.555076339170189,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.3625433694378143e-06,
+      "loss": 1.3202,
+      "step": 9014
+    },
+    {
+      "epoch": 1.555248857068921,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.3607866324883934e-06,
+      "loss": 1.4688,
+      "step": 9015
+    },
+    {
+      "epoch": 1.5554213749676529,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.3590304614996305e-06,
+      "loss": 1.4,
+      "step": 9016
+    },
+    {
+      "epoch": 1.5555938928663848,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.3572748566016345e-06,
+      "loss": 1.4108,
+      "step": 9017
+    },
+    {
+      "epoch": 1.5557664107651168,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.3555198179244707e-06,
+      "loss": 1.3687,
+      "step": 9018
+    },
+    {
+      "epoch": 1.5559389286638488,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.3537653455981655e-06,
+      "loss": 1.4195,
+      "step": 9019
+    },
+    {
+      "epoch": 1.5561114465625807,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.352011439752695e-06,
+      "loss": 1.3848,
+      "step": 9020
+    },
+    {
+      "epoch": 1.556283964461313,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.350258100518007e-06,
+      "loss": 1.4125,
+      "step": 9021
+    },
+    {
+      "epoch": 1.5564564823600449,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.3485053280239955e-06,
+      "loss": 1.4128,
+      "step": 9022
+    },
+    {
+      "epoch": 1.5566290002587768,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.3467531224005192e-06,
+      "loss": 1.4206,
+      "step": 9023
+    },
+    {
+      "epoch": 1.556801518157509,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.3450014837773937e-06,
+      "loss": 1.4219,
+      "step": 9024
+    },
+    {
+      "epoch": 1.556974036056241,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.3432504122843826e-06,
+      "loss": 1.4805,
+      "step": 9025
+    },
+    {
+      "epoch": 1.557146553954973,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.341499908051229e-06,
+      "loss": 1.3797,
+      "step": 9026
+    },
+    {
+      "epoch": 1.5573190718537049,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.3397499712076124e-06,
+      "loss": 1.3865,
+      "step": 9027
+    },
+    {
+      "epoch": 1.5574915897524368,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.3380006018831804e-06,
+      "loss": 1.3843,
+      "step": 9028
+    },
+    {
+      "epoch": 1.5576641076511688,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.3362518002075375e-06,
+      "loss": 1.3683,
+      "step": 9029
+    },
+    {
+      "epoch": 1.5578366255499008,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.3345035663102477e-06,
+      "loss": 1.5048,
+      "step": 9030
+    },
+    {
+      "epoch": 1.5580091434486327,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.33275590032083e-06,
+      "loss": 1.4577,
+      "step": 9031
+    },
+    {
+      "epoch": 1.5581816613473647,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.331008802368765e-06,
+      "loss": 1.3813,
+      "step": 9032
+    },
+    {
+      "epoch": 1.5583541792460966,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.32926227258348e-06,
+      "loss": 1.4236,
+      "step": 9033
+    },
+    {
+      "epoch": 1.5585266971448286,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.3275163110943812e-06,
+      "loss": 1.3342,
+      "step": 9034
+    },
+    {
+      "epoch": 1.5586992150435608,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.325770918030811e-06,
+      "loss": 1.4575,
+      "step": 9035
+    },
+    {
+      "epoch": 1.5588717329422928,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.3240260935220804e-06,
+      "loss": 1.4458,
+      "step": 9036
+    },
+    {
+      "epoch": 1.5590442508410247,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.32228183769746e-06,
+      "loss": 1.4072,
+      "step": 9037
+    },
+    {
+      "epoch": 1.559216768739757,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.3205381506861735e-06,
+      "loss": 1.3684,
+      "step": 9038
+    },
+    {
+      "epoch": 1.5593892866384889,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.3187950326174048e-06,
+      "loss": 1.4303,
+      "step": 9039
+    },
+    {
+      "epoch": 1.5595618045372208,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.3170524836202936e-06,
+      "loss": 1.4655,
+      "step": 9040
+    },
+    {
+      "epoch": 1.5597343224359528,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.31531050382394e-06,
+      "loss": 1.3948,
+      "step": 9041
+    },
+    {
+      "epoch": 1.5599068403346847,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.3135690933574007e-06,
+      "loss": 1.4346,
+      "step": 9042
+    },
+    {
+      "epoch": 1.5600793582334167,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.3118282523496917e-06,
+      "loss": 1.4503,
+      "step": 9043
+    },
+    {
+      "epoch": 1.5602518761321487,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.3100879809297793e-06,
+      "loss": 1.5228,
+      "step": 9044
+    },
+    {
+      "epoch": 1.5604243940308806,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.3083482792266032e-06,
+      "loss": 1.4708,
+      "step": 9045
+    },
+    {
+      "epoch": 1.5605969119296126,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.3066091473690433e-06,
+      "loss": 1.3849,
+      "step": 9046
+    },
+    {
+      "epoch": 1.5607694298283445,
+      "grad_norm": 0.625,
+      "learning_rate": 2.304870585485949e-06,
+      "loss": 1.4502,
+      "step": 9047
+    },
+    {
+      "epoch": 1.5609419477270767,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.3031325937061222e-06,
+      "loss": 1.3883,
+      "step": 9048
+    },
+    {
+      "epoch": 1.5611144656258087,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.301395172158325e-06,
+      "loss": 1.4215,
+      "step": 9049
+    },
+    {
+      "epoch": 1.5612869835245407,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.2996583209712776e-06,
+      "loss": 1.4415,
+      "step": 9050
+    },
+    {
+      "epoch": 1.5614595014232726,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.2979220402736547e-06,
+      "loss": 1.4267,
+      "step": 9051
+    },
+    {
+      "epoch": 1.5616320193220048,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.2961863301940924e-06,
+      "loss": 1.4367,
+      "step": 9052
+    },
+    {
+      "epoch": 1.5618045372207368,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.294451190861182e-06,
+      "loss": 1.4087,
+      "step": 9053
+    },
+    {
+      "epoch": 1.5619770551194687,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.292716622403475e-06,
+      "loss": 1.4869,
+      "step": 9054
+    },
+    {
+      "epoch": 1.5621495730182007,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.290982624949477e-06,
+      "loss": 1.4822,
+      "step": 9055
+    },
+    {
+      "epoch": 1.5623220909169326,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.2892491986276578e-06,
+      "loss": 1.437,
+      "step": 9056
+    },
+    {
+      "epoch": 1.5624946088156646,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.2875163435664306e-06,
+      "loss": 1.4063,
+      "step": 9057
+    },
+    {
+      "epoch": 1.5626671267143966,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.285784059894188e-06,
+      "loss": 1.4217,
+      "step": 9058
+    },
+    {
+      "epoch": 1.5628396446131285,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.2840523477392606e-06,
+      "loss": 1.3661,
+      "step": 9059
+    },
+    {
+      "epoch": 1.5630121625118605,
+      "grad_norm": 0.625,
+      "learning_rate": 2.2823212072299463e-06,
+      "loss": 1.4059,
+      "step": 9060
+    },
+    {
+      "epoch": 1.5631846804105924,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.2805906384945e-06,
+      "loss": 1.4256,
+      "step": 9061
+    },
+    {
+      "epoch": 1.5633571983093246,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.2788606416611314e-06,
+      "loss": 1.514,
+      "step": 9062
+    },
+    {
+      "epoch": 1.5635297162080566,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.277131216858011e-06,
+      "loss": 1.5115,
+      "step": 9063
+    },
+    {
+      "epoch": 1.5637022341067885,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.275402364213267e-06,
+      "loss": 1.4034,
+      "step": 9064
+    },
+    {
+      "epoch": 1.5638747520055207,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.2736740838549765e-06,
+      "loss": 1.3902,
+      "step": 9065
+    },
+    {
+      "epoch": 1.5640472699042527,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.2719463759111914e-06,
+      "loss": 1.5219,
+      "step": 9066
+    },
+    {
+      "epoch": 1.5642197878029847,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.2702192405099035e-06,
+      "loss": 1.414,
+      "step": 9067
+    },
+    {
+      "epoch": 1.5643923057017166,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.2684926777790717e-06,
+      "loss": 1.5138,
+      "step": 9068
+    },
+    {
+      "epoch": 1.5645648236004486,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.266766687846611e-06,
+      "loss": 1.4112,
+      "step": 9069
+    },
+    {
+      "epoch": 1.5647373414991805,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.2650412708403934e-06,
+      "loss": 1.4174,
+      "step": 9070
+    },
+    {
+      "epoch": 1.5649098593979125,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.2633164268882492e-06,
+      "loss": 1.4252,
+      "step": 9071
+    },
+    {
+      "epoch": 1.5650823772966445,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.261592156117964e-06,
+      "loss": 1.4543,
+      "step": 9072
+    },
+    {
+      "epoch": 1.5652548951953764,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.259868458657285e-06,
+      "loss": 1.4375,
+      "step": 9073
+    },
+    {
+      "epoch": 1.5654274130941084,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.258145334633912e-06,
+      "loss": 1.4361,
+      "step": 9074
+    },
+    {
+      "epoch": 1.5655999309928406,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.2564227841755105e-06,
+      "loss": 1.3977,
+      "step": 9075
+    },
+    {
+      "epoch": 1.5657724488915725,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.2547008074096864e-06,
+      "loss": 1.37,
+      "step": 9076
+    },
+    {
+      "epoch": 1.5659449667903045,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.252979404464027e-06,
+      "loss": 1.3545,
+      "step": 9077
+    },
+    {
+      "epoch": 1.5661174846890364,
+      "grad_norm": 0.625,
+      "learning_rate": 2.2512585754660533e-06,
+      "loss": 1.4416,
+      "step": 9078
+    },
+    {
+      "epoch": 1.5662900025877686,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.2495383205432665e-06,
+      "loss": 1.4559,
+      "step": 9079
+    },
+    {
+      "epoch": 1.5664625204865006,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.247818639823105e-06,
+      "loss": 1.4815,
+      "step": 9080
+    },
+    {
+      "epoch": 1.5666350383852325,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.2460995334329773e-06,
+      "loss": 1.4407,
+      "step": 9081
+    },
+    {
+      "epoch": 1.5668075562839645,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.244381001500244e-06,
+      "loss": 1.3868,
+      "step": 9082
+    },
+    {
+      "epoch": 1.5669800741826965,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.2426630441522257e-06,
+      "loss": 1.458,
+      "step": 9083
+    },
+    {
+      "epoch": 1.5671525920814284,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.2409456615161996e-06,
+      "loss": 1.406,
+      "step": 9084
+    },
+    {
+      "epoch": 1.5673251099801604,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.2392288537193987e-06,
+      "loss": 1.4267,
+      "step": 9085
+    },
+    {
+      "epoch": 1.5674976278788924,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.2375126208890164e-06,
+      "loss": 1.4312,
+      "step": 9086
+    },
+    {
+      "epoch": 1.5676701457776243,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.2357969631522016e-06,
+      "loss": 1.3354,
+      "step": 9087
+    },
+    {
+      "epoch": 1.5678426636763563,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.234081880636063e-06,
+      "loss": 1.4659,
+      "step": 9088
+    },
+    {
+      "epoch": 1.5680151815750885,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.2323673734676576e-06,
+      "loss": 1.3875,
+      "step": 9089
+    },
+    {
+      "epoch": 1.5681876994738204,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.2306534417740167e-06,
+      "loss": 1.5242,
+      "step": 9090
+    },
+    {
+      "epoch": 1.5683602173725524,
+      "grad_norm": 0.625,
+      "learning_rate": 2.228940085682111e-06,
+      "loss": 1.4999,
+      "step": 9091
+    },
+    {
+      "epoch": 1.5685327352712846,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.22722730531888e-06,
+      "loss": 1.3251,
+      "step": 9092
+    },
+    {
+      "epoch": 1.5687052531700165,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.2255151008112164e-06,
+      "loss": 1.4632,
+      "step": 9093
+    },
+    {
+      "epoch": 1.5688777710687485,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.2238034722859715e-06,
+      "loss": 1.4337,
+      "step": 9094
+    },
+    {
+      "epoch": 1.5690502889674804,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.2220924198699532e-06,
+      "loss": 1.511,
+      "step": 9095
+    },
+    {
+      "epoch": 1.5692228068662124,
+      "grad_norm": 0.625,
+      "learning_rate": 2.2203819436899297e-06,
+      "loss": 1.4009,
+      "step": 9096
+    },
+    {
+      "epoch": 1.5693953247649444,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.2186720438726163e-06,
+      "loss": 1.5326,
+      "step": 9097
+    },
+    {
+      "epoch": 1.5695678426636763,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.216962720544703e-06,
+      "loss": 1.3456,
+      "step": 9098
+    },
+    {
+      "epoch": 1.5697403605624083,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.2152539738328185e-06,
+      "loss": 1.533,
+      "step": 9099
+    },
+    {
+      "epoch": 1.5699128784611402,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.2135458038635614e-06,
+      "loss": 1.4004,
+      "step": 9100
+    },
+    {
+      "epoch": 1.5699128784611402,
+      "eval_loss": 1.4070823192596436,
+      "eval_runtime": 10.9402,
+      "eval_samples_per_second": 93.6,
+      "eval_steps_per_second": 23.4,
+      "step": 9100
+    },
+    {
+      "epoch": 1.5700853963598722,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.211838210763484e-06,
+      "loss": 1.4252,
+      "step": 9101
+    },
+    {
+      "epoch": 1.5702579142586042,
+      "grad_norm": 0.625,
+      "learning_rate": 2.2101311946590942e-06,
+      "loss": 1.4406,
+      "step": 9102
+    },
+    {
+      "epoch": 1.5704304321573364,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.2084247556768577e-06,
+      "loss": 1.4362,
+      "step": 9103
+    },
+    {
+      "epoch": 1.5706029500560683,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.2067188939432006e-06,
+      "loss": 1.4133,
+      "step": 9104
+    },
+    {
+      "epoch": 1.5707754679548003,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.205013609584502e-06,
+      "loss": 1.4302,
+      "step": 9105
+    },
+    {
+      "epoch": 1.5709479858535325,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.2033089027271003e-06,
+      "loss": 1.4432,
+      "step": 9106
+    },
+    {
+      "epoch": 1.5711205037522644,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.2016047734972923e-06,
+      "loss": 1.3627,
+      "step": 9107
+    },
+    {
+      "epoch": 1.5712930216509964,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.1999012220213277e-06,
+      "loss": 1.4397,
+      "step": 9108
+    },
+    {
+      "epoch": 1.5714655395497283,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.1981982484254214e-06,
+      "loss": 1.4597,
+      "step": 9109
+    },
+    {
+      "epoch": 1.5716380574484603,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.1964958528357317e-06,
+      "loss": 1.3625,
+      "step": 9110
+    },
+    {
+      "epoch": 1.5718105753471923,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.1947940353783927e-06,
+      "loss": 1.4335,
+      "step": 9111
+    },
+    {
+      "epoch": 1.5719830932459242,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.1930927961794777e-06,
+      "loss": 1.4665,
+      "step": 9112
+    },
+    {
+      "epoch": 1.5721556111446562,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.191392135365029e-06,
+      "loss": 1.3722,
+      "step": 9113
+    },
+    {
+      "epoch": 1.5723281290433881,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.189692053061041e-06,
+      "loss": 1.5526,
+      "step": 9114
+    },
+    {
+      "epoch": 1.57250064694212,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.1879925493934663e-06,
+      "loss": 1.4509,
+      "step": 9115
+    },
+    {
+      "epoch": 1.5726731648408523,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.186293624488216e-06,
+      "loss": 1.4767,
+      "step": 9116
+    },
+    {
+      "epoch": 1.5728456827395842,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.1845952784711555e-06,
+      "loss": 1.4349,
+      "step": 9117
+    },
+    {
+      "epoch": 1.5730182006383162,
+      "grad_norm": 0.625,
+      "learning_rate": 2.1828975114681097e-06,
+      "loss": 1.4376,
+      "step": 9118
+    },
+    {
+      "epoch": 1.5731907185370482,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.1812003236048607e-06,
+      "loss": 1.4136,
+      "step": 9119
+    },
+    {
+      "epoch": 1.5733632364357804,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.1795037150071473e-06,
+      "loss": 1.4266,
+      "step": 9120
+    },
+    {
+      "epoch": 1.5735357543345123,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.177807685800659e-06,
+      "loss": 1.4399,
+      "step": 9121
+    },
+    {
+      "epoch": 1.5737082722332443,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.1761122361110576e-06,
+      "loss": 1.4447,
+      "step": 9122
+    },
+    {
+      "epoch": 1.5738807901319762,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.1744173660639446e-06,
+      "loss": 1.4029,
+      "step": 9123
+    },
+    {
+      "epoch": 1.5740533080307082,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.1727230757848903e-06,
+      "loss": 1.3948,
+      "step": 9124
+    },
+    {
+      "epoch": 1.5742258259294402,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.1710293653994173e-06,
+      "loss": 1.4587,
+      "step": 9125
+    },
+    {
+      "epoch": 1.5743983438281721,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.169336235033007e-06,
+      "loss": 1.4837,
+      "step": 9126
+    },
+    {
+      "epoch": 1.574570861726904,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.167643684811096e-06,
+      "loss": 1.4112,
+      "step": 9127
+    },
+    {
+      "epoch": 1.574743379625636,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.1659517148590825e-06,
+      "loss": 1.4641,
+      "step": 9128
+    },
+    {
+      "epoch": 1.574915897524368,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.164260325302311e-06,
+      "loss": 1.4535,
+      "step": 9129
+    },
+    {
+      "epoch": 1.5750884154231002,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.1625695162660986e-06,
+      "loss": 1.2898,
+      "step": 9130
+    },
+    {
+      "epoch": 1.5752609333218321,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.1608792878757044e-06,
+      "loss": 1.4519,
+      "step": 9131
+    },
+    {
+      "epoch": 1.575433451220564,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.159189640256354e-06,
+      "loss": 1.5392,
+      "step": 9132
+    },
+    {
+      "epoch": 1.5756059691192963,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.1575005735332267e-06,
+      "loss": 1.3497,
+      "step": 9133
+    },
+    {
+      "epoch": 1.5757784870180283,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.155812087831457e-06,
+      "loss": 1.4987,
+      "step": 9134
+    },
+    {
+      "epoch": 1.5759510049167602,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.154124183276145e-06,
+      "loss": 1.4392,
+      "step": 9135
+    },
+    {
+      "epoch": 1.5761235228154922,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.152436859992335e-06,
+      "loss": 1.3764,
+      "step": 9136
+    },
+    {
+      "epoch": 1.5762960407142241,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.150750118105035e-06,
+      "loss": 1.37,
+      "step": 9137
+    },
+    {
+      "epoch": 1.576468558612956,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.1490639577392115e-06,
+      "loss": 1.3533,
+      "step": 9138
+    },
+    {
+      "epoch": 1.576641076511688,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.1473783790197854e-06,
+      "loss": 1.4118,
+      "step": 9139
+    },
+    {
+      "epoch": 1.57681359441042,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.145693382071633e-06,
+      "loss": 1.3812,
+      "step": 9140
+    },
+    {
+      "epoch": 1.576986112309152,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.144008967019595e-06,
+      "loss": 1.5229,
+      "step": 9141
+    },
+    {
+      "epoch": 1.577158630207884,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.1423251339884534e-06,
+      "loss": 1.4583,
+      "step": 9142
+    },
+    {
+      "epoch": 1.577331148106616,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.1406418831029673e-06,
+      "loss": 1.4862,
+      "step": 9143
+    },
+    {
+      "epoch": 1.577503666005348,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.138959214487837e-06,
+      "loss": 1.4565,
+      "step": 9144
+    },
+    {
+      "epoch": 1.57767618390408,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.137277128267725e-06,
+      "loss": 1.4422,
+      "step": 9145
+    },
+    {
+      "epoch": 1.577848701802812,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.1355956245672518e-06,
+      "loss": 1.388,
+      "step": 9146
+    },
+    {
+      "epoch": 1.5780212197015442,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.1339147035109943e-06,
+      "loss": 1.26,
+      "step": 9147
+    },
+    {
+      "epoch": 1.5781937376002761,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.1322343652234855e-06,
+      "loss": 1.5043,
+      "step": 9148
+    },
+    {
+      "epoch": 1.578366255499008,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.130554609829214e-06,
+      "loss": 1.4859,
+      "step": 9149
+    },
+    {
+      "epoch": 1.57853877339774,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.1288754374526275e-06,
+      "loss": 1.3498,
+      "step": 9150
+    },
+    {
+      "epoch": 1.578711291296472,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.1271968482181306e-06,
+      "loss": 1.4313,
+      "step": 9151
+    },
+    {
+      "epoch": 1.578883809195204,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.125518842250085e-06,
+      "loss": 1.4392,
+      "step": 9152
+    },
+    {
+      "epoch": 1.579056327093936,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.1238414196728007e-06,
+      "loss": 1.3872,
+      "step": 9153
+    },
+    {
+      "epoch": 1.579228844992668,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.1221645806105617e-06,
+      "loss": 1.4653,
+      "step": 9154
+    },
+    {
+      "epoch": 1.5794013628913999,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.120488325187591e-06,
+      "loss": 1.437,
+      "step": 9155
+    },
+    {
+      "epoch": 1.5795738807901318,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.118812653528077e-06,
+      "loss": 1.3615,
+      "step": 9156
+    },
+    {
+      "epoch": 1.579746398688864,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.1171375657561677e-06,
+      "loss": 1.5104,
+      "step": 9157
+    },
+    {
+      "epoch": 1.579918916587596,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.1154630619959605e-06,
+      "loss": 1.4759,
+      "step": 9158
+    },
+    {
+      "epoch": 1.580091434486328,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.113789142371515e-06,
+      "loss": 1.3506,
+      "step": 9159
+    },
+    {
+      "epoch": 1.58026395238506,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.112115807006848e-06,
+      "loss": 1.3673,
+      "step": 9160
+    },
+    {
+      "epoch": 1.580436470283792,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.1104430560259227e-06,
+      "loss": 1.4195,
+      "step": 9161
+    },
+    {
+      "epoch": 1.580608988182524,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.108770889552677e-06,
+      "loss": 1.5068,
+      "step": 9162
+    },
+    {
+      "epoch": 1.580781506081256,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.107099307710988e-06,
+      "loss": 1.3579,
+      "step": 9163
+    },
+    {
+      "epoch": 1.580954023979988,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.105428310624699e-06,
+      "loss": 1.4347,
+      "step": 9164
+    },
+    {
+      "epoch": 1.58112654187872,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.103757898417609e-06,
+      "loss": 1.316,
+      "step": 9165
+    },
+    {
+      "epoch": 1.5812990597774519,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.102088071213468e-06,
+      "loss": 1.3706,
+      "step": 9166
+    },
+    {
+      "epoch": 1.5814715776761838,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.1004188291359973e-06,
+      "loss": 1.4241,
+      "step": 9167
+    },
+    {
+      "epoch": 1.5816440955749158,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.0987501723088564e-06,
+      "loss": 1.4717,
+      "step": 9168
+    },
+    {
+      "epoch": 1.5818166134736478,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.0970821008556706e-06,
+      "loss": 1.489,
+      "step": 9169
+    },
+    {
+      "epoch": 1.5819891313723797,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.0954146149000243e-06,
+      "loss": 1.3503,
+      "step": 9170
+    },
+    {
+      "epoch": 1.582161649271112,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.093747714565453e-06,
+      "loss": 1.4744,
+      "step": 9171
+    },
+    {
+      "epoch": 1.5823341671698439,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.092081399975451e-06,
+      "loss": 1.4649,
+      "step": 9172
+    },
+    {
+      "epoch": 1.5825066850685758,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.0904156712534718e-06,
+      "loss": 1.3834,
+      "step": 9173
+    },
+    {
+      "epoch": 1.582679202967308,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.088750528522917e-06,
+      "loss": 1.5699,
+      "step": 9174
+    },
+    {
+      "epoch": 1.58285172086604,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.087085971907159e-06,
+      "loss": 1.438,
+      "step": 9175
+    },
+    {
+      "epoch": 1.583024238764772,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.0854220015295125e-06,
+      "loss": 1.3781,
+      "step": 9176
+    },
+    {
+      "epoch": 1.583196756663504,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.083758617513256e-06,
+      "loss": 1.39,
+      "step": 9177
+    },
+    {
+      "epoch": 1.5833692745622359,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.0820958199816253e-06,
+      "loss": 1.3723,
+      "step": 9178
+    },
+    {
+      "epoch": 1.5835417924609678,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.0804336090578094e-06,
+      "loss": 1.4139,
+      "step": 9179
+    },
+    {
+      "epoch": 1.5837143103596998,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.0787719848649544e-06,
+      "loss": 1.3735,
+      "step": 9180
+    },
+    {
+      "epoch": 1.5838868282584317,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.0771109475261654e-06,
+      "loss": 1.4363,
+      "step": 9181
+    },
+    {
+      "epoch": 1.5840593461571637,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.075450497164503e-06,
+      "loss": 1.4841,
+      "step": 9182
+    },
+    {
+      "epoch": 1.5842318640558957,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.0737906339029813e-06,
+      "loss": 1.4085,
+      "step": 9183
+    },
+    {
+      "epoch": 1.5844043819546276,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.0721313578645773e-06,
+      "loss": 1.3854,
+      "step": 9184
+    },
+    {
+      "epoch": 1.5845768998533598,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.0704726691722134e-06,
+      "loss": 1.4014,
+      "step": 9185
+    },
+    {
+      "epoch": 1.5847494177520918,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.068814567948785e-06,
+      "loss": 1.4236,
+      "step": 9186
+    },
+    {
+      "epoch": 1.5849219356508237,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.067157054317127e-06,
+      "loss": 1.4418,
+      "step": 9187
+    },
+    {
+      "epoch": 1.585094453549556,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.065500128400041e-06,
+      "loss": 1.546,
+      "step": 9188
+    },
+    {
+      "epoch": 1.5852669714482879,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.0638437903202825e-06,
+      "loss": 1.336,
+      "step": 9189
+    },
+    {
+      "epoch": 1.5854394893470198,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.0621880402005644e-06,
+      "loss": 1.4443,
+      "step": 9190
+    },
+    {
+      "epoch": 1.5856120072457518,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.0605328781635524e-06,
+      "loss": 1.5043,
+      "step": 9191
+    },
+    {
+      "epoch": 1.5857845251444838,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.058878304331877e-06,
+      "loss": 1.3014,
+      "step": 9192
+    },
+    {
+      "epoch": 1.5859570430432157,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.0572243188281093e-06,
+      "loss": 1.3639,
+      "step": 9193
+    },
+    {
+      "epoch": 1.5861295609419477,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.055570921774799e-06,
+      "loss": 1.2815,
+      "step": 9194
+    },
+    {
+      "epoch": 1.5863020788406796,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.0539181132944273e-06,
+      "loss": 1.4472,
+      "step": 9195
+    },
+    {
+      "epoch": 1.5864745967394116,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.0522658935094565e-06,
+      "loss": 1.3883,
+      "step": 9196
+    },
+    {
+      "epoch": 1.5866471146381436,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.0506142625422863e-06,
+      "loss": 1.3905,
+      "step": 9197
+    },
+    {
+      "epoch": 1.5868196325368757,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.0489632205152787e-06,
+      "loss": 1.5331,
+      "step": 9198
+    },
+    {
+      "epoch": 1.5869921504356077,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.047312767550761e-06,
+      "loss": 1.4573,
+      "step": 9199
+    },
+    {
+      "epoch": 1.5871646683343397,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.045662903771002e-06,
+      "loss": 1.4495,
+      "step": 9200
+    },
+    {
+      "epoch": 1.5871646683343397,
+      "eval_loss": 1.407097578048706,
+      "eval_runtime": 10.8771,
+      "eval_samples_per_second": 94.143,
+      "eval_steps_per_second": 23.536,
+      "step": 9200
+    },
+    {
+      "epoch": 1.5873371862330716,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.044013629298235e-06,
+      "loss": 1.4389,
+      "step": 9201
+    },
+    {
+      "epoch": 1.5875097041318038,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.042364944254651e-06,
+      "loss": 1.4526,
+      "step": 9202
+    },
+    {
+      "epoch": 1.5876822220305358,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.040716848762393e-06,
+      "loss": 1.3507,
+      "step": 9203
+    },
+    {
+      "epoch": 1.5878547399292677,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.0390693429435626e-06,
+      "loss": 1.4062,
+      "step": 9204
+    },
+    {
+      "epoch": 1.5880272578279997,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.0374224269202204e-06,
+      "loss": 1.4531,
+      "step": 9205
+    },
+    {
+      "epoch": 1.5881997757267317,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.0357761008143715e-06,
+      "loss": 1.451,
+      "step": 9206
+    },
+    {
+      "epoch": 1.5883722936254636,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.034130364747997e-06,
+      "loss": 1.4734,
+      "step": 9207
+    },
+    {
+      "epoch": 1.5885448115241956,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.0324852188430167e-06,
+      "loss": 1.339,
+      "step": 9208
+    },
+    {
+      "epoch": 1.5887173294229275,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.030840663221313e-06,
+      "loss": 1.4443,
+      "step": 9209
+    },
+    {
+      "epoch": 1.5888898473216595,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.0291966980047282e-06,
+      "loss": 1.4756,
+      "step": 9210
+    },
+    {
+      "epoch": 1.5890623652203915,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.027553323315055e-06,
+      "loss": 1.4799,
+      "step": 9211
+    },
+    {
+      "epoch": 1.5892348831191236,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.0259105392740462e-06,
+      "loss": 1.3639,
+      "step": 9212
+    },
+    {
+      "epoch": 1.5894074010178556,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.0242683460034084e-06,
+      "loss": 1.5483,
+      "step": 9213
+    },
+    {
+      "epoch": 1.5895799189165876,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.022626743624807e-06,
+      "loss": 1.5209,
+      "step": 9214
+    },
+    {
+      "epoch": 1.5897524368153197,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.02098573225986e-06,
+      "loss": 1.3927,
+      "step": 9215
+    },
+    {
+      "epoch": 1.5899249547140517,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.0193453120301496e-06,
+      "loss": 1.3861,
+      "step": 9216
+    },
+    {
+      "epoch": 1.5900974726127837,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.0177054830571974e-06,
+      "loss": 1.3874,
+      "step": 9217
+    },
+    {
+      "epoch": 1.5902699905115156,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.0160662454625045e-06,
+      "loss": 1.3126,
+      "step": 9218
+    },
+    {
+      "epoch": 1.5904425084102476,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.0144275993675077e-06,
+      "loss": 1.359,
+      "step": 9219
+    },
+    {
+      "epoch": 1.5906150263089796,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.01278954489361e-06,
+      "loss": 1.5007,
+      "step": 9220
+    },
+    {
+      "epoch": 1.5907875442077115,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.0111520821621686e-06,
+      "loss": 1.4547,
+      "step": 9221
+    },
+    {
+      "epoch": 1.5909600621064435,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.0095152112944983e-06,
+      "loss": 1.4354,
+      "step": 9222
+    },
+    {
+      "epoch": 1.5911325800051754,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.0078789324118686e-06,
+      "loss": 1.3445,
+      "step": 9223
+    },
+    {
+      "epoch": 1.5913050979039074,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.0062432456355064e-06,
+      "loss": 1.3494,
+      "step": 9224
+    },
+    {
+      "epoch": 1.5914776158026396,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.0046081510865865e-06,
+      "loss": 1.4539,
+      "step": 9225
+    },
+    {
+      "epoch": 1.5916501337013715,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.0029736488862585e-06,
+      "loss": 1.4198,
+      "step": 9226
+    },
+    {
+      "epoch": 1.5918226516001035,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.0013397391556042e-06,
+      "loss": 1.4261,
+      "step": 9227
+    },
+    {
+      "epoch": 1.5919951694988355,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.9997064220156857e-06,
+      "loss": 1.3818,
+      "step": 9228
+    },
+    {
+      "epoch": 1.5921676873975676,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.9980736975875005e-06,
+      "loss": 1.3872,
+      "step": 9229
+    },
+    {
+      "epoch": 1.5923402052962996,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.9964415659920123e-06,
+      "loss": 1.4362,
+      "step": 9230
+    },
+    {
+      "epoch": 1.5925127231950316,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.9948100273501456e-06,
+      "loss": 1.3703,
+      "step": 9231
+    },
+    {
+      "epoch": 1.5926852410937635,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.993179081782769e-06,
+      "loss": 1.3841,
+      "step": 9232
+    },
+    {
+      "epoch": 1.5928577589924955,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.991548729410715e-06,
+      "loss": 1.4737,
+      "step": 9233
+    },
+    {
+      "epoch": 1.5930302768912274,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.9899189703547694e-06,
+      "loss": 1.4497,
+      "step": 9234
+    },
+    {
+      "epoch": 1.5932027947899594,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.988289804735677e-06,
+      "loss": 1.337,
+      "step": 9235
+    },
+    {
+      "epoch": 1.5933753126886914,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.986661232674134e-06,
+      "loss": 1.4287,
+      "step": 9236
+    },
+    {
+      "epoch": 1.5935478305874233,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.9850332542908e-06,
+      "loss": 1.4264,
+      "step": 9237
+    },
+    {
+      "epoch": 1.5937203484861553,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.9834058697062776e-06,
+      "loss": 1.5309,
+      "step": 9238
+    },
+    {
+      "epoch": 1.5938928663848875,
+      "grad_norm": 0.625,
+      "learning_rate": 1.9817790790411428e-06,
+      "loss": 1.3928,
+      "step": 9239
+    },
+    {
+      "epoch": 1.5940653842836194,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.980152882415911e-06,
+      "loss": 1.342,
+      "step": 9240
+    },
+    {
+      "epoch": 1.5942379021823514,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.978527279951065e-06,
+      "loss": 1.3418,
+      "step": 9241
+    },
+    {
+      "epoch": 1.5944104200810834,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.976902271767037e-06,
+      "loss": 1.4355,
+      "step": 9242
+    },
+    {
+      "epoch": 1.5945829379798155,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.9752778579842214e-06,
+      "loss": 1.3677,
+      "step": 9243
+    },
+    {
+      "epoch": 1.5947554558785475,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.973654038722962e-06,
+      "loss": 1.45,
+      "step": 9244
+    },
+    {
+      "epoch": 1.5949279737772795,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.9720308141035647e-06,
+      "loss": 1.407,
+      "step": 9245
+    },
+    {
+      "epoch": 1.5951004916760114,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.9704081842462806e-06,
+      "loss": 1.3835,
+      "step": 9246
+    },
+    {
+      "epoch": 1.5952730095747434,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.9687861492713323e-06,
+      "loss": 1.2955,
+      "step": 9247
+    },
+    {
+      "epoch": 1.5954455274734753,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.967164709298889e-06,
+      "loss": 1.3696,
+      "step": 9248
+    },
+    {
+      "epoch": 1.5956180453722073,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.965543864449071e-06,
+      "loss": 1.3885,
+      "step": 9249
+    },
+    {
+      "epoch": 1.5957905632709393,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.9639236148419705e-06,
+      "loss": 1.4505,
+      "step": 9250
+    },
+    {
+      "epoch": 1.5959630811696712,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.962303960597618e-06,
+      "loss": 1.4014,
+      "step": 9251
+    },
+    {
+      "epoch": 1.5961355990684032,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.9606849018360096e-06,
+      "loss": 1.4158,
+      "step": 9252
+    },
+    {
+      "epoch": 1.5963081169671354,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.9590664386770953e-06,
+      "loss": 1.4112,
+      "step": 9253
+    },
+    {
+      "epoch": 1.5964806348658673,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.957448571240782e-06,
+      "loss": 1.3941,
+      "step": 9254
+    },
+    {
+      "epoch": 1.5966531527645993,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.9558312996469296e-06,
+      "loss": 1.4583,
+      "step": 9255
+    },
+    {
+      "epoch": 1.5968256706633315,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.954214624015358e-06,
+      "loss": 1.4152,
+      "step": 9256
+    },
+    {
+      "epoch": 1.5969981885620634,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.9525985444658393e-06,
+      "loss": 1.4532,
+      "step": 9257
+    },
+    {
+      "epoch": 1.5971707064607954,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.9509830611181047e-06,
+      "loss": 1.3964,
+      "step": 9258
+    },
+    {
+      "epoch": 1.5973432243595274,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.9493681740918335e-06,
+      "loss": 1.4545,
+      "step": 9259
+    },
+    {
+      "epoch": 1.5975157422582593,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.9477538835066747e-06,
+      "loss": 1.3786,
+      "step": 9260
+    },
+    {
+      "epoch": 1.5976882601569913,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.9461401894822187e-06,
+      "loss": 1.336,
+      "step": 9261
+    },
+    {
+      "epoch": 1.5978607780557232,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.944527092138018e-06,
+      "loss": 1.431,
+      "step": 9262
+    },
+    {
+      "epoch": 1.5980332959544552,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.9429145915935886e-06,
+      "loss": 1.4712,
+      "step": 9263
+    },
+    {
+      "epoch": 1.5982058138531872,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.941302687968386e-06,
+      "loss": 1.3915,
+      "step": 9264
+    },
+    {
+      "epoch": 1.5983783317519191,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.939691381381834e-06,
+      "loss": 1.3599,
+      "step": 9265
+    },
+    {
+      "epoch": 1.5985508496506513,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.938080671953307e-06,
+      "loss": 1.294,
+      "step": 9266
+    },
+    {
+      "epoch": 1.5987233675493833,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.9364705598021373e-06,
+      "loss": 1.407,
+      "step": 9267
+    },
+    {
+      "epoch": 1.5988958854481152,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.9348610450476112e-06,
+      "loss": 1.3775,
+      "step": 9268
+    },
+    {
+      "epoch": 1.5990684033468472,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.9332521278089756e-06,
+      "loss": 1.473,
+      "step": 9269
+    },
+    {
+      "epoch": 1.5992409212455794,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.9316438082054203e-06,
+      "loss": 1.4518,
+      "step": 9270
+    },
+    {
+      "epoch": 1.5994134391443113,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.930036086356111e-06,
+      "loss": 1.4212,
+      "step": 9271
+    },
+    {
+      "epoch": 1.5995859570430433,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.928428962380148e-06,
+      "loss": 1.4948,
+      "step": 9272
+    },
+    {
+      "epoch": 1.5997584749417753,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.9268224363966016e-06,
+      "loss": 1.3819,
+      "step": 9273
+    },
+    {
+      "epoch": 1.5999309928405072,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.925216508524492e-06,
+      "loss": 1.4281,
+      "step": 9274
+    },
+    {
+      "epoch": 1.6001035107392392,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.9236111788827983e-06,
+      "loss": 1.4988,
+      "step": 9275
+    },
+    {
+      "epoch": 1.6002760286379711,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.922006447590451e-06,
+      "loss": 1.4607,
+      "step": 9276
+    },
+    {
+      "epoch": 1.600448546536703,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.920402314766343e-06,
+      "loss": 1.3998,
+      "step": 9277
+    },
+    {
+      "epoch": 1.600621064435435,
+      "grad_norm": 0.625,
+      "learning_rate": 1.9187987805293096e-06,
+      "loss": 1.4442,
+      "step": 9278
+    },
+    {
+      "epoch": 1.600793582334167,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.9171958449981587e-06,
+      "loss": 1.3879,
+      "step": 9279
+    },
+    {
+      "epoch": 1.6009661002328992,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.9155935082916467e-06,
+      "loss": 1.3548,
+      "step": 9280
+    },
+    {
+      "epoch": 1.6011386181316312,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.913991770528475e-06,
+      "loss": 1.5053,
+      "step": 9281
+    },
+    {
+      "epoch": 1.6013111360303631,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.9123906318273234e-06,
+      "loss": 1.4098,
+      "step": 9282
+    },
+    {
+      "epoch": 1.6014836539290953,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.9107900923068013e-06,
+      "loss": 1.4627,
+      "step": 9283
+    },
+    {
+      "epoch": 1.6016561718278273,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.909190152085497e-06,
+      "loss": 1.4288,
+      "step": 9284
+    },
+    {
+      "epoch": 1.6018286897265592,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.9075908112819387e-06,
+      "loss": 1.376,
+      "step": 9285
+    },
+    {
+      "epoch": 1.6020012076252912,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.9059920700146163e-06,
+      "loss": 1.3245,
+      "step": 9286
+    },
+    {
+      "epoch": 1.6021737255240232,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.9043939284019753e-06,
+      "loss": 1.4066,
+      "step": 9287
+    },
+    {
+      "epoch": 1.6023462434227551,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.9027963865624166e-06,
+      "loss": 1.3476,
+      "step": 9288
+    },
+    {
+      "epoch": 1.602518761321487,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.9011994446142935e-06,
+      "loss": 1.3944,
+      "step": 9289
+    },
+    {
+      "epoch": 1.602691279220219,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.8996031026759232e-06,
+      "loss": 1.4003,
+      "step": 9290
+    },
+    {
+      "epoch": 1.602863797118951,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.8980073608655625e-06,
+      "loss": 1.4291,
+      "step": 9291
+    },
+    {
+      "epoch": 1.603036315017683,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.8964122193014456e-06,
+      "loss": 1.4436,
+      "step": 9292
+    },
+    {
+      "epoch": 1.603208832916415,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.8948176781017435e-06,
+      "loss": 1.3513,
+      "step": 9293
+    },
+    {
+      "epoch": 1.603381350815147,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.8932237373845874e-06,
+      "loss": 1.4303,
+      "step": 9294
+    },
+    {
+      "epoch": 1.603553868713879,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.891630397268076e-06,
+      "loss": 1.287,
+      "step": 9295
+    },
+    {
+      "epoch": 1.603726386612611,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.890037657870245e-06,
+      "loss": 1.4175,
+      "step": 9296
+    },
+    {
+      "epoch": 1.6038989045113432,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.8884455193090989e-06,
+      "loss": 1.4304,
+      "step": 9297
+    },
+    {
+      "epoch": 1.6040714224100752,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.886853981702591e-06,
+      "loss": 1.4971,
+      "step": 9298
+    },
+    {
+      "epoch": 1.6042439403088071,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.8852630451686337e-06,
+      "loss": 1.4024,
+      "step": 9299
+    },
+    {
+      "epoch": 1.604416458207539,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.883672709825094e-06,
+      "loss": 1.4162,
+      "step": 9300
+    },
+    {
+      "epoch": 1.604416458207539,
+      "eval_loss": 1.4071283340454102,
+      "eval_runtime": 10.92,
+      "eval_samples_per_second": 93.773,
+      "eval_steps_per_second": 23.443,
+      "step": 9300
+    },
+    {
+      "epoch": 1.604588976106271,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.8820829757897952e-06,
+      "loss": 1.3499,
+      "step": 9301
+    },
+    {
+      "epoch": 1.604761494005003,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.8804938431805064e-06,
+      "loss": 1.4917,
+      "step": 9302
+    },
+    {
+      "epoch": 1.604934011903735,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.8789053121149736e-06,
+      "loss": 1.3938,
+      "step": 9303
+    },
+    {
+      "epoch": 1.605106529802467,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.877317382710875e-06,
+      "loss": 1.4025,
+      "step": 9304
+    },
+    {
+      "epoch": 1.605279047701199,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.8757300550858571e-06,
+      "loss": 1.3692,
+      "step": 9305
+    },
+    {
+      "epoch": 1.6054515655999309,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.8741433293575196e-06,
+      "loss": 1.5194,
+      "step": 9306
+    },
+    {
+      "epoch": 1.605624083498663,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.8725572056434172e-06,
+      "loss": 1.4504,
+      "step": 9307
+    },
+    {
+      "epoch": 1.605796601397395,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.8709716840610592e-06,
+      "loss": 1.4591,
+      "step": 9308
+    },
+    {
+      "epoch": 1.605969119296127,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.869386764727914e-06,
+      "loss": 1.4517,
+      "step": 9309
+    },
+    {
+      "epoch": 1.606141637194859,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.8678024477613954e-06,
+      "loss": 1.4905,
+      "step": 9310
+    },
+    {
+      "epoch": 1.606314155093591,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.8662187332788861e-06,
+      "loss": 1.4565,
+      "step": 9311
+    },
+    {
+      "epoch": 1.606486672992323,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.8646356213977167e-06,
+      "loss": 1.3746,
+      "step": 9312
+    },
+    {
+      "epoch": 1.606659190891055,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.863053112235168e-06,
+      "loss": 1.4626,
+      "step": 9313
+    },
+    {
+      "epoch": 1.606831708789787,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.8614712059084927e-06,
+      "loss": 1.4173,
+      "step": 9314
+    },
+    {
+      "epoch": 1.607004226688519,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.8598899025348771e-06,
+      "loss": 1.3984,
+      "step": 9315
+    },
+    {
+      "epoch": 1.607176744587251,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.858309202231483e-06,
+      "loss": 1.4292,
+      "step": 9316
+    },
+    {
+      "epoch": 1.6073492624859829,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.8567291051154135e-06,
+      "loss": 1.4097,
+      "step": 9317
+    },
+    {
+      "epoch": 1.6075217803847148,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.8551496113037336e-06,
+      "loss": 1.4493,
+      "step": 9318
+    },
+    {
+      "epoch": 1.6076942982834468,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.8535707209134613e-06,
+      "loss": 1.2968,
+      "step": 9319
+    },
+    {
+      "epoch": 1.6078668161821787,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.8519924340615713e-06,
+      "loss": 1.4715,
+      "step": 9320
+    },
+    {
+      "epoch": 1.608039334080911,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.8504147508649928e-06,
+      "loss": 1.3962,
+      "step": 9321
+    },
+    {
+      "epoch": 1.608211851979643,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.8488376714406131e-06,
+      "loss": 1.4307,
+      "step": 9322
+    },
+    {
+      "epoch": 1.6083843698783749,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.8472611959052644e-06,
+      "loss": 1.4533,
+      "step": 9323
+    },
+    {
+      "epoch": 1.608556887777107,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.8456853243757522e-06,
+      "loss": 1.4285,
+      "step": 9324
+    },
+    {
+      "epoch": 1.608729405675839,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.8441100569688186e-06,
+      "loss": 1.4613,
+      "step": 9325
+    },
+    {
+      "epoch": 1.608901923574571,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.8425353938011702e-06,
+      "loss": 1.4717,
+      "step": 9326
+    },
+    {
+      "epoch": 1.609074441473303,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.8409613349894761e-06,
+      "loss": 1.4131,
+      "step": 9327
+    },
+    {
+      "epoch": 1.6092469593720349,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.839387880650343e-06,
+      "loss": 1.4449,
+      "step": 9328
+    },
+    {
+      "epoch": 1.6094194772707668,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.837815030900345e-06,
+      "loss": 1.3643,
+      "step": 9329
+    },
+    {
+      "epoch": 1.6095919951694988,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.8362427858560094e-06,
+      "loss": 1.379,
+      "step": 9330
+    },
+    {
+      "epoch": 1.6097645130682308,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.8346711456338185e-06,
+      "loss": 1.4487,
+      "step": 9331
+    },
+    {
+      "epoch": 1.6099370309669627,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.8331001103502077e-06,
+      "loss": 1.411,
+      "step": 9332
+    },
+    {
+      "epoch": 1.6101095488656947,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.8315296801215721e-06,
+      "loss": 1.3986,
+      "step": 9333
+    },
+    {
+      "epoch": 1.6102820667644266,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.8299598550642528e-06,
+      "loss": 1.5243,
+      "step": 9334
+    },
+    {
+      "epoch": 1.6104545846631588,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.82839063529456e-06,
+      "loss": 1.4422,
+      "step": 9335
+    },
+    {
+      "epoch": 1.6106271025618908,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.8268220209287457e-06,
+      "loss": 1.3994,
+      "step": 9336
+    },
+    {
+      "epoch": 1.6107996204606227,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.8252540120830253e-06,
+      "loss": 1.3916,
+      "step": 9337
+    },
+    {
+      "epoch": 1.610972138359355,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.823686608873565e-06,
+      "loss": 1.4315,
+      "step": 9338
+    },
+    {
+      "epoch": 1.611144656258087,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.8221198114164885e-06,
+      "loss": 1.3636,
+      "step": 9339
+    },
+    {
+      "epoch": 1.6113171741568189,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.8205536198278739e-06,
+      "loss": 1.4466,
+      "step": 9340
+    },
+    {
+      "epoch": 1.6114896920555508,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.8189880342237576e-06,
+      "loss": 1.4002,
+      "step": 9341
+    },
+    {
+      "epoch": 1.6116622099542828,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.81742305472012e-06,
+      "loss": 1.5221,
+      "step": 9342
+    },
+    {
+      "epoch": 1.6118347278530147,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.8158586814329126e-06,
+      "loss": 1.3755,
+      "step": 9343
+    },
+    {
+      "epoch": 1.6120072457517467,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.8142949144780297e-06,
+      "loss": 1.4148,
+      "step": 9344
+    },
+    {
+      "epoch": 1.6121797636504787,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.8127317539713273e-06,
+      "loss": 1.4302,
+      "step": 9345
+    },
+    {
+      "epoch": 1.6123522815492106,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.811169200028615e-06,
+      "loss": 1.4514,
+      "step": 9346
+    },
+    {
+      "epoch": 1.6125247994479426,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.8096072527656505e-06,
+      "loss": 1.4674,
+      "step": 9347
+    },
+    {
+      "epoch": 1.6126973173466748,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.8080459122981609e-06,
+      "loss": 1.5049,
+      "step": 9348
+    },
+    {
+      "epoch": 1.6128698352454067,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.8064851787418126e-06,
+      "loss": 1.4882,
+      "step": 9349
+    },
+    {
+      "epoch": 1.6130423531441387,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.8049250522122386e-06,
+      "loss": 1.4113,
+      "step": 9350
+    },
+    {
+      "epoch": 1.6132148710428706,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.8033655328250222e-06,
+      "loss": 1.5452,
+      "step": 9351
+    },
+    {
+      "epoch": 1.6133873889416028,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.8018066206957007e-06,
+      "loss": 1.5255,
+      "step": 9352
+    },
+    {
+      "epoch": 1.6135599068403348,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.800248315939771e-06,
+      "loss": 1.4721,
+      "step": 9353
+    },
+    {
+      "epoch": 1.6137324247390668,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.7986906186726815e-06,
+      "loss": 1.3949,
+      "step": 9354
+    },
+    {
+      "epoch": 1.6139049426377987,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.7971335290098314e-06,
+      "loss": 1.4231,
+      "step": 9355
+    },
+    {
+      "epoch": 1.6140774605365307,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.7955770470665879e-06,
+      "loss": 1.441,
+      "step": 9356
+    },
+    {
+      "epoch": 1.6142499784352626,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.794021172958258e-06,
+      "loss": 1.4214,
+      "step": 9357
+    },
+    {
+      "epoch": 1.6144224963339946,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.7924659068001094e-06,
+      "loss": 1.402,
+      "step": 9358
+    },
+    {
+      "epoch": 1.6145950142327266,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.7909112487073754e-06,
+      "loss": 1.4344,
+      "step": 9359
+    },
+    {
+      "epoch": 1.6147675321314585,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.7893571987952262e-06,
+      "loss": 1.5812,
+      "step": 9360
+    },
+    {
+      "epoch": 1.6149400500301905,
+      "grad_norm": 0.625,
+      "learning_rate": 1.7878037571787987e-06,
+      "loss": 1.4667,
+      "step": 9361
+    },
+    {
+      "epoch": 1.6151125679289227,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.7862509239731806e-06,
+      "loss": 1.4051,
+      "step": 9362
+    },
+    {
+      "epoch": 1.6152850858276546,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.7846986992934157e-06,
+      "loss": 1.3622,
+      "step": 9363
+    },
+    {
+      "epoch": 1.6154576037263866,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.7831470832545028e-06,
+      "loss": 1.4268,
+      "step": 9364
+    },
+    {
+      "epoch": 1.6156301216251188,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.781596075971398e-06,
+      "loss": 1.3411,
+      "step": 9365
+    },
+    {
+      "epoch": 1.6158026395238507,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.7800456775590014e-06,
+      "loss": 1.4521,
+      "step": 9366
+    },
+    {
+      "epoch": 1.6159751574225827,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.7784958881321868e-06,
+      "loss": 1.4027,
+      "step": 9367
+    },
+    {
+      "epoch": 1.6161476753213146,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.7769467078057644e-06,
+      "loss": 1.4094,
+      "step": 9368
+    },
+    {
+      "epoch": 1.6163201932200466,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.7753981366945105e-06,
+      "loss": 1.5092,
+      "step": 9369
+    },
+    {
+      "epoch": 1.6164927111187786,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.773850174913152e-06,
+      "loss": 1.3982,
+      "step": 9370
+    },
+    {
+      "epoch": 1.6166652290175105,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.7723028225763727e-06,
+      "loss": 1.3922,
+      "step": 9371
+    },
+    {
+      "epoch": 1.6168377469162425,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.770756079798809e-06,
+      "loss": 1.4047,
+      "step": 9372
+    },
+    {
+      "epoch": 1.6170102648149745,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.7692099466950551e-06,
+      "loss": 1.4964,
+      "step": 9373
+    },
+    {
+      "epoch": 1.6171827827137064,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.7676644233796525e-06,
+      "loss": 1.4487,
+      "step": 9374
+    },
+    {
+      "epoch": 1.6173553006124386,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.766119509967109e-06,
+      "loss": 1.5349,
+      "step": 9375
+    },
+    {
+      "epoch": 1.6175278185111706,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.7645752065718814e-06,
+      "loss": 1.4116,
+      "step": 9376
+    },
+    {
+      "epoch": 1.6177003364099025,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.7630315133083786e-06,
+      "loss": 1.3519,
+      "step": 9377
+    },
+    {
+      "epoch": 1.6178728543086345,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.7614884302909719e-06,
+      "loss": 1.3943,
+      "step": 9378
+    },
+    {
+      "epoch": 1.6180453722073667,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.7599459576339729e-06,
+      "loss": 1.5073,
+      "step": 9379
+    },
+    {
+      "epoch": 1.6182178901060986,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.758404095451669e-06,
+      "loss": 1.402,
+      "step": 9380
+    },
+    {
+      "epoch": 1.6183904080048306,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.7568628438582825e-06,
+      "loss": 1.4218,
+      "step": 9381
+    },
+    {
+      "epoch": 1.6185629259035625,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.7553222029680028e-06,
+      "loss": 1.4593,
+      "step": 9382
+    },
+    {
+      "epoch": 1.6187354438022945,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.7537821728949688e-06,
+      "loss": 1.3169,
+      "step": 9383
+    },
+    {
+      "epoch": 1.6189079617010265,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.752242753753276e-06,
+      "loss": 1.3619,
+      "step": 9384
+    },
+    {
+      "epoch": 1.6190804795997584,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.7507039456569753e-06,
+      "loss": 1.4082,
+      "step": 9385
+    },
+    {
+      "epoch": 1.6192529974984904,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.749165748720072e-06,
+      "loss": 1.4374,
+      "step": 9386
+    },
+    {
+      "epoch": 1.6194255153972223,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.7476281630565195e-06,
+      "loss": 1.4549,
+      "step": 9387
+    },
+    {
+      "epoch": 1.6195980332959543,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.74609118878024e-06,
+      "loss": 1.4547,
+      "step": 9388
+    },
+    {
+      "epoch": 1.6197705511946865,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.744554826005096e-06,
+      "loss": 1.4083,
+      "step": 9389
+    },
+    {
+      "epoch": 1.6199430690934185,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.7430190748449105e-06,
+      "loss": 1.4198,
+      "step": 9390
+    },
+    {
+      "epoch": 1.6201155869921504,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.7414839354134684e-06,
+      "loss": 1.3815,
+      "step": 9391
+    },
+    {
+      "epoch": 1.6202881048908824,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.739949407824496e-06,
+      "loss": 1.4645,
+      "step": 9392
+    },
+    {
+      "epoch": 1.6204606227896146,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.7384154921916819e-06,
+      "loss": 1.4625,
+      "step": 9393
+    },
+    {
+      "epoch": 1.6206331406883465,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.7368821886286691e-06,
+      "loss": 1.3335,
+      "step": 9394
+    },
+    {
+      "epoch": 1.6208056585870785,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.7353494972490548e-06,
+      "loss": 1.5648,
+      "step": 9395
+    },
+    {
+      "epoch": 1.6209781764858104,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.7338174181663891e-06,
+      "loss": 1.4175,
+      "step": 9396
+    },
+    {
+      "epoch": 1.6211506943845424,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.7322859514941813e-06,
+      "loss": 1.3912,
+      "step": 9397
+    },
+    {
+      "epoch": 1.6213232122832744,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.7307550973458843e-06,
+      "loss": 1.3587,
+      "step": 9398
+    },
+    {
+      "epoch": 1.6214957301820063,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.7292248558349233e-06,
+      "loss": 1.5153,
+      "step": 9399
+    },
+    {
+      "epoch": 1.6216682480807383,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.7276952270746606e-06,
+      "loss": 1.3662,
+      "step": 9400
+    },
+    {
+      "epoch": 1.6216682480807383,
+      "eval_loss": 1.4071524143218994,
+      "eval_runtime": 10.8335,
+      "eval_samples_per_second": 94.522,
+      "eval_steps_per_second": 23.631,
+      "step": 9400
+    },
+    {
+      "epoch": 1.6218407659794702,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.7261662111784229e-06,
+      "loss": 1.4412,
+      "step": 9401
+    },
+    {
+      "epoch": 1.6220132838782022,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.7246378082594906e-06,
+      "loss": 1.3786,
+      "step": 9402
+    },
+    {
+      "epoch": 1.6221858017769344,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.7231100184310955e-06,
+      "loss": 1.3919,
+      "step": 9403
+    },
+    {
+      "epoch": 1.6223583196756663,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.7215828418064263e-06,
+      "loss": 1.3885,
+      "step": 9404
+    },
+    {
+      "epoch": 1.6225308375743983,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.7200562784986263e-06,
+      "loss": 1.3885,
+      "step": 9405
+    },
+    {
+      "epoch": 1.6227033554731305,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.7185303286207932e-06,
+      "loss": 1.4058,
+      "step": 9406
+    },
+    {
+      "epoch": 1.6228758733718625,
+      "grad_norm": 0.625,
+      "learning_rate": 1.7170049922859788e-06,
+      "loss": 1.4768,
+      "step": 9407
+    },
+    {
+      "epoch": 1.6230483912705944,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.7154802696071882e-06,
+      "loss": 1.4446,
+      "step": 9408
+    },
+    {
+      "epoch": 1.6232209091693264,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.7139561606973832e-06,
+      "loss": 1.4344,
+      "step": 9409
+    },
+    {
+      "epoch": 1.6233934270680583,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.7124326656694823e-06,
+      "loss": 1.4298,
+      "step": 9410
+    },
+    {
+      "epoch": 1.6235659449667903,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.7109097846363476e-06,
+      "loss": 1.435,
+      "step": 9411
+    },
+    {
+      "epoch": 1.6237384628655223,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.7093875177108131e-06,
+      "loss": 1.4331,
+      "step": 9412
+    },
+    {
+      "epoch": 1.6239109807642542,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.7078658650056524e-06,
+      "loss": 1.405,
+      "step": 9413
+    },
+    {
+      "epoch": 1.6240834986629862,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.7063448266335991e-06,
+      "loss": 1.5126,
+      "step": 9414
+    },
+    {
+      "epoch": 1.6242560165617181,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.7048244027073424e-06,
+      "loss": 1.3834,
+      "step": 9415
+    },
+    {
+      "epoch": 1.6244285344604503,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.7033045933395242e-06,
+      "loss": 1.4099,
+      "step": 9416
+    },
+    {
+      "epoch": 1.6246010523591823,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.7017853986427423e-06,
+      "loss": 1.3877,
+      "step": 9417
+    },
+    {
+      "epoch": 1.6247735702579142,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.70026681872955e-06,
+      "loss": 1.3025,
+      "step": 9418
+    },
+    {
+      "epoch": 1.6249460881566462,
+      "grad_norm": 0.625,
+      "learning_rate": 1.6987488537124454e-06,
+      "loss": 1.4223,
+      "step": 9419
+    },
+    {
+      "epoch": 1.6251186060553784,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.6972315037039e-06,
+      "loss": 1.4389,
+      "step": 9420
+    },
+    {
+      "epoch": 1.6252911239541104,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.69571476881632e-06,
+      "loss": 1.4033,
+      "step": 9421
+    },
+    {
+      "epoch": 1.6254636418528423,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.6941986491620754e-06,
+      "loss": 1.3848,
+      "step": 9422
+    },
+    {
+      "epoch": 1.6256361597515743,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.6926831448534964e-06,
+      "loss": 1.448,
+      "step": 9423
+    },
+    {
+      "epoch": 1.6258086776503062,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.6911682560028552e-06,
+      "loss": 1.4198,
+      "step": 9424
+    },
+    {
+      "epoch": 1.6259811955490382,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.6896539827223845e-06,
+      "loss": 1.4735,
+      "step": 9425
+    },
+    {
+      "epoch": 1.6261537134477702,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.6881403251242723e-06,
+      "loss": 1.2885,
+      "step": 9426
+    },
+    {
+      "epoch": 1.6263262313465021,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.6866272833206598e-06,
+      "loss": 1.332,
+      "step": 9427
+    },
+    {
+      "epoch": 1.626498749245234,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.685114857423643e-06,
+      "loss": 1.3159,
+      "step": 9428
+    },
+    {
+      "epoch": 1.626671267143966,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.6836030475452737e-06,
+      "loss": 1.3439,
+      "step": 9429
+    },
+    {
+      "epoch": 1.6268437850426982,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.6820918537975484e-06,
+      "loss": 1.4719,
+      "step": 9430
+    },
+    {
+      "epoch": 1.6270163029414302,
+      "grad_norm": 0.625,
+      "learning_rate": 1.6805812762924367e-06,
+      "loss": 1.4097,
+      "step": 9431
+    },
+    {
+      "epoch": 1.6271888208401621,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.6790713151418403e-06,
+      "loss": 1.3939,
+      "step": 9432
+    },
+    {
+      "epoch": 1.6273613387388943,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.6775619704576385e-06,
+      "loss": 1.4855,
+      "step": 9433
+    },
+    {
+      "epoch": 1.6275338566376263,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.676053242351644e-06,
+      "loss": 1.3427,
+      "step": 9434
+    },
+    {
+      "epoch": 1.6277063745363582,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.6745451309356354e-06,
+      "loss": 1.4919,
+      "step": 9435
+    },
+    {
+      "epoch": 1.6278788924350902,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.673037636321344e-06,
+      "loss": 1.4468,
+      "step": 9436
+    },
+    {
+      "epoch": 1.6280514103338222,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.6715307586204533e-06,
+      "loss": 1.4331,
+      "step": 9437
+    },
+    {
+      "epoch": 1.6282239282325541,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.6700244979446034e-06,
+      "loss": 1.3679,
+      "step": 9438
+    },
+    {
+      "epoch": 1.628396446131286,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.6685188544053855e-06,
+      "loss": 1.317,
+      "step": 9439
+    },
+    {
+      "epoch": 1.628568964030018,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.66701382811435e-06,
+      "loss": 1.4306,
+      "step": 9440
+    },
+    {
+      "epoch": 1.62874148192875,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.6655094191829957e-06,
+      "loss": 1.3777,
+      "step": 9441
+    },
+    {
+      "epoch": 1.628913999827482,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.6640056277227824e-06,
+      "loss": 1.3518,
+      "step": 9442
+    },
+    {
+      "epoch": 1.629086517726214,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.662502453845114e-06,
+      "loss": 1.5547,
+      "step": 9443
+    },
+    {
+      "epoch": 1.6292590356249461,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.660999897661363e-06,
+      "loss": 1.4116,
+      "step": 9444
+    },
+    {
+      "epoch": 1.629431553523678,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.659497959282842e-06,
+      "loss": 1.3623,
+      "step": 9445
+    },
+    {
+      "epoch": 1.62960407142241,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.6579966388208257e-06,
+      "loss": 1.4435,
+      "step": 9446
+    },
+    {
+      "epoch": 1.6297765893211422,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.6564959363865418e-06,
+      "loss": 1.4531,
+      "step": 9447
+    },
+    {
+      "epoch": 1.6299491072198742,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.6549958520911713e-06,
+      "loss": 1.3742,
+      "step": 9448
+    },
+    {
+      "epoch": 1.6301216251186061,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.653496386045851e-06,
+      "loss": 1.341,
+      "step": 9449
+    },
+    {
+      "epoch": 1.630294143017338,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.6519975383616716e-06,
+      "loss": 1.3705,
+      "step": 9450
+    },
+    {
+      "epoch": 1.63046666091607,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.6504993091496701e-06,
+      "loss": 1.4054,
+      "step": 9451
+    },
+    {
+      "epoch": 1.630639178814802,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.6490016985208546e-06,
+      "loss": 1.4413,
+      "step": 9452
+    },
+    {
+      "epoch": 1.630811696713534,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.6475047065861705e-06,
+      "loss": 1.4356,
+      "step": 9453
+    },
+    {
+      "epoch": 1.630984214612266,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.6460083334565248e-06,
+      "loss": 1.3905,
+      "step": 9454
+    },
+    {
+      "epoch": 1.631156732510998,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.644512579242784e-06,
+      "loss": 1.3791,
+      "step": 9455
+    },
+    {
+      "epoch": 1.6313292504097299,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.6430174440557567e-06,
+      "loss": 1.4764,
+      "step": 9456
+    },
+    {
+      "epoch": 1.631501768308462,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.6415229280062139e-06,
+      "loss": 1.39,
+      "step": 9457
+    },
+    {
+      "epoch": 1.631674286207194,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.6400290312048794e-06,
+      "loss": 1.4338,
+      "step": 9458
+    },
+    {
+      "epoch": 1.631846804105926,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.638535753762429e-06,
+      "loss": 1.4187,
+      "step": 9459
+    },
+    {
+      "epoch": 1.632019322004658,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.6370430957894945e-06,
+      "loss": 1.4055,
+      "step": 9460
+    },
+    {
+      "epoch": 1.6321918399033901,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.6355510573966648e-06,
+      "loss": 1.5129,
+      "step": 9461
+    },
+    {
+      "epoch": 1.632364357802122,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.6340596386944707e-06,
+      "loss": 1.3567,
+      "step": 9462
+    },
+    {
+      "epoch": 1.632536875700854,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.632568839793417e-06,
+      "loss": 1.379,
+      "step": 9463
+    },
+    {
+      "epoch": 1.632709393599586,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.6310786608039408e-06,
+      "loss": 1.4042,
+      "step": 9464
+    },
+    {
+      "epoch": 1.632881911498318,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.629589101836454e-06,
+      "loss": 1.3955,
+      "step": 9465
+    },
+    {
+      "epoch": 1.63305442939705,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.6281001630013038e-06,
+      "loss": 1.3853,
+      "step": 9466
+    },
+    {
+      "epoch": 1.6332269472957819,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.6266118444088041e-06,
+      "loss": 1.3801,
+      "step": 9467
+    },
+    {
+      "epoch": 1.6333994651945138,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.625124146169218e-06,
+      "loss": 1.5522,
+      "step": 9468
+    },
+    {
+      "epoch": 1.6335719830932458,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.623637068392765e-06,
+      "loss": 1.4954,
+      "step": 9469
+    },
+    {
+      "epoch": 1.6337445009919778,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.6221506111896147e-06,
+      "loss": 1.4894,
+      "step": 9470
+    },
+    {
+      "epoch": 1.63391701889071,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.6206647746698945e-06,
+      "loss": 1.4969,
+      "step": 9471
+    },
+    {
+      "epoch": 1.634089536789442,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.6191795589436843e-06,
+      "loss": 1.4825,
+      "step": 9472
+    },
+    {
+      "epoch": 1.6342620546881739,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.6176949641210183e-06,
+      "loss": 1.4227,
+      "step": 9473
+    },
+    {
+      "epoch": 1.634434572586906,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.6162109903118872e-06,
+      "loss": 1.4529,
+      "step": 9474
+    },
+    {
+      "epoch": 1.634607090485638,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.6147276376262256e-06,
+      "loss": 1.4036,
+      "step": 9475
+    },
+    {
+      "epoch": 1.63477960838437,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.6132449061739386e-06,
+      "loss": 1.367,
+      "step": 9476
+    },
+    {
+      "epoch": 1.634952126283102,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.6117627960648684e-06,
+      "loss": 1.4574,
+      "step": 9477
+    },
+    {
+      "epoch": 1.635124644181834,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.6102813074088242e-06,
+      "loss": 1.4849,
+      "step": 9478
+    },
+    {
+      "epoch": 1.6352971620805659,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.6088004403155611e-06,
+      "loss": 1.3788,
+      "step": 9479
+    },
+    {
+      "epoch": 1.6354696799792978,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.6073201948947925e-06,
+      "loss": 1.3557,
+      "step": 9480
+    },
+    {
+      "epoch": 1.6356421978780298,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.605840571256183e-06,
+      "loss": 1.5011,
+      "step": 9481
+    },
+    {
+      "epoch": 1.6358147157767617,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.6043615695093561e-06,
+      "loss": 1.3889,
+      "step": 9482
+    },
+    {
+      "epoch": 1.6359872336754937,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.6028831897638774e-06,
+      "loss": 1.3383,
+      "step": 9483
+    },
+    {
+      "epoch": 1.6361597515742257,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.6014054321292848e-06,
+      "loss": 1.3963,
+      "step": 9484
+    },
+    {
+      "epoch": 1.6363322694729578,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.599928296715052e-06,
+      "loss": 1.4085,
+      "step": 9485
+    },
+    {
+      "epoch": 1.6365047873716898,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.5984517836306168e-06,
+      "loss": 1.4987,
+      "step": 9486
+    },
+    {
+      "epoch": 1.6366773052704218,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.5969758929853697e-06,
+      "loss": 1.4016,
+      "step": 9487
+    },
+    {
+      "epoch": 1.636849823169154,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.5955006248886519e-06,
+      "loss": 1.386,
+      "step": 9488
+    },
+    {
+      "epoch": 1.637022341067886,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.5940259794497614e-06,
+      "loss": 1.3809,
+      "step": 9489
+    },
+    {
+      "epoch": 1.6371948589666179,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.5925519567779502e-06,
+      "loss": 1.5296,
+      "step": 9490
+    },
+    {
+      "epoch": 1.6373673768653498,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.5910785569824217e-06,
+      "loss": 1.4201,
+      "step": 9491
+    },
+    {
+      "epoch": 1.6375398947640818,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.5896057801723352e-06,
+      "loss": 1.467,
+      "step": 9492
+    },
+    {
+      "epoch": 1.6377124126628138,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.5881336264568037e-06,
+      "loss": 1.4942,
+      "step": 9493
+    },
+    {
+      "epoch": 1.6378849305615457,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.586662095944892e-06,
+      "loss": 1.4433,
+      "step": 9494
+    },
+    {
+      "epoch": 1.6380574484602777,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.585191188745624e-06,
+      "loss": 1.3434,
+      "step": 9495
+    },
+    {
+      "epoch": 1.6382299663590096,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.5837209049679657e-06,
+      "loss": 1.5629,
+      "step": 9496
+    },
+    {
+      "epoch": 1.6384024842577416,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.582251244720855e-06,
+      "loss": 1.4401,
+      "step": 9497
+    },
+    {
+      "epoch": 1.6385750021564738,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.5807822081131664e-06,
+      "loss": 1.4043,
+      "step": 9498
+    },
+    {
+      "epoch": 1.6387475200552057,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.5793137952537374e-06,
+      "loss": 1.4337,
+      "step": 9499
+    },
+    {
+      "epoch": 1.6389200379539377,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.5778460062513578e-06,
+      "loss": 1.4094,
+      "step": 9500
+    },
+    {
+      "epoch": 1.6389200379539377,
+      "eval_loss": 1.4071273803710938,
+      "eval_runtime": 11.0018,
+      "eval_samples_per_second": 93.076,
+      "eval_steps_per_second": 23.269,
+      "step": 9500
+    },
+    {
+      "epoch": 1.6390925558526697,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.5763788412147695e-06,
+      "loss": 1.3746,
+      "step": 9501
+    },
+    {
+      "epoch": 1.6392650737514018,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.57491230025267e-06,
+      "loss": 1.4003,
+      "step": 9502
+    },
+    {
+      "epoch": 1.6394375916501338,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.5734463834737102e-06,
+      "loss": 1.4765,
+      "step": 9503
+    },
+    {
+      "epoch": 1.6396101095488658,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.5719810909864941e-06,
+      "loss": 1.4375,
+      "step": 9504
+    },
+    {
+      "epoch": 1.6397826274475977,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.5705164228995783e-06,
+      "loss": 1.4209,
+      "step": 9505
+    },
+    {
+      "epoch": 1.6399551453463297,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.5690523793214785e-06,
+      "loss": 1.4444,
+      "step": 9506
+    },
+    {
+      "epoch": 1.6401276632450617,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.567588960360653e-06,
+      "loss": 1.4155,
+      "step": 9507
+    },
+    {
+      "epoch": 1.6403001811437936,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.56612616612553e-06,
+      "loss": 1.5217,
+      "step": 9508
+    },
+    {
+      "epoch": 1.6404726990425256,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.5646639967244758e-06,
+      "loss": 1.5132,
+      "step": 9509
+    },
+    {
+      "epoch": 1.6406452169412575,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.5632024522658185e-06,
+      "loss": 1.3342,
+      "step": 9510
+    },
+    {
+      "epoch": 1.6408177348399895,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.56174153285784e-06,
+      "loss": 1.457,
+      "step": 9511
+    },
+    {
+      "epoch": 1.6409902527387217,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.560281238608773e-06,
+      "loss": 1.3921,
+      "step": 9512
+    },
+    {
+      "epoch": 1.6411627706374536,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.558821569626805e-06,
+      "loss": 1.4043,
+      "step": 9513
+    },
+    {
+      "epoch": 1.6413352885361856,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.557362526020081e-06,
+      "loss": 1.4605,
+      "step": 9514
+    },
+    {
+      "epoch": 1.6415078064349178,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.5559041078966885e-06,
+      "loss": 1.3708,
+      "step": 9515
+    },
+    {
+      "epoch": 1.6416803243336497,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.554446315364685e-06,
+      "loss": 1.35,
+      "step": 9516
+    },
+    {
+      "epoch": 1.6418528422323817,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.552989148532067e-06,
+      "loss": 1.529,
+      "step": 9517
+    },
+    {
+      "epoch": 1.6420253601311137,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.5515326075067916e-06,
+      "loss": 1.3917,
+      "step": 9518
+    },
+    {
+      "epoch": 1.6421978780298456,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.5500766923967692e-06,
+      "loss": 1.4447,
+      "step": 9519
+    },
+    {
+      "epoch": 1.6423703959285776,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.5486214033098613e-06,
+      "loss": 1.3875,
+      "step": 9520
+    },
+    {
+      "epoch": 1.6425429138273095,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.5471667403538904e-06,
+      "loss": 1.5287,
+      "step": 9521
+    },
+    {
+      "epoch": 1.6427154317260415,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.5457127036366216e-06,
+      "loss": 1.3733,
+      "step": 9522
+    },
+    {
+      "epoch": 1.6428879496247735,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.5442592932657797e-06,
+      "loss": 1.3913,
+      "step": 9523
+    },
+    {
+      "epoch": 1.6430604675235054,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.5428065093490441e-06,
+      "loss": 1.3989,
+      "step": 9524
+    },
+    {
+      "epoch": 1.6432329854222374,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.5413543519940443e-06,
+      "loss": 1.3217,
+      "step": 9525
+    },
+    {
+      "epoch": 1.6434055033209696,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.5399028213083666e-06,
+      "loss": 1.3971,
+      "step": 9526
+    },
+    {
+      "epoch": 1.6435780212197015,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.5384519173995515e-06,
+      "loss": 1.4748,
+      "step": 9527
+    },
+    {
+      "epoch": 1.6437505391184335,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.5370016403750843e-06,
+      "loss": 1.3221,
+      "step": 9528
+    },
+    {
+      "epoch": 1.6439230570171657,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.5355519903424187e-06,
+      "loss": 1.4242,
+      "step": 9529
+    },
+    {
+      "epoch": 1.6440955749158976,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.534102967408948e-06,
+      "loss": 1.4049,
+      "step": 9530
+    },
+    {
+      "epoch": 1.6442680928146296,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.5326545716820273e-06,
+      "loss": 1.2972,
+      "step": 9531
+    },
+    {
+      "epoch": 1.6444406107133616,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.531206803268962e-06,
+      "loss": 1.4253,
+      "step": 9532
+    },
+    {
+      "epoch": 1.6446131286120935,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.5297596622770117e-06,
+      "loss": 1.4365,
+      "step": 9533
+    },
+    {
+      "epoch": 1.6447856465108255,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.5283131488133906e-06,
+      "loss": 1.3451,
+      "step": 9534
+    },
+    {
+      "epoch": 1.6449581644095574,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.5268672629852655e-06,
+      "loss": 1.4688,
+      "step": 9535
+    },
+    {
+      "epoch": 1.6451306823082894,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.5254220048997548e-06,
+      "loss": 1.4259,
+      "step": 9536
+    },
+    {
+      "epoch": 1.6453032002070214,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.523977374663934e-06,
+      "loss": 1.3756,
+      "step": 9537
+    },
+    {
+      "epoch": 1.6454757181057533,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.5225333723848335e-06,
+      "loss": 1.4103,
+      "step": 9538
+    },
+    {
+      "epoch": 1.6456482360044855,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.5210899981694238e-06,
+      "loss": 1.5031,
+      "step": 9539
+    },
+    {
+      "epoch": 1.6458207539032175,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.5196472521246518e-06,
+      "loss": 1.405,
+      "step": 9540
+    },
+    {
+      "epoch": 1.6459932718019494,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.5182051343573968e-06,
+      "loss": 1.507,
+      "step": 9541
+    },
+    {
+      "epoch": 1.6461657897006814,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.5167636449745015e-06,
+      "loss": 1.4083,
+      "step": 9542
+    },
+    {
+      "epoch": 1.6463383075994136,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.515322784082761e-06,
+      "loss": 1.3217,
+      "step": 9543
+    },
+    {
+      "epoch": 1.6465108254981455,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.5138825517889232e-06,
+      "loss": 1.4405,
+      "step": 9544
+    },
+    {
+      "epoch": 1.6466833433968775,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.5124429481996904e-06,
+      "loss": 1.3475,
+      "step": 9545
+    },
+    {
+      "epoch": 1.6468558612956095,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.5110039734217186e-06,
+      "loss": 1.4315,
+      "step": 9546
+    },
+    {
+      "epoch": 1.6470283791943414,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.50956562756161e-06,
+      "loss": 1.3847,
+      "step": 9547
+    },
+    {
+      "epoch": 1.6472008970930734,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.508127910725934e-06,
+      "loss": 1.362,
+      "step": 9548
+    },
+    {
+      "epoch": 1.6473734149918053,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.5066908230212008e-06,
+      "loss": 1.448,
+      "step": 9549
+    },
+    {
+      "epoch": 1.6475459328905373,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.5052543645538798e-06,
+      "loss": 1.5037,
+      "step": 9550
+    },
+    {
+      "epoch": 1.6477184507892693,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.5038185354303924e-06,
+      "loss": 1.4538,
+      "step": 9551
+    },
+    {
+      "epoch": 1.6478909686880012,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.5023833357571128e-06,
+      "loss": 1.4369,
+      "step": 9552
+    },
+    {
+      "epoch": 1.6480634865867334,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.5009487656403765e-06,
+      "loss": 1.385,
+      "step": 9553
+    },
+    {
+      "epoch": 1.6482360044854654,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.4995148251864566e-06,
+      "loss": 1.331,
+      "step": 9554
+    },
+    {
+      "epoch": 1.6484085223841973,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.4980815145015925e-06,
+      "loss": 1.3528,
+      "step": 9555
+    },
+    {
+      "epoch": 1.6485810402829295,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.4966488336919728e-06,
+      "loss": 1.3618,
+      "step": 9556
+    },
+    {
+      "epoch": 1.6487535581816615,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.4952167828637377e-06,
+      "loss": 1.3758,
+      "step": 9557
+    },
+    {
+      "epoch": 1.6489260760803934,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.4937853621229848e-06,
+      "loss": 1.5034,
+      "step": 9558
+    },
+    {
+      "epoch": 1.6490985939791254,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.4923545715757625e-06,
+      "loss": 1.445,
+      "step": 9559
+    },
+    {
+      "epoch": 1.6492711118778574,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.4909244113280685e-06,
+      "loss": 1.4184,
+      "step": 9560
+    },
+    {
+      "epoch": 1.6494436297765893,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.4894948814858656e-06,
+      "loss": 1.4357,
+      "step": 9561
+    },
+    {
+      "epoch": 1.6496161476753213,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.4880659821550547e-06,
+      "loss": 1.4171,
+      "step": 9562
+    },
+    {
+      "epoch": 1.6497886655740532,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.4866377134415022e-06,
+      "loss": 1.4275,
+      "step": 9563
+    },
+    {
+      "epoch": 1.6499611834727852,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.4852100754510213e-06,
+      "loss": 1.4104,
+      "step": 9564
+    },
+    {
+      "epoch": 1.6501337013715172,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.4837830682893806e-06,
+      "loss": 1.4573,
+      "step": 9565
+    },
+    {
+      "epoch": 1.6503062192702493,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.4823566920623022e-06,
+      "loss": 1.3716,
+      "step": 9566
+    },
+    {
+      "epoch": 1.6504787371689813,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.4809309468754608e-06,
+      "loss": 1.3767,
+      "step": 9567
+    },
+    {
+      "epoch": 1.6506512550677133,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.4795058328344847e-06,
+      "loss": 1.4028,
+      "step": 9568
+    },
+    {
+      "epoch": 1.6508237729664452,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.4780813500449541e-06,
+      "loss": 1.3859,
+      "step": 9569
+    },
+    {
+      "epoch": 1.6509962908651774,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.4766574986124082e-06,
+      "loss": 1.4154,
+      "step": 9570
+    },
+    {
+      "epoch": 1.6511688087639094,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.4752342786423257e-06,
+      "loss": 1.4186,
+      "step": 9571
+    },
+    {
+      "epoch": 1.6513413266626413,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.4738116902401567e-06,
+      "loss": 1.3528,
+      "step": 9572
+    },
+    {
+      "epoch": 1.6515138445613733,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.4723897335112913e-06,
+      "loss": 1.397,
+      "step": 9573
+    },
+    {
+      "epoch": 1.6516863624601053,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.4709684085610765e-06,
+      "loss": 1.3941,
+      "step": 9574
+    },
+    {
+      "epoch": 1.6518588803588372,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.4695477154948134e-06,
+      "loss": 1.479,
+      "step": 9575
+    },
+    {
+      "epoch": 1.6520313982575692,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.468127654417757e-06,
+      "loss": 1.3031,
+      "step": 9576
+    },
+    {
+      "epoch": 1.6522039161563011,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.4667082254351127e-06,
+      "loss": 1.4715,
+      "step": 9577
+    },
+    {
+      "epoch": 1.652376434055033,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.4652894286520446e-06,
+      "loss": 1.4589,
+      "step": 9578
+    },
+    {
+      "epoch": 1.652548951953765,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.4638712641736585e-06,
+      "loss": 1.3271,
+      "step": 9579
+    },
+    {
+      "epoch": 1.6527214698524972,
+      "grad_norm": 0.625,
+      "learning_rate": 1.4624537321050292e-06,
+      "loss": 1.4969,
+      "step": 9580
+    },
+    {
+      "epoch": 1.6528939877512292,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.4610368325511693e-06,
+      "loss": 1.3658,
+      "step": 9581
+    },
+    {
+      "epoch": 1.6530665056499612,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.4596205656170581e-06,
+      "loss": 1.4521,
+      "step": 9582
+    },
+    {
+      "epoch": 1.6532390235486933,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.4582049314076163e-06,
+      "loss": 1.4213,
+      "step": 9583
+    },
+    {
+      "epoch": 1.6534115414474253,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.4567899300277222e-06,
+      "loss": 1.4668,
+      "step": 9584
+    },
+    {
+      "epoch": 1.6535840593461573,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.4553755615822151e-06,
+      "loss": 1.4904,
+      "step": 9585
+    },
+    {
+      "epoch": 1.6537565772448892,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.4539618261758737e-06,
+      "loss": 1.3653,
+      "step": 9586
+    },
+    {
+      "epoch": 1.6539290951436212,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.4525487239134373e-06,
+      "loss": 1.333,
+      "step": 9587
+    },
+    {
+      "epoch": 1.6541016130423531,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.4511362548996e-06,
+      "loss": 1.519,
+      "step": 9588
+    },
+    {
+      "epoch": 1.654274130941085,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.4497244192390037e-06,
+      "loss": 1.3489,
+      "step": 9589
+    },
+    {
+      "epoch": 1.654446648839817,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.4483132170362468e-06,
+      "loss": 1.4774,
+      "step": 9590
+    },
+    {
+      "epoch": 1.654619166738549,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.4469026483958837e-06,
+      "loss": 1.3916,
+      "step": 9591
+    },
+    {
+      "epoch": 1.654791684637281,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.4454927134224085e-06,
+      "loss": 1.4945,
+      "step": 9592
+    },
+    {
+      "epoch": 1.654964202536013,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.4440834122202896e-06,
+      "loss": 1.4431,
+      "step": 9593
+    },
+    {
+      "epoch": 1.6551367204347451,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.4426747448939294e-06,
+      "loss": 1.4583,
+      "step": 9594
+    },
+    {
+      "epoch": 1.655309238333477,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.4412667115476918e-06,
+      "loss": 1.3336,
+      "step": 9595
+    },
+    {
+      "epoch": 1.655481756232209,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.439859312285894e-06,
+      "loss": 1.387,
+      "step": 9596
+    },
+    {
+      "epoch": 1.6556542741309412,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.4384525472128052e-06,
+      "loss": 1.4272,
+      "step": 9597
+    },
+    {
+      "epoch": 1.6558267920296732,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.4370464164326458e-06,
+      "loss": 1.4258,
+      "step": 9598
+    },
+    {
+      "epoch": 1.6559993099284052,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.4356409200495924e-06,
+      "loss": 1.4453,
+      "step": 9599
+    },
+    {
+      "epoch": 1.6561718278271371,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.4342360581677717e-06,
+      "loss": 1.4543,
+      "step": 9600
+    },
+    {
+      "epoch": 1.6561718278271371,
+      "eval_loss": 1.4071323871612549,
+      "eval_runtime": 10.9314,
+      "eval_samples_per_second": 93.675,
+      "eval_steps_per_second": 23.419,
+      "step": 9600
+    },
+    {
+      "epoch": 1.656344345725869,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.4328318308912647e-06,
+      "loss": 1.3847,
+      "step": 9601
+    },
+    {
+      "epoch": 1.656516863624601,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.4314282383241097e-06,
+      "loss": 1.4592,
+      "step": 9602
+    },
+    {
+      "epoch": 1.656689381523333,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.430025280570284e-06,
+      "loss": 1.5076,
+      "step": 9603
+    },
+    {
+      "epoch": 1.656861899422065,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.4286229577337373e-06,
+      "loss": 1.3705,
+      "step": 9604
+    },
+    {
+      "epoch": 1.657034417320797,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.4272212699183574e-06,
+      "loss": 1.4033,
+      "step": 9605
+    },
+    {
+      "epoch": 1.6572069352195289,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.42582021722799e-06,
+      "loss": 1.4395,
+      "step": 9606
+    },
+    {
+      "epoch": 1.657379453118261,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.4244197997664354e-06,
+      "loss": 1.3369,
+      "step": 9607
+    },
+    {
+      "epoch": 1.657551971016993,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.4230200176374442e-06,
+      "loss": 1.4236,
+      "step": 9608
+    },
+    {
+      "epoch": 1.657724488915725,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.421620870944722e-06,
+      "loss": 1.4639,
+      "step": 9609
+    },
+    {
+      "epoch": 1.657897006814457,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.4202223597919285e-06,
+      "loss": 1.4062,
+      "step": 9610
+    },
+    {
+      "epoch": 1.6580695247131891,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.4188244842826659e-06,
+      "loss": 1.4875,
+      "step": 9611
+    },
+    {
+      "epoch": 1.658242042611921,
+      "grad_norm": 0.625,
+      "learning_rate": 1.4174272445205084e-06,
+      "loss": 1.3593,
+      "step": 9612
+    },
+    {
+      "epoch": 1.658414560510653,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.416030640608963e-06,
+      "loss": 1.4083,
+      "step": 9613
+    },
+    {
+      "epoch": 1.658587078409385,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.414634672651506e-06,
+      "loss": 1.391,
+      "step": 9614
+    },
+    {
+      "epoch": 1.658759596308117,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.4132393407515555e-06,
+      "loss": 1.5134,
+      "step": 9615
+    },
+    {
+      "epoch": 1.658932114206849,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.4118446450124834e-06,
+      "loss": 1.3834,
+      "step": 9616
+    },
+    {
+      "epoch": 1.659104632105581,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.4104505855376271e-06,
+      "loss": 1.4039,
+      "step": 9617
+    },
+    {
+      "epoch": 1.6592771500043129,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.4090571624302585e-06,
+      "loss": 1.5073,
+      "step": 9618
+    },
+    {
+      "epoch": 1.6594496679030448,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4076643757936147e-06,
+      "loss": 1.5064,
+      "step": 9619
+    },
+    {
+      "epoch": 1.6596221858017768,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.4062722257308803e-06,
+      "loss": 1.3995,
+      "step": 9620
+    },
+    {
+      "epoch": 1.659794703700509,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.4048807123451958e-06,
+      "loss": 1.4615,
+      "step": 9621
+    },
+    {
+      "epoch": 1.659967221599241,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.4034898357396532e-06,
+      "loss": 1.406,
+      "step": 9622
+    },
+    {
+      "epoch": 1.6601397394979729,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.4020995960173001e-06,
+      "loss": 1.4212,
+      "step": 9623
+    },
+    {
+      "epoch": 1.660312257396705,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.4007099932811253e-06,
+      "loss": 1.4811,
+      "step": 9624
+    },
+    {
+      "epoch": 1.660484775295437,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.3993210276340895e-06,
+      "loss": 1.4382,
+      "step": 9625
+    },
+    {
+      "epoch": 1.660657293194169,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.3979326991790898e-06,
+      "loss": 1.3531,
+      "step": 9626
+    },
+    {
+      "epoch": 1.660829811092901,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.3965450080189836e-06,
+      "loss": 1.4453,
+      "step": 9627
+    },
+    {
+      "epoch": 1.661002328991633,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.3951579542565808e-06,
+      "loss": 1.4147,
+      "step": 9628
+    },
+    {
+      "epoch": 1.6611748468903649,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.3937715379946414e-06,
+      "loss": 1.3453,
+      "step": 9629
+    },
+    {
+      "epoch": 1.6613473647890968,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.3923857593358813e-06,
+      "loss": 1.4904,
+      "step": 9630
+    },
+    {
+      "epoch": 1.6615198826878288,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.3910006183829671e-06,
+      "loss": 1.3955,
+      "step": 9631
+    },
+    {
+      "epoch": 1.6616924005865608,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.3896161152385178e-06,
+      "loss": 1.3752,
+      "step": 9632
+    },
+    {
+      "epoch": 1.6618649184852927,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.3882322500051072e-06,
+      "loss": 1.4852,
+      "step": 9633
+    },
+    {
+      "epoch": 1.6620374363840247,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.3868490227852626e-06,
+      "loss": 1.305,
+      "step": 9634
+    },
+    {
+      "epoch": 1.6622099542827569,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.385466433681456e-06,
+      "loss": 1.3817,
+      "step": 9635
+    },
+    {
+      "epoch": 1.6623824721814888,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.3840844827961263e-06,
+      "loss": 1.4308,
+      "step": 9636
+    },
+    {
+      "epoch": 1.6625549900802208,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.3827031702316495e-06,
+      "loss": 1.4287,
+      "step": 9637
+    },
+    {
+      "epoch": 1.662727507978953,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.381322496090367e-06,
+      "loss": 1.4174,
+      "step": 9638
+    },
+    {
+      "epoch": 1.662900025877685,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.3799424604745648e-06,
+      "loss": 1.4176,
+      "step": 9639
+    },
+    {
+      "epoch": 1.663072543776417,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.3785630634864855e-06,
+      "loss": 1.4426,
+      "step": 9640
+    },
+    {
+      "epoch": 1.6632450616751489,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.377184305228324e-06,
+      "loss": 1.4198,
+      "step": 9641
+    },
+    {
+      "epoch": 1.6634175795738808,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.3758061858022276e-06,
+      "loss": 1.3411,
+      "step": 9642
+    },
+    {
+      "epoch": 1.6635900974726128,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.3744287053102944e-06,
+      "loss": 1.4388,
+      "step": 9643
+    },
+    {
+      "epoch": 1.6637626153713447,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.3730518638545809e-06,
+      "loss": 1.4606,
+      "step": 9644
+    },
+    {
+      "epoch": 1.6639351332700767,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.3716756615370842e-06,
+      "loss": 1.3962,
+      "step": 9645
+    },
+    {
+      "epoch": 1.6641076511688087,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.3703000984597714e-06,
+      "loss": 1.3739,
+      "step": 9646
+    },
+    {
+      "epoch": 1.6642801690675406,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.3689251747245458e-06,
+      "loss": 1.4007,
+      "step": 9647
+    },
+    {
+      "epoch": 1.6644526869662728,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.3675508904332703e-06,
+      "loss": 1.3208,
+      "step": 9648
+    },
+    {
+      "epoch": 1.6646252048650048,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.3661772456877675e-06,
+      "loss": 1.3696,
+      "step": 9649
+    },
+    {
+      "epoch": 1.6647977227637367,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.3648042405897987e-06,
+      "loss": 1.3645,
+      "step": 9650
+    },
+    {
+      "epoch": 1.6649702406624687,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.3634318752410868e-06,
+      "loss": 1.4679,
+      "step": 9651
+    },
+    {
+      "epoch": 1.6651427585612009,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.362060149743305e-06,
+      "loss": 1.4031,
+      "step": 9652
+    },
+    {
+      "epoch": 1.6653152764599328,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.3606890641980809e-06,
+      "loss": 1.4893,
+      "step": 9653
+    },
+    {
+      "epoch": 1.6654877943586648,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.3593186187069907e-06,
+      "loss": 1.4081,
+      "step": 9654
+    },
+    {
+      "epoch": 1.6656603122573967,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.357948813371569e-06,
+      "loss": 1.3926,
+      "step": 9655
+    },
+    {
+      "epoch": 1.6658328301561287,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.3565796482932936e-06,
+      "loss": 1.3133,
+      "step": 9656
+    },
+    {
+      "epoch": 1.6660053480548607,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.3552111235736077e-06,
+      "loss": 1.3755,
+      "step": 9657
+    },
+    {
+      "epoch": 1.6661778659535926,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.3538432393138957e-06,
+      "loss": 1.4162,
+      "step": 9658
+    },
+    {
+      "epoch": 1.6663503838523246,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.3524759956155e-06,
+      "loss": 1.3238,
+      "step": 9659
+    },
+    {
+      "epoch": 1.6665229017510566,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.3511093925797136e-06,
+      "loss": 1.4309,
+      "step": 9660
+    },
+    {
+      "epoch": 1.6666954196497885,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.3497434303077861e-06,
+      "loss": 1.5292,
+      "step": 9661
+    },
+    {
+      "epoch": 1.6668679375485207,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.348378108900913e-06,
+      "loss": 1.3987,
+      "step": 9662
+    },
+    {
+      "epoch": 1.6670404554472527,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.3470134284602487e-06,
+      "loss": 1.4505,
+      "step": 9663
+    },
+    {
+      "epoch": 1.6672129733459846,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.345649389086895e-06,
+      "loss": 1.4911,
+      "step": 9664
+    },
+    {
+      "epoch": 1.6673854912447168,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.3442859908819095e-06,
+      "loss": 1.3742,
+      "step": 9665
+    },
+    {
+      "epoch": 1.6675580091434488,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.342923233946304e-06,
+      "loss": 1.3935,
+      "step": 9666
+    },
+    {
+      "epoch": 1.6677305270421807,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.3415611183810329e-06,
+      "loss": 1.3412,
+      "step": 9667
+    },
+    {
+      "epoch": 1.6679030449409127,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.340199644287018e-06,
+      "loss": 1.3944,
+      "step": 9668
+    },
+    {
+      "epoch": 1.6680755628396446,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.3388388117651186e-06,
+      "loss": 1.3959,
+      "step": 9669
+    },
+    {
+      "epoch": 1.6682480807383766,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.3374786209161617e-06,
+      "loss": 1.4506,
+      "step": 9670
+    },
+    {
+      "epoch": 1.6684205986371086,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.3361190718409123e-06,
+      "loss": 1.4873,
+      "step": 9671
+    },
+    {
+      "epoch": 1.6685931165358405,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.3347601646400965e-06,
+      "loss": 1.4537,
+      "step": 9672
+    },
+    {
+      "epoch": 1.6687656344345725,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.3334018994143916e-06,
+      "loss": 1.388,
+      "step": 9673
+    },
+    {
+      "epoch": 1.6689381523333044,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.3320442762644236e-06,
+      "loss": 1.4557,
+      "step": 9674
+    },
+    {
+      "epoch": 1.6691106702320364,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.3306872952907769e-06,
+      "loss": 1.4547,
+      "step": 9675
+    },
+    {
+      "epoch": 1.6692831881307686,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.329330956593985e-06,
+      "loss": 1.4649,
+      "step": 9676
+    },
+    {
+      "epoch": 1.6694557060295006,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.3279752602745287e-06,
+      "loss": 1.4182,
+      "step": 9677
+    },
+    {
+      "epoch": 1.6696282239282325,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.3266202064328548e-06,
+      "loss": 1.5169,
+      "step": 9678
+    },
+    {
+      "epoch": 1.6698007418269647,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.325265795169348e-06,
+      "loss": 1.4113,
+      "step": 9679
+    },
+    {
+      "epoch": 1.6699732597256967,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.3239120265843508e-06,
+      "loss": 1.4695,
+      "step": 9680
+    },
+    {
+      "epoch": 1.6701457776244286,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.3225589007781658e-06,
+      "loss": 1.4003,
+      "step": 9681
+    },
+    {
+      "epoch": 1.6703182955231606,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.3212064178510342e-06,
+      "loss": 1.373,
+      "step": 9682
+    },
+    {
+      "epoch": 1.6704908134218925,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.319854577903159e-06,
+      "loss": 1.3848,
+      "step": 9683
+    },
+    {
+      "epoch": 1.6706633313206245,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.3185033810346926e-06,
+      "loss": 1.4805,
+      "step": 9684
+    },
+    {
+      "epoch": 1.6708358492193565,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.3171528273457402e-06,
+      "loss": 1.4182,
+      "step": 9685
+    },
+    {
+      "epoch": 1.6710083671180884,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.3158029169363595e-06,
+      "loss": 1.4491,
+      "step": 9686
+    },
+    {
+      "epoch": 1.6711808850168204,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.314453649906562e-06,
+      "loss": 1.4682,
+      "step": 9687
+    },
+    {
+      "epoch": 1.6713534029155523,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.313105026356304e-06,
+      "loss": 1.4939,
+      "step": 9688
+    },
+    {
+      "epoch": 1.6715259208142845,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.311757046385508e-06,
+      "loss": 1.348,
+      "step": 9689
+    },
+    {
+      "epoch": 1.6716984387130165,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.310409710094035e-06,
+      "loss": 1.5174,
+      "step": 9690
+    },
+    {
+      "epoch": 1.6718709566117484,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.3090630175817042e-06,
+      "loss": 1.4019,
+      "step": 9691
+    },
+    {
+      "epoch": 1.6720434745104804,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.3077169689482893e-06,
+      "loss": 1.3465,
+      "step": 9692
+    },
+    {
+      "epoch": 1.6722159924092126,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.3063715642935137e-06,
+      "loss": 1.4102,
+      "step": 9693
+    },
+    {
+      "epoch": 1.6723885103079446,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.3050268037170533e-06,
+      "loss": 1.3047,
+      "step": 9694
+    },
+    {
+      "epoch": 1.6725610282066765,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.303682687318537e-06,
+      "loss": 1.4796,
+      "step": 9695
+    },
+    {
+      "epoch": 1.6727335461054085,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.30233921519754e-06,
+      "loss": 1.42,
+      "step": 9696
+    },
+    {
+      "epoch": 1.6729060640041404,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.3009963874536025e-06,
+      "loss": 1.4029,
+      "step": 9697
+    },
+    {
+      "epoch": 1.6730785819028724,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.299654204186208e-06,
+      "loss": 1.4715,
+      "step": 9698
+    },
+    {
+      "epoch": 1.6732510998016044,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.2983126654947886e-06,
+      "loss": 1.3625,
+      "step": 9699
+    },
+    {
+      "epoch": 1.6734236177003363,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.2969717714787411e-06,
+      "loss": 1.4229,
+      "step": 9700
+    },
+    {
+      "epoch": 1.6734236177003363,
+      "eval_loss": 1.407073736190796,
+      "eval_runtime": 10.8926,
+      "eval_samples_per_second": 94.009,
+      "eval_steps_per_second": 23.502,
+      "step": 9700
+    },
+    {
+      "epoch": 1.6735961355990683,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.2956315222374006e-06,
+      "loss": 1.4324,
+      "step": 9701
+    },
+    {
+      "epoch": 1.6737686534978002,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.2942919178700674e-06,
+      "loss": 1.4398,
+      "step": 9702
+    },
+    {
+      "epoch": 1.6739411713965324,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.2929529584759836e-06,
+      "loss": 1.5529,
+      "step": 9703
+    },
+    {
+      "epoch": 1.6741136892952644,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.2916146441543487e-06,
+      "loss": 1.4504,
+      "step": 9704
+    },
+    {
+      "epoch": 1.6742862071939963,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.290276975004312e-06,
+      "loss": 1.3953,
+      "step": 9705
+    },
+    {
+      "epoch": 1.6744587250927285,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.28893995112498e-06,
+      "loss": 1.4363,
+      "step": 9706
+    },
+    {
+      "epoch": 1.6746312429914605,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.2876035726154046e-06,
+      "loss": 1.3643,
+      "step": 9707
+    },
+    {
+      "epoch": 1.6748037608901925,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.2862678395745954e-06,
+      "loss": 1.4318,
+      "step": 9708
+    },
+    {
+      "epoch": 1.6749762787889244,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.2849327521015076e-06,
+      "loss": 1.4568,
+      "step": 9709
+    },
+    {
+      "epoch": 1.6751487966876564,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.283598310295059e-06,
+      "loss": 1.2942,
+      "step": 9710
+    },
+    {
+      "epoch": 1.6753213145863883,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.282264514254109e-06,
+      "loss": 1.3316,
+      "step": 9711
+    },
+    {
+      "epoch": 1.6754938324851203,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.2809313640774723e-06,
+      "loss": 1.4064,
+      "step": 9712
+    },
+    {
+      "epoch": 1.6756663503838523,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.279598859863923e-06,
+      "loss": 1.4209,
+      "step": 9713
+    },
+    {
+      "epoch": 1.6758388682825842,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.2782670017121768e-06,
+      "loss": 1.3198,
+      "step": 9714
+    },
+    {
+      "epoch": 1.6760113861813162,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.2769357897209056e-06,
+      "loss": 1.467,
+      "step": 9715
+    },
+    {
+      "epoch": 1.6761839040800484,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.2756052239887362e-06,
+      "loss": 1.3825,
+      "step": 9716
+    },
+    {
+      "epoch": 1.6763564219787803,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.2742753046142442e-06,
+      "loss": 1.4707,
+      "step": 9717
+    },
+    {
+      "epoch": 1.6765289398775123,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.2729460316959586e-06,
+      "loss": 1.4252,
+      "step": 9718
+    },
+    {
+      "epoch": 1.6767014577762442,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.2716174053323628e-06,
+      "loss": 1.4965,
+      "step": 9719
+    },
+    {
+      "epoch": 1.6768739756749764,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.2702894256218823e-06,
+      "loss": 1.482,
+      "step": 9720
+    },
+    {
+      "epoch": 1.6770464935737084,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.2689620926629108e-06,
+      "loss": 1.3392,
+      "step": 9721
+    },
+    {
+      "epoch": 1.6772190114724403,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.26763540655378e-06,
+      "loss": 1.4667,
+      "step": 9722
+    },
+    {
+      "epoch": 1.6773915293711723,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.2663093673927796e-06,
+      "loss": 1.3934,
+      "step": 9723
+    },
+    {
+      "epoch": 1.6775640472699043,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.2649839752781522e-06,
+      "loss": 1.4194,
+      "step": 9724
+    },
+    {
+      "epoch": 1.6777365651686362,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.263659230308092e-06,
+      "loss": 1.5728,
+      "step": 9725
+    },
+    {
+      "epoch": 1.6779090830673682,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.262335132580742e-06,
+      "loss": 1.4046,
+      "step": 9726
+    },
+    {
+      "epoch": 1.6780816009661002,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.2610116821942032e-06,
+      "loss": 1.4356,
+      "step": 9727
+    },
+    {
+      "epoch": 1.6782541188648321,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.259688879246519e-06,
+      "loss": 1.4939,
+      "step": 9728
+    },
+    {
+      "epoch": 1.678426636763564,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.2583667238356956e-06,
+      "loss": 1.4155,
+      "step": 9729
+    },
+    {
+      "epoch": 1.6785991546622963,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.2570452160596859e-06,
+      "loss": 1.3951,
+      "step": 9730
+    },
+    {
+      "epoch": 1.6787716725610282,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.2557243560163955e-06,
+      "loss": 1.419,
+      "step": 9731
+    },
+    {
+      "epoch": 1.6789441904597602,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.2544041438036836e-06,
+      "loss": 1.474,
+      "step": 9732
+    },
+    {
+      "epoch": 1.6791167083584921,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.2530845795193536e-06,
+      "loss": 1.3757,
+      "step": 9733
+    },
+    {
+      "epoch": 1.6792892262572243,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.2517656632611753e-06,
+      "loss": 1.4638,
+      "step": 9734
+    },
+    {
+      "epoch": 1.6794617441559563,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.2504473951268559e-06,
+      "loss": 1.3717,
+      "step": 9735
+    },
+    {
+      "epoch": 1.6796342620546882,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.249129775214064e-06,
+      "loss": 1.3255,
+      "step": 9736
+    },
+    {
+      "epoch": 1.6798067799534202,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.2478128036204151e-06,
+      "loss": 1.3849,
+      "step": 9737
+    },
+    {
+      "epoch": 1.6799792978521522,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.246496480443481e-06,
+      "loss": 1.4408,
+      "step": 9738
+    },
+    {
+      "epoch": 1.6801518157508841,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.2451808057807813e-06,
+      "loss": 1.4953,
+      "step": 9739
+    },
+    {
+      "epoch": 1.680324333649616,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.2438657797297927e-06,
+      "loss": 1.4011,
+      "step": 9740
+    },
+    {
+      "epoch": 1.680496851548348,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.2425514023879338e-06,
+      "loss": 1.3963,
+      "step": 9741
+    },
+    {
+      "epoch": 1.68066936944708,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.24123767385259e-06,
+      "loss": 1.3753,
+      "step": 9742
+    },
+    {
+      "epoch": 1.680841887345812,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.2399245942210847e-06,
+      "loss": 1.3894,
+      "step": 9743
+    },
+    {
+      "epoch": 1.6810144052445442,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.2386121635906978e-06,
+      "loss": 1.5035,
+      "step": 9744
+    },
+    {
+      "epoch": 1.6811869231432761,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.2373003820586715e-06,
+      "loss": 1.4379,
+      "step": 9745
+    },
+    {
+      "epoch": 1.681359441042008,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.2359892497221815e-06,
+      "loss": 1.4879,
+      "step": 9746
+    },
+    {
+      "epoch": 1.6815319589407403,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.2346787666783678e-06,
+      "loss": 1.434,
+      "step": 9747
+    },
+    {
+      "epoch": 1.6817044768394722,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.2333689330243204e-06,
+      "loss": 1.3453,
+      "step": 9748
+    },
+    {
+      "epoch": 1.6818769947382042,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.2320597488570774e-06,
+      "loss": 1.3516,
+      "step": 9749
+    },
+    {
+      "epoch": 1.6820495126369361,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.2307512142736344e-06,
+      "loss": 1.4055,
+      "step": 9750
+    },
+    {
+      "epoch": 1.682222030535668,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.2294433293709352e-06,
+      "loss": 1.4123,
+      "step": 9751
+    },
+    {
+      "epoch": 1.6823945484344,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.2281360942458708e-06,
+      "loss": 1.4203,
+      "step": 9752
+    },
+    {
+      "epoch": 1.682567066333132,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.2268295089952986e-06,
+      "loss": 1.4716,
+      "step": 9753
+    },
+    {
+      "epoch": 1.682739584231864,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.225523573716012e-06,
+      "loss": 1.4118,
+      "step": 9754
+    },
+    {
+      "epoch": 1.682912102130596,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.2242182885047638e-06,
+      "loss": 1.4714,
+      "step": 9755
+    },
+    {
+      "epoch": 1.683084620029328,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.2229136534582586e-06,
+      "loss": 1.4043,
+      "step": 9756
+    },
+    {
+      "epoch": 1.68325713792806,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.2216096686731515e-06,
+      "loss": 1.3979,
+      "step": 9757
+    },
+    {
+      "epoch": 1.683429655826792,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.2203063342460496e-06,
+      "loss": 1.2974,
+      "step": 9758
+    },
+    {
+      "epoch": 1.683602173725524,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.2190036502735158e-06,
+      "loss": 1.3953,
+      "step": 9759
+    },
+    {
+      "epoch": 1.683774691624256,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.2177016168520528e-06,
+      "loss": 1.4185,
+      "step": 9760
+    },
+    {
+      "epoch": 1.6839472095229882,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.216400234078131e-06,
+      "loss": 1.5038,
+      "step": 9761
+    },
+    {
+      "epoch": 1.6841197274217201,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.2150995020481616e-06,
+      "loss": 1.4263,
+      "step": 9762
+    },
+    {
+      "epoch": 1.684292245320452,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.2137994208585125e-06,
+      "loss": 1.4546,
+      "step": 9763
+    },
+    {
+      "epoch": 1.684464763219184,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.2124999906055024e-06,
+      "loss": 1.3523,
+      "step": 9764
+    },
+    {
+      "epoch": 1.684637281117916,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.2112012113853955e-06,
+      "loss": 1.2748,
+      "step": 9765
+    },
+    {
+      "epoch": 1.684809799016648,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.2099030832944224e-06,
+      "loss": 1.4521,
+      "step": 9766
+    },
+    {
+      "epoch": 1.68498231691538,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.2086056064287498e-06,
+      "loss": 1.3665,
+      "step": 9767
+    },
+    {
+      "epoch": 1.6851548348141119,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.2073087808845052e-06,
+      "loss": 1.433,
+      "step": 9768
+    },
+    {
+      "epoch": 1.6853273527128438,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.206012606757765e-06,
+      "loss": 1.4158,
+      "step": 9769
+    },
+    {
+      "epoch": 1.6854998706115758,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.2047170841445589e-06,
+      "loss": 1.47,
+      "step": 9770
+    },
+    {
+      "epoch": 1.685672388510308,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.2034222131408669e-06,
+      "loss": 1.4627,
+      "step": 9771
+    },
+    {
+      "epoch": 1.68584490640904,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.2021279938426223e-06,
+      "loss": 1.4755,
+      "step": 9772
+    },
+    {
+      "epoch": 1.686017424307772,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.2008344263457029e-06,
+      "loss": 1.4738,
+      "step": 9773
+    },
+    {
+      "epoch": 1.686189942206504,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.199541510745954e-06,
+      "loss": 1.3455,
+      "step": 9774
+    },
+    {
+      "epoch": 1.686362460105236,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.1982492471391549e-06,
+      "loss": 1.4601,
+      "step": 9775
+    },
+    {
+      "epoch": 1.686534978003968,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.1969576356210466e-06,
+      "loss": 1.4357,
+      "step": 9776
+    },
+    {
+      "epoch": 1.6867074959027,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.1956666762873236e-06,
+      "loss": 1.4608,
+      "step": 9777
+    },
+    {
+      "epoch": 1.686880013801432,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.194376369233624e-06,
+      "loss": 1.4586,
+      "step": 9778
+    },
+    {
+      "epoch": 1.687052531700164,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.1930867145555424e-06,
+      "loss": 1.4273,
+      "step": 9779
+    },
+    {
+      "epoch": 1.6872250495988959,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.1917977123486258e-06,
+      "loss": 1.3533,
+      "step": 9780
+    },
+    {
+      "epoch": 1.6873975674976278,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.1905093627083698e-06,
+      "loss": 1.3954,
+      "step": 9781
+    },
+    {
+      "epoch": 1.6875700853963598,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.1892216657302247e-06,
+      "loss": 1.4152,
+      "step": 9782
+    },
+    {
+      "epoch": 1.6877426032950917,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.1879346215095932e-06,
+      "loss": 1.3656,
+      "step": 9783
+    },
+    {
+      "epoch": 1.6879151211938237,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.1866482301418213e-06,
+      "loss": 1.4807,
+      "step": 9784
+    },
+    {
+      "epoch": 1.6880876390925559,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.1853624917222217e-06,
+      "loss": 1.3826,
+      "step": 9785
+    },
+    {
+      "epoch": 1.6882601569912878,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.1840774063460403e-06,
+      "loss": 1.4644,
+      "step": 9786
+    },
+    {
+      "epoch": 1.6884326748900198,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.1827929741084931e-06,
+      "loss": 1.4286,
+      "step": 9787
+    },
+    {
+      "epoch": 1.688605192788752,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.1815091951047331e-06,
+      "loss": 1.4263,
+      "step": 9788
+    },
+    {
+      "epoch": 1.688777710687484,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.1802260694298717e-06,
+      "loss": 1.4465,
+      "step": 9789
+    },
+    {
+      "epoch": 1.688950228586216,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.178943597178972e-06,
+      "loss": 1.398,
+      "step": 9790
+    },
+    {
+      "epoch": 1.6891227464849479,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.1776617784470469e-06,
+      "loss": 1.4605,
+      "step": 9791
+    },
+    {
+      "epoch": 1.6892952643836798,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.1763806133290623e-06,
+      "loss": 1.4018,
+      "step": 9792
+    },
+    {
+      "epoch": 1.6894677822824118,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.1751001019199337e-06,
+      "loss": 1.4055,
+      "step": 9793
+    },
+    {
+      "epoch": 1.6896403001811438,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.1738202443145307e-06,
+      "loss": 1.4657,
+      "step": 9794
+    },
+    {
+      "epoch": 1.6898128180798757,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.172541040607672e-06,
+      "loss": 1.4122,
+      "step": 9795
+    },
+    {
+      "epoch": 1.6899853359786077,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.1712624908941318e-06,
+      "loss": 1.3335,
+      "step": 9796
+    },
+    {
+      "epoch": 1.6901578538773396,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.1699845952686273e-06,
+      "loss": 1.3836,
+      "step": 9797
+    },
+    {
+      "epoch": 1.6903303717760718,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.1687073538258398e-06,
+      "loss": 1.4557,
+      "step": 9798
+    },
+    {
+      "epoch": 1.6905028896748038,
+      "grad_norm": 0.625,
+      "learning_rate": 1.1674307666603901e-06,
+      "loss": 1.4109,
+      "step": 9799
+    },
+    {
+      "epoch": 1.6906754075735357,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.1661548338668572e-06,
+      "loss": 1.3921,
+      "step": 9800
+    },
+    {
+      "epoch": 1.6906754075735357,
+      "eval_loss": 1.4071011543273926,
+      "eval_runtime": 10.8438,
+      "eval_samples_per_second": 94.432,
+      "eval_steps_per_second": 23.608,
+      "step": 9800
+    },
+    {
+      "epoch": 1.6908479254722677,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.1648795555397719e-06,
+      "loss": 1.3999,
+      "step": 9801
+    },
+    {
+      "epoch": 1.6910204433709999,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.163604931773612e-06,
+      "loss": 1.5643,
+      "step": 9802
+    },
+    {
+      "epoch": 1.6911929612697318,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.1623309626628121e-06,
+      "loss": 1.4941,
+      "step": 9803
+    },
+    {
+      "epoch": 1.6913654791684638,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.1610576483017566e-06,
+      "loss": 1.5024,
+      "step": 9804
+    },
+    {
+      "epoch": 1.6915379970671958,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.1597849887847746e-06,
+      "loss": 1.4233,
+      "step": 9805
+    },
+    {
+      "epoch": 1.6917105149659277,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.1585129842061605e-06,
+      "loss": 1.3673,
+      "step": 9806
+    },
+    {
+      "epoch": 1.6918830328646597,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.1572416346601467e-06,
+      "loss": 1.4561,
+      "step": 9807
+    },
+    {
+      "epoch": 1.6920555507633916,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.1559709402409236e-06,
+      "loss": 1.3763,
+      "step": 9808
+    },
+    {
+      "epoch": 1.6922280686621236,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.1547009010426368e-06,
+      "loss": 1.423,
+      "step": 9809
+    },
+    {
+      "epoch": 1.6924005865608556,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.1534315171593736e-06,
+      "loss": 1.5238,
+      "step": 9810
+    },
+    {
+      "epoch": 1.6925731044595875,
+      "grad_norm": 0.625,
+      "learning_rate": 1.1521627886851794e-06,
+      "loss": 1.4278,
+      "step": 9811
+    },
+    {
+      "epoch": 1.6927456223583197,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.1508947157140493e-06,
+      "loss": 1.3774,
+      "step": 9812
+    },
+    {
+      "epoch": 1.6929181402570517,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.1496272983399303e-06,
+      "loss": 1.439,
+      "step": 9813
+    },
+    {
+      "epoch": 1.6930906581557836,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.1483605366567208e-06,
+      "loss": 1.3043,
+      "step": 9814
+    },
+    {
+      "epoch": 1.6932631760545158,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.147094430758272e-06,
+      "loss": 1.4669,
+      "step": 9815
+    },
+    {
+      "epoch": 1.6934356939532478,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.1458289807383804e-06,
+      "loss": 1.4208,
+      "step": 9816
+    },
+    {
+      "epoch": 1.6936082118519797,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.1445641866908042e-06,
+      "loss": 1.4478,
+      "step": 9817
+    },
+    {
+      "epoch": 1.6937807297507117,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.1433000487092415e-06,
+      "loss": 1.4366,
+      "step": 9818
+    },
+    {
+      "epoch": 1.6939532476494437,
+      "grad_norm": 0.625,
+      "learning_rate": 1.1420365668873535e-06,
+      "loss": 1.5139,
+      "step": 9819
+    },
+    {
+      "epoch": 1.6941257655481756,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.1407737413187426e-06,
+      "loss": 1.4421,
+      "step": 9820
+    },
+    {
+      "epoch": 1.6942982834469076,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.1395115720969663e-06,
+      "loss": 1.3685,
+      "step": 9821
+    },
+    {
+      "epoch": 1.6944708013456395,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.1382500593155377e-06,
+      "loss": 1.443,
+      "step": 9822
+    },
+    {
+      "epoch": 1.6946433192443715,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.1369892030679142e-06,
+      "loss": 1.4063,
+      "step": 9823
+    },
+    {
+      "epoch": 1.6948158371431035,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.1357290034475087e-06,
+      "loss": 1.3731,
+      "step": 9824
+    },
+    {
+      "epoch": 1.6949883550418354,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.1344694605476859e-06,
+      "loss": 1.3671,
+      "step": 9825
+    },
+    {
+      "epoch": 1.6951608729405676,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.1332105744617605e-06,
+      "loss": 1.4597,
+      "step": 9826
+    },
+    {
+      "epoch": 1.6953333908392996,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.1319523452829973e-06,
+      "loss": 1.3508,
+      "step": 9827
+    },
+    {
+      "epoch": 1.6955059087380315,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.1306947731046169e-06,
+      "loss": 1.4448,
+      "step": 9828
+    },
+    {
+      "epoch": 1.6956784266367637,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.1294378580197806e-06,
+      "loss": 1.3837,
+      "step": 9829
+    },
+    {
+      "epoch": 1.6958509445354957,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.1281816001216183e-06,
+      "loss": 1.3462,
+      "step": 9830
+    },
+    {
+      "epoch": 1.6960234624342276,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.1269259995031955e-06,
+      "loss": 1.4266,
+      "step": 9831
+    },
+    {
+      "epoch": 1.6961959803329596,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.1256710562575346e-06,
+      "loss": 1.4333,
+      "step": 9832
+    },
+    {
+      "epoch": 1.6963684982316916,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.124416770477612e-06,
+      "loss": 1.3963,
+      "step": 9833
+    },
+    {
+      "epoch": 1.6965410161304235,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.1231631422563526e-06,
+      "loss": 1.4522,
+      "step": 9834
+    },
+    {
+      "epoch": 1.6967135340291555,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.1219101716866332e-06,
+      "loss": 1.3365,
+      "step": 9835
+    },
+    {
+      "epoch": 1.6968860519278874,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.1206578588612815e-06,
+      "loss": 1.4551,
+      "step": 9836
+    },
+    {
+      "epoch": 1.6970585698266194,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.1194062038730735e-06,
+      "loss": 1.5414,
+      "step": 9837
+    },
+    {
+      "epoch": 1.6972310877253514,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.118155206814746e-06,
+      "loss": 1.4245,
+      "step": 9838
+    },
+    {
+      "epoch": 1.6974036056240835,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.1169048677789751e-06,
+      "loss": 1.4564,
+      "step": 9839
+    },
+    {
+      "epoch": 1.6975761235228155,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.1156551868583942e-06,
+      "loss": 1.4433,
+      "step": 9840
+    },
+    {
+      "epoch": 1.6977486414215475,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.1144061641455927e-06,
+      "loss": 1.4318,
+      "step": 9841
+    },
+    {
+      "epoch": 1.6979211593202794,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.1131577997331e-06,
+      "loss": 1.4245,
+      "step": 9842
+    },
+    {
+      "epoch": 1.6980936772190116,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.1119100937134052e-06,
+      "loss": 1.5177,
+      "step": 9843
+    },
+    {
+      "epoch": 1.6982661951177436,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.1106630461789459e-06,
+      "loss": 1.4323,
+      "step": 9844
+    },
+    {
+      "epoch": 1.6984387130164755,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.109416657222112e-06,
+      "loss": 1.38,
+      "step": 9845
+    },
+    {
+      "epoch": 1.6986112309152075,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.1081709269352426e-06,
+      "loss": 1.5297,
+      "step": 9846
+    },
+    {
+      "epoch": 1.6987837488139395,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.1069258554106289e-06,
+      "loss": 1.3105,
+      "step": 9847
+    },
+    {
+      "epoch": 1.6989562667126714,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.1056814427405148e-06,
+      "loss": 1.4188,
+      "step": 9848
+    },
+    {
+      "epoch": 1.6991287846114034,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.1044376890170971e-06,
+      "loss": 1.4763,
+      "step": 9849
+    },
+    {
+      "epoch": 1.6993013025101353,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.1031945943325118e-06,
+      "loss": 1.3321,
+      "step": 9850
+    },
+    {
+      "epoch": 1.6994738204088673,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.101952158778865e-06,
+      "loss": 1.3957,
+      "step": 9851
+    },
+    {
+      "epoch": 1.6996463383075993,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.100710382448198e-06,
+      "loss": 1.5045,
+      "step": 9852
+    },
+    {
+      "epoch": 1.6998188562063314,
+      "grad_norm": 1.75,
+      "learning_rate": 1.0994692654325111e-06,
+      "loss": 1.5065,
+      "step": 9853
+    },
+    {
+      "epoch": 1.6999913741050634,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.0982288078237547e-06,
+      "loss": 1.3886,
+      "step": 9854
+    },
+    {
+      "epoch": 1.7001638920037954,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.096989009713828e-06,
+      "loss": 1.4662,
+      "step": 9855
+    },
+    {
+      "epoch": 1.7003364099025275,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.0957498711945858e-06,
+      "loss": 1.3959,
+      "step": 9856
+    },
+    {
+      "epoch": 1.7005089278012595,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.0945113923578277e-06,
+      "loss": 1.5277,
+      "step": 9857
+    },
+    {
+      "epoch": 1.7006814456999915,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.0932735732953103e-06,
+      "loss": 1.4183,
+      "step": 9858
+    },
+    {
+      "epoch": 1.7008539635987234,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.0920364140987383e-06,
+      "loss": 1.3968,
+      "step": 9859
+    },
+    {
+      "epoch": 1.7010264814974554,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.0907999148597703e-06,
+      "loss": 1.4088,
+      "step": 9860
+    },
+    {
+      "epoch": 1.7011989993961874,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.0895640756700087e-06,
+      "loss": 1.3208,
+      "step": 9861
+    },
+    {
+      "epoch": 1.7013715172949193,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.0883288966210181e-06,
+      "loss": 1.4109,
+      "step": 9862
+    },
+    {
+      "epoch": 1.7015440351936513,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.0870943778043041e-06,
+      "loss": 1.4039,
+      "step": 9863
+    },
+    {
+      "epoch": 1.7017165530923832,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.0858605193113292e-06,
+      "loss": 1.3308,
+      "step": 9864
+    },
+    {
+      "epoch": 1.7018890709911152,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.0846273212335046e-06,
+      "loss": 1.3447,
+      "step": 9865
+    },
+    {
+      "epoch": 1.7020615888898474,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.083394783662194e-06,
+      "loss": 1.4377,
+      "step": 9866
+    },
+    {
+      "epoch": 1.7022341067885793,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.0821629066887118e-06,
+      "loss": 1.4277,
+      "step": 9867
+    },
+    {
+      "epoch": 1.7024066246873113,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.0809316904043243e-06,
+      "loss": 1.3596,
+      "step": 9868
+    },
+    {
+      "epoch": 1.7025791425860433,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.0797011349002418e-06,
+      "loss": 1.3799,
+      "step": 9869
+    },
+    {
+      "epoch": 1.7027516604847754,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.0784712402676412e-06,
+      "loss": 1.5748,
+      "step": 9870
+    },
+    {
+      "epoch": 1.7029241783835074,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.0772420065976319e-06,
+      "loss": 1.4519,
+      "step": 9871
+    },
+    {
+      "epoch": 1.7030966962822394,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.0760134339812856e-06,
+      "loss": 1.4332,
+      "step": 9872
+    },
+    {
+      "epoch": 1.7032692141809713,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.0747855225096272e-06,
+      "loss": 1.376,
+      "step": 9873
+    },
+    {
+      "epoch": 1.7034417320797033,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.0735582722736205e-06,
+      "loss": 1.3404,
+      "step": 9874
+    },
+    {
+      "epoch": 1.7036142499784352,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.0723316833641961e-06,
+      "loss": 1.3909,
+      "step": 9875
+    },
+    {
+      "epoch": 1.7037867678771672,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.0711057558722216e-06,
+      "loss": 1.3385,
+      "step": 9876
+    },
+    {
+      "epoch": 1.7039592857758992,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.069880489888523e-06,
+      "loss": 1.4349,
+      "step": 9877
+    },
+    {
+      "epoch": 1.7041318036746311,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.0686558855038753e-06,
+      "loss": 1.4143,
+      "step": 9878
+    },
+    {
+      "epoch": 1.704304321573363,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.0674319428090052e-06,
+      "loss": 1.489,
+      "step": 9879
+    },
+    {
+      "epoch": 1.7044768394720953,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.0662086618945911e-06,
+      "loss": 1.4414,
+      "step": 9880
+    },
+    {
+      "epoch": 1.7046493573708272,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.0649860428512604e-06,
+      "loss": 1.3912,
+      "step": 9881
+    },
+    {
+      "epoch": 1.7048218752695592,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.0637640857695897e-06,
+      "loss": 1.408,
+      "step": 9882
+    },
+    {
+      "epoch": 1.7049943931682912,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.0625427907401154e-06,
+      "loss": 1.3399,
+      "step": 9883
+    },
+    {
+      "epoch": 1.7051669110670233,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.0613221578533128e-06,
+      "loss": 1.3745,
+      "step": 9884
+    },
+    {
+      "epoch": 1.7053394289657553,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.0601021871996154e-06,
+      "loss": 1.302,
+      "step": 9885
+    },
+    {
+      "epoch": 1.7055119468644873,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.0588828788694082e-06,
+      "loss": 1.4234,
+      "step": 9886
+    },
+    {
+      "epoch": 1.7056844647632192,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.057664232953024e-06,
+      "loss": 1.436,
+      "step": 9887
+    },
+    {
+      "epoch": 1.7058569826619512,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.0564462495407468e-06,
+      "loss": 1.3744,
+      "step": 9888
+    },
+    {
+      "epoch": 1.7060295005606831,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.0552289287228145e-06,
+      "loss": 1.5192,
+      "step": 9889
+    },
+    {
+      "epoch": 1.706202018459415,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.0540122705894117e-06,
+      "loss": 1.4054,
+      "step": 9890
+    },
+    {
+      "epoch": 1.706374536358147,
+      "grad_norm": 0.625,
+      "learning_rate": 1.0527962752306776e-06,
+      "loss": 1.4689,
+      "step": 9891
+    },
+    {
+      "epoch": 1.706547054256879,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.051580942736702e-06,
+      "loss": 1.4161,
+      "step": 9892
+    },
+    {
+      "epoch": 1.706719572155611,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.0503662731975184e-06,
+      "loss": 1.3389,
+      "step": 9893
+    },
+    {
+      "epoch": 1.7068920900543432,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.049152266703124e-06,
+      "loss": 1.4677,
+      "step": 9894
+    },
+    {
+      "epoch": 1.7070646079530751,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.0479389233434566e-06,
+      "loss": 1.3986,
+      "step": 9895
+    },
+    {
+      "epoch": 1.707237125851807,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.0467262432084092e-06,
+      "loss": 1.497,
+      "step": 9896
+    },
+    {
+      "epoch": 1.7074096437505393,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.0455142263878226e-06,
+      "loss": 1.4305,
+      "step": 9897
+    },
+    {
+      "epoch": 1.7075821616492712,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.0443028729714944e-06,
+      "loss": 1.4552,
+      "step": 9898
+    },
+    {
+      "epoch": 1.7077546795480032,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.0430921830491657e-06,
+      "loss": 1.3979,
+      "step": 9899
+    },
+    {
+      "epoch": 1.7079271974467352,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.0418821567105353e-06,
+      "loss": 1.3842,
+      "step": 9900
+    },
+    {
+      "epoch": 1.7079271974467352,
+      "eval_loss": 1.4070851802825928,
+      "eval_runtime": 10.7355,
+      "eval_samples_per_second": 95.384,
+      "eval_steps_per_second": 23.846,
+      "step": 9900
+    },
+    {
+      "epoch": 1.7080997153454671,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.0406727940452443e-06,
+      "loss": 1.5596,
+      "step": 9901
+    },
+    {
+      "epoch": 1.708272233244199,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.0394640951428968e-06,
+      "loss": 1.416,
+      "step": 9902
+    },
+    {
+      "epoch": 1.708444751142931,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.038256060093036e-06,
+      "loss": 1.4251,
+      "step": 9903
+    },
+    {
+      "epoch": 1.708617269041663,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.0370486889851616e-06,
+      "loss": 1.3457,
+      "step": 9904
+    },
+    {
+      "epoch": 1.708789786940395,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.0358419819087228e-06,
+      "loss": 1.434,
+      "step": 9905
+    },
+    {
+      "epoch": 1.708962304839127,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.034635938953119e-06,
+      "loss": 1.3788,
+      "step": 9906
+    },
+    {
+      "epoch": 1.709134822737859,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.0334305602077067e-06,
+      "loss": 1.4269,
+      "step": 9907
+    },
+    {
+      "epoch": 1.709307340636591,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.0322258457617828e-06,
+      "loss": 1.4744,
+      "step": 9908
+    },
+    {
+      "epoch": 1.709479858535323,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.0310217957046009e-06,
+      "loss": 1.3707,
+      "step": 9909
+    },
+    {
+      "epoch": 1.709652376434055,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.029818410125365e-06,
+      "loss": 1.4436,
+      "step": 9910
+    },
+    {
+      "epoch": 1.7098248943327872,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.0286156891132303e-06,
+      "loss": 1.3644,
+      "step": 9911
+    },
+    {
+      "epoch": 1.7099974122315191,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.0274136327573004e-06,
+      "loss": 1.4127,
+      "step": 9912
+    },
+    {
+      "epoch": 1.710169930130251,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.0262122411466346e-06,
+      "loss": 1.3324,
+      "step": 9913
+    },
+    {
+      "epoch": 1.710342448028983,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.0250115143702321e-06,
+      "loss": 1.5323,
+      "step": 9914
+    },
+    {
+      "epoch": 1.710514965927715,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.0238114525170595e-06,
+      "loss": 1.4241,
+      "step": 9915
+    },
+    {
+      "epoch": 1.710687483826447,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.0226120556760178e-06,
+      "loss": 1.4542,
+      "step": 9916
+    },
+    {
+      "epoch": 1.710860001725179,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.021413323935969e-06,
+      "loss": 1.4219,
+      "step": 9917
+    },
+    {
+      "epoch": 1.711032519623911,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.0202152573857216e-06,
+      "loss": 1.3185,
+      "step": 9918
+    },
+    {
+      "epoch": 1.7112050375226429,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.0190178561140363e-06,
+      "loss": 1.4635,
+      "step": 9919
+    },
+    {
+      "epoch": 1.7113775554213748,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.0178211202096233e-06,
+      "loss": 1.3594,
+      "step": 9920
+    },
+    {
+      "epoch": 1.711550073320107,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.0166250497611462e-06,
+      "loss": 1.4801,
+      "step": 9921
+    },
+    {
+      "epoch": 1.711722591218839,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.0154296448572154e-06,
+      "loss": 1.392,
+      "step": 9922
+    },
+    {
+      "epoch": 1.711895109117571,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.0142349055863942e-06,
+      "loss": 1.434,
+      "step": 9923
+    },
+    {
+      "epoch": 1.712067627016303,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.0130408320371998e-06,
+      "loss": 1.4354,
+      "step": 9924
+    },
+    {
+      "epoch": 1.712240144915035,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.0118474242980892e-06,
+      "loss": 1.3889,
+      "step": 9925
+    },
+    {
+      "epoch": 1.712412662813767,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.010654682457486e-06,
+      "loss": 1.4636,
+      "step": 9926
+    },
+    {
+      "epoch": 1.712585180712499,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.0094626066037506e-06,
+      "loss": 1.4264,
+      "step": 9927
+    },
+    {
+      "epoch": 1.712757698611231,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.0082711968252e-06,
+      "loss": 1.4446,
+      "step": 9928
+    },
+    {
+      "epoch": 1.712930216509963,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.0070804532101019e-06,
+      "loss": 1.4269,
+      "step": 9929
+    },
+    {
+      "epoch": 1.7131027344086949,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.0058903758466743e-06,
+      "loss": 1.4745,
+      "step": 9930
+    },
+    {
+      "epoch": 1.7132752523074268,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.0047009648230865e-06,
+      "loss": 1.5025,
+      "step": 9931
+    },
+    {
+      "epoch": 1.7134477702061588,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.0035122202274572e-06,
+      "loss": 1.5489,
+      "step": 9932
+    },
+    {
+      "epoch": 1.7136202881048908,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.0023241421478513e-06,
+      "loss": 1.3831,
+      "step": 9933
+    },
+    {
+      "epoch": 1.7137928060036227,
+      "grad_norm": 0.625,
+      "learning_rate": 1.0011367306722975e-06,
+      "loss": 1.3894,
+      "step": 9934
+    },
+    {
+      "epoch": 1.713965323902355,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.999499858887585e-07,
+      "loss": 1.3981,
+      "step": 9935
+    },
+    {
+      "epoch": 1.7141378418010869,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.987639078851629e-07,
+      "loss": 1.3585,
+      "step": 9936
+    },
+    {
+      "epoch": 1.7143103596998188,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.975784967493774e-07,
+      "loss": 1.4269,
+      "step": 9937
+    },
+    {
+      "epoch": 1.714482877598551,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.963937525692247e-07,
+      "loss": 1.4254,
+      "step": 9938
+    },
+    {
+      "epoch": 1.714655395497283,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.952096754324847e-07,
+      "loss": 1.5058,
+      "step": 9939
+    },
+    {
+      "epoch": 1.714827913396015,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.940262654268729e-07,
+      "loss": 1.4928,
+      "step": 9940
+    },
+    {
+      "epoch": 1.7150004312947469,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.928435226400678e-07,
+      "loss": 1.3425,
+      "step": 9941
+    },
+    {
+      "epoch": 1.7151729491934788,
+      "grad_norm": 0.71875,
+      "learning_rate": 9.916614471596953e-07,
+      "loss": 1.456,
+      "step": 9942
+    },
+    {
+      "epoch": 1.7153454670922108,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.904800390733283e-07,
+      "loss": 1.4231,
+      "step": 9943
+    },
+    {
+      "epoch": 1.7155179849909428,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.892992984684935e-07,
+      "loss": 1.3841,
+      "step": 9944
+    },
+    {
+      "epoch": 1.7156905028896747,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.88119225432671e-07,
+      "loss": 1.4788,
+      "step": 9945
+    },
+    {
+      "epoch": 1.7158630207884067,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.86939820053281e-07,
+      "loss": 1.4304,
+      "step": 9946
+    },
+    {
+      "epoch": 1.7160355386871387,
+      "grad_norm": 0.66796875,
+      "learning_rate": 9.857610824177088e-07,
+      "loss": 1.3979,
+      "step": 9947
+    },
+    {
+      "epoch": 1.7162080565858708,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.845830126132782e-07,
+      "loss": 1.3864,
+      "step": 9948
+    },
+    {
+      "epoch": 1.7163805744846028,
+      "grad_norm": 0.78125,
+      "learning_rate": 9.834056107272682e-07,
+      "loss": 1.4578,
+      "step": 9949
+    },
+    {
+      "epoch": 1.7165530923833348,
+      "grad_norm": 0.78515625,
+      "learning_rate": 9.82228876846909e-07,
+      "loss": 1.4952,
+      "step": 9950
+    },
+    {
+      "epoch": 1.7167256102820667,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.810528110593798e-07,
+      "loss": 1.4029,
+      "step": 9951
+    },
+    {
+      "epoch": 1.716898128180799,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.798774134518119e-07,
+      "loss": 1.446,
+      "step": 9952
+    },
+    {
+      "epoch": 1.7170706460795309,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.787026841112856e-07,
+      "loss": 1.2539,
+      "step": 9953
+    },
+    {
+      "epoch": 1.7172431639782628,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.775286231248316e-07,
+      "loss": 1.4176,
+      "step": 9954
+    },
+    {
+      "epoch": 1.7174156818769948,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.763552305794323e-07,
+      "loss": 1.3271,
+      "step": 9955
+    },
+    {
+      "epoch": 1.7175881997757267,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.751825065620223e-07,
+      "loss": 1.5245,
+      "step": 9956
+    },
+    {
+      "epoch": 1.7177607176744587,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.740104511594783e-07,
+      "loss": 1.5093,
+      "step": 9957
+    },
+    {
+      "epoch": 1.7179332355731907,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.728390644586394e-07,
+      "loss": 1.4256,
+      "step": 9958
+    },
+    {
+      "epoch": 1.7181057534719226,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.716683465462862e-07,
+      "loss": 1.3404,
+      "step": 9959
+    },
+    {
+      "epoch": 1.7182782713706546,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.704982975091538e-07,
+      "loss": 1.4166,
+      "step": 9960
+    },
+    {
+      "epoch": 1.7184507892693865,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.69328917433926e-07,
+      "loss": 1.3496,
+      "step": 9961
+    },
+    {
+      "epoch": 1.7186233071681187,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.681602064072382e-07,
+      "loss": 1.3654,
+      "step": 9962
+    },
+    {
+      "epoch": 1.7187958250668507,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.669921645156755e-07,
+      "loss": 1.4074,
+      "step": 9963
+    },
+    {
+      "epoch": 1.7189683429655827,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.658247918457763e-07,
+      "loss": 1.3444,
+      "step": 9964
+    },
+    {
+      "epoch": 1.7191408608643148,
+      "grad_norm": 0.8125,
+      "learning_rate": 9.646580884840207e-07,
+      "loss": 1.5314,
+      "step": 9965
+    },
+    {
+      "epoch": 1.7193133787630468,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.634920545168535e-07,
+      "loss": 1.4694,
+      "step": 9966
+    },
+    {
+      "epoch": 1.7194858966617788,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.623266900306538e-07,
+      "loss": 1.421,
+      "step": 9967
+    },
+    {
+      "epoch": 1.7196584145605107,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.611619951117657e-07,
+      "loss": 1.4543,
+      "step": 9968
+    },
+    {
+      "epoch": 1.7198309324592427,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.599979698464733e-07,
+      "loss": 1.4106,
+      "step": 9969
+    },
+    {
+      "epoch": 1.7200034503579746,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.588346143210137e-07,
+      "loss": 1.4361,
+      "step": 9970
+    },
+    {
+      "epoch": 1.7201759682567066,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.576719286215818e-07,
+      "loss": 1.3925,
+      "step": 9971
+    },
+    {
+      "epoch": 1.7203484861554386,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.565099128343103e-07,
+      "loss": 1.4657,
+      "step": 9972
+    },
+    {
+      "epoch": 1.7205210040541705,
+      "grad_norm": 0.96484375,
+      "learning_rate": 9.553485670452911e-07,
+      "loss": 1.3915,
+      "step": 9973
+    },
+    {
+      "epoch": 1.7206935219529025,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.541878913405633e-07,
+      "loss": 1.4143,
+      "step": 9974
+    },
+    {
+      "epoch": 1.7208660398516344,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.530278858061171e-07,
+      "loss": 1.4653,
+      "step": 9975
+    },
+    {
+      "epoch": 1.7210385577503666,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.518685505278935e-07,
+      "loss": 1.4554,
+      "step": 9976
+    },
+    {
+      "epoch": 1.7212110756490986,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.507098855917851e-07,
+      "loss": 1.5149,
+      "step": 9977
+    },
+    {
+      "epoch": 1.7213835935478305,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.495518910836276e-07,
+      "loss": 1.555,
+      "step": 9978
+    },
+    {
+      "epoch": 1.7215561114465627,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.483945670892191e-07,
+      "loss": 1.4126,
+      "step": 9979
+    },
+    {
+      "epoch": 1.7217286293452947,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.472379136942955e-07,
+      "loss": 1.4291,
+      "step": 9980
+    },
+    {
+      "epoch": 1.7219011472440267,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.460819309845526e-07,
+      "loss": 1.3575,
+      "step": 9981
+    },
+    {
+      "epoch": 1.7220736651427586,
+      "grad_norm": 0.625,
+      "learning_rate": 9.449266190456318e-07,
+      "loss": 1.4072,
+      "step": 9982
+    },
+    {
+      "epoch": 1.7222461830414906,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.437719779631249e-07,
+      "loss": 1.4664,
+      "step": 9983
+    },
+    {
+      "epoch": 1.7224187009402225,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.426180078225766e-07,
+      "loss": 1.4289,
+      "step": 9984
+    },
+    {
+      "epoch": 1.7225912188389545,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.414647087094786e-07,
+      "loss": 1.3887,
+      "step": 9985
+    },
+    {
+      "epoch": 1.7227637367376865,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.403120807092759e-07,
+      "loss": 1.4219,
+      "step": 9986
+    },
+    {
+      "epoch": 1.7229362546364184,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.391601239073611e-07,
+      "loss": 1.3415,
+      "step": 9987
+    },
+    {
+      "epoch": 1.7231087725351504,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.380088383890818e-07,
+      "loss": 1.3889,
+      "step": 9988
+    },
+    {
+      "epoch": 1.7232812904338826,
+      "grad_norm": 0.6328125,
+      "learning_rate": 9.368582242397262e-07,
+      "loss": 1.3631,
+      "step": 9989
+    },
+    {
+      "epoch": 1.7234538083326145,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.357082815445451e-07,
+      "loss": 1.4397,
+      "step": 9990
+    },
+    {
+      "epoch": 1.7236263262313465,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.345590103887292e-07,
+      "loss": 1.4168,
+      "step": 9991
+    },
+    {
+      "epoch": 1.7237988441300784,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.33410410857426e-07,
+      "loss": 1.4388,
+      "step": 9992
+    },
+    {
+      "epoch": 1.7239713620288106,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.322624830357297e-07,
+      "loss": 1.3092,
+      "step": 9993
+    },
+    {
+      "epoch": 1.7241438799275426,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.311152270086865e-07,
+      "loss": 1.3909,
+      "step": 9994
+    },
+    {
+      "epoch": 1.7243163978262745,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.29968642861293e-07,
+      "loss": 1.4871,
+      "step": 9995
+    },
+    {
+      "epoch": 1.7244889157250065,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.288227306784936e-07,
+      "loss": 1.403,
+      "step": 9996
+    },
+    {
+      "epoch": 1.7246614336237385,
+      "grad_norm": 0.71484375,
+      "learning_rate": 9.276774905451868e-07,
+      "loss": 1.4458,
+      "step": 9997
+    },
+    {
+      "epoch": 1.7248339515224704,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.265329225462183e-07,
+      "loss": 1.4917,
+      "step": 9998
+    },
+    {
+      "epoch": 1.7250064694212024,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.253890267663824e-07,
+      "loss": 1.4892,
+      "step": 9999
+    },
+    {
+      "epoch": 1.7251789873199344,
+      "grad_norm": 0.70703125,
+      "learning_rate": 9.242458032904311e-07,
+      "loss": 1.5056,
+      "step": 10000
+    },
+    {
+      "epoch": 1.7251789873199344,
+      "eval_loss": 1.4070346355438232,
+      "eval_runtime": 10.8088,
+      "eval_samples_per_second": 94.738,
+      "eval_steps_per_second": 23.684,
+      "step": 10000
+    },
+    {
+      "epoch": 1.7253515052186663,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.231032522030569e-07,
+      "loss": 1.508,
+      "step": 10001
+    },
+    {
+      "epoch": 1.7255240231173983,
+      "grad_norm": 0.6171875,
+      "learning_rate": 9.219613735889066e-07,
+      "loss": 1.4105,
+      "step": 10002
+    },
+    {
+      "epoch": 1.7256965410161305,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.208201675325834e-07,
+      "loss": 1.4321,
+      "step": 10003
+    },
+    {
+      "epoch": 1.7258690589148624,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.19679634118631e-07,
+      "loss": 1.3757,
+      "step": 10004
+    },
+    {
+      "epoch": 1.7260415768135944,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.185397734315471e-07,
+      "loss": 1.4183,
+      "step": 10005
+    },
+    {
+      "epoch": 1.7262140947123266,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.174005855557799e-07,
+      "loss": 1.3735,
+      "step": 10006
+    },
+    {
+      "epoch": 1.7263866126110585,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.162620705757286e-07,
+      "loss": 1.4336,
+      "step": 10007
+    },
+    {
+      "epoch": 1.7265591305097905,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.1512422857574e-07,
+      "loss": 1.4212,
+      "step": 10008
+    },
+    {
+      "epoch": 1.7267316484085224,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.139870596401168e-07,
+      "loss": 1.493,
+      "step": 10009
+    },
+    {
+      "epoch": 1.7269041663072544,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.128505638530993e-07,
+      "loss": 1.4634,
+      "step": 10010
+    },
+    {
+      "epoch": 1.7270766842059864,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.117147412988958e-07,
+      "loss": 1.3975,
+      "step": 10011
+    },
+    {
+      "epoch": 1.7272492021047183,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.105795920616478e-07,
+      "loss": 1.4103,
+      "step": 10012
+    },
+    {
+      "epoch": 1.7274217200034503,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.09445116225458e-07,
+      "loss": 1.3861,
+      "step": 10013
+    },
+    {
+      "epoch": 1.7275942379021823,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.083113138743738e-07,
+      "loss": 1.3988,
+      "step": 10014
+    },
+    {
+      "epoch": 1.7277667558009142,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.071781850923944e-07,
+      "loss": 1.3857,
+      "step": 10015
+    },
+    {
+      "epoch": 1.7279392736996462,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.060457299634706e-07,
+      "loss": 1.4469,
+      "step": 10016
+    },
+    {
+      "epoch": 1.7281117915983784,
+      "grad_norm": 0.58984375,
+      "learning_rate": 9.049139485714997e-07,
+      "loss": 1.4281,
+      "step": 10017
+    },
+    {
+      "epoch": 1.7282843094971103,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.037828410003336e-07,
+      "loss": 1.4445,
+      "step": 10018
+    },
+    {
+      "epoch": 1.7284568273958423,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.026524073337695e-07,
+      "loss": 1.3679,
+      "step": 10019
+    },
+    {
+      "epoch": 1.7286293452945745,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.015226476555594e-07,
+      "loss": 1.4261,
+      "step": 10020
+    },
+    {
+      "epoch": 1.7288018631933064,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.003935620493987e-07,
+      "loss": 1.5309,
+      "step": 10021
+    },
+    {
+      "epoch": 1.7289743810920384,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.992651505989425e-07,
+      "loss": 1.3974,
+      "step": 10022
+    },
+    {
+      "epoch": 1.7291468989907703,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.981374133877851e-07,
+      "loss": 1.4646,
+      "step": 10023
+    },
+    {
+      "epoch": 1.7293194168895023,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.970103504994832e-07,
+      "loss": 1.5218,
+      "step": 10024
+    },
+    {
+      "epoch": 1.7294919347882343,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.958839620175297e-07,
+      "loss": 1.4581,
+      "step": 10025
+    },
+    {
+      "epoch": 1.7296644526869662,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.94758248025378e-07,
+      "loss": 1.3696,
+      "step": 10026
+    },
+    {
+      "epoch": 1.7298369705856982,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.936332086064281e-07,
+      "loss": 1.4554,
+      "step": 10027
+    },
+    {
+      "epoch": 1.7300094884844301,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.925088438440288e-07,
+      "loss": 1.4383,
+      "step": 10028
+    },
+    {
+      "epoch": 1.730182006383162,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.913851538214802e-07,
+      "loss": 1.4926,
+      "step": 10029
+    },
+    {
+      "epoch": 1.7303545242818943,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.902621386220355e-07,
+      "loss": 1.3719,
+      "step": 10030
+    },
+    {
+      "epoch": 1.7305270421806263,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.891397983288874e-07,
+      "loss": 1.4848,
+      "step": 10031
+    },
+    {
+      "epoch": 1.7306995600793582,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.880181330251935e-07,
+      "loss": 1.4134,
+      "step": 10032
+    },
+    {
+      "epoch": 1.7308720779780902,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.868971427940498e-07,
+      "loss": 1.4186,
+      "step": 10033
+    },
+    {
+      "epoch": 1.7310445958768224,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.857768277185041e-07,
+      "loss": 1.4711,
+      "step": 10034
+    },
+    {
+      "epoch": 1.7312171137755543,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.846571878815647e-07,
+      "loss": 1.4353,
+      "step": 10035
+    },
+    {
+      "epoch": 1.7313896316742863,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.835382233661727e-07,
+      "loss": 1.4302,
+      "step": 10036
+    },
+    {
+      "epoch": 1.7315621495730182,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.82419934255232e-07,
+      "loss": 1.4478,
+      "step": 10037
+    },
+    {
+      "epoch": 1.7317346674717502,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.813023206315918e-07,
+      "loss": 1.3949,
+      "step": 10038
+    },
+    {
+      "epoch": 1.7319071853704822,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.801853825780516e-07,
+      "loss": 1.3945,
+      "step": 10039
+    },
+    {
+      "epoch": 1.7320797032692141,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.790691201773616e-07,
+      "loss": 1.4858,
+      "step": 10040
+    },
+    {
+      "epoch": 1.732252221167946,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.779535335122236e-07,
+      "loss": 1.4134,
+      "step": 10041
+    },
+    {
+      "epoch": 1.732424739066678,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.768386226652814e-07,
+      "loss": 1.5136,
+      "step": 10042
+    },
+    {
+      "epoch": 1.73259725696541,
+      "grad_norm": 1.5234375,
+      "learning_rate": 8.757243877191412e-07,
+      "loss": 1.4558,
+      "step": 10043
+    },
+    {
+      "epoch": 1.7327697748641422,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.746108287563482e-07,
+      "loss": 1.5203,
+      "step": 10044
+    },
+    {
+      "epoch": 1.7329422927628741,
+      "grad_norm": 0.80078125,
+      "learning_rate": 8.734979458594028e-07,
+      "loss": 1.5167,
+      "step": 10045
+    },
+    {
+      "epoch": 1.733114810661606,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.723857391107549e-07,
+      "loss": 1.5184,
+      "step": 10046
+    },
+    {
+      "epoch": 1.7332873285603383,
+      "grad_norm": 0.734375,
+      "learning_rate": 8.712742085928027e-07,
+      "loss": 1.5099,
+      "step": 10047
+    },
+    {
+      "epoch": 1.7334598464590703,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.701633543878974e-07,
+      "loss": 1.4642,
+      "step": 10048
+    },
+    {
+      "epoch": 1.7336323643578022,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.69053176578335e-07,
+      "loss": 1.3658,
+      "step": 10049
+    },
+    {
+      "epoch": 1.7338048822565342,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.679436752463677e-07,
+      "loss": 1.467,
+      "step": 10050
+    },
+    {
+      "epoch": 1.7339774001552661,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.668348504741919e-07,
+      "loss": 1.4236,
+      "step": 10051
+    },
+    {
+      "epoch": 1.734149918053998,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.657267023439586e-07,
+      "loss": 1.3282,
+      "step": 10052
+    },
+    {
+      "epoch": 1.73432243595273,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.64619230937761e-07,
+      "loss": 1.5454,
+      "step": 10053
+    },
+    {
+      "epoch": 1.734494953851462,
+      "grad_norm": 0.66015625,
+      "learning_rate": 8.635124363376557e-07,
+      "loss": 1.3545,
+      "step": 10054
+    },
+    {
+      "epoch": 1.734667471750194,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.624063186256327e-07,
+      "loss": 1.3744,
+      "step": 10055
+    },
+    {
+      "epoch": 1.734839989648926,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.613008778836463e-07,
+      "loss": 1.3671,
+      "step": 10056
+    },
+    {
+      "epoch": 1.7350125075476581,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.601961141935922e-07,
+      "loss": 1.4429,
+      "step": 10057
+    },
+    {
+      "epoch": 1.73518502544639,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.590920276373172e-07,
+      "loss": 1.4882,
+      "step": 10058
+    },
+    {
+      "epoch": 1.735357543345122,
+      "grad_norm": 0.89453125,
+      "learning_rate": 8.579886182966191e-07,
+      "loss": 1.4334,
+      "step": 10059
+    },
+    {
+      "epoch": 1.735530061243854,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.56885886253247e-07,
+      "loss": 1.4344,
+      "step": 10060
+    },
+    {
+      "epoch": 1.7357025791425862,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8.557838315888966e-07,
+      "loss": 1.3999,
+      "step": 10061
+    },
+    {
+      "epoch": 1.7358750970413181,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.546824543852183e-07,
+      "loss": 1.4345,
+      "step": 10062
+    },
+    {
+      "epoch": 1.73604761494005,
+      "grad_norm": 0.625,
+      "learning_rate": 8.535817547238023e-07,
+      "loss": 1.4524,
+      "step": 10063
+    },
+    {
+      "epoch": 1.736220132838782,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.524817326862034e-07,
+      "loss": 1.5366,
+      "step": 10064
+    },
+    {
+      "epoch": 1.736392650737514,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.513823883539118e-07,
+      "loss": 1.2663,
+      "step": 10065
+    },
+    {
+      "epoch": 1.736565168636246,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.502837218083737e-07,
+      "loss": 1.3951,
+      "step": 10066
+    },
+    {
+      "epoch": 1.736737686534978,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.491857331309928e-07,
+      "loss": 1.3688,
+      "step": 10067
+    },
+    {
+      "epoch": 1.73691020443371,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.480884224031061e-07,
+      "loss": 1.378,
+      "step": 10068
+    },
+    {
+      "epoch": 1.7370827223324419,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.469917897060142e-07,
+      "loss": 1.3862,
+      "step": 10069
+    },
+    {
+      "epoch": 1.7372552402311738,
+      "grad_norm": 0.55078125,
+      "learning_rate": 8.458958351209601e-07,
+      "loss": 1.4146,
+      "step": 10070
+    },
+    {
+      "epoch": 1.737427758129906,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.448005587291408e-07,
+      "loss": 1.3553,
+      "step": 10071
+    },
+    {
+      "epoch": 1.737600276028638,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.437059606117004e-07,
+      "loss": 1.5589,
+      "step": 10072
+    },
+    {
+      "epoch": 1.73777279392737,
+      "grad_norm": 0.6015625,
+      "learning_rate": 8.426120408497351e-07,
+      "loss": 1.3337,
+      "step": 10073
+    },
+    {
+      "epoch": 1.7379453118261021,
+      "grad_norm": 0.87109375,
+      "learning_rate": 8.415187995242846e-07,
+      "loss": 1.4074,
+      "step": 10074
+    },
+    {
+      "epoch": 1.738117829724834,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.404262367163495e-07,
+      "loss": 1.3617,
+      "step": 10075
+    },
+    {
+      "epoch": 1.738290347623566,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.393343525068687e-07,
+      "loss": 1.4669,
+      "step": 10076
+    },
+    {
+      "epoch": 1.738462865522298,
+      "grad_norm": 0.6796875,
+      "learning_rate": 8.382431469767372e-07,
+      "loss": 1.4331,
+      "step": 10077
+    },
+    {
+      "epoch": 1.73863538342103,
+      "grad_norm": 0.98828125,
+      "learning_rate": 8.371526202067993e-07,
+      "loss": 1.3406,
+      "step": 10078
+    },
+    {
+      "epoch": 1.738807901319762,
+      "grad_norm": 0.62890625,
+      "learning_rate": 8.360627722778469e-07,
+      "loss": 1.3818,
+      "step": 10079
+    },
+    {
+      "epoch": 1.738980419218494,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.349736032706234e-07,
+      "loss": 1.3747,
+      "step": 10080
+    },
+    {
+      "epoch": 1.7391529371172259,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.338851132658221e-07,
+      "loss": 1.5443,
+      "step": 10081
+    },
+    {
+      "epoch": 1.7393254550159578,
+      "grad_norm": 0.640625,
+      "learning_rate": 8.327973023440827e-07,
+      "loss": 1.4706,
+      "step": 10082
+    },
+    {
+      "epoch": 1.7394979729146898,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.317101705859986e-07,
+      "loss": 1.4077,
+      "step": 10083
+    },
+    {
+      "epoch": 1.7396704908134217,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.306237180721121e-07,
+      "loss": 1.3999,
+      "step": 10084
+    },
+    {
+      "epoch": 1.739843008712154,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.295379448829133e-07,
+      "loss": 1.422,
+      "step": 10085
+    },
+    {
+      "epoch": 1.7400155266108859,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.284528510988444e-07,
+      "loss": 1.3776,
+      "step": 10086
+    },
+    {
+      "epoch": 1.7401880445096178,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.273684368002922e-07,
+      "loss": 1.4917,
+      "step": 10087
+    },
+    {
+      "epoch": 1.74036056240835,
+      "grad_norm": 0.71875,
+      "learning_rate": 8.262847020676024e-07,
+      "loss": 1.5477,
+      "step": 10088
+    },
+    {
+      "epoch": 1.740533080307082,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.252016469810597e-07,
+      "loss": 1.4891,
+      "step": 10089
+    },
+    {
+      "epoch": 1.740705598205814,
+      "grad_norm": 0.6640625,
+      "learning_rate": 8.241192716209056e-07,
+      "loss": 1.404,
+      "step": 10090
+    },
+    {
+      "epoch": 1.740878116104546,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.230375760673304e-07,
+      "loss": 1.3418,
+      "step": 10091
+    },
+    {
+      "epoch": 1.7410506340032779,
+      "grad_norm": 0.625,
+      "learning_rate": 8.219565604004709e-07,
+      "loss": 1.4332,
+      "step": 10092
+    },
+    {
+      "epoch": 1.7412231519020098,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.208762247004176e-07,
+      "loss": 1.3749,
+      "step": 10093
+    },
+    {
+      "epoch": 1.7413956698007418,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.197965690472088e-07,
+      "loss": 1.4559,
+      "step": 10094
+    },
+    {
+      "epoch": 1.7415681876994737,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.187175935208269e-07,
+      "loss": 1.3704,
+      "step": 10095
+    },
+    {
+      "epoch": 1.7417407055982057,
+      "grad_norm": 0.54296875,
+      "learning_rate": 8.17639298201216e-07,
+      "loss": 1.4201,
+      "step": 10096
+    },
+    {
+      "epoch": 1.7419132234969377,
+      "grad_norm": 0.55859375,
+      "learning_rate": 8.165616831682588e-07,
+      "loss": 1.3645,
+      "step": 10097
+    },
+    {
+      "epoch": 1.7420857413956699,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.154847485017913e-07,
+      "loss": 1.5445,
+      "step": 10098
+    },
+    {
+      "epoch": 1.7422582592944018,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.144084942816043e-07,
+      "loss": 1.4455,
+      "step": 10099
+    },
+    {
+      "epoch": 1.7424307771931338,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.133329205874286e-07,
+      "loss": 1.4443,
+      "step": 10100
+    },
+    {
+      "epoch": 1.7424307771931338,
+      "eval_loss": 1.4070847034454346,
+      "eval_runtime": 11.2683,
+      "eval_samples_per_second": 90.874,
+      "eval_steps_per_second": 22.719,
+      "step": 10100
+    },
+    {
+      "epoch": 1.7426032950918657,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.122580274989511e-07,
+      "loss": 1.3178,
+      "step": 10101
+    },
+    {
+      "epoch": 1.742775812990598,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.111838150958062e-07,
+      "loss": 1.4056,
+      "step": 10102
+    },
+    {
+      "epoch": 1.7429483308893299,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.101102834575792e-07,
+      "loss": 1.4306,
+      "step": 10103
+    },
+    {
+      "epoch": 1.7431208487880618,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.090374326638028e-07,
+      "loss": 1.3321,
+      "step": 10104
+    },
+    {
+      "epoch": 1.7432933666867938,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.079652627939638e-07,
+      "loss": 1.4098,
+      "step": 10105
+    },
+    {
+      "epoch": 1.7434658845855258,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.068937739274885e-07,
+      "loss": 1.4515,
+      "step": 10106
+    },
+    {
+      "epoch": 1.7436384024842577,
+      "grad_norm": 0.70703125,
+      "learning_rate": 8.058229661437677e-07,
+      "loss": 1.4637,
+      "step": 10107
+    },
+    {
+      "epoch": 1.7438109203829897,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.047528395221271e-07,
+      "loss": 1.4579,
+      "step": 10108
+    },
+    {
+      "epoch": 1.7439834382817216,
+      "grad_norm": 0.625,
+      "learning_rate": 8.03683394141851e-07,
+      "loss": 1.4756,
+      "step": 10109
+    },
+    {
+      "epoch": 1.7441559561804536,
+      "grad_norm": 0.5546875,
+      "learning_rate": 8.026146300821702e-07,
+      "loss": 1.3693,
+      "step": 10110
+    },
+    {
+      "epoch": 1.7443284740791856,
+      "grad_norm": 0.68359375,
+      "learning_rate": 8.015465474222661e-07,
+      "loss": 1.3065,
+      "step": 10111
+    },
+    {
+      "epoch": 1.7445009919779177,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.004791462412675e-07,
+      "loss": 1.4108,
+      "step": 10112
+    },
+    {
+      "epoch": 1.7446735098766497,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.994124266182568e-07,
+      "loss": 1.4297,
+      "step": 10113
+    },
+    {
+      "epoch": 1.7448460277753817,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.983463886322584e-07,
+      "loss": 1.4389,
+      "step": 10114
+    },
+    {
+      "epoch": 1.7450185456741139,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.97281032362256e-07,
+      "loss": 1.4527,
+      "step": 10115
+    },
+    {
+      "epoch": 1.7451910635728458,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.962163578871751e-07,
+      "loss": 1.3734,
+      "step": 10116
+    },
+    {
+      "epoch": 1.7453635814715778,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.95152365285895e-07,
+      "loss": 1.4244,
+      "step": 10117
+    },
+    {
+      "epoch": 1.7455360993703097,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.940890546372437e-07,
+      "loss": 1.408,
+      "step": 10118
+    },
+    {
+      "epoch": 1.7457086172690417,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.930264260199938e-07,
+      "loss": 1.4781,
+      "step": 10119
+    },
+    {
+      "epoch": 1.7458811351677737,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.919644795128767e-07,
+      "loss": 1.393,
+      "step": 10120
+    },
+    {
+      "epoch": 1.7460536530665056,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.909032151945639e-07,
+      "loss": 1.5179,
+      "step": 10121
+    },
+    {
+      "epoch": 1.7462261709652376,
+      "grad_norm": 0.796875,
+      "learning_rate": 7.898426331436815e-07,
+      "loss": 1.4342,
+      "step": 10122
+    },
+    {
+      "epoch": 1.7463986888639695,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.887827334388054e-07,
+      "loss": 1.4451,
+      "step": 10123
+    },
+    {
+      "epoch": 1.7465712067627015,
+      "grad_norm": 0.625,
+      "learning_rate": 7.877235161584584e-07,
+      "loss": 1.4059,
+      "step": 10124
+    },
+    {
+      "epoch": 1.7467437246614335,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.866649813811145e-07,
+      "loss": 1.4599,
+      "step": 10125
+    },
+    {
+      "epoch": 1.7469162425601656,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7.856071291851975e-07,
+      "loss": 1.4546,
+      "step": 10126
+    },
+    {
+      "epoch": 1.7470887604588976,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.845499596490758e-07,
+      "loss": 1.3923,
+      "step": 10127
+    },
+    {
+      "epoch": 1.7472612783576296,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.834934728510768e-07,
+      "loss": 1.3865,
+      "step": 10128
+    },
+    {
+      "epoch": 1.7474337962563617,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.824376688694668e-07,
+      "loss": 1.3667,
+      "step": 10129
+    },
+    {
+      "epoch": 1.7476063141550937,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.813825477824665e-07,
+      "loss": 1.5013,
+      "step": 10130
+    },
+    {
+      "epoch": 1.7477788320538257,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.803281096682524e-07,
+      "loss": 1.4107,
+      "step": 10131
+    },
+    {
+      "epoch": 1.7479513499525576,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.792743546049364e-07,
+      "loss": 1.4919,
+      "step": 10132
+    },
+    {
+      "epoch": 1.7481238678512896,
+      "grad_norm": 0.625,
+      "learning_rate": 7.782212826705892e-07,
+      "loss": 1.3758,
+      "step": 10133
+    },
+    {
+      "epoch": 1.7482963857500216,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.771688939432309e-07,
+      "loss": 1.5687,
+      "step": 10134
+    },
+    {
+      "epoch": 1.7484689036487535,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.761171885008279e-07,
+      "loss": 1.3382,
+      "step": 10135
+    },
+    {
+      "epoch": 1.7486414215474855,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.750661664212966e-07,
+      "loss": 1.3175,
+      "step": 10136
+    },
+    {
+      "epoch": 1.7488139394462174,
+      "grad_norm": 0.62109375,
+      "learning_rate": 7.74015827782505e-07,
+      "loss": 1.522,
+      "step": 10137
+    },
+    {
+      "epoch": 1.7489864573449494,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.72966172662265e-07,
+      "loss": 1.5441,
+      "step": 10138
+    },
+    {
+      "epoch": 1.7491589752436816,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.719172011383468e-07,
+      "loss": 1.3934,
+      "step": 10139
+    },
+    {
+      "epoch": 1.7493314931424135,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.708689132884606e-07,
+      "loss": 1.4005,
+      "step": 10140
+    },
+    {
+      "epoch": 1.7495040110411455,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.698213091902718e-07,
+      "loss": 1.5491,
+      "step": 10141
+    },
+    {
+      "epoch": 1.7496765289398775,
+      "grad_norm": 0.890625,
+      "learning_rate": 7.687743889213939e-07,
+      "loss": 1.431,
+      "step": 10142
+    },
+    {
+      "epoch": 1.7498490468386096,
+      "grad_norm": 0.8984375,
+      "learning_rate": 7.677281525593871e-07,
+      "loss": 1.4588,
+      "step": 10143
+    },
+    {
+      "epoch": 1.7500215647373416,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.666826001817651e-07,
+      "loss": 1.4738,
+      "step": 10144
+    },
+    {
+      "epoch": 1.7501940826360736,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.656377318659891e-07,
+      "loss": 1.3973,
+      "step": 10145
+    },
+    {
+      "epoch": 1.7503666005348055,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.645935476894684e-07,
+      "loss": 1.4568,
+      "step": 10146
+    },
+    {
+      "epoch": 1.7505391184335375,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.635500477295632e-07,
+      "loss": 1.3341,
+      "step": 10147
+    },
+    {
+      "epoch": 1.7507116363322694,
+      "grad_norm": 0.7421875,
+      "learning_rate": 7.625072320635829e-07,
+      "loss": 1.3969,
+      "step": 10148
+    },
+    {
+      "epoch": 1.7508841542310014,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.614651007687857e-07,
+      "loss": 1.5004,
+      "step": 10149
+    },
+    {
+      "epoch": 1.7510566721297334,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.604236539223797e-07,
+      "loss": 1.4903,
+      "step": 10150
+    },
+    {
+      "epoch": 1.7512291900284653,
+      "grad_norm": 0.5546875,
+      "learning_rate": 7.593828916015178e-07,
+      "loss": 1.5149,
+      "step": 10151
+    },
+    {
+      "epoch": 1.7514017079271973,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.583428138833126e-07,
+      "loss": 1.3694,
+      "step": 10152
+    },
+    {
+      "epoch": 1.7515742258259295,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.573034208448149e-07,
+      "loss": 1.3829,
+      "step": 10153
+    },
+    {
+      "epoch": 1.7517467437246614,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.562647125630307e-07,
+      "loss": 1.4432,
+      "step": 10154
+    },
+    {
+      "epoch": 1.7519192616233934,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.55226689114914e-07,
+      "loss": 1.4537,
+      "step": 10155
+    },
+    {
+      "epoch": 1.7520917795221256,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.541893505773679e-07,
+      "loss": 1.4597,
+      "step": 10156
+    },
+    {
+      "epoch": 1.7522642974208575,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.531526970272463e-07,
+      "loss": 1.3658,
+      "step": 10157
+    },
+    {
+      "epoch": 1.7524368153195895,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.5211672854135e-07,
+      "loss": 1.4581,
+      "step": 10158
+    },
+    {
+      "epoch": 1.7526093332183215,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.510814451964277e-07,
+      "loss": 1.4452,
+      "step": 10159
+    },
+    {
+      "epoch": 1.7527818511170534,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.500468470691846e-07,
+      "loss": 1.4134,
+      "step": 10160
+    },
+    {
+      "epoch": 1.7529543690157854,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.490129342362662e-07,
+      "loss": 1.3958,
+      "step": 10161
+    },
+    {
+      "epoch": 1.7531268869145173,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.479797067742711e-07,
+      "loss": 1.4252,
+      "step": 10162
+    },
+    {
+      "epoch": 1.7532994048132493,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.469471647597515e-07,
+      "loss": 1.4293,
+      "step": 10163
+    },
+    {
+      "epoch": 1.7534719227119813,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.459153082691994e-07,
+      "loss": 1.4295,
+      "step": 10164
+    },
+    {
+      "epoch": 1.7536444406107132,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.448841373790639e-07,
+      "loss": 1.3745,
+      "step": 10165
+    },
+    {
+      "epoch": 1.7538169585094452,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.438536521657402e-07,
+      "loss": 1.401,
+      "step": 10166
+    },
+    {
+      "epoch": 1.7539894764081774,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.428238527055731e-07,
+      "loss": 1.4449,
+      "step": 10167
+    },
+    {
+      "epoch": 1.7541619943069093,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.41794739074857e-07,
+      "loss": 1.3888,
+      "step": 10168
+    },
+    {
+      "epoch": 1.7543345122056413,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.407663113498353e-07,
+      "loss": 1.4028,
+      "step": 10169
+    },
+    {
+      "epoch": 1.7545070301043735,
+      "grad_norm": 0.5390625,
+      "learning_rate": 7.397385696066972e-07,
+      "loss": 1.3718,
+      "step": 10170
+    },
+    {
+      "epoch": 1.7546795480031054,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.387115139215895e-07,
+      "loss": 1.3943,
+      "step": 10171
+    },
+    {
+      "epoch": 1.7548520659018374,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.376851443705968e-07,
+      "loss": 1.5253,
+      "step": 10172
+    },
+    {
+      "epoch": 1.7550245838005694,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.36659461029765e-07,
+      "loss": 1.4728,
+      "step": 10173
+    },
+    {
+      "epoch": 1.7551971016993013,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.35634463975079e-07,
+      "loss": 1.4457,
+      "step": 10174
+    },
+    {
+      "epoch": 1.7553696195980333,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.346101532824789e-07,
+      "loss": 1.398,
+      "step": 10175
+    },
+    {
+      "epoch": 1.7555421374967652,
+      "grad_norm": 0.9921875,
+      "learning_rate": 7.33586529027851e-07,
+      "loss": 1.4072,
+      "step": 10176
+    },
+    {
+      "epoch": 1.7557146553954972,
+      "grad_norm": 0.640625,
+      "learning_rate": 7.325635912870321e-07,
+      "loss": 1.4049,
+      "step": 10177
+    },
+    {
+      "epoch": 1.7558871732942292,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.315413401358084e-07,
+      "loss": 1.529,
+      "step": 10178
+    },
+    {
+      "epoch": 1.7560596911929611,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.305197756499139e-07,
+      "loss": 1.4471,
+      "step": 10179
+    },
+    {
+      "epoch": 1.7562322090916933,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.294988979050333e-07,
+      "loss": 1.4572,
+      "step": 10180
+    },
+    {
+      "epoch": 1.7564047269904253,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.284787069767984e-07,
+      "loss": 1.3271,
+      "step": 10181
+    },
+    {
+      "epoch": 1.7565772448891572,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.274592029407946e-07,
+      "loss": 1.4513,
+      "step": 10182
+    },
+    {
+      "epoch": 1.7567497627878892,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.264403858725466e-07,
+      "loss": 1.504,
+      "step": 10183
+    },
+    {
+      "epoch": 1.7569222806866214,
+      "grad_norm": 0.55078125,
+      "learning_rate": 7.254222558475421e-07,
+      "loss": 1.3852,
+      "step": 10184
+    },
+    {
+      "epoch": 1.7570947985853533,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.244048129412051e-07,
+      "loss": 1.5207,
+      "step": 10185
+    },
+    {
+      "epoch": 1.7572673164840853,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.233880572289165e-07,
+      "loss": 1.4163,
+      "step": 10186
+    },
+    {
+      "epoch": 1.7574398343828173,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.223719887860037e-07,
+      "loss": 1.4681,
+      "step": 10187
+    },
+    {
+      "epoch": 1.7576123522815492,
+      "grad_norm": 0.625,
+      "learning_rate": 7.213566076877431e-07,
+      "loss": 1.3599,
+      "step": 10188
+    },
+    {
+      "epoch": 1.7577848701802812,
+      "grad_norm": 0.6796875,
+      "learning_rate": 7.203419140093604e-07,
+      "loss": 1.3668,
+      "step": 10189
+    },
+    {
+      "epoch": 1.7579573880790131,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.193279078260329e-07,
+      "loss": 1.4273,
+      "step": 10190
+    },
+    {
+      "epoch": 1.758129905977745,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.183145892128785e-07,
+      "loss": 1.4715,
+      "step": 10191
+    },
+    {
+      "epoch": 1.758302423876477,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.17301958244978e-07,
+      "loss": 1.4836,
+      "step": 10192
+    },
+    {
+      "epoch": 1.758474941775209,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.162900149973473e-07,
+      "loss": 1.362,
+      "step": 10193
+    },
+    {
+      "epoch": 1.7586474596739412,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.152787595449573e-07,
+      "loss": 1.3473,
+      "step": 10194
+    },
+    {
+      "epoch": 1.7588199775726732,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.142681919627348e-07,
+      "loss": 1.4041,
+      "step": 10195
+    },
+    {
+      "epoch": 1.7589924954714051,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.132583123255421e-07,
+      "loss": 1.3846,
+      "step": 10196
+    },
+    {
+      "epoch": 1.7591650133701373,
+      "grad_norm": 0.72265625,
+      "learning_rate": 7.122491207082006e-07,
+      "loss": 1.3585,
+      "step": 10197
+    },
+    {
+      "epoch": 1.7593375312688693,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.112406171854758e-07,
+      "loss": 1.4366,
+      "step": 10198
+    },
+    {
+      "epoch": 1.7595100491676012,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.102328018320859e-07,
+      "loss": 1.4385,
+      "step": 10199
+    },
+    {
+      "epoch": 1.7596825670663332,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.092256747226944e-07,
+      "loss": 1.4074,
+      "step": 10200
+    },
+    {
+      "epoch": 1.7596825670663332,
+      "eval_loss": 1.4070675373077393,
+      "eval_runtime": 10.9117,
+      "eval_samples_per_second": 93.844,
+      "eval_steps_per_second": 23.461,
+      "step": 10200
+    },
+    {
+      "epoch": 1.7598550849650652,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.082192359319184e-07,
+      "loss": 1.3849,
+      "step": 10201
+    },
+    {
+      "epoch": 1.7600276028637971,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.07213485534316e-07,
+      "loss": 1.3763,
+      "step": 10202
+    },
+    {
+      "epoch": 1.760200120762529,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.062084236044065e-07,
+      "loss": 1.4973,
+      "step": 10203
+    },
+    {
+      "epoch": 1.760372638661261,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.052040502166424e-07,
+      "loss": 1.359,
+      "step": 10204
+    },
+    {
+      "epoch": 1.760545156559993,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.042003654454432e-07,
+      "loss": 1.4431,
+      "step": 10205
+    },
+    {
+      "epoch": 1.760717674458725,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.031973693651617e-07,
+      "loss": 1.4034,
+      "step": 10206
+    },
+    {
+      "epoch": 1.7608901923574571,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.021950620501084e-07,
+      "loss": 1.5532,
+      "step": 10207
+    },
+    {
+      "epoch": 1.761062710256189,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.011934435745404e-07,
+      "loss": 1.3747,
+      "step": 10208
+    },
+    {
+      "epoch": 1.761235228154921,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.001925140126631e-07,
+      "loss": 1.4181,
+      "step": 10209
+    },
+    {
+      "epoch": 1.761407746053653,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.991922734386336e-07,
+      "loss": 1.3301,
+      "step": 10210
+    },
+    {
+      "epoch": 1.7615802639523852,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.981927219265527e-07,
+      "loss": 1.5361,
+      "step": 10211
+    },
+    {
+      "epoch": 1.7617527818511172,
+      "grad_norm": 0.7265625,
+      "learning_rate": 6.971938595504768e-07,
+      "loss": 1.4583,
+      "step": 10212
+    },
+    {
+      "epoch": 1.7619252997498491,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.961956863844055e-07,
+      "loss": 1.4323,
+      "step": 10213
+    },
+    {
+      "epoch": 1.762097817648581,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.951982025022929e-07,
+      "loss": 1.3942,
+      "step": 10214
+    },
+    {
+      "epoch": 1.762270335547313,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.942014079780335e-07,
+      "loss": 1.4092,
+      "step": 10215
+    },
+    {
+      "epoch": 1.762442853446045,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.932053028854813e-07,
+      "loss": 1.4492,
+      "step": 10216
+    },
+    {
+      "epoch": 1.762615371344777,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.922098872984317e-07,
+      "loss": 1.4895,
+      "step": 10217
+    },
+    {
+      "epoch": 1.762787889243509,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.912151612906303e-07,
+      "loss": 1.4247,
+      "step": 10218
+    },
+    {
+      "epoch": 1.762960407142241,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.902211249357738e-07,
+      "loss": 1.4815,
+      "step": 10219
+    },
+    {
+      "epoch": 1.7631329250409729,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.892277783075063e-07,
+      "loss": 1.4463,
+      "step": 10220
+    },
+    {
+      "epoch": 1.763305442939705,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.882351214794225e-07,
+      "loss": 1.422,
+      "step": 10221
+    },
+    {
+      "epoch": 1.763477960838437,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.872431545250636e-07,
+      "loss": 1.5413,
+      "step": 10222
+    },
+    {
+      "epoch": 1.763650478737169,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.862518775179183e-07,
+      "loss": 1.4846,
+      "step": 10223
+    },
+    {
+      "epoch": 1.763822996635901,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.852612905314326e-07,
+      "loss": 1.4419,
+      "step": 10224
+    },
+    {
+      "epoch": 1.763995514534633,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.842713936389889e-07,
+      "loss": 1.3773,
+      "step": 10225
+    },
+    {
+      "epoch": 1.764168032433365,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.832821869139272e-07,
+      "loss": 1.4648,
+      "step": 10226
+    },
+    {
+      "epoch": 1.764340550332097,
+      "grad_norm": 0.890625,
+      "learning_rate": 6.822936704295369e-07,
+      "loss": 1.3654,
+      "step": 10227
+    },
+    {
+      "epoch": 1.764513068230829,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.813058442590504e-07,
+      "loss": 1.4976,
+      "step": 10228
+    },
+    {
+      "epoch": 1.764685586129561,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.803187084756524e-07,
+      "loss": 1.371,
+      "step": 10229
+    },
+    {
+      "epoch": 1.764858104028293,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.793322631524768e-07,
+      "loss": 1.4333,
+      "step": 10230
+    },
+    {
+      "epoch": 1.7650306219270249,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.78346508362605e-07,
+      "loss": 1.4147,
+      "step": 10231
+    },
+    {
+      "epoch": 1.7652031398257568,
+      "grad_norm": 0.62890625,
+      "learning_rate": 6.773614441790677e-07,
+      "loss": 1.3943,
+      "step": 10232
+    },
+    {
+      "epoch": 1.7653756577244888,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.763770706748462e-07,
+      "loss": 1.4275,
+      "step": 10233
+    },
+    {
+      "epoch": 1.7655481756232208,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.75393387922868e-07,
+      "loss": 1.3645,
+      "step": 10234
+    },
+    {
+      "epoch": 1.765720693521953,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.744103959960113e-07,
+      "loss": 1.3754,
+      "step": 10235
+    },
+    {
+      "epoch": 1.765893211420685,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.734280949670991e-07,
+      "loss": 1.3763,
+      "step": 10236
+    },
+    {
+      "epoch": 1.7660657293194169,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.724464849089107e-07,
+      "loss": 1.4541,
+      "step": 10237
+    },
+    {
+      "epoch": 1.766238247218149,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.714655658941671e-07,
+      "loss": 1.3945,
+      "step": 10238
+    },
+    {
+      "epoch": 1.766410765116881,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.704853379955423e-07,
+      "loss": 1.3877,
+      "step": 10239
+    },
+    {
+      "epoch": 1.766583283015613,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.69505801285657e-07,
+      "loss": 1.4581,
+      "step": 10240
+    },
+    {
+      "epoch": 1.766755800914345,
+      "grad_norm": 0.77734375,
+      "learning_rate": 6.68526955837081e-07,
+      "loss": 1.4175,
+      "step": 10241
+    },
+    {
+      "epoch": 1.7669283188130769,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.675488017223342e-07,
+      "loss": 1.4392,
+      "step": 10242
+    },
+    {
+      "epoch": 1.7671008367118088,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.665713390138839e-07,
+      "loss": 1.3622,
+      "step": 10243
+    },
+    {
+      "epoch": 1.7672733546105408,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.655945677841457e-07,
+      "loss": 1.3939,
+      "step": 10244
+    },
+    {
+      "epoch": 1.7674458725092728,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.646184881054874e-07,
+      "loss": 1.3592,
+      "step": 10245
+    },
+    {
+      "epoch": 1.7676183904080047,
+      "grad_norm": 0.86328125,
+      "learning_rate": 6.636431000502231e-07,
+      "loss": 1.4977,
+      "step": 10246
+    },
+    {
+      "epoch": 1.7677909083067367,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.626684036906106e-07,
+      "loss": 1.4954,
+      "step": 10247
+    },
+    {
+      "epoch": 1.7679634262054689,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.616943990988689e-07,
+      "loss": 1.4395,
+      "step": 10248
+    },
+    {
+      "epoch": 1.7681359441042008,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.607210863471525e-07,
+      "loss": 1.4819,
+      "step": 10249
+    },
+    {
+      "epoch": 1.7683084620029328,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.597484655075726e-07,
+      "loss": 1.3285,
+      "step": 10250
+    },
+    {
+      "epoch": 1.7684809799016648,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.58776536652187e-07,
+      "loss": 1.4535,
+      "step": 10251
+    },
+    {
+      "epoch": 1.768653497800397,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.578052998530016e-07,
+      "loss": 1.539,
+      "step": 10252
+    },
+    {
+      "epoch": 1.768826015699129,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.568347551819731e-07,
+      "loss": 1.4043,
+      "step": 10253
+    },
+    {
+      "epoch": 1.7689985335978609,
+      "grad_norm": 0.7109375,
+      "learning_rate": 6.558649027110054e-07,
+      "loss": 1.3375,
+      "step": 10254
+    },
+    {
+      "epoch": 1.7691710514965928,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.548957425119484e-07,
+      "loss": 1.4756,
+      "step": 10255
+    },
+    {
+      "epoch": 1.7693435693953248,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.539272746566083e-07,
+      "loss": 1.4516,
+      "step": 10256
+    },
+    {
+      "epoch": 1.7695160872940567,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.529594992167321e-07,
+      "loss": 1.3337,
+      "step": 10257
+    },
+    {
+      "epoch": 1.7696886051927887,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.519924162640168e-07,
+      "loss": 1.5003,
+      "step": 10258
+    },
+    {
+      "epoch": 1.7698611230915207,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.510260258701151e-07,
+      "loss": 1.476,
+      "step": 10259
+    },
+    {
+      "epoch": 1.7700336409902526,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.500603281066175e-07,
+      "loss": 1.3975,
+      "step": 10260
+    },
+    {
+      "epoch": 1.7702061588889846,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.490953230450758e-07,
+      "loss": 1.4206,
+      "step": 10261
+    },
+    {
+      "epoch": 1.7703786767877168,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6.481310107569772e-07,
+      "loss": 1.4545,
+      "step": 10262
+    },
+    {
+      "epoch": 1.7705511946864487,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.471673913137666e-07,
+      "loss": 1.5057,
+      "step": 10263
+    },
+    {
+      "epoch": 1.7707237125851807,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.46204464786836e-07,
+      "loss": 1.3801,
+      "step": 10264
+    },
+    {
+      "epoch": 1.7708962304839129,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.452422312475226e-07,
+      "loss": 1.4318,
+      "step": 10265
+    },
+    {
+      "epoch": 1.7710687483826448,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.442806907671162e-07,
+      "loss": 1.4907,
+      "step": 10266
+    },
+    {
+      "epoch": 1.7712412662813768,
+      "grad_norm": 0.65625,
+      "learning_rate": 6.433198434168552e-07,
+      "loss": 1.512,
+      "step": 10267
+    },
+    {
+      "epoch": 1.7714137841801088,
+      "grad_norm": 0.53125,
+      "learning_rate": 6.423596892679207e-07,
+      "loss": 1.3861,
+      "step": 10268
+    },
+    {
+      "epoch": 1.7715863020788407,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.414002283914522e-07,
+      "loss": 1.4541,
+      "step": 10269
+    },
+    {
+      "epoch": 1.7717588199775727,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.404414608585285e-07,
+      "loss": 1.5342,
+      "step": 10270
+    },
+    {
+      "epoch": 1.7719313378763046,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.394833867401829e-07,
+      "loss": 1.4425,
+      "step": 10271
+    },
+    {
+      "epoch": 1.7721038557750366,
+      "grad_norm": 0.73828125,
+      "learning_rate": 6.385260061073962e-07,
+      "loss": 1.4565,
+      "step": 10272
+    },
+    {
+      "epoch": 1.7722763736737686,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.37569319031095e-07,
+      "loss": 1.4543,
+      "step": 10273
+    },
+    {
+      "epoch": 1.7724488915725005,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.366133255821572e-07,
+      "loss": 1.4817,
+      "step": 10274
+    },
+    {
+      "epoch": 1.7726214094712325,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.356580258314105e-07,
+      "loss": 1.3968,
+      "step": 10275
+    },
+    {
+      "epoch": 1.7727939273699647,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.347034198496271e-07,
+      "loss": 1.4356,
+      "step": 10276
+    },
+    {
+      "epoch": 1.7729664452686966,
+      "grad_norm": 0.70703125,
+      "learning_rate": 6.337495077075328e-07,
+      "loss": 1.4766,
+      "step": 10277
+    },
+    {
+      "epoch": 1.7731389631674286,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.327962894757988e-07,
+      "loss": 1.5474,
+      "step": 10278
+    },
+    {
+      "epoch": 1.7733114810661608,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.31843765225042e-07,
+      "loss": 1.4506,
+      "step": 10279
+    },
+    {
+      "epoch": 1.7734839989648927,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.308919350258369e-07,
+      "loss": 1.4491,
+      "step": 10280
+    },
+    {
+      "epoch": 1.7736565168636247,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.29940798948696e-07,
+      "loss": 1.4158,
+      "step": 10281
+    },
+    {
+      "epoch": 1.7738290347623566,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.289903570640887e-07,
+      "loss": 1.419,
+      "step": 10282
+    },
+    {
+      "epoch": 1.7740015526610886,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.280406094424285e-07,
+      "loss": 1.4394,
+      "step": 10283
+    },
+    {
+      "epoch": 1.7741740705598206,
+      "grad_norm": 1.4296875,
+      "learning_rate": 6.27091556154078e-07,
+      "loss": 1.5351,
+      "step": 10284
+    },
+    {
+      "epoch": 1.7743465884585525,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.261431972693499e-07,
+      "loss": 1.3564,
+      "step": 10285
+    },
+    {
+      "epoch": 1.7745191063572845,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.251955328585057e-07,
+      "loss": 1.3911,
+      "step": 10286
+    },
+    {
+      "epoch": 1.7746916242560165,
+      "grad_norm": 0.5546875,
+      "learning_rate": 6.242485629917494e-07,
+      "loss": 1.4094,
+      "step": 10287
+    },
+    {
+      "epoch": 1.7748641421547484,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.233022877392458e-07,
+      "loss": 1.4178,
+      "step": 10288
+    },
+    {
+      "epoch": 1.7750366600534806,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.223567071710946e-07,
+      "loss": 1.3358,
+      "step": 10289
+    },
+    {
+      "epoch": 1.7752091779522126,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.214118213573517e-07,
+      "loss": 1.4563,
+      "step": 10290
+    },
+    {
+      "epoch": 1.7753816958509445,
+      "grad_norm": 0.62109375,
+      "learning_rate": 6.204676303680246e-07,
+      "loss": 1.373,
+      "step": 10291
+    },
+    {
+      "epoch": 1.7755542137496765,
+      "grad_norm": 0.59375,
+      "learning_rate": 6.195241342730585e-07,
+      "loss": 1.5764,
+      "step": 10292
+    },
+    {
+      "epoch": 1.7757267316484087,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.185813331423584e-07,
+      "loss": 1.4124,
+      "step": 10293
+    },
+    {
+      "epoch": 1.7758992495471406,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.176392270457709e-07,
+      "loss": 1.3338,
+      "step": 10294
+    },
+    {
+      "epoch": 1.7760717674458726,
+      "grad_norm": 0.6640625,
+      "learning_rate": 6.166978160530923e-07,
+      "loss": 1.4685,
+      "step": 10295
+    },
+    {
+      "epoch": 1.7762442853446045,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.15757100234069e-07,
+      "loss": 1.4325,
+      "step": 10296
+    },
+    {
+      "epoch": 1.7764168032433365,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.148170796583963e-07,
+      "loss": 1.3772,
+      "step": 10297
+    },
+    {
+      "epoch": 1.7765893211420685,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.138777543957141e-07,
+      "loss": 1.4324,
+      "step": 10298
+    },
+    {
+      "epoch": 1.7767618390408004,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.129391245156168e-07,
+      "loss": 1.5199,
+      "step": 10299
+    },
+    {
+      "epoch": 1.7769343569395324,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.1200119008764e-07,
+      "loss": 1.4584,
+      "step": 10300
+    },
+    {
+      "epoch": 1.7769343569395324,
+      "eval_loss": 1.4070483446121216,
+      "eval_runtime": 11.0405,
+      "eval_samples_per_second": 92.75,
+      "eval_steps_per_second": 23.187,
+      "step": 10300
+    },
+    {
+      "epoch": 1.7771068748382643,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.110639511812765e-07,
+      "loss": 1.3487,
+      "step": 10301
+    },
+    {
+      "epoch": 1.7772793927369963,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.101274078659591e-07,
+      "loss": 1.4138,
+      "step": 10302
+    },
+    {
+      "epoch": 1.7774519106357285,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.091915602110743e-07,
+      "loss": 1.4851,
+      "step": 10303
+    },
+    {
+      "epoch": 1.7776244285344605,
+      "grad_norm": 0.66796875,
+      "learning_rate": 6.082564082859543e-07,
+      "loss": 1.4141,
+      "step": 10304
+    },
+    {
+      "epoch": 1.7777969464331924,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.073219521598828e-07,
+      "loss": 1.369,
+      "step": 10305
+    },
+    {
+      "epoch": 1.7779694643319246,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.063881919020887e-07,
+      "loss": 1.3713,
+      "step": 10306
+    },
+    {
+      "epoch": 1.7781419822306566,
+      "grad_norm": 0.84765625,
+      "learning_rate": 6.05455127581751e-07,
+      "loss": 1.5123,
+      "step": 10307
+    },
+    {
+      "epoch": 1.7783145001293885,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.04522759267997e-07,
+      "loss": 1.4518,
+      "step": 10308
+    },
+    {
+      "epoch": 1.7784870180281205,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.035910870299033e-07,
+      "loss": 1.435,
+      "step": 10309
+    },
+    {
+      "epoch": 1.7786595359268524,
+      "grad_norm": 0.640625,
+      "learning_rate": 6.026601109364949e-07,
+      "loss": 1.5177,
+      "step": 10310
+    },
+    {
+      "epoch": 1.7788320538255844,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.017298310567399e-07,
+      "loss": 1.4735,
+      "step": 10311
+    },
+    {
+      "epoch": 1.7790045717243164,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.008002474595653e-07,
+      "loss": 1.4801,
+      "step": 10312
+    },
+    {
+      "epoch": 1.7791770896230483,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.998713602138351e-07,
+      "loss": 1.3985,
+      "step": 10313
+    },
+    {
+      "epoch": 1.7793496075217803,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.989431693883696e-07,
+      "loss": 1.4779,
+      "step": 10314
+    },
+    {
+      "epoch": 1.7795221254205122,
+      "grad_norm": 0.69921875,
+      "learning_rate": 5.98015675051935e-07,
+      "loss": 1.415,
+      "step": 10315
+    },
+    {
+      "epoch": 1.7796946433192442,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.970888772732453e-07,
+      "loss": 1.4775,
+      "step": 10316
+    },
+    {
+      "epoch": 1.7798671612179764,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.961627761209632e-07,
+      "loss": 1.4578,
+      "step": 10317
+    },
+    {
+      "epoch": 1.7800396791167084,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.952373716637016e-07,
+      "loss": 1.415,
+      "step": 10318
+    },
+    {
+      "epoch": 1.7802121970154403,
+      "grad_norm": 0.54296875,
+      "learning_rate": 5.94312663970017e-07,
+      "loss": 1.3462,
+      "step": 10319
+    },
+    {
+      "epoch": 1.7803847149141725,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.933886531084232e-07,
+      "loss": 1.4286,
+      "step": 10320
+    },
+    {
+      "epoch": 1.7805572328129045,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.924653391473689e-07,
+      "loss": 1.4845,
+      "step": 10321
+    },
+    {
+      "epoch": 1.7807297507116364,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.915427221552672e-07,
+      "loss": 1.4518,
+      "step": 10322
+    },
+    {
+      "epoch": 1.7809022686103684,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.906208022004656e-07,
+      "loss": 1.4583,
+      "step": 10323
+    },
+    {
+      "epoch": 1.7810747865091003,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.89699579351266e-07,
+      "loss": 1.3691,
+      "step": 10324
+    },
+    {
+      "epoch": 1.7812473044078323,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.88779053675923e-07,
+      "loss": 1.4326,
+      "step": 10325
+    },
+    {
+      "epoch": 1.7814198223065643,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.878592252426296e-07,
+      "loss": 1.4378,
+      "step": 10326
+    },
+    {
+      "epoch": 1.7815923402052962,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.869400941195357e-07,
+      "loss": 1.488,
+      "step": 10327
+    },
+    {
+      "epoch": 1.7817648581040282,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.860216603747349e-07,
+      "loss": 1.3521,
+      "step": 10328
+    },
+    {
+      "epoch": 1.7819373760027601,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.851039240762702e-07,
+      "loss": 1.4167,
+      "step": 10329
+    },
+    {
+      "epoch": 1.7821098939014923,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.84186885292134e-07,
+      "loss": 1.5135,
+      "step": 10330
+    },
+    {
+      "epoch": 1.7822824118002243,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.832705440902675e-07,
+      "loss": 1.4273,
+      "step": 10331
+    },
+    {
+      "epoch": 1.7824549296989562,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.823549005385543e-07,
+      "loss": 1.5056,
+      "step": 10332
+    },
+    {
+      "epoch": 1.7826274475976882,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.814399547048378e-07,
+      "loss": 1.4058,
+      "step": 10333
+    },
+    {
+      "epoch": 1.7827999654964204,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.80525706656897e-07,
+      "loss": 1.4132,
+      "step": 10334
+    },
+    {
+      "epoch": 1.7829724833951524,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.796121564624679e-07,
+      "loss": 1.4934,
+      "step": 10335
+    },
+    {
+      "epoch": 1.7831450012938843,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.786993041892319e-07,
+      "loss": 1.3976,
+      "step": 10336
+    },
+    {
+      "epoch": 1.7833175191926163,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.777871499048182e-07,
+      "loss": 1.5118,
+      "step": 10337
+    },
+    {
+      "epoch": 1.7834900370913482,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.76875693676805e-07,
+      "loss": 1.3587,
+      "step": 10338
+    },
+    {
+      "epoch": 1.7836625549900802,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.759649355727182e-07,
+      "loss": 1.4272,
+      "step": 10339
+    },
+    {
+      "epoch": 1.7838350728888122,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.750548756600338e-07,
+      "loss": 1.4136,
+      "step": 10340
+    },
+    {
+      "epoch": 1.7840075907875441,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.741455140061747e-07,
+      "loss": 1.3687,
+      "step": 10341
+    },
+    {
+      "epoch": 1.784180108686276,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.732368506785113e-07,
+      "loss": 1.471,
+      "step": 10342
+    },
+    {
+      "epoch": 1.784352626585008,
+      "grad_norm": 0.6953125,
+      "learning_rate": 5.72328885744361e-07,
+      "loss": 1.3807,
+      "step": 10343
+    },
+    {
+      "epoch": 1.7845251444837402,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.714216192709976e-07,
+      "loss": 1.4743,
+      "step": 10344
+    },
+    {
+      "epoch": 1.7846976623824722,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.705150513256297e-07,
+      "loss": 1.3449,
+      "step": 10345
+    },
+    {
+      "epoch": 1.7848701802812041,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.696091819754268e-07,
+      "loss": 1.4331,
+      "step": 10346
+    },
+    {
+      "epoch": 1.7850426981799363,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.687040112874986e-07,
+      "loss": 1.3974,
+      "step": 10347
+    },
+    {
+      "epoch": 1.7852152160786683,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.67799539328906e-07,
+      "loss": 1.4973,
+      "step": 10348
+    },
+    {
+      "epoch": 1.7853877339774002,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.668957661666597e-07,
+      "loss": 1.46,
+      "step": 10349
+    },
+    {
+      "epoch": 1.7855602518761322,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.659926918677172e-07,
+      "loss": 1.3491,
+      "step": 10350
+    },
+    {
+      "epoch": 1.7857327697748642,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.650903164989784e-07,
+      "loss": 1.4681,
+      "step": 10351
+    },
+    {
+      "epoch": 1.7859052876735961,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.641886401273056e-07,
+      "loss": 1.4334,
+      "step": 10352
+    },
+    {
+      "epoch": 1.786077805572328,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.632876628194917e-07,
+      "loss": 1.3925,
+      "step": 10353
+    },
+    {
+      "epoch": 1.78625032347106,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.623873846422945e-07,
+      "loss": 1.4241,
+      "step": 10354
+    },
+    {
+      "epoch": 1.786422841369792,
+      "grad_norm": 0.7421875,
+      "learning_rate": 5.614878056624074e-07,
+      "loss": 1.3811,
+      "step": 10355
+    },
+    {
+      "epoch": 1.786595359268524,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.60588925946477e-07,
+      "loss": 1.4115,
+      "step": 10356
+    },
+    {
+      "epoch": 1.7867678771672562,
+      "grad_norm": 0.6953125,
+      "learning_rate": 5.596907455611011e-07,
+      "loss": 1.3852,
+      "step": 10357
+    },
+    {
+      "epoch": 1.7869403950659881,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.5879326457282e-07,
+      "loss": 1.4252,
+      "step": 10358
+    },
+    {
+      "epoch": 1.78711291296472,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.578964830481249e-07,
+      "loss": 1.4649,
+      "step": 10359
+    },
+    {
+      "epoch": 1.787285430863452,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.570004010534557e-07,
+      "loss": 1.436,
+      "step": 10360
+    },
+    {
+      "epoch": 1.7874579487621842,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.561050186551986e-07,
+      "loss": 1.457,
+      "step": 10361
+    },
+    {
+      "epoch": 1.7876304666609162,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.552103359196914e-07,
+      "loss": 1.4263,
+      "step": 10362
+    },
+    {
+      "epoch": 1.7878029845596481,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.543163529132168e-07,
+      "loss": 1.4285,
+      "step": 10363
+    },
+    {
+      "epoch": 1.78797550245838,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.534230697020027e-07,
+      "loss": 1.4432,
+      "step": 10364
+    },
+    {
+      "epoch": 1.788148020357112,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.525304863522363e-07,
+      "loss": 1.394,
+      "step": 10365
+    },
+    {
+      "epoch": 1.788320538255844,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.516386029300391e-07,
+      "loss": 1.4578,
+      "step": 10366
+    },
+    {
+      "epoch": 1.788493056154576,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.507474195014917e-07,
+      "loss": 1.3135,
+      "step": 10367
+    },
+    {
+      "epoch": 1.788665574053308,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.498569361326168e-07,
+      "loss": 1.402,
+      "step": 10368
+    },
+    {
+      "epoch": 1.78883809195204,
+      "grad_norm": 0.71484375,
+      "learning_rate": 5.489671528893869e-07,
+      "loss": 1.4344,
+      "step": 10369
+    },
+    {
+      "epoch": 1.7890106098507719,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.480780698377241e-07,
+      "loss": 1.3027,
+      "step": 10370
+    },
+    {
+      "epoch": 1.789183127749504,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.471896870434957e-07,
+      "loss": 1.4547,
+      "step": 10371
+    },
+    {
+      "epoch": 1.789355645648236,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.463020045725187e-07,
+      "loss": 1.3488,
+      "step": 10372
+    },
+    {
+      "epoch": 1.789528163546968,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.454150224905586e-07,
+      "loss": 1.4106,
+      "step": 10373
+    },
+    {
+      "epoch": 1.7897006814457,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.445287408633304e-07,
+      "loss": 1.3909,
+      "step": 10374
+    },
+    {
+      "epoch": 1.7898731993444321,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.436431597564907e-07,
+      "loss": 1.5508,
+      "step": 10375
+    },
+    {
+      "epoch": 1.790045717243164,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.427582792356545e-07,
+      "loss": 1.4281,
+      "step": 10376
+    },
+    {
+      "epoch": 1.790218235141896,
+      "grad_norm": 0.66796875,
+      "learning_rate": 5.418740993663751e-07,
+      "loss": 1.4708,
+      "step": 10377
+    },
+    {
+      "epoch": 1.790390753040628,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.409906202141602e-07,
+      "loss": 1.4345,
+      "step": 10378
+    },
+    {
+      "epoch": 1.79056327093936,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.401078418444617e-07,
+      "loss": 1.4302,
+      "step": 10379
+    },
+    {
+      "epoch": 1.790735788838092,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.392257643226828e-07,
+      "loss": 1.4859,
+      "step": 10380
+    },
+    {
+      "epoch": 1.7909083067368239,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.383443877141737e-07,
+      "loss": 1.4114,
+      "step": 10381
+    },
+    {
+      "epoch": 1.7910808246355558,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.374637120842308e-07,
+      "loss": 1.476,
+      "step": 10382
+    },
+    {
+      "epoch": 1.7912533425342878,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.365837374981009e-07,
+      "loss": 1.531,
+      "step": 10383
+    },
+    {
+      "epoch": 1.7914258604330198,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.357044640209796e-07,
+      "loss": 1.4339,
+      "step": 10384
+    },
+    {
+      "epoch": 1.791598378331752,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.348258917180038e-07,
+      "loss": 1.3312,
+      "step": 10385
+    },
+    {
+      "epoch": 1.791770896230484,
+      "grad_norm": 0.83203125,
+      "learning_rate": 5.339480206542702e-07,
+      "loss": 1.4726,
+      "step": 10386
+    },
+    {
+      "epoch": 1.7919434141292159,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.330708508948123e-07,
+      "loss": 1.3928,
+      "step": 10387
+    },
+    {
+      "epoch": 1.792115932027948,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.321943825046171e-07,
+      "loss": 1.4194,
+      "step": 10388
+    },
+    {
+      "epoch": 1.79228844992668,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.313186155486216e-07,
+      "loss": 1.4375,
+      "step": 10389
+    },
+    {
+      "epoch": 1.792460967825412,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.304435500917049e-07,
+      "loss": 1.4913,
+      "step": 10390
+    },
+    {
+      "epoch": 1.792633485724144,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.295691861986985e-07,
+      "loss": 1.3996,
+      "step": 10391
+    },
+    {
+      "epoch": 1.792806003622876,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.286955239343816e-07,
+      "loss": 1.4013,
+      "step": 10392
+    },
+    {
+      "epoch": 1.7929785215216079,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.278225633634793e-07,
+      "loss": 1.3898,
+      "step": 10393
+    },
+    {
+      "epoch": 1.7931510394203398,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.269503045506652e-07,
+      "loss": 1.4631,
+      "step": 10394
+    },
+    {
+      "epoch": 1.7933235573190718,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.260787475605656e-07,
+      "loss": 1.4925,
+      "step": 10395
+    },
+    {
+      "epoch": 1.7934960752178037,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.252078924577453e-07,
+      "loss": 1.5018,
+      "step": 10396
+    },
+    {
+      "epoch": 1.7936685931165357,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.243377393067284e-07,
+      "loss": 1.3756,
+      "step": 10397
+    },
+    {
+      "epoch": 1.7938411110152679,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.234682881719766e-07,
+      "loss": 1.4001,
+      "step": 10398
+    },
+    {
+      "epoch": 1.7940136289139998,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.225995391179061e-07,
+      "loss": 1.419,
+      "step": 10399
+    },
+    {
+      "epoch": 1.7941861468127318,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.21731492208879e-07,
+      "loss": 1.3098,
+      "step": 10400
+    },
+    {
+      "epoch": 1.7941861468127318,
+      "eval_loss": 1.4070560932159424,
+      "eval_runtime": 10.8247,
+      "eval_samples_per_second": 94.598,
+      "eval_steps_per_second": 23.65,
+      "step": 10400
+    },
+    {
+      "epoch": 1.7943586647114638,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.208641475092069e-07,
+      "loss": 1.3357,
+      "step": 10401
+    },
+    {
+      "epoch": 1.794531182610196,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.199975050831463e-07,
+      "loss": 1.5601,
+      "step": 10402
+    },
+    {
+      "epoch": 1.794703700508928,
+      "grad_norm": 0.80859375,
+      "learning_rate": 5.191315649949046e-07,
+      "loss": 1.363,
+      "step": 10403
+    },
+    {
+      "epoch": 1.7948762184076599,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.182663273086364e-07,
+      "loss": 1.38,
+      "step": 10404
+    },
+    {
+      "epoch": 1.7950487363063918,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.174017920884423e-07,
+      "loss": 1.3737,
+      "step": 10405
+    },
+    {
+      "epoch": 1.7952212542051238,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.165379593983755e-07,
+      "loss": 1.5047,
+      "step": 10406
+    },
+    {
+      "epoch": 1.7953937721038558,
+      "grad_norm": 0.73046875,
+      "learning_rate": 5.156748293024283e-07,
+      "loss": 1.4924,
+      "step": 10407
+    },
+    {
+      "epoch": 1.7955662900025877,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.148124018645539e-07,
+      "loss": 1.3867,
+      "step": 10408
+    },
+    {
+      "epoch": 1.7957388079013197,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.139506771486414e-07,
+      "loss": 1.3976,
+      "step": 10409
+    },
+    {
+      "epoch": 1.7959113258000516,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.130896552185349e-07,
+      "loss": 1.4595,
+      "step": 10410
+    },
+    {
+      "epoch": 1.7960838436987836,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.122293361380238e-07,
+      "loss": 1.4081,
+      "step": 10411
+    },
+    {
+      "epoch": 1.7962563615975158,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.113697199708456e-07,
+      "loss": 1.4327,
+      "step": 10412
+    },
+    {
+      "epoch": 1.7964288794962477,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.105108067806863e-07,
+      "loss": 1.3137,
+      "step": 10413
+    },
+    {
+      "epoch": 1.7966013973949797,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.096525966311794e-07,
+      "loss": 1.4744,
+      "step": 10414
+    },
+    {
+      "epoch": 1.7967739152937119,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.087950895859062e-07,
+      "loss": 1.5153,
+      "step": 10415
+    },
+    {
+      "epoch": 1.7969464331924438,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.079382857083981e-07,
+      "loss": 1.4555,
+      "step": 10416
+    },
+    {
+      "epoch": 1.7971189510911758,
+      "grad_norm": 0.5546875,
+      "learning_rate": 5.070821850621277e-07,
+      "loss": 1.3925,
+      "step": 10417
+    },
+    {
+      "epoch": 1.7972914689899078,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.062267877105275e-07,
+      "loss": 1.4293,
+      "step": 10418
+    },
+    {
+      "epoch": 1.7974639868886397,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.053720937169648e-07,
+      "loss": 1.3991,
+      "step": 10419
+    },
+    {
+      "epoch": 1.7976365047873717,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.04518103144761e-07,
+      "loss": 1.4303,
+      "step": 10420
+    },
+    {
+      "epoch": 1.7978090226861037,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.036648160571889e-07,
+      "loss": 1.4,
+      "step": 10421
+    },
+    {
+      "epoch": 1.7979815405848356,
+      "grad_norm": 0.640625,
+      "learning_rate": 5.028122325174623e-07,
+      "loss": 1.4848,
+      "step": 10422
+    },
+    {
+      "epoch": 1.7981540584835676,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.019603525887462e-07,
+      "loss": 1.4521,
+      "step": 10423
+    },
+    {
+      "epoch": 1.7983265763822995,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.011091763341547e-07,
+      "loss": 1.4574,
+      "step": 10424
+    },
+    {
+      "epoch": 1.7984990942810315,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.002587038167461e-07,
+      "loss": 1.3833,
+      "step": 10425
+    },
+    {
+      "epoch": 1.7986716121797637,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.994089350995301e-07,
+      "loss": 1.5584,
+      "step": 10426
+    },
+    {
+      "epoch": 1.7988441300784956,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.985598702454653e-07,
+      "loss": 1.3979,
+      "step": 10427
+    },
+    {
+      "epoch": 1.7990166479772276,
+      "grad_norm": 0.625,
+      "learning_rate": 4.977115093174489e-07,
+      "loss": 1.5396,
+      "step": 10428
+    },
+    {
+      "epoch": 1.7991891658759598,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.968638523783398e-07,
+      "loss": 1.5011,
+      "step": 10429
+    },
+    {
+      "epoch": 1.7993616837746917,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.960168994909343e-07,
+      "loss": 1.3286,
+      "step": 10430
+    },
+    {
+      "epoch": 1.7995342016734237,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.951706507179788e-07,
+      "loss": 1.3573,
+      "step": 10431
+    },
+    {
+      "epoch": 1.7997067195721557,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.943251061221721e-07,
+      "loss": 1.4431,
+      "step": 10432
+    },
+    {
+      "epoch": 1.7998792374708876,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.934802657661553e-07,
+      "loss": 1.3382,
+      "step": 10433
+    },
+    {
+      "epoch": 1.8000517553696196,
+      "grad_norm": 0.6640625,
+      "learning_rate": 4.926361297125193e-07,
+      "loss": 1.516,
+      "step": 10434
+    },
+    {
+      "epoch": 1.8002242732683515,
+      "grad_norm": 0.55859375,
+      "learning_rate": 4.917926980238041e-07,
+      "loss": 1.3845,
+      "step": 10435
+    },
+    {
+      "epoch": 1.8003967911670835,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.909499707624966e-07,
+      "loss": 1.3371,
+      "step": 10436
+    },
+    {
+      "epoch": 1.8005693090658155,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.901079479910287e-07,
+      "loss": 1.462,
+      "step": 10437
+    },
+    {
+      "epoch": 1.8007418269645474,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.892666297717874e-07,
+      "loss": 1.4135,
+      "step": 10438
+    },
+    {
+      "epoch": 1.8009143448632796,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.884260161670972e-07,
+      "loss": 1.4871,
+      "step": 10439
+    },
+    {
+      "epoch": 1.8010868627620116,
+      "grad_norm": 0.6328125,
+      "learning_rate": 4.875861072392408e-07,
+      "loss": 1.3797,
+      "step": 10440
+    },
+    {
+      "epoch": 1.8012593806607435,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.86746903050439e-07,
+      "loss": 1.4785,
+      "step": 10441
+    },
+    {
+      "epoch": 1.8014318985594755,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.859084036628714e-07,
+      "loss": 1.5303,
+      "step": 10442
+    },
+    {
+      "epoch": 1.8016044164582077,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.850706091386548e-07,
+      "loss": 1.4003,
+      "step": 10443
+    },
+    {
+      "epoch": 1.8017769343569396,
+      "grad_norm": 1.3046875,
+      "learning_rate": 4.842335195398595e-07,
+      "loss": 1.3405,
+      "step": 10444
+    },
+    {
+      "epoch": 1.8019494522556716,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.833971349285027e-07,
+      "loss": 1.3997,
+      "step": 10445
+    },
+    {
+      "epoch": 1.8021219701544036,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.825614553665481e-07,
+      "loss": 1.4876,
+      "step": 10446
+    },
+    {
+      "epoch": 1.8022944880531355,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.817264809159084e-07,
+      "loss": 1.41,
+      "step": 10447
+    },
+    {
+      "epoch": 1.8024670059518675,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.808922116384451e-07,
+      "loss": 1.4368,
+      "step": 10448
+    },
+    {
+      "epoch": 1.8026395238505994,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.800586475959623e-07,
+      "loss": 1.4177,
+      "step": 10449
+    },
+    {
+      "epoch": 1.8028120417493314,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.792257888502217e-07,
+      "loss": 1.4365,
+      "step": 10450
+    },
+    {
+      "epoch": 1.8029845596480634,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.783936354629215e-07,
+      "loss": 1.3633,
+      "step": 10451
+    },
+    {
+      "epoch": 1.8031570775467953,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.775621874957126e-07,
+      "loss": 1.4407,
+      "step": 10452
+    },
+    {
+      "epoch": 1.8033295954455275,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.7673144501019897e-07,
+      "loss": 1.3938,
+      "step": 10453
+    },
+    {
+      "epoch": 1.8035021133442595,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.759014080679225e-07,
+      "loss": 1.4431,
+      "step": 10454
+    },
+    {
+      "epoch": 1.8036746312429914,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.7507207673037956e-07,
+      "loss": 1.4637,
+      "step": 10455
+    },
+    {
+      "epoch": 1.8038471491417236,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.7424345105901105e-07,
+      "loss": 1.459,
+      "step": 10456
+    },
+    {
+      "epoch": 1.8040196670404556,
+      "grad_norm": 0.625,
+      "learning_rate": 4.7341553111520776e-07,
+      "loss": 1.4838,
+      "step": 10457
+    },
+    {
+      "epoch": 1.8041921849391875,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.7258831696030624e-07,
+      "loss": 1.3738,
+      "step": 10458
+    },
+    {
+      "epoch": 1.8043647028379195,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.7176180865559416e-07,
+      "loss": 1.393,
+      "step": 10459
+    },
+    {
+      "epoch": 1.8045372207366515,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.709360062622992e-07,
+      "loss": 1.4474,
+      "step": 10460
+    },
+    {
+      "epoch": 1.8047097386353834,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.701109098416079e-07,
+      "loss": 1.5006,
+      "step": 10461
+    },
+    {
+      "epoch": 1.8048822565341154,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.6928651945464585e-07,
+      "loss": 1.4166,
+      "step": 10462
+    },
+    {
+      "epoch": 1.8050547744328473,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.6846283516248756e-07,
+      "loss": 1.5211,
+      "step": 10463
+    },
+    {
+      "epoch": 1.8052272923315793,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.6763985702615864e-07,
+      "loss": 1.4285,
+      "step": 10464
+    },
+    {
+      "epoch": 1.8053998102303113,
+      "grad_norm": 0.546875,
+      "learning_rate": 4.6681758510663035e-07,
+      "loss": 1.4389,
+      "step": 10465
+    },
+    {
+      "epoch": 1.8055723281290432,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.6599601946482164e-07,
+      "loss": 1.4591,
+      "step": 10466
+    },
+    {
+      "epoch": 1.8057448460277754,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.651751601615984e-07,
+      "loss": 1.4747,
+      "step": 10467
+    },
+    {
+      "epoch": 1.8059173639265074,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.6435500725777627e-07,
+      "loss": 1.3947,
+      "step": 10468
+    },
+    {
+      "epoch": 1.8060898818252393,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.635355608141168e-07,
+      "loss": 1.4092,
+      "step": 10469
+    },
+    {
+      "epoch": 1.8062623997239715,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.6271682089132796e-07,
+      "loss": 1.3991,
+      "step": 10470
+    },
+    {
+      "epoch": 1.8064349176227035,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.618987875500702e-07,
+      "loss": 1.4512,
+      "step": 10471
+    },
+    {
+      "epoch": 1.8066074355214354,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.610814608509484e-07,
+      "loss": 1.3993,
+      "step": 10472
+    },
+    {
+      "epoch": 1.8067799534201674,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.602648408545107e-07,
+      "loss": 1.4226,
+      "step": 10473
+    },
+    {
+      "epoch": 1.8069524713188994,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.594489276212633e-07,
+      "loss": 1.3826,
+      "step": 10474
+    },
+    {
+      "epoch": 1.8071249892176313,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.5863372121165095e-07,
+      "loss": 1.4057,
+      "step": 10475
+    },
+    {
+      "epoch": 1.8072975071163633,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.5781922168606883e-07,
+      "loss": 1.5323,
+      "step": 10476
+    },
+    {
+      "epoch": 1.8074700250150952,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.570054291048609e-07,
+      "loss": 1.3534,
+      "step": 10477
+    },
+    {
+      "epoch": 1.8076425429138272,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.561923435283189e-07,
+      "loss": 1.4982,
+      "step": 10478
+    },
+    {
+      "epoch": 1.8078150608125592,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.553799650166801e-07,
+      "loss": 1.376,
+      "step": 10479
+    },
+    {
+      "epoch": 1.8079875787112913,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.545682936301321e-07,
+      "loss": 1.4337,
+      "step": 10480
+    },
+    {
+      "epoch": 1.8081600966100233,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.5375732942880557e-07,
+      "loss": 1.4383,
+      "step": 10481
+    },
+    {
+      "epoch": 1.8083326145087553,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.529470724727858e-07,
+      "loss": 1.4463,
+      "step": 10482
+    },
+    {
+      "epoch": 1.8085051324074872,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.5213752282209924e-07,
+      "loss": 1.4528,
+      "step": 10483
+    },
+    {
+      "epoch": 1.8086776503062194,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.513286805367212e-07,
+      "loss": 1.4562,
+      "step": 10484
+    },
+    {
+      "epoch": 1.8088501682049514,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.505205456765793e-07,
+      "loss": 1.384,
+      "step": 10485
+    },
+    {
+      "epoch": 1.8090226861036833,
+      "grad_norm": 0.625,
+      "learning_rate": 4.497131183015424e-07,
+      "loss": 1.3599,
+      "step": 10486
+    },
+    {
+      "epoch": 1.8091952040024153,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.4890639847143035e-07,
+      "loss": 1.4611,
+      "step": 10487
+    },
+    {
+      "epoch": 1.8093677219011473,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.4810038624601095e-07,
+      "loss": 1.4362,
+      "step": 10488
+    },
+    {
+      "epoch": 1.8095402397998792,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.472950816849975e-07,
+      "loss": 1.4141,
+      "step": 10489
+    },
+    {
+      "epoch": 1.8097127576986112,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.464904848480522e-07,
+      "loss": 1.4283,
+      "step": 10490
+    },
+    {
+      "epoch": 1.8098852755973431,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.4568659579478647e-07,
+      "loss": 1.4313,
+      "step": 10491
+    },
+    {
+      "epoch": 1.810057793496075,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.4488341458475247e-07,
+      "loss": 1.4505,
+      "step": 10492
+    },
+    {
+      "epoch": 1.810230311394807,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.440809412774616e-07,
+      "loss": 1.3547,
+      "step": 10493
+    },
+    {
+      "epoch": 1.8104028292935392,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.432791759323618e-07,
+      "loss": 1.5555,
+      "step": 10494
+    },
+    {
+      "epoch": 1.8105753471922712,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.4247811860885335e-07,
+      "loss": 1.394,
+      "step": 10495
+    },
+    {
+      "epoch": 1.8107478650910032,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.416777693662844e-07,
+      "loss": 1.3825,
+      "step": 10496
+    },
+    {
+      "epoch": 1.8109203829897353,
+      "grad_norm": 0.625,
+      "learning_rate": 4.408781282639485e-07,
+      "loss": 1.586,
+      "step": 10497
+    },
+    {
+      "epoch": 1.8110929008884673,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.4007919536109057e-07,
+      "loss": 1.392,
+      "step": 10498
+    },
+    {
+      "epoch": 1.8112654187871993,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.392809707168977e-07,
+      "loss": 1.4427,
+      "step": 10499
+    },
+    {
+      "epoch": 1.8114379366859312,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.384834543905092e-07,
+      "loss": 1.4485,
+      "step": 10500
+    },
+    {
+      "epoch": 1.8114379366859312,
+      "eval_loss": 1.4070926904678345,
+      "eval_runtime": 11.9829,
+      "eval_samples_per_second": 85.455,
+      "eval_steps_per_second": 21.364,
+      "step": 10500
+    },
+    {
+      "epoch": 1.8116104545846632,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.37686646441009e-07,
+      "loss": 1.3746,
+      "step": 10501
+    },
+    {
+      "epoch": 1.8117829724833951,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.3689054692743094e-07,
+      "loss": 1.4008,
+      "step": 10502
+    },
+    {
+      "epoch": 1.811955490382127,
+      "grad_norm": 0.625,
+      "learning_rate": 4.360951559087534e-07,
+      "loss": 1.4617,
+      "step": 10503
+    },
+    {
+      "epoch": 1.812128008280859,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.353004734439059e-07,
+      "loss": 1.3902,
+      "step": 10504
+    },
+    {
+      "epoch": 1.812300526179591,
+      "grad_norm": 0.625,
+      "learning_rate": 4.345064995917603e-07,
+      "loss": 1.3388,
+      "step": 10505
+    },
+    {
+      "epoch": 1.812473044078323,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.337132344111439e-07,
+      "loss": 1.4352,
+      "step": 10506
+    },
+    {
+      "epoch": 1.812645561977055,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.3292067796082304e-07,
+      "loss": 1.4476,
+      "step": 10507
+    },
+    {
+      "epoch": 1.8128180798757871,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.3212883029951523e-07,
+      "loss": 1.3818,
+      "step": 10508
+    },
+    {
+      "epoch": 1.812990597774519,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.313376914858869e-07,
+      "loss": 1.4204,
+      "step": 10509
+    },
+    {
+      "epoch": 1.813163115673251,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.3054726157855e-07,
+      "loss": 1.2998,
+      "step": 10510
+    },
+    {
+      "epoch": 1.8133356335719832,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.2975754063606544e-07,
+      "loss": 1.376,
+      "step": 10511
+    },
+    {
+      "epoch": 1.8135081514707152,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.289685287169398e-07,
+      "loss": 1.3711,
+      "step": 10512
+    },
+    {
+      "epoch": 1.8136806693694472,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.281802258796275e-07,
+      "loss": 1.429,
+      "step": 10513
+    },
+    {
+      "epoch": 1.8138531872681791,
+      "grad_norm": 1.8125,
+      "learning_rate": 4.273926321825328e-07,
+      "loss": 1.4026,
+      "step": 10514
+    },
+    {
+      "epoch": 1.814025705166911,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.266057476840024e-07,
+      "loss": 1.5232,
+      "step": 10515
+    },
+    {
+      "epoch": 1.814198223065643,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.258195724423353e-07,
+      "loss": 1.4122,
+      "step": 10516
+    },
+    {
+      "epoch": 1.814370740964375,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.250341065157793e-07,
+      "loss": 1.4027,
+      "step": 10517
+    },
+    {
+      "epoch": 1.814543258863107,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.242493499625222e-07,
+      "loss": 1.3615,
+      "step": 10518
+    },
+    {
+      "epoch": 1.814715776761839,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.234653028407054e-07,
+      "loss": 1.4586,
+      "step": 10519
+    },
+    {
+      "epoch": 1.8148882946605709,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.2268196520841574e-07,
+      "loss": 1.4106,
+      "step": 10520
+    },
+    {
+      "epoch": 1.815060812559303,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.218993371236879e-07,
+      "loss": 1.4377,
+      "step": 10521
+    },
+    {
+      "epoch": 1.815233330458035,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.211174186445033e-07,
+      "loss": 1.4085,
+      "step": 10522
+    },
+    {
+      "epoch": 1.815405848356767,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.203362098287944e-07,
+      "loss": 1.3718,
+      "step": 10523
+    },
+    {
+      "epoch": 1.815578366255499,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.195557107344328e-07,
+      "loss": 1.4049,
+      "step": 10524
+    },
+    {
+      "epoch": 1.8157508841542311,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.187759214192477e-07,
+      "loss": 1.5663,
+      "step": 10525
+    },
+    {
+      "epoch": 1.815923402052963,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.179968419410063e-07,
+      "loss": 1.41,
+      "step": 10526
+    },
+    {
+      "epoch": 1.816095919951695,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.172184723574324e-07,
+      "loss": 1.3764,
+      "step": 10527
+    },
+    {
+      "epoch": 1.816268437850427,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.1644081272618874e-07,
+      "loss": 1.406,
+      "step": 10528
+    },
+    {
+      "epoch": 1.816440955749159,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.1566386310489035e-07,
+      "loss": 1.4075,
+      "step": 10529
+    },
+    {
+      "epoch": 1.816613473647891,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.14887623551099e-07,
+      "loss": 1.3793,
+      "step": 10530
+    },
+    {
+      "epoch": 1.816785991546623,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.141120941223231e-07,
+      "loss": 1.441,
+      "step": 10531
+    },
+    {
+      "epoch": 1.8169585094453549,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.133372748760178e-07,
+      "loss": 1.4461,
+      "step": 10532
+    },
+    {
+      "epoch": 1.8171310273440868,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.1256316586958835e-07,
+      "loss": 1.3624,
+      "step": 10533
+    },
+    {
+      "epoch": 1.8173035452428188,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.117897671603843e-07,
+      "loss": 1.3699,
+      "step": 10534
+    },
+    {
+      "epoch": 1.817476063141551,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.110170788057044e-07,
+      "loss": 1.3984,
+      "step": 10535
+    },
+    {
+      "epoch": 1.817648581040283,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.10245100862795e-07,
+      "loss": 1.4331,
+      "step": 10536
+    },
+    {
+      "epoch": 1.817821098939015,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.0947383338884594e-07,
+      "loss": 1.4868,
+      "step": 10537
+    },
+    {
+      "epoch": 1.817993616837747,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.087032764410015e-07,
+      "loss": 1.423,
+      "step": 10538
+    },
+    {
+      "epoch": 1.818166134736479,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.0793343007634713e-07,
+      "loss": 1.4135,
+      "step": 10539
+    },
+    {
+      "epoch": 1.818338652635211,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.071642943519183e-07,
+      "loss": 1.404,
+      "step": 10540
+    },
+    {
+      "epoch": 1.818511170533943,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.063958693246961e-07,
+      "loss": 1.4487,
+      "step": 10541
+    },
+    {
+      "epoch": 1.818683688432675,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.056281550516128e-07,
+      "loss": 1.4732,
+      "step": 10542
+    },
+    {
+      "epoch": 1.8188562063314069,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.0486115158954396e-07,
+      "loss": 1.4363,
+      "step": 10543
+    },
+    {
+      "epoch": 1.8190287242301388,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.040948589953153e-07,
+      "loss": 1.4726,
+      "step": 10544
+    },
+    {
+      "epoch": 1.8192012421288708,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.0332927732569473e-07,
+      "loss": 1.4246,
+      "step": 10545
+    },
+    {
+      "epoch": 1.8193737600276028,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.025644066374068e-07,
+      "loss": 1.2917,
+      "step": 10546
+    },
+    {
+      "epoch": 1.8195462779263347,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.0180024698711404e-07,
+      "loss": 1.3544,
+      "step": 10547
+    },
+    {
+      "epoch": 1.819718795825067,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.0103679843142895e-07,
+      "loss": 1.4888,
+      "step": 10548
+    },
+    {
+      "epoch": 1.8198913137237989,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.002740610269185e-07,
+      "loss": 1.3407,
+      "step": 10549
+    },
+    {
+      "epoch": 1.8200638316225308,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.995120348300863e-07,
+      "loss": 1.3905,
+      "step": 10550
+    },
+    {
+      "epoch": 1.8202363495212628,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.9875071989738943e-07,
+      "loss": 1.432,
+      "step": 10551
+    },
+    {
+      "epoch": 1.820408867419995,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.9799011628522953e-07,
+      "loss": 1.4171,
+      "step": 10552
+    },
+    {
+      "epoch": 1.820581385318727,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.9723022404995926e-07,
+      "loss": 1.5395,
+      "step": 10553
+    },
+    {
+      "epoch": 1.820753903217459,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.964710432478736e-07,
+      "loss": 1.436,
+      "step": 10554
+    },
+    {
+      "epoch": 1.8209264211161909,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.957125739352208e-07,
+      "loss": 1.5376,
+      "step": 10555
+    },
+    {
+      "epoch": 1.8210989390149228,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.949548161681882e-07,
+      "loss": 1.4383,
+      "step": 10556
+    },
+    {
+      "epoch": 1.8212714569136548,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.941977700029198e-07,
+      "loss": 1.5523,
+      "step": 10557
+    },
+    {
+      "epoch": 1.8214439748123867,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.934414354954985e-07,
+      "loss": 1.4397,
+      "step": 10558
+    },
+    {
+      "epoch": 1.8216164927111187,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.9268581270196284e-07,
+      "loss": 1.4584,
+      "step": 10559
+    },
+    {
+      "epoch": 1.8217890106098507,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.919309016782902e-07,
+      "loss": 1.3746,
+      "step": 10560
+    },
+    {
+      "epoch": 1.8219615285085826,
+      "grad_norm": 0.625,
+      "learning_rate": 3.911767024804092e-07,
+      "loss": 1.3883,
+      "step": 10561
+    },
+    {
+      "epoch": 1.8221340464073148,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.9042321516419844e-07,
+      "loss": 1.4587,
+      "step": 10562
+    },
+    {
+      "epoch": 1.8223065643060468,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.896704397854778e-07,
+      "loss": 1.2624,
+      "step": 10563
+    },
+    {
+      "epoch": 1.8224790822047787,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.889183764000204e-07,
+      "loss": 1.3084,
+      "step": 10564
+    },
+    {
+      "epoch": 1.822651600103511,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.8816702506354163e-07,
+      "loss": 1.3891,
+      "step": 10565
+    },
+    {
+      "epoch": 1.8228241180022429,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.8741638583170814e-07,
+      "loss": 1.4988,
+      "step": 10566
+    },
+    {
+      "epoch": 1.8229966359009748,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.866664587601299e-07,
+      "loss": 1.4436,
+      "step": 10567
+    },
+    {
+      "epoch": 1.8231691537997068,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.8591724390436904e-07,
+      "loss": 1.4547,
+      "step": 10568
+    },
+    {
+      "epoch": 1.8233416716984387,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.851687413199279e-07,
+      "loss": 1.3426,
+      "step": 10569
+    },
+    {
+      "epoch": 1.8235141895971707,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.844209510622643e-07,
+      "loss": 1.4252,
+      "step": 10570
+    },
+    {
+      "epoch": 1.8236867074959027,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.836738731867773e-07,
+      "loss": 1.4587,
+      "step": 10571
+    },
+    {
+      "epoch": 1.8238592253946346,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.8292750774881483e-07,
+      "loss": 1.5709,
+      "step": 10572
+    },
+    {
+      "epoch": 1.8240317432933666,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.8218185480367264e-07,
+      "loss": 1.3629,
+      "step": 10573
+    },
+    {
+      "epoch": 1.8242042611920986,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.814369144065944e-07,
+      "loss": 1.3875,
+      "step": 10574
+    },
+    {
+      "epoch": 1.8243767790908305,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.8069268661276916e-07,
+      "loss": 1.452,
+      "step": 10575
+    },
+    {
+      "epoch": 1.8245492969895627,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.79949171477334e-07,
+      "loss": 1.4527,
+      "step": 10576
+    },
+    {
+      "epoch": 1.8247218148882947,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.792063690553716e-07,
+      "loss": 1.4349,
+      "step": 10577
+    },
+    {
+      "epoch": 1.8248943327870266,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.7846427940191663e-07,
+      "loss": 1.4183,
+      "step": 10578
+    },
+    {
+      "epoch": 1.8250668506857588,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.777229025719453e-07,
+      "loss": 1.467,
+      "step": 10579
+    },
+    {
+      "epoch": 1.8252393685844908,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.7698223862038254e-07,
+      "loss": 1.3971,
+      "step": 10580
+    },
+    {
+      "epoch": 1.8254118864832227,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.7624228760210545e-07,
+      "loss": 1.3448,
+      "step": 10581
+    },
+    {
+      "epoch": 1.8255844043819547,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.755030495719303e-07,
+      "loss": 1.3441,
+      "step": 10582
+    },
+    {
+      "epoch": 1.8257569222806866,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.7476452458462654e-07,
+      "loss": 1.328,
+      "step": 10583
+    },
+    {
+      "epoch": 1.8259294401794186,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.740267126949071e-07,
+      "loss": 1.3725,
+      "step": 10584
+    },
+    {
+      "epoch": 1.8261019580781506,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.732896139574349e-07,
+      "loss": 1.4206,
+      "step": 10585
+    },
+    {
+      "epoch": 1.8262744759768825,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.7255322842681963e-07,
+      "loss": 1.5134,
+      "step": 10586
+    },
+    {
+      "epoch": 1.8264469938756145,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.718175561576154e-07,
+      "loss": 1.4835,
+      "step": 10587
+    },
+    {
+      "epoch": 1.8266195117743464,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.710825972043264e-07,
+      "loss": 1.4358,
+      "step": 10588
+    },
+    {
+      "epoch": 1.8267920296730786,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.7034835162140347e-07,
+      "loss": 1.5118,
+      "step": 10589
+    },
+    {
+      "epoch": 1.8269645475718106,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.6961481946324205e-07,
+      "loss": 1.5324,
+      "step": 10590
+    },
+    {
+      "epoch": 1.8271370654705426,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.688820007841898e-07,
+      "loss": 1.4644,
+      "step": 10591
+    },
+    {
+      "epoch": 1.8273095833692745,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.6814989563853654e-07,
+      "loss": 1.4278,
+      "step": 10592
+    },
+    {
+      "epoch": 1.8274821012680067,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.6741850408052114e-07,
+      "loss": 1.4077,
+      "step": 10593
+    },
+    {
+      "epoch": 1.8276546191667387,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.666878261643303e-07,
+      "loss": 1.4508,
+      "step": 10594
+    },
+    {
+      "epoch": 1.8278271370654706,
+      "grad_norm": 0.625,
+      "learning_rate": 3.659578619440962e-07,
+      "loss": 1.4561,
+      "step": 10595
+    },
+    {
+      "epoch": 1.8279996549642026,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.652286114739012e-07,
+      "loss": 1.376,
+      "step": 10596
+    },
+    {
+      "epoch": 1.8281721728629345,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.645000748077709e-07,
+      "loss": 1.4252,
+      "step": 10597
+    },
+    {
+      "epoch": 1.8283446907616665,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.6377225199968003e-07,
+      "loss": 1.4235,
+      "step": 10598
+    },
+    {
+      "epoch": 1.8285172086603985,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.6304514310355086e-07,
+      "loss": 1.4719,
+      "step": 10599
+    },
+    {
+      "epoch": 1.8286897265591304,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.6231874817325375e-07,
+      "loss": 1.4056,
+      "step": 10600
+    },
+    {
+      "epoch": 1.8286897265591304,
+      "eval_loss": 1.4070792198181152,
+      "eval_runtime": 10.8107,
+      "eval_samples_per_second": 94.721,
+      "eval_steps_per_second": 23.68,
+      "step": 10600
+    },
+    {
+      "epoch": 1.8288622444578624,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.615930672626e-07,
+      "loss": 1.4744,
+      "step": 10601
+    },
+    {
+      "epoch": 1.8290347623565943,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.608681004253578e-07,
+      "loss": 1.3409,
+      "step": 10602
+    },
+    {
+      "epoch": 1.8292072802553265,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.60143847715233e-07,
+      "loss": 1.4295,
+      "step": 10603
+    },
+    {
+      "epoch": 1.8293797981540585,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.594203091858861e-07,
+      "loss": 1.4565,
+      "step": 10604
+    },
+    {
+      "epoch": 1.8295523160527905,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.586974848909186e-07,
+      "loss": 1.3717,
+      "step": 10605
+    },
+    {
+      "epoch": 1.8297248339515226,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.5797537488388326e-07,
+      "loss": 1.5116,
+      "step": 10606
+    },
+    {
+      "epoch": 1.8298973518502546,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.572539792182783e-07,
+      "loss": 1.4081,
+      "step": 10607
+    },
+    {
+      "epoch": 1.8300698697489866,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.5653329794755e-07,
+      "loss": 1.4567,
+      "step": 10608
+    },
+    {
+      "epoch": 1.8302423876477185,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.558133311250889e-07,
+      "loss": 1.3465,
+      "step": 10609
+    },
+    {
+      "epoch": 1.8304149055464505,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.5509407880423783e-07,
+      "loss": 1.4702,
+      "step": 10610
+    },
+    {
+      "epoch": 1.8305874234451824,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.5437554103827985e-07,
+      "loss": 1.4834,
+      "step": 10611
+    },
+    {
+      "epoch": 1.8307599413439144,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.53657717880449e-07,
+      "loss": 1.4094,
+      "step": 10612
+    },
+    {
+      "epoch": 1.8309324592426464,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.5294060938393046e-07,
+      "loss": 1.4454,
+      "step": 10613
+    },
+    {
+      "epoch": 1.8311049771413783,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.522242156018474e-07,
+      "loss": 1.48,
+      "step": 10614
+    },
+    {
+      "epoch": 1.8312774950401103,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.5150853658727837e-07,
+      "loss": 1.3688,
+      "step": 10615
+    },
+    {
+      "epoch": 1.8314500129388422,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.5079357239324205e-07,
+      "loss": 1.3982,
+      "step": 10616
+    },
+    {
+      "epoch": 1.8316225308375744,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.500793230727095e-07,
+      "loss": 1.4082,
+      "step": 10617
+    },
+    {
+      "epoch": 1.8317950487363064,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.493657886785962e-07,
+      "loss": 1.3345,
+      "step": 10618
+    },
+    {
+      "epoch": 1.8319675666350383,
+      "grad_norm": 0.54296875,
+      "learning_rate": 3.486529692637641e-07,
+      "loss": 1.3584,
+      "step": 10619
+    },
+    {
+      "epoch": 1.8321400845337705,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.4794086488102564e-07,
+      "loss": 1.3406,
+      "step": 10620
+    },
+    {
+      "epoch": 1.8323126024325025,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.4722947558313737e-07,
+      "loss": 1.4438,
+      "step": 10621
+    },
+    {
+      "epoch": 1.8324851203312345,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.4651880142280047e-07,
+      "loss": 1.3952,
+      "step": 10622
+    },
+    {
+      "epoch": 1.8326576382299664,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.4580884245267064e-07,
+      "loss": 1.3912,
+      "step": 10623
+    },
+    {
+      "epoch": 1.8328301561286984,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.450995987253436e-07,
+      "loss": 1.4326,
+      "step": 10624
+    },
+    {
+      "epoch": 1.8330026740274303,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.443910702933639e-07,
+      "loss": 1.4591,
+      "step": 10625
+    },
+    {
+      "epoch": 1.8331751919261623,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.436832572092264e-07,
+      "loss": 1.357,
+      "step": 10626
+    },
+    {
+      "epoch": 1.8333477098248943,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.429761595253667e-07,
+      "loss": 1.3664,
+      "step": 10627
+    },
+    {
+      "epoch": 1.8335202277236262,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.422697772941741e-07,
+      "loss": 1.433,
+      "step": 10628
+    },
+    {
+      "epoch": 1.8336927456223582,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.415641105679801e-07,
+      "loss": 1.4772,
+      "step": 10629
+    },
+    {
+      "epoch": 1.8338652635210904,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.408591593990662e-07,
+      "loss": 1.5794,
+      "step": 10630
+    },
+    {
+      "epoch": 1.8340377814198223,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.401549238396584e-07,
+      "loss": 1.3741,
+      "step": 10631
+    },
+    {
+      "epoch": 1.8342102993185543,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.3945140394193276e-07,
+      "loss": 1.4539,
+      "step": 10632
+    },
+    {
+      "epoch": 1.8343828172172862,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.3874859975800644e-07,
+      "loss": 1.3585,
+      "step": 10633
+    },
+    {
+      "epoch": 1.8345553351160184,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.3804651133995226e-07,
+      "loss": 1.4456,
+      "step": 10634
+    },
+    {
+      "epoch": 1.8347278530147504,
+      "grad_norm": 0.671875,
+      "learning_rate": 3.3734513873978193e-07,
+      "loss": 1.4126,
+      "step": 10635
+    },
+    {
+      "epoch": 1.8349003709134823,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.3664448200945943e-07,
+      "loss": 1.3672,
+      "step": 10636
+    },
+    {
+      "epoch": 1.8350728888122143,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.3594454120089216e-07,
+      "loss": 1.4352,
+      "step": 10637
+    },
+    {
+      "epoch": 1.8352454067109463,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.352453163659386e-07,
+      "loss": 1.3739,
+      "step": 10638
+    },
+    {
+      "epoch": 1.8354179246096782,
+      "grad_norm": 0.58984375,
+      "learning_rate": 3.3454680755639847e-07,
+      "loss": 1.5121,
+      "step": 10639
+    },
+    {
+      "epoch": 1.8355904425084102,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.3384901482402585e-07,
+      "loss": 1.3696,
+      "step": 10640
+    },
+    {
+      "epoch": 1.8357629604071422,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.3315193822051283e-07,
+      "loss": 1.3752,
+      "step": 10641
+    },
+    {
+      "epoch": 1.8359354783058741,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.3245557779750693e-07,
+      "loss": 1.4764,
+      "step": 10642
+    },
+    {
+      "epoch": 1.836107996204606,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.3175993360659684e-07,
+      "loss": 1.4615,
+      "step": 10643
+    },
+    {
+      "epoch": 1.8362805141033383,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.310650056993192e-07,
+      "loss": 1.4907,
+      "step": 10644
+    },
+    {
+      "epoch": 1.8364530320020702,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.3037079412716276e-07,
+      "loss": 1.4502,
+      "step": 10645
+    },
+    {
+      "epoch": 1.8366255499008022,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.296772989415542e-07,
+      "loss": 1.461,
+      "step": 10646
+    },
+    {
+      "epoch": 1.8367980677995344,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.2898452019387685e-07,
+      "loss": 1.4372,
+      "step": 10647
+    },
+    {
+      "epoch": 1.8369705856982663,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.2829245793545186e-07,
+      "loss": 1.5597,
+      "step": 10648
+    },
+    {
+      "epoch": 1.8371431035969983,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.2760111221755375e-07,
+      "loss": 1.4422,
+      "step": 10649
+    },
+    {
+      "epoch": 1.8373156214957302,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.269104830914016e-07,
+      "loss": 1.4101,
+      "step": 10650
+    },
+    {
+      "epoch": 1.8374881393944622,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.2622057060816004e-07,
+      "loss": 1.4499,
+      "step": 10651
+    },
+    {
+      "epoch": 1.8376606572931942,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.255313748189437e-07,
+      "loss": 1.4314,
+      "step": 10652
+    },
+    {
+      "epoch": 1.8378331751919261,
+      "grad_norm": 1.2578125,
+      "learning_rate": 3.2484289577481286e-07,
+      "loss": 1.34,
+      "step": 10653
+    },
+    {
+      "epoch": 1.838005693090658,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.241551335267712e-07,
+      "loss": 1.3133,
+      "step": 10654
+    },
+    {
+      "epoch": 1.83817821098939,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.234680881257779e-07,
+      "loss": 1.438,
+      "step": 10655
+    },
+    {
+      "epoch": 1.838350728888122,
+      "grad_norm": 1.1953125,
+      "learning_rate": 3.2278175962272783e-07,
+      "loss": 1.4366,
+      "step": 10656
+    },
+    {
+      "epoch": 1.838523246786854,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.2209614806847256e-07,
+      "loss": 1.4324,
+      "step": 10657
+    },
+    {
+      "epoch": 1.8386957646855862,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.2141125351380363e-07,
+      "loss": 1.4163,
+      "step": 10658
+    },
+    {
+      "epoch": 1.8388682825843181,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.20727076009465e-07,
+      "loss": 1.4913,
+      "step": 10659
+    },
+    {
+      "epoch": 1.83904080048305,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.200436156061426e-07,
+      "loss": 1.4095,
+      "step": 10660
+    },
+    {
+      "epoch": 1.8392133183817823,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.1936087235447165e-07,
+      "loss": 1.3978,
+      "step": 10661
+    },
+    {
+      "epoch": 1.8393858362805142,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.186788463050361e-07,
+      "loss": 1.4577,
+      "step": 10662
+    },
+    {
+      "epoch": 1.8395583541792462,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.1799753750836215e-07,
+      "loss": 1.4572,
+      "step": 10663
+    },
+    {
+      "epoch": 1.8397308720779781,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.1731694601492834e-07,
+      "loss": 1.464,
+      "step": 10664
+    },
+    {
+      "epoch": 1.83990338997671,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.1663707187515325e-07,
+      "loss": 1.4048,
+      "step": 10665
+    },
+    {
+      "epoch": 1.840075907875442,
+      "grad_norm": 0.55078125,
+      "learning_rate": 3.1595791513941097e-07,
+      "loss": 1.5117,
+      "step": 10666
+    },
+    {
+      "epoch": 1.840248425774174,
+      "grad_norm": 0.671875,
+      "learning_rate": 3.1527947585801246e-07,
+      "loss": 1.394,
+      "step": 10667
+    },
+    {
+      "epoch": 1.840420943672906,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.146017540812241e-07,
+      "loss": 1.4003,
+      "step": 10668
+    },
+    {
+      "epoch": 1.840593461571638,
+      "grad_norm": 0.640625,
+      "learning_rate": 3.139247498592557e-07,
+      "loss": 1.4011,
+      "step": 10669
+    },
+    {
+      "epoch": 1.84076597947037,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.1324846324226165e-07,
+      "loss": 1.3679,
+      "step": 10670
+    },
+    {
+      "epoch": 1.840938497369102,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.1257289428034854e-07,
+      "loss": 1.437,
+      "step": 10671
+    },
+    {
+      "epoch": 1.841111015267834,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.1189804302356517e-07,
+      "loss": 1.4845,
+      "step": 10672
+    },
+    {
+      "epoch": 1.841283533166566,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.112239095219072e-07,
+      "loss": 1.4396,
+      "step": 10673
+    },
+    {
+      "epoch": 1.841456051065298,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.105504938253223e-07,
+      "loss": 1.4616,
+      "step": 10674
+    },
+    {
+      "epoch": 1.8416285689640302,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.098777959836974e-07,
+      "loss": 1.4047,
+      "step": 10675
+    },
+    {
+      "epoch": 1.8418010868627621,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.092058160468736e-07,
+      "loss": 1.3966,
+      "step": 10676
+    },
+    {
+      "epoch": 1.841973604761494,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.085345540646345e-07,
+      "loss": 1.4355,
+      "step": 10677
+    },
+    {
+      "epoch": 1.842146122660226,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.0786401008670806e-07,
+      "loss": 1.4064,
+      "step": 10678
+    },
+    {
+      "epoch": 1.842318640558958,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.071941841627779e-07,
+      "loss": 1.3451,
+      "step": 10679
+    },
+    {
+      "epoch": 1.84249115845769,
+      "grad_norm": 0.5390625,
+      "learning_rate": 3.065250763424643e-07,
+      "loss": 1.4058,
+      "step": 10680
+    },
+    {
+      "epoch": 1.842663676356422,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.0585668667534097e-07,
+      "loss": 1.3844,
+      "step": 10681
+    },
+    {
+      "epoch": 1.8428361942551539,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.0518901521092605e-07,
+      "loss": 1.4213,
+      "step": 10682
+    },
+    {
+      "epoch": 1.8430087121538858,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.045220619986844e-07,
+      "loss": 1.5024,
+      "step": 10683
+    },
+    {
+      "epoch": 1.8431812300526178,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.038558270880287e-07,
+      "loss": 1.3808,
+      "step": 10684
+    },
+    {
+      "epoch": 1.84335374795135,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.031903105283196e-07,
+      "loss": 1.3525,
+      "step": 10685
+    },
+    {
+      "epoch": 1.843526265850082,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.025255123688575e-07,
+      "loss": 1.4255,
+      "step": 10686
+    },
+    {
+      "epoch": 1.843698783748814,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.018614326589009e-07,
+      "loss": 1.424,
+      "step": 10687
+    },
+    {
+      "epoch": 1.843871301647546,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.011980714476448e-07,
+      "loss": 1.4183,
+      "step": 10688
+    },
+    {
+      "epoch": 1.844043819546278,
+      "grad_norm": 0.6640625,
+      "learning_rate": 3.0053542878423657e-07,
+      "loss": 1.4809,
+      "step": 10689
+    },
+    {
+      "epoch": 1.84421633744501,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.998735047177692e-07,
+      "loss": 1.4014,
+      "step": 10690
+    },
+    {
+      "epoch": 1.844388855343742,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.992122992972812e-07,
+      "loss": 1.3376,
+      "step": 10691
+    },
+    {
+      "epoch": 1.844561373242474,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.9855181257176015e-07,
+      "loss": 1.4756,
+      "step": 10692
+    },
+    {
+      "epoch": 1.844733891141206,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.978920445901379e-07,
+      "loss": 1.3947,
+      "step": 10693
+    },
+    {
+      "epoch": 1.8449064090399379,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.9723299540129423e-07,
+      "loss": 1.4171,
+      "step": 10694
+    },
+    {
+      "epoch": 1.8450789269386698,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.9657466505405573e-07,
+      "loss": 1.4056,
+      "step": 10695
+    },
+    {
+      "epoch": 1.8452514448374018,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.959170535971978e-07,
+      "loss": 1.4627,
+      "step": 10696
+    },
+    {
+      "epoch": 1.8454239627361337,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.952601610794359e-07,
+      "loss": 1.4116,
+      "step": 10697
+    },
+    {
+      "epoch": 1.845596480634866,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.9460398754944106e-07,
+      "loss": 1.4963,
+      "step": 10698
+    },
+    {
+      "epoch": 1.8457689985335979,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.9394853305582337e-07,
+      "loss": 1.4373,
+      "step": 10699
+    },
+    {
+      "epoch": 1.8459415164323298,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.9329379764714615e-07,
+      "loss": 1.3966,
+      "step": 10700
+    },
+    {
+      "epoch": 1.8459415164323298,
+      "eval_loss": 1.4070720672607422,
+      "eval_runtime": 12.0233,
+      "eval_samples_per_second": 85.168,
+      "eval_steps_per_second": 21.292,
+      "step": 10700
+    },
+    {
+      "epoch": 1.8461140343310618,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.9263978137191396e-07,
+      "loss": 1.5311,
+      "step": 10701
+    },
+    {
+      "epoch": 1.846286552229794,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.919864842785802e-07,
+      "loss": 1.4487,
+      "step": 10702
+    },
+    {
+      "epoch": 1.846459070128526,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.9133390641554736e-07,
+      "loss": 1.4173,
+      "step": 10703
+    },
+    {
+      "epoch": 1.846631588027258,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.9068204783116227e-07,
+      "loss": 1.4152,
+      "step": 10704
+    },
+    {
+      "epoch": 1.8468041059259899,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.900309085737152e-07,
+      "loss": 1.4875,
+      "step": 10705
+    },
+    {
+      "epoch": 1.8469766238247218,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.8938048869145087e-07,
+      "loss": 1.4071,
+      "step": 10706
+    },
+    {
+      "epoch": 1.8471491417234538,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.8873078823255297e-07,
+      "loss": 1.4231,
+      "step": 10707
+    },
+    {
+      "epoch": 1.8473216596221858,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.8808180724515856e-07,
+      "loss": 1.3817,
+      "step": 10708
+    },
+    {
+      "epoch": 1.8474941775209177,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.8743354577734805e-07,
+      "loss": 1.3916,
+      "step": 10709
+    },
+    {
+      "epoch": 1.8476666954196497,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.8678600387714417e-07,
+      "loss": 1.3462,
+      "step": 10710
+    },
+    {
+      "epoch": 1.8478392133183816,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.8613918159252627e-07,
+      "loss": 1.5002,
+      "step": 10711
+    },
+    {
+      "epoch": 1.8480117312171138,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.8549307897141274e-07,
+      "loss": 1.3669,
+      "step": 10712
+    },
+    {
+      "epoch": 1.8481842491158458,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.8484769606167085e-07,
+      "loss": 1.4325,
+      "step": 10713
+    },
+    {
+      "epoch": 1.8483567670145777,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.842030329111134e-07,
+      "loss": 1.4155,
+      "step": 10714
+    },
+    {
+      "epoch": 1.8485292849133097,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.8355908956750335e-07,
+      "loss": 1.4419,
+      "step": 10715
+    },
+    {
+      "epoch": 1.8487018028120419,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.8291586607854693e-07,
+      "loss": 1.3997,
+      "step": 10716
+    },
+    {
+      "epoch": 1.8488743207107738,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.822733624918994e-07,
+      "loss": 1.3909,
+      "step": 10717
+    },
+    {
+      "epoch": 1.8490468386095058,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.8163157885515824e-07,
+      "loss": 1.4904,
+      "step": 10718
+    },
+    {
+      "epoch": 1.8492193565082378,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.809905152158754e-07,
+      "loss": 1.435,
+      "step": 10719
+    },
+    {
+      "epoch": 1.8493918744069697,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.8035017162154063e-07,
+      "loss": 1.4916,
+      "step": 10720
+    },
+    {
+      "epoch": 1.8495643923057017,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.7971054811959717e-07,
+      "loss": 1.481,
+      "step": 10721
+    },
+    {
+      "epoch": 1.8497369102044336,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.790716447574304e-07,
+      "loss": 1.4353,
+      "step": 10722
+    },
+    {
+      "epoch": 1.8499094281031656,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.7843346158237586e-07,
+      "loss": 1.3761,
+      "step": 10723
+    },
+    {
+      "epoch": 1.8500819460018976,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.777959986417134e-07,
+      "loss": 1.4107,
+      "step": 10724
+    },
+    {
+      "epoch": 1.8502544639006295,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.771592559826708e-07,
+      "loss": 1.4496,
+      "step": 10725
+    },
+    {
+      "epoch": 1.8504269817993617,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.765232336524215e-07,
+      "loss": 1.4275,
+      "step": 10726
+    },
+    {
+      "epoch": 1.8505994996980937,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.758879316980867e-07,
+      "loss": 1.4414,
+      "step": 10727
+    },
+    {
+      "epoch": 1.8507720175968256,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.7525335016673315e-07,
+      "loss": 1.4265,
+      "step": 10728
+    },
+    {
+      "epoch": 1.8509445354955578,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.746194891053733e-07,
+      "loss": 1.4261,
+      "step": 10729
+    },
+    {
+      "epoch": 1.8511170533942898,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.739863485609695e-07,
+      "loss": 1.4396,
+      "step": 10730
+    },
+    {
+      "epoch": 1.8512895712930217,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.7335392858042764e-07,
+      "loss": 1.4106,
+      "step": 10731
+    },
+    {
+      "epoch": 1.8514620891917537,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.727222292106024e-07,
+      "loss": 1.3709,
+      "step": 10732
+    },
+    {
+      "epoch": 1.8516346070904857,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.72091250498292e-07,
+      "loss": 1.4935,
+      "step": 10733
+    },
+    {
+      "epoch": 1.8518071249892176,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.714609924902445e-07,
+      "loss": 1.4064,
+      "step": 10734
+    },
+    {
+      "epoch": 1.8519796428879496,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.708314552331548e-07,
+      "loss": 1.5138,
+      "step": 10735
+    },
+    {
+      "epoch": 1.8521521607866815,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.7020263877366005e-07,
+      "loss": 1.3863,
+      "step": 10736
+    },
+    {
+      "epoch": 1.8523246786854135,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.6957454315834965e-07,
+      "loss": 1.4406,
+      "step": 10737
+    },
+    {
+      "epoch": 1.8524971965841455,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.6894716843375523e-07,
+      "loss": 1.5074,
+      "step": 10738
+    },
+    {
+      "epoch": 1.8526697144828776,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.6832051464635636e-07,
+      "loss": 1.3569,
+      "step": 10739
+    },
+    {
+      "epoch": 1.8528422323816096,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.6769458184258134e-07,
+      "loss": 1.4224,
+      "step": 10740
+    },
+    {
+      "epoch": 1.8530147502803416,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.6706937006880095e-07,
+      "loss": 1.4153,
+      "step": 10741
+    },
+    {
+      "epoch": 1.8531872681790735,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.664448793713348e-07,
+      "loss": 1.5044,
+      "step": 10742
+    },
+    {
+      "epoch": 1.8533597860778057,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.6582110979645246e-07,
+      "loss": 1.5223,
+      "step": 10743
+    },
+    {
+      "epoch": 1.8535323039765377,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.651980613903626e-07,
+      "loss": 1.4345,
+      "step": 10744
+    },
+    {
+      "epoch": 1.8537048218752696,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.645757341992261e-07,
+      "loss": 1.3272,
+      "step": 10745
+    },
+    {
+      "epoch": 1.8538773397740016,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.6395412826915046e-07,
+      "loss": 1.3986,
+      "step": 10746
+    },
+    {
+      "epoch": 1.8540498576727336,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.6333324364618553e-07,
+      "loss": 1.3335,
+      "step": 10747
+    },
+    {
+      "epoch": 1.8542223755714655,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.6271308037633113e-07,
+      "loss": 1.3878,
+      "step": 10748
+    },
+    {
+      "epoch": 1.8543948934701975,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.6209363850553393e-07,
+      "loss": 1.3827,
+      "step": 10749
+    },
+    {
+      "epoch": 1.8545674113689294,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.6147491807968385e-07,
+      "loss": 1.4321,
+      "step": 10750
+    },
+    {
+      "epoch": 1.8547399292676614,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.6085691914462306e-07,
+      "loss": 1.4484,
+      "step": 10751
+    },
+    {
+      "epoch": 1.8549124471663934,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.6023964174613393e-07,
+      "loss": 1.3907,
+      "step": 10752
+    },
+    {
+      "epoch": 1.8550849650651255,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.596230859299487e-07,
+      "loss": 1.3456,
+      "step": 10753
+    },
+    {
+      "epoch": 1.8552574829638575,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.5900725174174524e-07,
+      "loss": 1.347,
+      "step": 10754
+    },
+    {
+      "epoch": 1.8554300008625895,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.5839213922714936e-07,
+      "loss": 1.4167,
+      "step": 10755
+    },
+    {
+      "epoch": 1.8556025187613217,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.5777774843173233e-07,
+      "loss": 1.3874,
+      "step": 10756
+    },
+    {
+      "epoch": 1.8557750366600536,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.5716407940101217e-07,
+      "loss": 1.4425,
+      "step": 10757
+    },
+    {
+      "epoch": 1.8559475545587856,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.565511321804537e-07,
+      "loss": 1.2792,
+      "step": 10758
+    },
+    {
+      "epoch": 1.8561200724575175,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.559389068154661e-07,
+      "loss": 1.4789,
+      "step": 10759
+    },
+    {
+      "epoch": 1.8562925903562495,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.553274033514097e-07,
+      "loss": 1.5425,
+      "step": 10760
+    },
+    {
+      "epoch": 1.8564651082549815,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.5471662183358394e-07,
+      "loss": 1.4059,
+      "step": 10761
+    },
+    {
+      "epoch": 1.8566376261537134,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.5410656230724475e-07,
+      "loss": 1.3846,
+      "step": 10762
+    },
+    {
+      "epoch": 1.8568101440524454,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.5349722481758487e-07,
+      "loss": 1.427,
+      "step": 10763
+    },
+    {
+      "epoch": 1.8569826619511773,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.5288860940975046e-07,
+      "loss": 1.476,
+      "step": 10764
+    },
+    {
+      "epoch": 1.8571551798499093,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.52280716128831e-07,
+      "loss": 1.2507,
+      "step": 10765
+    },
+    {
+      "epoch": 1.8573276977486413,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.516735450198615e-07,
+      "loss": 1.4192,
+      "step": 10766
+    },
+    {
+      "epoch": 1.8575002156473734,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.510670961278272e-07,
+      "loss": 1.477,
+      "step": 10767
+    },
+    {
+      "epoch": 1.8576727335461054,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.5046136949765544e-07,
+      "loss": 1.5125,
+      "step": 10768
+    },
+    {
+      "epoch": 1.8578452514448374,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.4985636517422365e-07,
+      "loss": 1.3664,
+      "step": 10769
+    },
+    {
+      "epoch": 1.8580177693435695,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.4925208320235593e-07,
+      "loss": 1.428,
+      "step": 10770
+    },
+    {
+      "epoch": 1.8581902872423015,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.486485236268166e-07,
+      "loss": 1.4213,
+      "step": 10771
+    },
+    {
+      "epoch": 1.8583628051410335,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.4804568649232643e-07,
+      "loss": 1.4684,
+      "step": 10772
+    },
+    {
+      "epoch": 1.8585353230397654,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.4744357184354305e-07,
+      "loss": 1.3504,
+      "step": 10773
+    },
+    {
+      "epoch": 1.8587078409384974,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.468421797250764e-07,
+      "loss": 1.4598,
+      "step": 10774
+    },
+    {
+      "epoch": 1.8588803588372294,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.46241510181483e-07,
+      "loss": 1.4404,
+      "step": 10775
+    },
+    {
+      "epoch": 1.8590528767359613,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.456415632572617e-07,
+      "loss": 1.4462,
+      "step": 10776
+    },
+    {
+      "epoch": 1.8592253946346933,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.450423389968626e-07,
+      "loss": 1.4246,
+      "step": 10777
+    },
+    {
+      "epoch": 1.8593979125334252,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.444438374446778e-07,
+      "loss": 1.4082,
+      "step": 10778
+    },
+    {
+      "epoch": 1.8595704304321572,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.4384605864504973e-07,
+      "loss": 1.4425,
+      "step": 10779
+    },
+    {
+      "epoch": 1.8597429483308894,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.4324900264226405e-07,
+      "loss": 1.4271,
+      "step": 10780
+    },
+    {
+      "epoch": 1.8599154662296213,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.426526694805564e-07,
+      "loss": 1.5036,
+      "step": 10781
+    },
+    {
+      "epoch": 1.8600879841283533,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.420570592041038e-07,
+      "loss": 1.3449,
+      "step": 10782
+    },
+    {
+      "epoch": 1.8602605020270853,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.4146217185703755e-07,
+      "loss": 1.4024,
+      "step": 10783
+    },
+    {
+      "epoch": 1.8604330199258174,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.4086800748342577e-07,
+      "loss": 1.5128,
+      "step": 10784
+    },
+    {
+      "epoch": 1.8606055378245494,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.4027456612728985e-07,
+      "loss": 1.537,
+      "step": 10785
+    },
+    {
+      "epoch": 1.8607780557232814,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.396818478325968e-07,
+      "loss": 1.362,
+      "step": 10786
+    },
+    {
+      "epoch": 1.8609505736220133,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.3908985264325614e-07,
+      "loss": 1.4572,
+      "step": 10787
+    },
+    {
+      "epoch": 1.8611230915207453,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.3849858060312924e-07,
+      "loss": 1.4654,
+      "step": 10788
+    },
+    {
+      "epoch": 1.8612956094194772,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.3790803175602007e-07,
+      "loss": 1.4219,
+      "step": 10789
+    },
+    {
+      "epoch": 1.8614681273182092,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.3731820614568023e-07,
+      "loss": 1.396,
+      "step": 10790
+    },
+    {
+      "epoch": 1.8616406452169412,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.3672910381580817e-07,
+      "loss": 1.4431,
+      "step": 10791
+    },
+    {
+      "epoch": 1.8618131631156731,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.3614072481004778e-07,
+      "loss": 1.4451,
+      "step": 10792
+    },
+    {
+      "epoch": 1.861985681014405,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.3555306917198872e-07,
+      "loss": 1.2986,
+      "step": 10793
+    },
+    {
+      "epoch": 1.8621581989131373,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.3496613694517056e-07,
+      "loss": 1.4842,
+      "step": 10794
+    },
+    {
+      "epoch": 1.8623307168118692,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.343799281730741e-07,
+      "loss": 1.3365,
+      "step": 10795
+    },
+    {
+      "epoch": 1.8625032347106012,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.3379444289913344e-07,
+      "loss": 1.4792,
+      "step": 10796
+    },
+    {
+      "epoch": 1.8626757526093334,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.3320968116672172e-07,
+      "loss": 1.3473,
+      "step": 10797
+    },
+    {
+      "epoch": 1.8628482705080653,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.32625643019162e-07,
+      "loss": 1.4253,
+      "step": 10798
+    },
+    {
+      "epoch": 1.8630207884067973,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.320423284997242e-07,
+      "loss": 1.4237,
+      "step": 10799
+    },
+    {
+      "epoch": 1.8631933063055293,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.3145973765162367e-07,
+      "loss": 1.4938,
+      "step": 10800
+    },
+    {
+      "epoch": 1.8631933063055293,
+      "eval_loss": 1.4070444107055664,
+      "eval_runtime": 10.8998,
+      "eval_samples_per_second": 93.946,
+      "eval_steps_per_second": 23.487,
+      "step": 10800
+    },
+    {
+      "epoch": 1.8633658242042612,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.3087787051802146e-07,
+      "loss": 1.4087,
+      "step": 10801
+    },
+    {
+      "epoch": 1.8635383421029932,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.3029672714202866e-07,
+      "loss": 1.3573,
+      "step": 10802
+    },
+    {
+      "epoch": 1.8637108600017251,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.2971630756669637e-07,
+      "loss": 1.3931,
+      "step": 10803
+    },
+    {
+      "epoch": 1.863883377900457,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.291366118350302e-07,
+      "loss": 1.4099,
+      "step": 10804
+    },
+    {
+      "epoch": 1.864055895799189,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.285576399899736e-07,
+      "loss": 1.4181,
+      "step": 10805
+    },
+    {
+      "epoch": 1.864228413697921,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.2797939207442e-07,
+      "loss": 1.3713,
+      "step": 10806
+    },
+    {
+      "epoch": 1.864400931596653,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.2740186813121402e-07,
+      "loss": 1.46,
+      "step": 10807
+    },
+    {
+      "epoch": 1.8645734494953852,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.268250682031392e-07,
+      "loss": 1.3744,
+      "step": 10808
+    },
+    {
+      "epoch": 1.8647459673941171,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.2624899233292807e-07,
+      "loss": 1.3474,
+      "step": 10809
+    },
+    {
+      "epoch": 1.864918485292849,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.256736405632609e-07,
+      "loss": 1.5073,
+      "step": 10810
+    },
+    {
+      "epoch": 1.8650910031915813,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.250990129367636e-07,
+      "loss": 1.427,
+      "step": 10811
+    },
+    {
+      "epoch": 1.8652635210903132,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.245251094960077e-07,
+      "loss": 1.3456,
+      "step": 10812
+    },
+    {
+      "epoch": 1.8654360389890452,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.2395193028351247e-07,
+      "loss": 1.5013,
+      "step": 10813
+    },
+    {
+      "epoch": 1.8656085568877772,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.2337947534174064e-07,
+      "loss": 1.5743,
+      "step": 10814
+    },
+    {
+      "epoch": 1.8657810747865091,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.2280774471310496e-07,
+      "loss": 1.3436,
+      "step": 10815
+    },
+    {
+      "epoch": 1.865953592685241,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.2223673843996263e-07,
+      "loss": 1.4284,
+      "step": 10816
+    },
+    {
+      "epoch": 1.866126110583973,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.216664565646165e-07,
+      "loss": 1.449,
+      "step": 10817
+    },
+    {
+      "epoch": 1.866298628482705,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.210968991293172e-07,
+      "loss": 1.39,
+      "step": 10818
+    },
+    {
+      "epoch": 1.866471146381437,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.20528066176261e-07,
+      "loss": 1.4709,
+      "step": 10819
+    },
+    {
+      "epoch": 1.866643664280169,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.1995995774759082e-07,
+      "loss": 1.3929,
+      "step": 10820
+    },
+    {
+      "epoch": 1.866816182178901,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.1939257388539525e-07,
+      "loss": 1.4292,
+      "step": 10821
+    },
+    {
+      "epoch": 1.866988700077633,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.1882591463170956e-07,
+      "loss": 1.3822,
+      "step": 10822
+    },
+    {
+      "epoch": 1.867161217976365,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.1825998002851566e-07,
+      "loss": 1.4464,
+      "step": 10823
+    },
+    {
+      "epoch": 1.867333735875097,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.1769477011774232e-07,
+      "loss": 1.3908,
+      "step": 10824
+    },
+    {
+      "epoch": 1.8675062537738292,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.1713028494126265e-07,
+      "loss": 1.3778,
+      "step": 10825
+    },
+    {
+      "epoch": 1.8676787716725611,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.1656652454089878e-07,
+      "loss": 1.4023,
+      "step": 10826
+    },
+    {
+      "epoch": 1.867851289571293,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.1600348895841394e-07,
+      "loss": 1.5256,
+      "step": 10827
+    },
+    {
+      "epoch": 1.868023807470025,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.1544117823552592e-07,
+      "loss": 1.4666,
+      "step": 10828
+    },
+    {
+      "epoch": 1.868196325368757,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.1487959241389244e-07,
+      "loss": 1.4394,
+      "step": 10829
+    },
+    {
+      "epoch": 1.868368843267489,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.1431873153511807e-07,
+      "loss": 1.4317,
+      "step": 10830
+    },
+    {
+      "epoch": 1.868541361166221,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.1375859564075508e-07,
+      "loss": 1.5176,
+      "step": 10831
+    },
+    {
+      "epoch": 1.868713879064953,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.131991847723036e-07,
+      "loss": 1.3555,
+      "step": 10832
+    },
+    {
+      "epoch": 1.8688863969636849,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.126404989712072e-07,
+      "loss": 1.4658,
+      "step": 10833
+    },
+    {
+      "epoch": 1.8690589148624168,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.1208253827885828e-07,
+      "loss": 1.4349,
+      "step": 10834
+    },
+    {
+      "epoch": 1.869231432761149,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.1152530273658932e-07,
+      "loss": 1.4306,
+      "step": 10835
+    },
+    {
+      "epoch": 1.869403950659881,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.1096879238569068e-07,
+      "loss": 1.4883,
+      "step": 10836
+    },
+    {
+      "epoch": 1.869576468558613,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.10413007267386e-07,
+      "loss": 1.3261,
+      "step": 10837
+    },
+    {
+      "epoch": 1.869748986457345,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.098579474228546e-07,
+      "loss": 1.4961,
+      "step": 10838
+    },
+    {
+      "epoch": 1.869921504356077,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.093036128932191e-07,
+      "loss": 1.4159,
+      "step": 10839
+    },
+    {
+      "epoch": 1.870094022254809,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.0875000371954557e-07,
+      "loss": 1.4955,
+      "step": 10840
+    },
+    {
+      "epoch": 1.870266540153541,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.081971199428512e-07,
+      "loss": 1.4283,
+      "step": 10841
+    },
+    {
+      "epoch": 1.870439058052273,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.0764496160409653e-07,
+      "loss": 1.3329,
+      "step": 10842
+    },
+    {
+      "epoch": 1.870611575951005,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.0709352874418777e-07,
+      "loss": 1.5467,
+      "step": 10843
+    },
+    {
+      "epoch": 1.8707840938497369,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.0654282140397996e-07,
+      "loss": 1.4522,
+      "step": 10844
+    },
+    {
+      "epoch": 1.8709566117484688,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.0599283962427274e-07,
+      "loss": 1.4411,
+      "step": 10845
+    },
+    {
+      "epoch": 1.8711291296472008,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.0544358344580905e-07,
+      "loss": 1.363,
+      "step": 10846
+    },
+    {
+      "epoch": 1.8713016475459328,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.0489505290928747e-07,
+      "loss": 1.4794,
+      "step": 10847
+    },
+    {
+      "epoch": 1.871474165444665,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.04347248055341e-07,
+      "loss": 1.4492,
+      "step": 10848
+    },
+    {
+      "epoch": 1.871646683343397,
+      "grad_norm": 1.1796875,
+      "learning_rate": 2.0380016892455611e-07,
+      "loss": 1.4276,
+      "step": 10849
+    },
+    {
+      "epoch": 1.8718192012421289,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.0325381555746483e-07,
+      "loss": 1.5,
+      "step": 10850
+    },
+    {
+      "epoch": 1.8719917191408608,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.0270818799454472e-07,
+      "loss": 1.4219,
+      "step": 10851
+    },
+    {
+      "epoch": 1.872164237039593,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.0216328627621685e-07,
+      "loss": 1.5137,
+      "step": 10852
+    },
+    {
+      "epoch": 1.872336754938325,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.016191104428533e-07,
+      "loss": 1.3509,
+      "step": 10853
+    },
+    {
+      "epoch": 1.872509272837057,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.0107566053476856e-07,
+      "loss": 1.4113,
+      "step": 10854
+    },
+    {
+      "epoch": 1.8726817907357889,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.0053293659222595e-07,
+      "loss": 1.4339,
+      "step": 10855
+    },
+    {
+      "epoch": 1.8728543086345208,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.999909386554333e-07,
+      "loss": 1.5374,
+      "step": 10856
+    },
+    {
+      "epoch": 1.8730268265332528,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.9944966676454402e-07,
+      "loss": 1.4512,
+      "step": 10857
+    },
+    {
+      "epoch": 1.8731993444319848,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.9890912095966274e-07,
+      "loss": 1.3596,
+      "step": 10858
+    },
+    {
+      "epoch": 1.8733718623307167,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.9836930128083076e-07,
+      "loss": 1.3449,
+      "step": 10859
+    },
+    {
+      "epoch": 1.8735443802294487,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.9783020776804718e-07,
+      "loss": 1.3641,
+      "step": 10860
+    },
+    {
+      "epoch": 1.8737168981281807,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.9729184046124673e-07,
+      "loss": 1.4446,
+      "step": 10861
+    },
+    {
+      "epoch": 1.8738894160269128,
+      "grad_norm": 0.625,
+      "learning_rate": 1.9675419940031748e-07,
+      "loss": 1.4485,
+      "step": 10862
+    },
+    {
+      "epoch": 1.8740619339256448,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.9621728462508981e-07,
+      "loss": 1.4549,
+      "step": 10863
+    },
+    {
+      "epoch": 1.8742344518243768,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.9568109617534304e-07,
+      "loss": 1.4242,
+      "step": 10864
+    },
+    {
+      "epoch": 1.8744069697231087,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.951456340908009e-07,
+      "loss": 1.4486,
+      "step": 10865
+    },
+    {
+      "epoch": 1.874579487621841,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.9461089841113502e-07,
+      "loss": 1.4479,
+      "step": 10866
+    },
+    {
+      "epoch": 1.8747520055205729,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.9407688917595923e-07,
+      "loss": 1.4525,
+      "step": 10867
+    },
+    {
+      "epoch": 1.8749245234193048,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.9354360642483862e-07,
+      "loss": 1.3623,
+      "step": 10868
+    },
+    {
+      "epoch": 1.8750970413180368,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.9301105019728038e-07,
+      "loss": 1.4158,
+      "step": 10869
+    },
+    {
+      "epoch": 1.8752695592167687,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.9247922053273972e-07,
+      "loss": 1.4427,
+      "step": 10870
+    },
+    {
+      "epoch": 1.8754420771155007,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.9194811747062058e-07,
+      "loss": 1.3908,
+      "step": 10871
+    },
+    {
+      "epoch": 1.8756145950142327,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.914177410502671e-07,
+      "loss": 1.4076,
+      "step": 10872
+    },
+    {
+      "epoch": 1.8757871129129646,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.9088809131097562e-07,
+      "loss": 1.4296,
+      "step": 10873
+    },
+    {
+      "epoch": 1.8759596308116966,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.9035916829198255e-07,
+      "loss": 1.4531,
+      "step": 10874
+    },
+    {
+      "epoch": 1.8761321487104285,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.8983097203247647e-07,
+      "loss": 1.4507,
+      "step": 10875
+    },
+    {
+      "epoch": 1.8763046666091607,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.8930350257158946e-07,
+      "loss": 1.4777,
+      "step": 10876
+    },
+    {
+      "epoch": 1.8764771845078927,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.8877675994839918e-07,
+      "loss": 1.513,
+      "step": 10877
+    },
+    {
+      "epoch": 1.8766497024066247,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.882507442019288e-07,
+      "loss": 1.4396,
+      "step": 10878
+    },
+    {
+      "epoch": 1.8768222203053568,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.877254553711505e-07,
+      "loss": 1.4054,
+      "step": 10879
+    },
+    {
+      "epoch": 1.8769947382040888,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.8720089349498093e-07,
+      "loss": 1.4699,
+      "step": 10880
+    },
+    {
+      "epoch": 1.8771672561028208,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.8667705861228126e-07,
+      "loss": 1.3528,
+      "step": 10881
+    },
+    {
+      "epoch": 1.8773397740015527,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.861539507618626e-07,
+      "loss": 1.412,
+      "step": 10882
+    },
+    {
+      "epoch": 1.8775122919002847,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.8563156998247844e-07,
+      "loss": 1.3684,
+      "step": 10883
+    },
+    {
+      "epoch": 1.8776848097990166,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.8510991631283003e-07,
+      "loss": 1.3471,
+      "step": 10884
+    },
+    {
+      "epoch": 1.8778573276977486,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.8458898979156536e-07,
+      "loss": 1.5138,
+      "step": 10885
+    },
+    {
+      "epoch": 1.8780298455964806,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.840687904572791e-07,
+      "loss": 1.422,
+      "step": 10886
+    },
+    {
+      "epoch": 1.8782023634952125,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.835493183485082e-07,
+      "loss": 1.4165,
+      "step": 10887
+    },
+    {
+      "epoch": 1.8783748813939445,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.8303057350374077e-07,
+      "loss": 1.4187,
+      "step": 10888
+    },
+    {
+      "epoch": 1.8785473992926767,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.8251255596140716e-07,
+      "loss": 1.3732,
+      "step": 10889
+    },
+    {
+      "epoch": 1.8787199171914086,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.819952657598867e-07,
+      "loss": 1.3429,
+      "step": 10890
+    },
+    {
+      "epoch": 1.8788924350901406,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.8147870293750092e-07,
+      "loss": 1.3351,
+      "step": 10891
+    },
+    {
+      "epoch": 1.8790649529888725,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.8096286753252368e-07,
+      "loss": 1.3576,
+      "step": 10892
+    },
+    {
+      "epoch": 1.8792374708876047,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.8044775958316884e-07,
+      "loss": 1.4096,
+      "step": 10893
+    },
+    {
+      "epoch": 1.8794099887863367,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.7993337912759924e-07,
+      "loss": 1.4544,
+      "step": 10894
+    },
+    {
+      "epoch": 1.8795825066850687,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.7941972620392322e-07,
+      "loss": 1.4095,
+      "step": 10895
+    },
+    {
+      "epoch": 1.8797550245838006,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.7890680085019597e-07,
+      "loss": 1.4747,
+      "step": 10896
+    },
+    {
+      "epoch": 1.8799275424825326,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.7839460310441814e-07,
+      "loss": 1.3743,
+      "step": 10897
+    },
+    {
+      "epoch": 1.8801000603812645,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.778831330045372e-07,
+      "loss": 1.3766,
+      "step": 10898
+    },
+    {
+      "epoch": 1.8802725782799965,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.773723905884428e-07,
+      "loss": 1.4824,
+      "step": 10899
+    },
+    {
+      "epoch": 1.8804450961787285,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.7686237589397914e-07,
+      "loss": 1.5453,
+      "step": 10900
+    },
+    {
+      "epoch": 1.8804450961787285,
+      "eval_loss": 1.4070632457733154,
+      "eval_runtime": 10.8553,
+      "eval_samples_per_second": 94.332,
+      "eval_steps_per_second": 23.583,
+      "step": 10900
+    },
+    {
+      "epoch": 1.8806176140774604,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.76353088958926e-07,
+      "loss": 1.4266,
+      "step": 10901
+    },
+    {
+      "epoch": 1.8807901319761924,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.7584452982101764e-07,
+      "loss": 1.5084,
+      "step": 10902
+    },
+    {
+      "epoch": 1.8809626498749246,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.7533669851793166e-07,
+      "loss": 1.3585,
+      "step": 10903
+    },
+    {
+      "epoch": 1.8811351677736565,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.7482959508729024e-07,
+      "loss": 1.4781,
+      "step": 10904
+    },
+    {
+      "epoch": 1.8813076856723885,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.743232195666622e-07,
+      "loss": 1.4678,
+      "step": 10905
+    },
+    {
+      "epoch": 1.8814802035711207,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.738175719935642e-07,
+      "loss": 1.4632,
+      "step": 10906
+    },
+    {
+      "epoch": 1.8816527214698526,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.7331265240545624e-07,
+      "loss": 1.3286,
+      "step": 10907
+    },
+    {
+      "epoch": 1.8818252393685846,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.7280846083974735e-07,
+      "loss": 1.4574,
+      "step": 10908
+    },
+    {
+      "epoch": 1.8819977572673166,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.7230499733379202e-07,
+      "loss": 1.4725,
+      "step": 10909
+    },
+    {
+      "epoch": 1.8821702751660485,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.7180226192488715e-07,
+      "loss": 1.4448,
+      "step": 10910
+    },
+    {
+      "epoch": 1.8823427930647805,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.7130025465028178e-07,
+      "loss": 1.4886,
+      "step": 10911
+    },
+    {
+      "epoch": 1.8825153109635124,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.7079897554716508e-07,
+      "loss": 1.4193,
+      "step": 10912
+    },
+    {
+      "epoch": 1.8826878288622444,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.7029842465267622e-07,
+      "loss": 1.4011,
+      "step": 10913
+    },
+    {
+      "epoch": 1.8828603467609764,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.6979860200389885e-07,
+      "loss": 1.4604,
+      "step": 10914
+    },
+    {
+      "epoch": 1.8830328646597083,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.692995076378634e-07,
+      "loss": 1.4078,
+      "step": 10915
+    },
+    {
+      "epoch": 1.8832053825584403,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.6880114159154471e-07,
+      "loss": 1.4332,
+      "step": 10916
+    },
+    {
+      "epoch": 1.8833779004571725,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.6830350390186546e-07,
+      "loss": 1.3531,
+      "step": 10917
+    },
+    {
+      "epoch": 1.8835504183559044,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.6780659460569505e-07,
+      "loss": 1.3574,
+      "step": 10918
+    },
+    {
+      "epoch": 1.8837229362546364,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.6731041373984513e-07,
+      "loss": 1.308,
+      "step": 10919
+    },
+    {
+      "epoch": 1.8838954541533686,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.6681496134107856e-07,
+      "loss": 1.4647,
+      "step": 10920
+    },
+    {
+      "epoch": 1.8840679720521005,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.6632023744609925e-07,
+      "loss": 1.5234,
+      "step": 10921
+    },
+    {
+      "epoch": 1.8842404899508325,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.658262420915613e-07,
+      "loss": 1.4465,
+      "step": 10922
+    },
+    {
+      "epoch": 1.8844130078495644,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.653329753140609e-07,
+      "loss": 1.3919,
+      "step": 10923
+    },
+    {
+      "epoch": 1.8845855257482964,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.648404371501444e-07,
+      "loss": 1.3689,
+      "step": 10924
+    },
+    {
+      "epoch": 1.8847580436470284,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.6434862763630156e-07,
+      "loss": 1.4015,
+      "step": 10925
+    },
+    {
+      "epoch": 1.8849305615457603,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.6385754680896758e-07,
+      "loss": 1.452,
+      "step": 10926
+    },
+    {
+      "epoch": 1.8851030794444923,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.6336719470452566e-07,
+      "loss": 1.4592,
+      "step": 10927
+    },
+    {
+      "epoch": 1.8852755973432243,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.6287757135930448e-07,
+      "loss": 1.5216,
+      "step": 10928
+    },
+    {
+      "epoch": 1.8854481152419562,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.6238867680957726e-07,
+      "loss": 1.3362,
+      "step": 10929
+    },
+    {
+      "epoch": 1.8856206331406884,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.6190051109156613e-07,
+      "loss": 1.4932,
+      "step": 10930
+    },
+    {
+      "epoch": 1.8857931510394204,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.6141307424143549e-07,
+      "loss": 1.3702,
+      "step": 10931
+    },
+    {
+      "epoch": 1.8859656689381523,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.609263662952998e-07,
+      "loss": 1.383,
+      "step": 10932
+    },
+    {
+      "epoch": 1.8861381868368843,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.6044038728921575e-07,
+      "loss": 1.3771,
+      "step": 10933
+    },
+    {
+      "epoch": 1.8863107047356165,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.5995513725918676e-07,
+      "loss": 1.4453,
+      "step": 10934
+    },
+    {
+      "epoch": 1.8864832226343484,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.5947061624116634e-07,
+      "loss": 1.4363,
+      "step": 10935
+    },
+    {
+      "epoch": 1.8866557405330804,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.5898682427104905e-07,
+      "loss": 1.4706,
+      "step": 10936
+    },
+    {
+      "epoch": 1.8868282584318123,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.5850376138467626e-07,
+      "loss": 1.2775,
+      "step": 10937
+    },
+    {
+      "epoch": 1.8870007763305443,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.5802142761783824e-07,
+      "loss": 1.5225,
+      "step": 10938
+    },
+    {
+      "epoch": 1.8871732942292763,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.5753982300626859e-07,
+      "loss": 1.3915,
+      "step": 10939
+    },
+    {
+      "epoch": 1.8873458121280082,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.5705894758564654e-07,
+      "loss": 1.4447,
+      "step": 10940
+    },
+    {
+      "epoch": 1.8875183300267402,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.565788013916003e-07,
+      "loss": 1.3441,
+      "step": 10941
+    },
+    {
+      "epoch": 1.8876908479254721,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.560993844596992e-07,
+      "loss": 1.3908,
+      "step": 10942
+    },
+    {
+      "epoch": 1.887863365824204,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.5562069682546587e-07,
+      "loss": 1.4749,
+      "step": 10943
+    },
+    {
+      "epoch": 1.8880358837229363,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.551427385243587e-07,
+      "loss": 1.3988,
+      "step": 10944
+    },
+    {
+      "epoch": 1.8882084016216683,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.546655095917937e-07,
+      "loss": 1.5236,
+      "step": 10945
+    },
+    {
+      "epoch": 1.8883809195204002,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.541890100631227e-07,
+      "loss": 1.4393,
+      "step": 10946
+    },
+    {
+      "epoch": 1.8885534374191324,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.5371323997364962e-07,
+      "loss": 1.4745,
+      "step": 10947
+    },
+    {
+      "epoch": 1.8887259553178644,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.5323819935862183e-07,
+      "loss": 1.491,
+      "step": 10948
+    },
+    {
+      "epoch": 1.8888984732165963,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.5276388825323452e-07,
+      "loss": 1.3892,
+      "step": 10949
+    },
+    {
+      "epoch": 1.8890709911153283,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.5229030669262622e-07,
+      "loss": 1.3417,
+      "step": 10950
+    },
+    {
+      "epoch": 1.8892435090140602,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.5181745471188336e-07,
+      "loss": 1.4163,
+      "step": 10951
+    },
+    {
+      "epoch": 1.8894160269127922,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.5134533234603786e-07,
+      "loss": 1.4501,
+      "step": 10952
+    },
+    {
+      "epoch": 1.8895885448115242,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.5087393963006736e-07,
+      "loss": 1.4131,
+      "step": 10953
+    },
+    {
+      "epoch": 1.8897610627102561,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.504032765988961e-07,
+      "loss": 1.4359,
+      "step": 10954
+    },
+    {
+      "epoch": 1.889933580608988,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.4993334328739174e-07,
+      "loss": 1.3766,
+      "step": 10955
+    },
+    {
+      "epoch": 1.89010609850772,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.494641397303731e-07,
+      "loss": 1.4124,
+      "step": 10956
+    },
+    {
+      "epoch": 1.890278616406452,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.4899566596259907e-07,
+      "loss": 1.4417,
+      "step": 10957
+    },
+    {
+      "epoch": 1.8904511343051842,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.4852792201877742e-07,
+      "loss": 1.3079,
+      "step": 10958
+    },
+    {
+      "epoch": 1.8906236522039161,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.4806090793356266e-07,
+      "loss": 1.5389,
+      "step": 10959
+    },
+    {
+      "epoch": 1.890796170102648,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.4759462374155376e-07,
+      "loss": 1.4332,
+      "step": 10960
+    },
+    {
+      "epoch": 1.8909686880013803,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.4712906947729643e-07,
+      "loss": 1.3728,
+      "step": 10961
+    },
+    {
+      "epoch": 1.8911412059001123,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.4666424517528088e-07,
+      "loss": 1.3431,
+      "step": 10962
+    },
+    {
+      "epoch": 1.8913137237988442,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.4620015086994398e-07,
+      "loss": 1.3682,
+      "step": 10963
+    },
+    {
+      "epoch": 1.8914862416975762,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.457367865956705e-07,
+      "loss": 1.4846,
+      "step": 10964
+    },
+    {
+      "epoch": 1.8916587595963081,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.4527415238678622e-07,
+      "loss": 1.4709,
+      "step": 10965
+    },
+    {
+      "epoch": 1.89183127749504,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.4481224827756824e-07,
+      "loss": 1.3932,
+      "step": 10966
+    },
+    {
+      "epoch": 1.892003795393772,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.4435107430223806e-07,
+      "loss": 1.3974,
+      "step": 10967
+    },
+    {
+      "epoch": 1.892176313292504,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.4389063049496165e-07,
+      "loss": 1.4036,
+      "step": 10968
+    },
+    {
+      "epoch": 1.892348831191236,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.4343091688984956e-07,
+      "loss": 1.406,
+      "step": 10969
+    },
+    {
+      "epoch": 1.892521349089968,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.4297193352096228e-07,
+      "loss": 1.3486,
+      "step": 10970
+    },
+    {
+      "epoch": 1.8926938669887001,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.4251368042230485e-07,
+      "loss": 1.4483,
+      "step": 10971
+    },
+    {
+      "epoch": 1.892866384887432,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.4205615762782566e-07,
+      "loss": 1.4674,
+      "step": 10972
+    },
+    {
+      "epoch": 1.893038902786164,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.415993651714209e-07,
+      "loss": 1.3881,
+      "step": 10973
+    },
+    {
+      "epoch": 1.893211420684896,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.4114330308693358e-07,
+      "loss": 1.3689,
+      "step": 10974
+    },
+    {
+      "epoch": 1.8933839385836282,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.4068797140815217e-07,
+      "loss": 1.4367,
+      "step": 10975
+    },
+    {
+      "epoch": 1.8935564564823602,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.4023337016880856e-07,
+      "loss": 1.3465,
+      "step": 10976
+    },
+    {
+      "epoch": 1.8937289743810921,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.3977949940258474e-07,
+      "loss": 1.4499,
+      "step": 10977
+    },
+    {
+      "epoch": 1.893901492279824,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.3932635914310488e-07,
+      "loss": 1.4315,
+      "step": 10978
+    },
+    {
+      "epoch": 1.894074010178556,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.3887394942393994e-07,
+      "loss": 1.5568,
+      "step": 10979
+    },
+    {
+      "epoch": 1.894246528077288,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.3842227027860866e-07,
+      "loss": 1.4174,
+      "step": 10980
+    },
+    {
+      "epoch": 1.89441904597602,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.3797132174057315e-07,
+      "loss": 1.3474,
+      "step": 10981
+    },
+    {
+      "epoch": 1.894591563874752,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.3752110384324336e-07,
+      "loss": 1.4749,
+      "step": 10982
+    },
+    {
+      "epoch": 1.8947640817734839,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.370716166199726e-07,
+      "loss": 1.4026,
+      "step": 10983
+    },
+    {
+      "epoch": 1.8949365996722158,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.3662286010406423e-07,
+      "loss": 1.3766,
+      "step": 10984
+    },
+    {
+      "epoch": 1.895109117570948,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.3617483432876278e-07,
+      "loss": 1.4039,
+      "step": 10985
+    },
+    {
+      "epoch": 1.89528163546968,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.357275393272628e-07,
+      "loss": 1.3893,
+      "step": 10986
+    },
+    {
+      "epoch": 1.895454153368412,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.352809751327e-07,
+      "loss": 1.4591,
+      "step": 10987
+    },
+    {
+      "epoch": 1.8956266712671441,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.3483514177816127e-07,
+      "loss": 1.4423,
+      "step": 10988
+    },
+    {
+      "epoch": 1.895799189165876,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.343900392966757e-07,
+      "loss": 1.3756,
+      "step": 10989
+    },
+    {
+      "epoch": 1.895971707064608,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.339456677212192e-07,
+      "loss": 1.4249,
+      "step": 10990
+    },
+    {
+      "epoch": 1.89614422496334,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.3350202708471316e-07,
+      "loss": 1.4022,
+      "step": 10991
+    },
+    {
+      "epoch": 1.896316742862072,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.3305911742002575e-07,
+      "loss": 1.3189,
+      "step": 10992
+    },
+    {
+      "epoch": 1.896489260760804,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.3261693875996962e-07,
+      "loss": 1.3691,
+      "step": 10993
+    },
+    {
+      "epoch": 1.896661778659536,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.3217549113730633e-07,
+      "loss": 1.4177,
+      "step": 10994
+    },
+    {
+      "epoch": 1.8968342965582679,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.317347745847386e-07,
+      "loss": 1.4069,
+      "step": 10995
+    },
+    {
+      "epoch": 1.8970068144569998,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.3129478913491923e-07,
+      "loss": 1.4167,
+      "step": 10996
+    },
+    {
+      "epoch": 1.8971793323557318,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.3085553482044434e-07,
+      "loss": 1.4617,
+      "step": 10997
+    },
+    {
+      "epoch": 1.8973518502544637,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.3041701167385567e-07,
+      "loss": 1.4375,
+      "step": 10998
+    },
+    {
+      "epoch": 1.897524368153196,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.2997921972764395e-07,
+      "loss": 1.4244,
+      "step": 10999
+    },
+    {
+      "epoch": 1.8976968860519279,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.2954215901424204e-07,
+      "loss": 1.4471,
+      "step": 11000
+    },
+    {
+      "epoch": 1.8976968860519279,
+      "eval_loss": 1.4070650339126587,
+      "eval_runtime": 10.8027,
+      "eval_samples_per_second": 94.791,
+      "eval_steps_per_second": 23.698,
+      "step": 11000
+    },
+    {
+      "epoch": 1.8978694039506598,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.291058295660308e-07,
+      "loss": 1.426,
+      "step": 11001
+    },
+    {
+      "epoch": 1.898041921849392,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.2867023141533542e-07,
+      "loss": 1.4994,
+      "step": 11002
+    },
+    {
+      "epoch": 1.898214439748124,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.2823536459442788e-07,
+      "loss": 1.4619,
+      "step": 11003
+    },
+    {
+      "epoch": 1.898386957646856,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.2780122913552684e-07,
+      "loss": 1.4617,
+      "step": 11004
+    },
+    {
+      "epoch": 1.898559475545588,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.2736782507079436e-07,
+      "loss": 1.3247,
+      "step": 11005
+    },
+    {
+      "epoch": 1.8987319934443199,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.2693515243234146e-07,
+      "loss": 1.5786,
+      "step": 11006
+    },
+    {
+      "epoch": 1.8989045113430518,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.2650321125222243e-07,
+      "loss": 1.4376,
+      "step": 11007
+    },
+    {
+      "epoch": 1.8990770292417838,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.2607200156243615e-07,
+      "loss": 1.4922,
+      "step": 11008
+    },
+    {
+      "epoch": 1.8992495471405157,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.2564152339493263e-07,
+      "loss": 1.3807,
+      "step": 11009
+    },
+    {
+      "epoch": 1.8994220650392477,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.2521177678160302e-07,
+      "loss": 1.3754,
+      "step": 11010
+    },
+    {
+      "epoch": 1.8995945829379797,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.2478276175428516e-07,
+      "loss": 1.451,
+      "step": 11011
+    },
+    {
+      "epoch": 1.8997671008367119,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.2435447834476254e-07,
+      "loss": 1.472,
+      "step": 11012
+    },
+    {
+      "epoch": 1.8999396187354438,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.2392692658476758e-07,
+      "loss": 1.4268,
+      "step": 11013
+    },
+    {
+      "epoch": 1.9001121366341758,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.2350010650597378e-07,
+      "loss": 1.4491,
+      "step": 11014
+    },
+    {
+      "epoch": 1.9002846545329077,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.2307401814000252e-07,
+      "loss": 1.4408,
+      "step": 11015
+    },
+    {
+      "epoch": 1.90045717243164,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.2264866151842304e-07,
+      "loss": 1.3987,
+      "step": 11016
+    },
+    {
+      "epoch": 1.9006296903303719,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.2222403667274673e-07,
+      "loss": 1.3893,
+      "step": 11017
+    },
+    {
+      "epoch": 1.9008022082291038,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.2180014363443404e-07,
+      "loss": 1.3289,
+      "step": 11018
+    },
+    {
+      "epoch": 1.9009747261278358,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.213769824348865e-07,
+      "loss": 1.3839,
+      "step": 11019
+    },
+    {
+      "epoch": 1.9011472440265678,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.2095455310545788e-07,
+      "loss": 1.3243,
+      "step": 11020
+    },
+    {
+      "epoch": 1.9013197619252997,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.2053285567744323e-07,
+      "loss": 1.4218,
+      "step": 11021
+    },
+    {
+      "epoch": 1.9014922798240317,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.2011189018208414e-07,
+      "loss": 1.4655,
+      "step": 11022
+    },
+    {
+      "epoch": 1.9016647977227636,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.1969165665056904e-07,
+      "loss": 1.4107,
+      "step": 11023
+    },
+    {
+      "epoch": 1.9018373156214956,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.1927215511402968e-07,
+      "loss": 1.4179,
+      "step": 11024
+    },
+    {
+      "epoch": 1.9020098335202276,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.1885338560354787e-07,
+      "loss": 1.3512,
+      "step": 11025
+    },
+    {
+      "epoch": 1.9021823514189597,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.1843534815014767e-07,
+      "loss": 1.4255,
+      "step": 11026
+    },
+    {
+      "epoch": 1.9023548693176917,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.1801804278479877e-07,
+      "loss": 1.3959,
+      "step": 11027
+    },
+    {
+      "epoch": 1.9025273872164237,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.1760146953841978e-07,
+      "loss": 1.4338,
+      "step": 11028
+    },
+    {
+      "epoch": 1.9026999051151559,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.1718562844187153e-07,
+      "loss": 1.4376,
+      "step": 11029
+    },
+    {
+      "epoch": 1.9028724230138878,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.1677051952596164e-07,
+      "loss": 1.4145,
+      "step": 11030
+    },
+    {
+      "epoch": 1.9030449409126198,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.1635614282144658e-07,
+      "loss": 1.3891,
+      "step": 11031
+    },
+    {
+      "epoch": 1.9032174588113517,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.1594249835902294e-07,
+      "loss": 1.4681,
+      "step": 11032
+    },
+    {
+      "epoch": 1.9033899767100837,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.155295861693384e-07,
+      "loss": 1.4878,
+      "step": 11033
+    },
+    {
+      "epoch": 1.9035624946088157,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.1511740628298296e-07,
+      "loss": 1.343,
+      "step": 11034
+    },
+    {
+      "epoch": 1.9037350125075476,
+      "grad_norm": 0.625,
+      "learning_rate": 1.147059587304944e-07,
+      "loss": 1.3924,
+      "step": 11035
+    },
+    {
+      "epoch": 1.9039075304062796,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.1429524354235277e-07,
+      "loss": 1.4782,
+      "step": 11036
+    },
+    {
+      "epoch": 1.9040800483050115,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.1388526074898931e-07,
+      "loss": 1.4173,
+      "step": 11037
+    },
+    {
+      "epoch": 1.9042525662037435,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.134760103807775e-07,
+      "loss": 1.3841,
+      "step": 11038
+    },
+    {
+      "epoch": 1.9044250841024757,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.130674924680364e-07,
+      "loss": 1.4633,
+      "step": 11039
+    },
+    {
+      "epoch": 1.9045976020012076,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.1265970704103069e-07,
+      "loss": 1.3245,
+      "step": 11040
+    },
+    {
+      "epoch": 1.9047701198999396,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.122526541299751e-07,
+      "loss": 1.4491,
+      "step": 11041
+    },
+    {
+      "epoch": 1.9049426377986716,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.1184633376502218e-07,
+      "loss": 1.4918,
+      "step": 11042
+    },
+    {
+      "epoch": 1.9051151556974038,
+      "grad_norm": 0.625,
+      "learning_rate": 1.1144074597627785e-07,
+      "loss": 1.4963,
+      "step": 11043
+    },
+    {
+      "epoch": 1.9052876735961357,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.1103589079378918e-07,
+      "loss": 1.3569,
+      "step": 11044
+    },
+    {
+      "epoch": 1.9054601914948677,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.1063176824755107e-07,
+      "loss": 1.4771,
+      "step": 11045
+    },
+    {
+      "epoch": 1.9056327093935996,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.1022837836750399e-07,
+      "loss": 1.3965,
+      "step": 11046
+    },
+    {
+      "epoch": 1.9058052272923316,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.0982572118353186e-07,
+      "loss": 1.5245,
+      "step": 11047
+    },
+    {
+      "epoch": 1.9059777451910636,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.0942379672546743e-07,
+      "loss": 1.3126,
+      "step": 11048
+    },
+    {
+      "epoch": 1.9061502630897955,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.0902260502308692e-07,
+      "loss": 1.4799,
+      "step": 11049
+    },
+    {
+      "epoch": 1.9063227809885275,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.0862214610611432e-07,
+      "loss": 1.3847,
+      "step": 11050
+    },
+    {
+      "epoch": 1.9064952988872594,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.082224200042159e-07,
+      "loss": 1.5114,
+      "step": 11051
+    },
+    {
+      "epoch": 1.9066678167859914,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.0782342674700907e-07,
+      "loss": 1.3262,
+      "step": 11052
+    },
+    {
+      "epoch": 1.9068403346847236,
+      "grad_norm": 4.6875,
+      "learning_rate": 1.074251663640502e-07,
+      "loss": 1.3784,
+      "step": 11053
+    },
+    {
+      "epoch": 1.9070128525834555,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.0702763888484791e-07,
+      "loss": 1.4332,
+      "step": 11054
+    },
+    {
+      "epoch": 1.9071853704821875,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.0663084433885196e-07,
+      "loss": 1.3191,
+      "step": 11055
+    },
+    {
+      "epoch": 1.9073578883809197,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.0623478275545884e-07,
+      "loss": 1.4464,
+      "step": 11056
+    },
+    {
+      "epoch": 1.9075304062796516,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.0583945416401286e-07,
+      "loss": 1.4451,
+      "step": 11057
+    },
+    {
+      "epoch": 1.9077029241783836,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.0544485859380172e-07,
+      "loss": 1.4179,
+      "step": 11058
+    },
+    {
+      "epoch": 1.9078754420771156,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.0505099607405866e-07,
+      "loss": 1.4735,
+      "step": 11059
+    },
+    {
+      "epoch": 1.9080479599758475,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.0465786663396593e-07,
+      "loss": 1.4257,
+      "step": 11060
+    },
+    {
+      "epoch": 1.9082204778745795,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.0426547030264577e-07,
+      "loss": 1.4477,
+      "step": 11061
+    },
+    {
+      "epoch": 1.9083929957733115,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.0387380710917161e-07,
+      "loss": 1.4187,
+      "step": 11062
+    },
+    {
+      "epoch": 1.9085655136720434,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.0348287708256021e-07,
+      "loss": 1.4474,
+      "step": 11063
+    },
+    {
+      "epoch": 1.9087380315707754,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.0309268025177288e-07,
+      "loss": 1.4031,
+      "step": 11064
+    },
+    {
+      "epoch": 1.9089105494695073,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.0270321664571981e-07,
+      "loss": 1.4469,
+      "step": 11065
+    },
+    {
+      "epoch": 1.9090830673682393,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.0231448629325236e-07,
+      "loss": 1.2811,
+      "step": 11066
+    },
+    {
+      "epoch": 1.9092555852669715,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.0192648922317084e-07,
+      "loss": 1.4864,
+      "step": 11067
+    },
+    {
+      "epoch": 1.9094281031657034,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.0153922546422223e-07,
+      "loss": 1.4214,
+      "step": 11068
+    },
+    {
+      "epoch": 1.9096006210644354,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.0115269504509583e-07,
+      "loss": 1.4991,
+      "step": 11069
+    },
+    {
+      "epoch": 1.9097731389631676,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.0076689799442874e-07,
+      "loss": 1.3097,
+      "step": 11070
+    },
+    {
+      "epoch": 1.9099456568618995,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.0038183434080363e-07,
+      "loss": 1.3634,
+      "step": 11071
+    },
+    {
+      "epoch": 1.9101181747606315,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.999750411274655e-08,
+      "loss": 1.3467,
+      "step": 11072
+    },
+    {
+      "epoch": 1.9102906926593635,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.961390733873366e-08,
+      "loss": 1.456,
+      "step": 11073
+    },
+    {
+      "epoch": 1.9104632105580954,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.92310440471822e-08,
+      "loss": 1.3618,
+      "step": 11074
+    },
+    {
+      "epoch": 1.9106357284568274,
+      "grad_norm": 0.61328125,
+      "learning_rate": 9.884891426645837e-08,
+      "loss": 1.4833,
+      "step": 11075
+    },
+    {
+      "epoch": 1.9108082463555593,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.846751802487175e-08,
+      "loss": 1.4058,
+      "step": 11076
+    },
+    {
+      "epoch": 1.9109807642542913,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.808685535067863e-08,
+      "loss": 1.5136,
+      "step": 11077
+    },
+    {
+      "epoch": 1.9111532821530233,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.7706926272082e-08,
+      "loss": 1.4108,
+      "step": 11078
+    },
+    {
+      "epoch": 1.9113258000517552,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.732773081722824e-08,
+      "loss": 1.4449,
+      "step": 11079
+    },
+    {
+      "epoch": 1.9114983179504874,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.694926901421153e-08,
+      "loss": 1.4676,
+      "step": 11080
+    },
+    {
+      "epoch": 1.9116708358492194,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.657154089106946e-08,
+      "loss": 1.4309,
+      "step": 11081
+    },
+    {
+      "epoch": 1.9118433537479513,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.619454647578852e-08,
+      "loss": 1.3618,
+      "step": 11082
+    },
+    {
+      "epoch": 1.9120158716466833,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.581828579629526e-08,
+      "loss": 1.3586,
+      "step": 11083
+    },
+    {
+      "epoch": 1.9121883895454155,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.544275888046961e-08,
+      "loss": 1.3806,
+      "step": 11084
+    },
+    {
+      "epoch": 1.9123609074441474,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.506796575612931e-08,
+      "loss": 1.3433,
+      "step": 11085
+    },
+    {
+      "epoch": 1.9125334253428794,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.469390645104437e-08,
+      "loss": 1.4082,
+      "step": 11086
+    },
+    {
+      "epoch": 1.9127059432416114,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.432058099292484e-08,
+      "loss": 1.3976,
+      "step": 11087
+    },
+    {
+      "epoch": 1.9128784611403433,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.394798940943083e-08,
+      "loss": 1.353,
+      "step": 11088
+    },
+    {
+      "epoch": 1.9130509790390753,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.357613172816471e-08,
+      "loss": 1.377,
+      "step": 11089
+    },
+    {
+      "epoch": 1.9132234969378072,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.320500797667886e-08,
+      "loss": 1.4407,
+      "step": 11090
+    },
+    {
+      "epoch": 1.9133960148365392,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.283461818246464e-08,
+      "loss": 1.3338,
+      "step": 11091
+    },
+    {
+      "epoch": 1.9135685327352712,
+      "grad_norm": 0.6015625,
+      "learning_rate": 9.246496237296565e-08,
+      "loss": 1.5407,
+      "step": 11092
+    },
+    {
+      "epoch": 1.9137410506340031,
+      "grad_norm": 0.62890625,
+      "learning_rate": 9.209604057556665e-08,
+      "loss": 1.448,
+      "step": 11093
+    },
+    {
+      "epoch": 1.9139135685327353,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.172785281760132e-08,
+      "loss": 1.381,
+      "step": 11094
+    },
+    {
+      "epoch": 1.9140860864314673,
+      "grad_norm": 0.640625,
+      "learning_rate": 9.136039912634675e-08,
+      "loss": 1.4523,
+      "step": 11095
+    },
+    {
+      "epoch": 1.9142586043301992,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.099367952902449e-08,
+      "loss": 1.4942,
+      "step": 11096
+    },
+    {
+      "epoch": 1.9144311222289314,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.062769405280614e-08,
+      "loss": 1.369,
+      "step": 11097
+    },
+    {
+      "epoch": 1.9146036401276634,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.026244272480445e-08,
+      "loss": 1.3735,
+      "step": 11098
+    },
+    {
+      "epoch": 1.9147761580263953,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.989792557207889e-08,
+      "loss": 1.4239,
+      "step": 11099
+    },
+    {
+      "epoch": 1.9149486759251273,
+      "grad_norm": 0.62109375,
+      "learning_rate": 8.953414262163674e-08,
+      "loss": 1.3988,
+      "step": 11100
+    },
+    {
+      "epoch": 1.9149486759251273,
+      "eval_loss": 1.4070982933044434,
+      "eval_runtime": 10.8142,
+      "eval_samples_per_second": 94.69,
+      "eval_steps_per_second": 23.673,
+      "step": 11100
+    },
+    {
+      "epoch": 1.9151211938238593,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.917109390042866e-08,
+      "loss": 1.4235,
+      "step": 11101
+    },
+    {
+      "epoch": 1.9152937117225912,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.880877943535204e-08,
+      "loss": 1.3556,
+      "step": 11102
+    },
+    {
+      "epoch": 1.9154662296213232,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.844719925324985e-08,
+      "loss": 1.46,
+      "step": 11103
+    },
+    {
+      "epoch": 1.9156387475200551,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.808635338090732e-08,
+      "loss": 1.4678,
+      "step": 11104
+    },
+    {
+      "epoch": 1.915811265418787,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.772624184506196e-08,
+      "loss": 1.4166,
+      "step": 11105
+    },
+    {
+      "epoch": 1.915983783317519,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.736686467239131e-08,
+      "loss": 1.435,
+      "step": 11106
+    },
+    {
+      "epoch": 1.916156301216251,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.700822188951963e-08,
+      "loss": 1.5098,
+      "step": 11107
+    },
+    {
+      "epoch": 1.9163288191149832,
+      "grad_norm": 0.66796875,
+      "learning_rate": 8.665031352301789e-08,
+      "loss": 1.3868,
+      "step": 11108
+    },
+    {
+      "epoch": 1.9165013370137152,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.629313959940266e-08,
+      "loss": 1.3686,
+      "step": 11109
+    },
+    {
+      "epoch": 1.9166738549124471,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.5936700145135e-08,
+      "loss": 1.4824,
+      "step": 11110
+    },
+    {
+      "epoch": 1.9168463728111793,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.558099518662378e-08,
+      "loss": 1.4923,
+      "step": 11111
+    },
+    {
+      "epoch": 1.9170188907099113,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.522602475021902e-08,
+      "loss": 1.3338,
+      "step": 11112
+    },
+    {
+      "epoch": 1.9171914086086432,
+      "grad_norm": 0.59765625,
+      "learning_rate": 8.487178886222192e-08,
+      "loss": 1.3928,
+      "step": 11113
+    },
+    {
+      "epoch": 1.9173639265073752,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.451828754887481e-08,
+      "loss": 1.4036,
+      "step": 11114
+    },
+    {
+      "epoch": 1.9175364444061072,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.416552083636676e-08,
+      "loss": 1.4697,
+      "step": 11115
+    },
+    {
+      "epoch": 1.9177089623048391,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.381348875083573e-08,
+      "loss": 1.3408,
+      "step": 11116
+    },
+    {
+      "epoch": 1.917881480203571,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.346219131835976e-08,
+      "loss": 1.5091,
+      "step": 11117
+    },
+    {
+      "epoch": 1.918053998102303,
+      "grad_norm": 0.6171875,
+      "learning_rate": 8.31116285649658e-08,
+      "loss": 1.3755,
+      "step": 11118
+    },
+    {
+      "epoch": 1.918226516001035,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.276180051662641e-08,
+      "loss": 1.4617,
+      "step": 11119
+    },
+    {
+      "epoch": 1.918399033899767,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.241270719925865e-08,
+      "loss": 1.414,
+      "step": 11120
+    },
+    {
+      "epoch": 1.9185715517984991,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.206434863872514e-08,
+      "loss": 1.4393,
+      "step": 11121
+    },
+    {
+      "epoch": 1.918744069697231,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.171672486083526e-08,
+      "loss": 1.3922,
+      "step": 11122
+    },
+    {
+      "epoch": 1.918916587595963,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.136983589134173e-08,
+      "loss": 1.4418,
+      "step": 11123
+    },
+    {
+      "epoch": 1.919089105494695,
+      "grad_norm": 0.5703125,
+      "learning_rate": 8.102368175594733e-08,
+      "loss": 1.4394,
+      "step": 11124
+    },
+    {
+      "epoch": 1.9192616233934272,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.067826248029264e-08,
+      "loss": 1.3664,
+      "step": 11125
+    },
+    {
+      "epoch": 1.9194341412921592,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.033357808997278e-08,
+      "loss": 1.4185,
+      "step": 11126
+    },
+    {
+      "epoch": 1.9196066591908911,
+      "grad_norm": 0.61328125,
+      "learning_rate": 7.998962861052173e-08,
+      "loss": 1.3817,
+      "step": 11127
+    },
+    {
+      "epoch": 1.919779177089623,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.964641406742135e-08,
+      "loss": 1.3643,
+      "step": 11128
+    },
+    {
+      "epoch": 1.919951694988355,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.93039344861013e-08,
+      "loss": 1.4467,
+      "step": 11129
+    },
+    {
+      "epoch": 1.920124212887087,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.896218989193239e-08,
+      "loss": 1.4712,
+      "step": 11130
+    },
+    {
+      "epoch": 1.920296730785819,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.862118031023436e-08,
+      "loss": 1.4224,
+      "step": 11131
+    },
+    {
+      "epoch": 1.920469248684551,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.828090576627034e-08,
+      "loss": 1.4686,
+      "step": 11132
+    },
+    {
+      "epoch": 1.920641766583283,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.794136628525129e-08,
+      "loss": 1.4157,
+      "step": 11133
+    },
+    {
+      "epoch": 1.9208142844820149,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.760256189233151e-08,
+      "loss": 1.4661,
+      "step": 11134
+    },
+    {
+      "epoch": 1.920986802380747,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.726449261261205e-08,
+      "loss": 1.4879,
+      "step": 11135
+    },
+    {
+      "epoch": 1.921159320279479,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.69271584711384e-08,
+      "loss": 1.424,
+      "step": 11136
+    },
+    {
+      "epoch": 1.921331838178211,
+      "grad_norm": 0.6328125,
+      "learning_rate": 7.659055949290395e-08,
+      "loss": 1.4365,
+      "step": 11137
+    },
+    {
+      "epoch": 1.9215043560769431,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.625469570284427e-08,
+      "loss": 1.4613,
+      "step": 11138
+    },
+    {
+      "epoch": 1.921676873975675,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.591956712584392e-08,
+      "loss": 1.5346,
+      "step": 11139
+    },
+    {
+      "epoch": 1.921849391874407,
+      "grad_norm": 1.28125,
+      "learning_rate": 7.55851737867297e-08,
+      "loss": 1.4781,
+      "step": 11140
+    },
+    {
+      "epoch": 1.922021909773139,
+      "grad_norm": 0.57421875,
+      "learning_rate": 7.525151571027734e-08,
+      "loss": 1.4212,
+      "step": 11141
+    },
+    {
+      "epoch": 1.922194427671871,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.491859292120484e-08,
+      "loss": 1.4737,
+      "step": 11142
+    },
+    {
+      "epoch": 1.922366945570603,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.458640544417806e-08,
+      "loss": 1.4486,
+      "step": 11143
+    },
+    {
+      "epoch": 1.922539463469335,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.425495330380617e-08,
+      "loss": 1.4661,
+      "step": 11144
+    },
+    {
+      "epoch": 1.9227119813680669,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.392423652464731e-08,
+      "loss": 1.4051,
+      "step": 11145
+    },
+    {
+      "epoch": 1.9228844992667988,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.3594255131203e-08,
+      "loss": 1.4672,
+      "step": 11146
+    },
+    {
+      "epoch": 1.9230570171655308,
+      "grad_norm": 0.625,
+      "learning_rate": 7.326500914791701e-08,
+      "loss": 1.3742,
+      "step": 11147
+    },
+    {
+      "epoch": 1.9232295350642628,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.29364985991865e-08,
+      "loss": 1.3044,
+      "step": 11148
+    },
+    {
+      "epoch": 1.923402052962995,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.260872350934533e-08,
+      "loss": 1.3507,
+      "step": 11149
+    },
+    {
+      "epoch": 1.923574570861727,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.228168390268075e-08,
+      "loss": 1.3799,
+      "step": 11150
+    },
+    {
+      "epoch": 1.9237470887604589,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.195537980341894e-08,
+      "loss": 1.4878,
+      "step": 11151
+    },
+    {
+      "epoch": 1.923919606659191,
+      "grad_norm": 0.5703125,
+      "learning_rate": 7.162981123573609e-08,
+      "loss": 1.4538,
+      "step": 11152
+    },
+    {
+      "epoch": 1.924092124557923,
+      "grad_norm": 0.59375,
+      "learning_rate": 7.130497822375293e-08,
+      "loss": 1.5007,
+      "step": 11153
+    },
+    {
+      "epoch": 1.924264642456655,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.098088079153353e-08,
+      "loss": 1.5421,
+      "step": 11154
+    },
+    {
+      "epoch": 1.924437160355387,
+      "grad_norm": 1.125,
+      "learning_rate": 7.065751896309092e-08,
+      "loss": 1.3665,
+      "step": 11155
+    },
+    {
+      "epoch": 1.9246096782541189,
+      "grad_norm": 0.625,
+      "learning_rate": 7.033489276238037e-08,
+      "loss": 1.4511,
+      "step": 11156
+    },
+    {
+      "epoch": 1.9247821961528508,
+      "grad_norm": 0.609375,
+      "learning_rate": 7.001300221330387e-08,
+      "loss": 1.398,
+      "step": 11157
+    },
+    {
+      "epoch": 1.9249547140515828,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.969184733970902e-08,
+      "loss": 1.3415,
+      "step": 11158
+    },
+    {
+      "epoch": 1.9251272319503148,
+      "grad_norm": 0.63671875,
+      "learning_rate": 6.937142816539121e-08,
+      "loss": 1.3978,
+      "step": 11159
+    },
+    {
+      "epoch": 1.9252997498490467,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.905174471408594e-08,
+      "loss": 1.4171,
+      "step": 11160
+    },
+    {
+      "epoch": 1.9254722677477787,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.873279700947977e-08,
+      "loss": 1.4101,
+      "step": 11161
+    },
+    {
+      "epoch": 1.9256447856465109,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.84145850752016e-08,
+      "loss": 1.4081,
+      "step": 11162
+    },
+    {
+      "epoch": 1.9258173035452428,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.809710893482591e-08,
+      "loss": 1.367,
+      "step": 11163
+    },
+    {
+      "epoch": 1.9259898214439748,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.778036861187277e-08,
+      "loss": 1.4203,
+      "step": 11164
+    },
+    {
+      "epoch": 1.9261623393427068,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.746436412981117e-08,
+      "loss": 1.5098,
+      "step": 11165
+    },
+    {
+      "epoch": 1.926334857241439,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.714909551204907e-08,
+      "loss": 1.3379,
+      "step": 11166
+    },
+    {
+      "epoch": 1.926507375140171,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.683456278194666e-08,
+      "loss": 1.4345,
+      "step": 11167
+    },
+    {
+      "epoch": 1.9266798930389029,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.652076596280422e-08,
+      "loss": 1.4645,
+      "step": 11168
+    },
+    {
+      "epoch": 1.9268524109376348,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.6207705077872e-08,
+      "loss": 1.4089,
+      "step": 11169
+    },
+    {
+      "epoch": 1.9270249288363668,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.589538015034148e-08,
+      "loss": 1.3834,
+      "step": 11170
+    },
+    {
+      "epoch": 1.9271974467350987,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.55837912033519e-08,
+      "loss": 1.4337,
+      "step": 11171
+    },
+    {
+      "epoch": 1.9273699646338307,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.527293825998815e-08,
+      "loss": 1.4587,
+      "step": 11172
+    },
+    {
+      "epoch": 1.9275424825325627,
+      "grad_norm": 0.60546875,
+      "learning_rate": 6.496282134328069e-08,
+      "loss": 1.4673,
+      "step": 11173
+    },
+    {
+      "epoch": 1.9277150004312946,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.465344047620336e-08,
+      "loss": 1.416,
+      "step": 11174
+    },
+    {
+      "epoch": 1.9278875183300266,
+      "grad_norm": 0.61328125,
+      "learning_rate": 6.434479568167896e-08,
+      "loss": 1.4711,
+      "step": 11175
+    },
+    {
+      "epoch": 1.9280600362287588,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.40368869825736e-08,
+      "loss": 1.4945,
+      "step": 11176
+    },
+    {
+      "epoch": 1.9282325541274907,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.372971440169684e-08,
+      "loss": 1.3634,
+      "step": 11177
+    },
+    {
+      "epoch": 1.9284050720262227,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.342327796180936e-08,
+      "loss": 1.4011,
+      "step": 11178
+    },
+    {
+      "epoch": 1.9285775899249549,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.311757768560967e-08,
+      "loss": 1.4638,
+      "step": 11179
+    },
+    {
+      "epoch": 1.9287501078236868,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.281261359575074e-08,
+      "loss": 1.4616,
+      "step": 11180
+    },
+    {
+      "epoch": 1.9289226257224188,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.250838571482231e-08,
+      "loss": 1.5161,
+      "step": 11181
+    },
+    {
+      "epoch": 1.9290951436211508,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.220489406536523e-08,
+      "loss": 1.3275,
+      "step": 11182
+    },
+    {
+      "epoch": 1.9292676615198827,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.190213866986483e-08,
+      "loss": 1.4386,
+      "step": 11183
+    },
+    {
+      "epoch": 1.9294401794186147,
+      "grad_norm": 0.6171875,
+      "learning_rate": 6.160011955074874e-08,
+      "loss": 1.4496,
+      "step": 11184
+    },
+    {
+      "epoch": 1.9296126973173466,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.12988367303946e-08,
+      "loss": 1.3581,
+      "step": 11185
+    },
+    {
+      "epoch": 1.9297852152160786,
+      "grad_norm": 0.5703125,
+      "learning_rate": 6.099829023112236e-08,
+      "loss": 1.4737,
+      "step": 11186
+    },
+    {
+      "epoch": 1.9299577331148106,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.069848007519863e-08,
+      "loss": 1.3849,
+      "step": 11187
+    },
+    {
+      "epoch": 1.9301302510135425,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.039940628483454e-08,
+      "loss": 1.4742,
+      "step": 11188
+    },
+    {
+      "epoch": 1.9303027689122747,
+      "grad_norm": 0.6875,
+      "learning_rate": 6.010106888218792e-08,
+      "loss": 1.4461,
+      "step": 11189
+    },
+    {
+      "epoch": 1.9304752868110067,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.98034678893622e-08,
+      "loss": 1.4375,
+      "step": 11190
+    },
+    {
+      "epoch": 1.9306478047097386,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.950660332840419e-08,
+      "loss": 1.3679,
+      "step": 11191
+    },
+    {
+      "epoch": 1.9308203226084706,
+      "grad_norm": 0.81640625,
+      "learning_rate": 5.9210475221308515e-08,
+      "loss": 1.4385,
+      "step": 11192
+    },
+    {
+      "epoch": 1.9309928405072028,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.8915083590013186e-08,
+      "loss": 1.4601,
+      "step": 11193
+    },
+    {
+      "epoch": 1.9311653584059347,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.862042845640403e-08,
+      "loss": 1.3369,
+      "step": 11194
+    },
+    {
+      "epoch": 1.9313378763046667,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.832650984231025e-08,
+      "loss": 1.4013,
+      "step": 11195
+    },
+    {
+      "epoch": 1.9315103942033987,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.8033327769505546e-08,
+      "loss": 1.4503,
+      "step": 11196
+    },
+    {
+      "epoch": 1.9316829121021306,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.774088225971364e-08,
+      "loss": 1.264,
+      "step": 11197
+    },
+    {
+      "epoch": 1.9318554300008626,
+      "grad_norm": 0.83984375,
+      "learning_rate": 5.7449173334598316e-08,
+      "loss": 1.4681,
+      "step": 11198
+    },
+    {
+      "epoch": 1.9320279478995945,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.7158201015773404e-08,
+      "loss": 1.5123,
+      "step": 11199
+    },
+    {
+      "epoch": 1.9322004657983265,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.6867965324793886e-08,
+      "loss": 1.3687,
+      "step": 11200
+    },
+    {
+      "epoch": 1.9322004657983265,
+      "eval_loss": 1.4070696830749512,
+      "eval_runtime": 10.8729,
+      "eval_samples_per_second": 94.179,
+      "eval_steps_per_second": 23.545,
+      "step": 11200
+    },
+    {
+      "epoch": 1.9323729836970585,
+      "grad_norm": 0.65234375,
+      "learning_rate": 5.657846628316366e-08,
+      "loss": 1.3929,
+      "step": 11201
+    },
+    {
+      "epoch": 1.9325455015957904,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.628970391232891e-08,
+      "loss": 1.5419,
+      "step": 11202
+    },
+    {
+      "epoch": 1.9327180194945226,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.600167823368474e-08,
+      "loss": 1.3657,
+      "step": 11203
+    },
+    {
+      "epoch": 1.9328905373932546,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.571438926856964e-08,
+      "loss": 1.4119,
+      "step": 11204
+    },
+    {
+      "epoch": 1.9330630552919865,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.5427837038266595e-08,
+      "loss": 1.409,
+      "step": 11205
+    },
+    {
+      "epoch": 1.9332355731907187,
+      "grad_norm": 0.6640625,
+      "learning_rate": 5.5142021564006386e-08,
+      "loss": 1.6224,
+      "step": 11206
+    },
+    {
+      "epoch": 1.9334080910894507,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.485694286696319e-08,
+      "loss": 1.4667,
+      "step": 11207
+    },
+    {
+      "epoch": 1.9335806089881826,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.4572600968257894e-08,
+      "loss": 1.3575,
+      "step": 11208
+    },
+    {
+      "epoch": 1.9337531268869146,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.428899588895586e-08,
+      "loss": 1.4499,
+      "step": 11209
+    },
+    {
+      "epoch": 1.9339256447856465,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.4006127650069185e-08,
+      "loss": 1.3383,
+      "step": 11210
+    },
+    {
+      "epoch": 1.9340981626843785,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.372399627255442e-08,
+      "loss": 1.5185,
+      "step": 11211
+    },
+    {
+      "epoch": 1.9342706805831105,
+      "grad_norm": 0.6015625,
+      "learning_rate": 5.344260177731264e-08,
+      "loss": 1.4495,
+      "step": 11212
+    },
+    {
+      "epoch": 1.9344431984818424,
+      "grad_norm": 0.58203125,
+      "learning_rate": 5.31619441851916e-08,
+      "loss": 1.4181,
+      "step": 11213
+    },
+    {
+      "epoch": 1.9346157163805744,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.288202351698468e-08,
+      "loss": 1.4742,
+      "step": 11214
+    },
+    {
+      "epoch": 1.9347882342793064,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.260283979343084e-08,
+      "loss": 1.4408,
+      "step": 11215
+    },
+    {
+      "epoch": 1.9349607521780383,
+      "grad_norm": 0.5625,
+      "learning_rate": 5.232439303521131e-08,
+      "loss": 1.5229,
+      "step": 11216
+    },
+    {
+      "epoch": 1.9351332700767705,
+      "grad_norm": 0.64453125,
+      "learning_rate": 5.2046683262957366e-08,
+      "loss": 1.441,
+      "step": 11217
+    },
+    {
+      "epoch": 1.9353057879755025,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.1769710497243664e-08,
+      "loss": 1.4484,
+      "step": 11218
+    },
+    {
+      "epoch": 1.9354783058742344,
+      "grad_norm": 0.61328125,
+      "learning_rate": 5.149347475858824e-08,
+      "loss": 1.4355,
+      "step": 11219
+    },
+    {
+      "epoch": 1.9356508237729666,
+      "grad_norm": 0.5859375,
+      "learning_rate": 5.121797606745804e-08,
+      "loss": 1.3858,
+      "step": 11220
+    },
+    {
+      "epoch": 1.9358233416716986,
+      "grad_norm": 0.59375,
+      "learning_rate": 5.094321444426231e-08,
+      "loss": 1.4759,
+      "step": 11221
+    },
+    {
+      "epoch": 1.9359958595704305,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.0669189909358094e-08,
+      "loss": 1.4333,
+      "step": 11222
+    },
+    {
+      "epoch": 1.9361683774691625,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.0395902483046934e-08,
+      "loss": 1.4954,
+      "step": 11223
+    },
+    {
+      "epoch": 1.9363408953678944,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.012335218557374e-08,
+      "loss": 1.3914,
+      "step": 11224
+    },
+    {
+      "epoch": 1.9365134132666264,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.985153903713458e-08,
+      "loss": 1.3743,
+      "step": 11225
+    },
+    {
+      "epoch": 1.9366859311653584,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.9580463057863345e-08,
+      "loss": 1.5157,
+      "step": 11226
+    },
+    {
+      "epoch": 1.9368584490640903,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.9310124267845095e-08,
+      "loss": 1.4125,
+      "step": 11227
+    },
+    {
+      "epoch": 1.9370309669628223,
+      "grad_norm": 0.55078125,
+      "learning_rate": 4.904052268710713e-08,
+      "loss": 1.4332,
+      "step": 11228
+    },
+    {
+      "epoch": 1.9372034848615542,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.8771658335623476e-08,
+      "loss": 1.4508,
+      "step": 11229
+    },
+    {
+      "epoch": 1.9373760027602864,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.850353123331486e-08,
+      "loss": 1.4796,
+      "step": 11230
+    },
+    {
+      "epoch": 1.9375485206590184,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.823614140004429e-08,
+      "loss": 1.4209,
+      "step": 11231
+    },
+    {
+      "epoch": 1.9377210385577504,
+      "grad_norm": 0.58984375,
+      "learning_rate": 4.796948885562036e-08,
+      "loss": 1.3682,
+      "step": 11232
+    },
+    {
+      "epoch": 1.9378935564564823,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.7703573619800604e-08,
+      "loss": 1.3958,
+      "step": 11233
+    },
+    {
+      "epoch": 1.9380660743552145,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.743839571228592e-08,
+      "loss": 1.4617,
+      "step": 11234
+    },
+    {
+      "epoch": 1.9382385922539465,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.7173955152719496e-08,
+      "loss": 1.4021,
+      "step": 11235
+    },
+    {
+      "epoch": 1.9384111101526784,
+      "grad_norm": 4.59375,
+      "learning_rate": 4.6910251960695655e-08,
+      "loss": 1.43,
+      "step": 11236
+    },
+    {
+      "epoch": 1.9385836280514104,
+      "grad_norm": 0.640625,
+      "learning_rate": 4.664728615574987e-08,
+      "loss": 1.5197,
+      "step": 11237
+    },
+    {
+      "epoch": 1.9387561459501423,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.638505775736546e-08,
+      "loss": 1.4641,
+      "step": 11238
+    },
+    {
+      "epoch": 1.9389286638488743,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.612356678496799e-08,
+      "loss": 1.4546,
+      "step": 11239
+    },
+    {
+      "epoch": 1.9391011817476063,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.5862813257931957e-08,
+      "loss": 1.4626,
+      "step": 11240
+    },
+    {
+      "epoch": 1.9392736996463382,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.5602797195574143e-08,
+      "loss": 1.4488,
+      "step": 11241
+    },
+    {
+      "epoch": 1.9394462175450702,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.534351861716024e-08,
+      "loss": 1.4772,
+      "step": 11242
+    },
+    {
+      "epoch": 1.9396187354438021,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.5084977541897116e-08,
+      "loss": 1.3595,
+      "step": 11243
+    },
+    {
+      "epoch": 1.9397912533425343,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.482717398894165e-08,
+      "loss": 1.303,
+      "step": 11244
+    },
+    {
+      "epoch": 1.9399637712412663,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.4570107977389696e-08,
+      "loss": 1.3287,
+      "step": 11245
+    },
+    {
+      "epoch": 1.9401362891399982,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.4313779526290457e-08,
+      "loss": 1.4473,
+      "step": 11246
+    },
+    {
+      "epoch": 1.9403088070387304,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.4058188654630965e-08,
+      "loss": 1.3413,
+      "step": 11247
+    },
+    {
+      "epoch": 1.9404813249374624,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.3803335381349396e-08,
+      "loss": 1.4993,
+      "step": 11248
+    },
+    {
+      "epoch": 1.9406538428361944,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.354921972532511e-08,
+      "loss": 1.3871,
+      "step": 11249
+    },
+    {
+      "epoch": 1.9408263607349263,
+      "grad_norm": 0.625,
+      "learning_rate": 4.3295841705386365e-08,
+      "loss": 1.3752,
+      "step": 11250
+    },
+    {
+      "epoch": 1.9409988786336583,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.30432013403026e-08,
+      "loss": 1.4766,
+      "step": 11251
+    },
+    {
+      "epoch": 1.9411713965323902,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.279129864879439e-08,
+      "loss": 1.3649,
+      "step": 11252
+    },
+    {
+      "epoch": 1.9413439144311222,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.2540133649520145e-08,
+      "loss": 1.4277,
+      "step": 11253
+    },
+    {
+      "epoch": 1.9415164323298542,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.2289706361091643e-08,
+      "loss": 1.4751,
+      "step": 11254
+    },
+    {
+      "epoch": 1.9416889502285861,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.2040016802059604e-08,
+      "loss": 1.4691,
+      "step": 11255
+    },
+    {
+      "epoch": 1.941861468127318,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.179106499092367e-08,
+      "loss": 1.4345,
+      "step": 11256
+    },
+    {
+      "epoch": 1.94203398602605,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.1542850946126864e-08,
+      "loss": 1.359,
+      "step": 11257
+    },
+    {
+      "epoch": 1.9422065039247822,
+      "grad_norm": 0.578125,
+      "learning_rate": 4.129537468605893e-08,
+      "loss": 1.4108,
+      "step": 11258
+    },
+    {
+      "epoch": 1.9423790218235142,
+      "grad_norm": 0.59375,
+      "learning_rate": 4.1048636229055194e-08,
+      "loss": 1.4584,
+      "step": 11259
+    },
+    {
+      "epoch": 1.9425515397222461,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.080263559339437e-08,
+      "loss": 1.4231,
+      "step": 11260
+    },
+    {
+      "epoch": 1.9427240576209783,
+      "grad_norm": 1.4453125,
+      "learning_rate": 4.0557372797302984e-08,
+      "loss": 1.3745,
+      "step": 11261
+    },
+    {
+      "epoch": 1.9428965755197103,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.0312847858949846e-08,
+      "loss": 1.4831,
+      "step": 11262
+    },
+    {
+      "epoch": 1.9430690934184423,
+      "grad_norm": 0.53515625,
+      "learning_rate": 4.006906079645267e-08,
+      "loss": 1.3649,
+      "step": 11263
+    },
+    {
+      "epoch": 1.9432416113171742,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.982601162787147e-08,
+      "loss": 1.4526,
+      "step": 11264
+    },
+    {
+      "epoch": 1.9434141292159062,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.9583700371214064e-08,
+      "loss": 1.4766,
+      "step": 11265
+    },
+    {
+      "epoch": 1.9435866471146381,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.9342127044430524e-08,
+      "loss": 1.4487,
+      "step": 11266
+    },
+    {
+      "epoch": 1.94375916501337,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.910129166541987e-08,
+      "loss": 1.4204,
+      "step": 11267
+    },
+    {
+      "epoch": 1.943931682912102,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.8861194252024504e-08,
+      "loss": 1.3871,
+      "step": 11268
+    },
+    {
+      "epoch": 1.944104200810834,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.86218348220313e-08,
+      "loss": 1.4365,
+      "step": 11269
+    },
+    {
+      "epoch": 1.944276718709566,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.8383213393174965e-08,
+      "loss": 1.4831,
+      "step": 11270
+    },
+    {
+      "epoch": 1.9444492366082982,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.814532998313247e-08,
+      "loss": 1.4162,
+      "step": 11271
+    },
+    {
+      "epoch": 1.9446217545070301,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.790818460952861e-08,
+      "loss": 1.356,
+      "step": 11272
+    },
+    {
+      "epoch": 1.944794272405762,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.7671777289932654e-08,
+      "loss": 1.4233,
+      "step": 11273
+    },
+    {
+      "epoch": 1.944966790304494,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.743610804185949e-08,
+      "loss": 1.4831,
+      "step": 11274
+    },
+    {
+      "epoch": 1.9451393082032262,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.720117688276737e-08,
+      "loss": 1.3817,
+      "step": 11275
+    },
+    {
+      "epoch": 1.9453118261019582,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.6966983830063477e-08,
+      "loss": 1.3679,
+      "step": 11276
+    },
+    {
+      "epoch": 1.9454843440006901,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.673352890109616e-08,
+      "loss": 1.4826,
+      "step": 11277
+    },
+    {
+      "epoch": 1.945656861899422,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.650081211316381e-08,
+      "loss": 1.4326,
+      "step": 11278
+    },
+    {
+      "epoch": 1.945829379798154,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.626883348350485e-08,
+      "loss": 1.4831,
+      "step": 11279
+    },
+    {
+      "epoch": 1.946001897696886,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.603759302930776e-08,
+      "loss": 1.4117,
+      "step": 11280
+    },
+    {
+      "epoch": 1.946174415595618,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.5807090767703276e-08,
+      "loss": 1.4233,
+      "step": 11281
+    },
+    {
+      "epoch": 1.94634693349435,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.557732671576885e-08,
+      "loss": 1.4643,
+      "step": 11282
+    },
+    {
+      "epoch": 1.946519451393082,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.534830089052532e-08,
+      "loss": 1.4645,
+      "step": 11283
+    },
+    {
+      "epoch": 1.9466919692918139,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.512001330894355e-08,
+      "loss": 1.3985,
+      "step": 11284
+    },
+    {
+      "epoch": 1.946864487190546,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.4892463987933335e-08,
+      "loss": 1.4634,
+      "step": 11285
+    },
+    {
+      "epoch": 1.947037005089278,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.4665652944355646e-08,
+      "loss": 1.4396,
+      "step": 11286
+    },
+    {
+      "epoch": 1.94720952298801,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.443958019501148e-08,
+      "loss": 1.5846,
+      "step": 11287
+    },
+    {
+      "epoch": 1.9473820408867422,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.421424575665078e-08,
+      "loss": 1.4767,
+      "step": 11288
+    },
+    {
+      "epoch": 1.9475545587854741,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.398964964596907e-08,
+      "loss": 1.524,
+      "step": 11289
+    },
+    {
+      "epoch": 1.947727076684206,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.376579187960305e-08,
+      "loss": 1.3998,
+      "step": 11290
+    },
+    {
+      "epoch": 1.947899594582938,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.354267247414056e-08,
+      "loss": 1.4296,
+      "step": 11291
+    },
+    {
+      "epoch": 1.94807211248167,
+      "grad_norm": 0.66796875,
+      "learning_rate": 3.332029144610949e-08,
+      "loss": 1.4121,
+      "step": 11292
+    },
+    {
+      "epoch": 1.948244630380402,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.309864881198555e-08,
+      "loss": 1.4809,
+      "step": 11293
+    },
+    {
+      "epoch": 1.948417148279134,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.2877744588190044e-08,
+      "loss": 1.3625,
+      "step": 11294
+    },
+    {
+      "epoch": 1.9485896661778659,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.2657578791088775e-08,
+      "loss": 1.4392,
+      "step": 11295
+    },
+    {
+      "epoch": 1.9487621840765978,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.243815143699314e-08,
+      "loss": 1.3932,
+      "step": 11296
+    },
+    {
+      "epoch": 1.9489347019753298,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.2219462542159016e-08,
+      "loss": 1.4316,
+      "step": 11297
+    },
+    {
+      "epoch": 1.9491072198740618,
+      "grad_norm": 0.55859375,
+      "learning_rate": 3.2001512122789014e-08,
+      "loss": 1.5107,
+      "step": 11298
+    },
+    {
+      "epoch": 1.949279737772794,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.17843001950302e-08,
+      "loss": 1.417,
+      "step": 11299
+    },
+    {
+      "epoch": 1.949452255671526,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.1567826774974166e-08,
+      "loss": 1.4208,
+      "step": 11300
+    },
+    {
+      "epoch": 1.949452255671526,
+      "eval_loss": 1.4070621728897095,
+      "eval_runtime": 10.985,
+      "eval_samples_per_second": 93.218,
+      "eval_steps_per_second": 23.305,
+      "step": 11300
+    },
+    {
+      "epoch": 1.9496247735702579,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.135209187865917e-08,
+      "loss": 1.4143,
+      "step": 11301
+    },
+    {
+      "epoch": 1.94979729146899,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.1137095522068006e-08,
+      "loss": 1.4008,
+      "step": 11302
+    },
+    {
+      "epoch": 1.949969809367722,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.092283772113014e-08,
+      "loss": 1.5115,
+      "step": 11303
+    },
+    {
+      "epoch": 1.950142327266454,
+      "grad_norm": 0.60546875,
+      "learning_rate": 3.070931849171732e-08,
+      "loss": 1.3597,
+      "step": 11304
+    },
+    {
+      "epoch": 1.950314845165186,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.049653784964912e-08,
+      "loss": 1.505,
+      "step": 11305
+    },
+    {
+      "epoch": 1.950487363063918,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.028449581068959e-08,
+      "loss": 1.4266,
+      "step": 11306
+    },
+    {
+      "epoch": 1.9506598809626499,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.00731923905484e-08,
+      "loss": 1.4253,
+      "step": 11307
+    },
+    {
+      "epoch": 1.9508323988613818,
+      "grad_norm": 0.625,
+      "learning_rate": 2.986262760488079e-08,
+      "loss": 1.5135,
+      "step": 11308
+    },
+    {
+      "epoch": 1.9510049167601138,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.9652801469285396e-08,
+      "loss": 1.4114,
+      "step": 11309
+    },
+    {
+      "epoch": 1.9511774346588457,
+      "grad_norm": 2.265625,
+      "learning_rate": 2.9443713999308676e-08,
+      "loss": 1.4138,
+      "step": 11310
+    },
+    {
+      "epoch": 1.9513499525575777,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.923536521044046e-08,
+      "loss": 1.3957,
+      "step": 11311
+    },
+    {
+      "epoch": 1.9515224704563099,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.9027755118116175e-08,
+      "loss": 1.447,
+      "step": 11312
+    },
+    {
+      "epoch": 1.9516949883550418,
+      "grad_norm": 0.54296875,
+      "learning_rate": 2.8820883737716853e-08,
+      "loss": 1.3001,
+      "step": 11313
+    },
+    {
+      "epoch": 1.9518675062537738,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.8614751084570236e-08,
+      "loss": 1.4411,
+      "step": 11314
+    },
+    {
+      "epoch": 1.9520400241525058,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.8409357173946327e-08,
+      "loss": 1.3716,
+      "step": 11315
+    },
+    {
+      "epoch": 1.952212542051238,
+      "grad_norm": 1.4140625,
+      "learning_rate": 2.8204702021062958e-08,
+      "loss": 1.5538,
+      "step": 11316
+    },
+    {
+      "epoch": 1.95238505994997,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.800078564108133e-08,
+      "loss": 1.3402,
+      "step": 11317
+    },
+    {
+      "epoch": 1.9525575778487019,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.779760804911047e-08,
+      "loss": 1.3476,
+      "step": 11318
+    },
+    {
+      "epoch": 1.9527300957474338,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.759516926020056e-08,
+      "loss": 1.3268,
+      "step": 11319
+    },
+    {
+      "epoch": 1.9529026136461658,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.7393469289351825e-08,
+      "loss": 1.4073,
+      "step": 11320
+    },
+    {
+      "epoch": 1.9530751315448978,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.7192508151506758e-08,
+      "loss": 1.412,
+      "step": 11321
+    },
+    {
+      "epoch": 1.9532476494436297,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.6992285861553447e-08,
+      "loss": 1.5015,
+      "step": 11322
+    },
+    {
+      "epoch": 1.9534201673423617,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.6792802434326692e-08,
+      "loss": 1.5071,
+      "step": 11323
+    },
+    {
+      "epoch": 1.9535926852410936,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.6594057884603565e-08,
+      "loss": 1.3677,
+      "step": 11324
+    },
+    {
+      "epoch": 1.9537652031398256,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.639605222710895e-08,
+      "loss": 1.4214,
+      "step": 11325
+    },
+    {
+      "epoch": 1.9539377210385578,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.6198785476513333e-08,
+      "loss": 1.3839,
+      "step": 11326
+    },
+    {
+      "epoch": 1.9541102389372897,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.6002257647431694e-08,
+      "loss": 1.5533,
+      "step": 11327
+    },
+    {
+      "epoch": 1.9542827568360217,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.5806468754422388e-08,
+      "loss": 1.3973,
+      "step": 11328
+    },
+    {
+      "epoch": 1.954455274734754,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.5611418811991586e-08,
+      "loss": 1.4482,
+      "step": 11329
+    },
+    {
+      "epoch": 1.9546277926334858,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.541710783458884e-08,
+      "loss": 1.4284,
+      "step": 11330
+    },
+    {
+      "epoch": 1.9548003105322178,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.5223535836612634e-08,
+      "loss": 1.4421,
+      "step": 11331
+    },
+    {
+      "epoch": 1.9549728284309498,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.503070283240039e-08,
+      "loss": 1.4014,
+      "step": 11332
+    },
+    {
+      "epoch": 1.9551453463296817,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.4838608836241783e-08,
+      "loss": 1.3639,
+      "step": 11333
+    },
+    {
+      "epoch": 1.9553178642284137,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.4647253862365438e-08,
+      "loss": 1.4609,
+      "step": 11334
+    },
+    {
+      "epoch": 1.9554903821271457,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.445663792495001e-08,
+      "loss": 1.3816,
+      "step": 11335
+    },
+    {
+      "epoch": 1.9556629000258776,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.4266761038116428e-08,
+      "loss": 1.418,
+      "step": 11336
+    },
+    {
+      "epoch": 1.9558354179246096,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.4077623215933432e-08,
+      "loss": 1.495,
+      "step": 11337
+    },
+    {
+      "epoch": 1.9560079358233415,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.388922447241204e-08,
+      "loss": 1.4245,
+      "step": 11338
+    },
+    {
+      "epoch": 1.9561804537220737,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.370156482150998e-08,
+      "loss": 1.4573,
+      "step": 11339
+    },
+    {
+      "epoch": 1.9563529716208057,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.3514644277131682e-08,
+      "loss": 1.4742,
+      "step": 11340
+    },
+    {
+      "epoch": 1.9565254895195376,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.3328462853123846e-08,
+      "loss": 1.5041,
+      "step": 11341
+    },
+    {
+      "epoch": 1.9566980074182696,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.3143020563280993e-08,
+      "loss": 1.4036,
+      "step": 11342
+    },
+    {
+      "epoch": 1.9568705253170018,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.2958317421341026e-08,
+      "loss": 1.3942,
+      "step": 11343
+    },
+    {
+      "epoch": 1.9570430432157337,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.277435344098855e-08,
+      "loss": 1.4601,
+      "step": 11344
+    },
+    {
+      "epoch": 1.9572155611144657,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.2591128635852666e-08,
+      "loss": 1.3396,
+      "step": 11345
+    },
+    {
+      "epoch": 1.9573880790131977,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.240864301950807e-08,
+      "loss": 1.4657,
+      "step": 11346
+    },
+    {
+      "epoch": 1.9575605969119296,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.2226896605473945e-08,
+      "loss": 1.4744,
+      "step": 11347
+    },
+    {
+      "epoch": 1.9577331148106616,
+      "grad_norm": 1.578125,
+      "learning_rate": 2.2045889407215082e-08,
+      "loss": 1.4355,
+      "step": 11348
+    },
+    {
+      "epoch": 1.9579056327093936,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.186562143814186e-08,
+      "loss": 1.3912,
+      "step": 11349
+    },
+    {
+      "epoch": 1.9580781506081255,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.1686092711609154e-08,
+      "loss": 1.4397,
+      "step": 11350
+    },
+    {
+      "epoch": 1.9582506685068575,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.1507303240918543e-08,
+      "loss": 1.4747,
+      "step": 11351
+    },
+    {
+      "epoch": 1.9584231864055894,
+      "grad_norm": 0.625,
+      "learning_rate": 2.132925303931499e-08,
+      "loss": 1.4209,
+      "step": 11352
+    },
+    {
+      "epoch": 1.9585957043043216,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.1151942119991274e-08,
+      "loss": 1.3934,
+      "step": 11353
+    },
+    {
+      "epoch": 1.9587682222030536,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.0975370496081336e-08,
+      "loss": 1.4919,
+      "step": 11354
+    },
+    {
+      "epoch": 1.9589407401017855,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.0799538180668044e-08,
+      "loss": 1.4422,
+      "step": 11355
+    },
+    {
+      "epoch": 1.9591132580005175,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.0624445186777643e-08,
+      "loss": 1.4383,
+      "step": 11356
+    },
+    {
+      "epoch": 1.9592857758992497,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.045009152738309e-08,
+      "loss": 1.4113,
+      "step": 11357
+    },
+    {
+      "epoch": 1.9594582937979816,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.0276477215399604e-08,
+      "loss": 1.3592,
+      "step": 11358
+    },
+    {
+      "epoch": 1.9596308116967136,
+      "grad_norm": 0.5625,
+      "learning_rate": 2.0103602263692455e-08,
+      "loss": 1.4196,
+      "step": 11359
+    },
+    {
+      "epoch": 1.9598033295954456,
+      "grad_norm": 0.65234375,
+      "learning_rate": 1.993146668506585e-08,
+      "loss": 1.4562,
+      "step": 11360
+    },
+    {
+      "epoch": 1.9599758474941775,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.976007049227624e-08,
+      "loss": 1.5406,
+      "step": 11361
+    },
+    {
+      "epoch": 1.9601483653929095,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.9589413698019034e-08,
+      "loss": 1.4636,
+      "step": 11362
+    },
+    {
+      "epoch": 1.9603208832916414,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.9419496314939667e-08,
+      "loss": 1.4362,
+      "step": 11363
+    },
+    {
+      "epoch": 1.9604934011903734,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.925031835562474e-08,
+      "loss": 1.5589,
+      "step": 11364
+    },
+    {
+      "epoch": 1.9606659190891054,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.9081879832608674e-08,
+      "loss": 1.4037,
+      "step": 11365
+    },
+    {
+      "epoch": 1.9608384369878373,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.891418075837037e-08,
+      "loss": 1.5149,
+      "step": 11366
+    },
+    {
+      "epoch": 1.9610109548865695,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.8747221145334337e-08,
+      "loss": 1.4395,
+      "step": 11367
+    },
+    {
+      "epoch": 1.9611834727853015,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.858100100587068e-08,
+      "loss": 1.3103,
+      "step": 11368
+    },
+    {
+      "epoch": 1.9613559906840334,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.841552035229288e-08,
+      "loss": 1.4143,
+      "step": 11369
+    },
+    {
+      "epoch": 1.9615285085827656,
+      "grad_norm": 0.68359375,
+      "learning_rate": 1.8250779196861136e-08,
+      "loss": 1.31,
+      "step": 11370
+    },
+    {
+      "epoch": 1.9617010264814976,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.8086777551780122e-08,
+      "loss": 1.4315,
+      "step": 11371
+    },
+    {
+      "epoch": 1.9618735443802295,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.7923515429201232e-08,
+      "loss": 1.4189,
+      "step": 11372
+    },
+    {
+      "epoch": 1.9620460622789615,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.7760992841219237e-08,
+      "loss": 1.3975,
+      "step": 11373
+    },
+    {
+      "epoch": 1.9622185801776935,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.7599209799874505e-08,
+      "loss": 1.4508,
+      "step": 11374
+    },
+    {
+      "epoch": 1.9623910980764254,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.7438166317153005e-08,
+      "loss": 1.4656,
+      "step": 11375
+    },
+    {
+      "epoch": 1.9625636159751574,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.72778624049863e-08,
+      "loss": 1.3995,
+      "step": 11376
+    },
+    {
+      "epoch": 1.9627361338738893,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.711829807525045e-08,
+      "loss": 1.3984,
+      "step": 11377
+    },
+    {
+      "epoch": 1.9629086517726213,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.6959473339765997e-08,
+      "loss": 1.4435,
+      "step": 11378
+    },
+    {
+      "epoch": 1.9630811696713533,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.6801388210302416e-08,
+      "loss": 1.3658,
+      "step": 11379
+    },
+    {
+      "epoch": 1.9632536875700854,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.6644042698569228e-08,
+      "loss": 1.4498,
+      "step": 11380
+    },
+    {
+      "epoch": 1.9634262054688174,
+      "grad_norm": 0.625,
+      "learning_rate": 1.648743681622378e-08,
+      "loss": 1.4733,
+      "step": 11381
+    },
+    {
+      "epoch": 1.9635987233675494,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.6331570574869005e-08,
+      "loss": 1.4511,
+      "step": 11382
+    },
+    {
+      "epoch": 1.9637712412662813,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.6176443986052337e-08,
+      "loss": 1.4096,
+      "step": 11383
+    },
+    {
+      "epoch": 1.9639437591650135,
+      "grad_norm": 0.625,
+      "learning_rate": 1.6022057061266804e-08,
+      "loss": 1.4136,
+      "step": 11384
+    },
+    {
+      "epoch": 1.9641162770637455,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.5868409811949926e-08,
+      "loss": 1.435,
+      "step": 11385
+    },
+    {
+      "epoch": 1.9642887949624774,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.5715502249484816e-08,
+      "loss": 1.3491,
+      "step": 11386
+    },
+    {
+      "epoch": 1.9644613128612094,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.55633343852013e-08,
+      "loss": 1.339,
+      "step": 11387
+    },
+    {
+      "epoch": 1.9646338307599414,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.5411906230370366e-08,
+      "loss": 1.3539,
+      "step": 11388
+    },
+    {
+      "epoch": 1.9648063486586733,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.5261217796211923e-08,
+      "loss": 1.4352,
+      "step": 11389
+    },
+    {
+      "epoch": 1.9649788665574053,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.5111269093890378e-08,
+      "loss": 1.4615,
+      "step": 11390
+    },
+    {
+      "epoch": 1.9651513844561372,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.4962060134513512e-08,
+      "loss": 1.3839,
+      "step": 11391
+    },
+    {
+      "epoch": 1.9653239023548692,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.4813590929138032e-08,
+      "loss": 1.5242,
+      "step": 11392
+    },
+    {
+      "epoch": 1.9654964202536012,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.4665861488761813e-08,
+      "loss": 1.3994,
+      "step": 11393
+    },
+    {
+      "epoch": 1.9656689381523333,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.4518871824329428e-08,
+      "loss": 1.354,
+      "step": 11394
+    },
+    {
+      "epoch": 1.9658414560510653,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.4372621946731058e-08,
+      "loss": 1.4321,
+      "step": 11395
+    },
+    {
+      "epoch": 1.9660139739497973,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.4227111866802479e-08,
+      "loss": 1.307,
+      "step": 11396
+    },
+    {
+      "epoch": 1.9661864918485294,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.4082341595322846e-08,
+      "loss": 1.4024,
+      "step": 11397
+    },
+    {
+      "epoch": 1.9663590097472614,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.3938311143018024e-08,
+      "loss": 1.3452,
+      "step": 11398
+    },
+    {
+      "epoch": 1.9665315276459934,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.3795020520559477e-08,
+      "loss": 1.3352,
+      "step": 11399
+    },
+    {
+      "epoch": 1.9667040455447253,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.3652469738562046e-08,
+      "loss": 1.3326,
+      "step": 11400
+    },
+    {
+      "epoch": 1.9667040455447253,
+      "eval_loss": 1.407050371170044,
+      "eval_runtime": 10.8811,
+      "eval_samples_per_second": 94.108,
+      "eval_steps_per_second": 23.527,
+      "step": 11400
+    },
+    {
+      "epoch": 1.9668765634434573,
+      "grad_norm": 0.625,
+      "learning_rate": 1.3510658807588394e-08,
+      "loss": 1.3218,
+      "step": 11401
+    },
+    {
+      "epoch": 1.9670490813421893,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.3369587738142343e-08,
+      "loss": 1.3314,
+      "step": 11402
+    },
+    {
+      "epoch": 1.9672215992409212,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.3229256540676638e-08,
+      "loss": 1.4445,
+      "step": 11403
+    },
+    {
+      "epoch": 1.9673941171396532,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.3089665225588522e-08,
+      "loss": 1.3972,
+      "step": 11404
+    },
+    {
+      "epoch": 1.9675666350383851,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.295081380321861e-08,
+      "loss": 1.4608,
+      "step": 11405
+    },
+    {
+      "epoch": 1.967739152937117,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.2812702283855338e-08,
+      "loss": 1.4053,
+      "step": 11406
+    },
+    {
+      "epoch": 1.967911670835849,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.2675330677729413e-08,
+      "loss": 1.4719,
+      "step": 11407
+    },
+    {
+      "epoch": 1.9680841887345812,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.253869899501825e-08,
+      "loss": 1.523,
+      "step": 11408
+    },
+    {
+      "epoch": 1.9682567066333132,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.2402807245844861e-08,
+      "loss": 1.4111,
+      "step": 11409
+    },
+    {
+      "epoch": 1.9684292245320452,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.226765544027675e-08,
+      "loss": 1.4143,
+      "step": 11410
+    },
+    {
+      "epoch": 1.9686017424307773,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.2133243588327014e-08,
+      "loss": 1.4456,
+      "step": 11411
+    },
+    {
+      "epoch": 1.9687742603295093,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.199957169995436e-08,
+      "loss": 1.4577,
+      "step": 11412
+    },
+    {
+      "epoch": 1.9689467782282413,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.1866639785060862e-08,
+      "loss": 1.3764,
+      "step": 11413
+    },
+    {
+      "epoch": 1.9691192961269732,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.1734447853495312e-08,
+      "loss": 1.3514,
+      "step": 11414
+    },
+    {
+      "epoch": 1.9692918140257052,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.1602995915050985e-08,
+      "loss": 1.5606,
+      "step": 11415
+    },
+    {
+      "epoch": 1.9694643319244372,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.1472283979467868e-08,
+      "loss": 1.4479,
+      "step": 11416
+    },
+    {
+      "epoch": 1.9696368498231691,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.1342312056429328e-08,
+      "loss": 1.5718,
+      "step": 11417
+    },
+    {
+      "epoch": 1.969809367721901,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.1213080155564327e-08,
+      "loss": 1.5088,
+      "step": 11418
+    },
+    {
+      "epoch": 1.969981885620633,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.1084588286446319e-08,
+      "loss": 1.467,
+      "step": 11419
+    },
+    {
+      "epoch": 1.970154403519365,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.0956836458596576e-08,
+      "loss": 1.3432,
+      "step": 11420
+    },
+    {
+      "epoch": 1.9703269214180972,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.082982468147864e-08,
+      "loss": 1.4005,
+      "step": 11421
+    },
+    {
+      "epoch": 1.9704994393168291,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.070355296450165e-08,
+      "loss": 1.3914,
+      "step": 11422
+    },
+    {
+      "epoch": 1.970671957215561,
+      "grad_norm": 0.56640625,
+      "learning_rate": 1.0578021317022569e-08,
+      "loss": 1.315,
+      "step": 11423
+    },
+    {
+      "epoch": 1.970844475114293,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.045322974833951e-08,
+      "loss": 1.4263,
+      "step": 11424
+    },
+    {
+      "epoch": 1.9710169930130252,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.0329178267699525e-08,
+      "loss": 1.4316,
+      "step": 11425
+    },
+    {
+      "epoch": 1.9711895109117572,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.0205866884291926e-08,
+      "loss": 1.4727,
+      "step": 11426
+    },
+    {
+      "epoch": 1.9713620288104892,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.0083295607252741e-08,
+      "loss": 1.3613,
+      "step": 11427
+    },
+    {
+      "epoch": 1.9715345467092211,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.961464445663594e-09,
+      "loss": 1.4517,
+      "step": 11428
+    },
+    {
+      "epoch": 1.971707064607953,
+      "grad_norm": 0.65625,
+      "learning_rate": 9.840373408548376e-09,
+      "loss": 1.3807,
+      "step": 11429
+    },
+    {
+      "epoch": 1.971879582506685,
+      "grad_norm": 0.5859375,
+      "learning_rate": 9.720022504881022e-09,
+      "loss": 1.36,
+      "step": 11430
+    },
+    {
+      "epoch": 1.972052100405417,
+      "grad_norm": 0.59765625,
+      "learning_rate": 9.600411743576621e-09,
+      "loss": 1.4093,
+      "step": 11431
+    },
+    {
+      "epoch": 1.972224618304149,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.481541133495864e-09,
+      "loss": 1.4134,
+      "step": 11432
+    },
+    {
+      "epoch": 1.972397136202881,
+      "grad_norm": 1.5625,
+      "learning_rate": 9.36341068344615e-09,
+      "loss": 1.4477,
+      "step": 11433
+    },
+    {
+      "epoch": 1.972569654101613,
+      "grad_norm": 0.56640625,
+      "learning_rate": 9.246020402179368e-09,
+      "loss": 1.4009,
+      "step": 11434
+    },
+    {
+      "epoch": 1.972742172000345,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.129370298393004e-09,
+      "loss": 1.4254,
+      "step": 11435
+    },
+    {
+      "epoch": 1.972914689899077,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.013460380729033e-09,
+      "loss": 1.414,
+      "step": 11436
+    },
+    {
+      "epoch": 1.973087207797809,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.898290657773923e-09,
+      "loss": 1.4608,
+      "step": 11437
+    },
+    {
+      "epoch": 1.9732597256965412,
+      "grad_norm": 0.64453125,
+      "learning_rate": 8.783861138060845e-09,
+      "loss": 1.5263,
+      "step": 11438
+    },
+    {
+      "epoch": 1.9734322435952731,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.670171830067464e-09,
+      "loss": 1.3921,
+      "step": 11439
+    },
+    {
+      "epoch": 1.973604761494005,
+      "grad_norm": 0.61328125,
+      "learning_rate": 8.557222742215932e-09,
+      "loss": 1.3992,
+      "step": 11440
+    },
+    {
+      "epoch": 1.973777279392737,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.445013882875108e-09,
+      "loss": 1.547,
+      "step": 11441
+    },
+    {
+      "epoch": 1.973949797291469,
+      "grad_norm": 0.58984375,
+      "learning_rate": 8.333545260357235e-09,
+      "loss": 1.4586,
+      "step": 11442
+    },
+    {
+      "epoch": 1.974122315190201,
+      "grad_norm": 0.57421875,
+      "learning_rate": 8.222816882922368e-09,
+      "loss": 1.3983,
+      "step": 11443
+    },
+    {
+      "epoch": 1.974294833088933,
+      "grad_norm": 2.140625,
+      "learning_rate": 8.11282875877173e-09,
+      "loss": 1.557,
+      "step": 11444
+    },
+    {
+      "epoch": 1.974467350987665,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.003580896055462e-09,
+      "loss": 1.363,
+      "step": 11445
+    },
+    {
+      "epoch": 1.9746398688863969,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.895073302865985e-09,
+      "loss": 1.3117,
+      "step": 11446
+    },
+    {
+      "epoch": 1.9748123867851288,
+      "grad_norm": 0.55859375,
+      "learning_rate": 7.787305987243532e-09,
+      "loss": 1.329,
+      "step": 11447
+    },
+    {
+      "epoch": 1.9749849046838608,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.680278957171716e-09,
+      "loss": 1.2998,
+      "step": 11448
+    },
+    {
+      "epoch": 1.975157422582593,
+      "grad_norm": 0.58203125,
+      "learning_rate": 7.573992220580862e-09,
+      "loss": 1.3532,
+      "step": 11449
+    },
+    {
+      "epoch": 1.975329940481325,
+      "grad_norm": 0.64453125,
+      "learning_rate": 7.468445785342448e-09,
+      "loss": 1.341,
+      "step": 11450
+    },
+    {
+      "epoch": 1.975502458380057,
+      "grad_norm": 0.62890625,
+      "learning_rate": 7.363639659279109e-09,
+      "loss": 1.4146,
+      "step": 11451
+    },
+    {
+      "epoch": 1.975674976278789,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.259573850153523e-09,
+      "loss": 1.5061,
+      "step": 11452
+    },
+    {
+      "epoch": 1.975847494177521,
+      "grad_norm": 0.58984375,
+      "learning_rate": 7.156248365676188e-09,
+      "loss": 1.323,
+      "step": 11453
+    },
+    {
+      "epoch": 1.976020012076253,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.053663213502093e-09,
+      "loss": 1.3581,
+      "step": 11454
+    },
+    {
+      "epoch": 1.976192529974985,
+      "grad_norm": 0.58203125,
+      "learning_rate": 6.951818401231825e-09,
+      "loss": 1.372,
+      "step": 11455
+    },
+    {
+      "epoch": 1.976365047873717,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.850713936410458e-09,
+      "loss": 1.3757,
+      "step": 11456
+    },
+    {
+      "epoch": 1.9765375657724489,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.750349826527558e-09,
+      "loss": 1.3695,
+      "step": 11457
+    },
+    {
+      "epoch": 1.9767100836711808,
+      "grad_norm": 0.57421875,
+      "learning_rate": 6.650726079019398e-09,
+      "loss": 1.3879,
+      "step": 11458
+    },
+    {
+      "epoch": 1.9768826015699128,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.551842701267852e-09,
+      "loss": 1.5441,
+      "step": 11459
+    },
+    {
+      "epoch": 1.9770551194686448,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.45369970059706e-09,
+      "loss": 1.3529,
+      "step": 11460
+    },
+    {
+      "epoch": 1.9772276373673767,
+      "grad_norm": 0.68359375,
+      "learning_rate": 6.356297084278984e-09,
+      "loss": 1.5121,
+      "step": 11461
+    },
+    {
+      "epoch": 1.977400155266109,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.259634859528962e-09,
+      "loss": 1.3742,
+      "step": 11462
+    },
+    {
+      "epoch": 1.9775726731648409,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.1637130335090446e-09,
+      "loss": 1.4024,
+      "step": 11463
+    },
+    {
+      "epoch": 1.9777451910635728,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.068531613326878e-09,
+      "loss": 1.4148,
+      "step": 11464
+    },
+    {
+      "epoch": 1.9779177089623048,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.9740906060312685e-09,
+      "loss": 1.4551,
+      "step": 11465
+    },
+    {
+      "epoch": 1.978090226861037,
+      "grad_norm": 0.63671875,
+      "learning_rate": 5.880390018621063e-09,
+      "loss": 1.4255,
+      "step": 11466
+    },
+    {
+      "epoch": 1.978262744759769,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.787429858038485e-09,
+      "loss": 1.4293,
+      "step": 11467
+    },
+    {
+      "epoch": 1.978435262658501,
+      "grad_norm": 0.625,
+      "learning_rate": 5.695210131169137e-09,
+      "loss": 1.4679,
+      "step": 11468
+    },
+    {
+      "epoch": 1.9786077805572329,
+      "grad_norm": 0.57421875,
+      "learning_rate": 5.603730844846444e-09,
+      "loss": 1.3792,
+      "step": 11469
+    },
+    {
+      "epoch": 1.9787802984559648,
+      "grad_norm": 0.6328125,
+      "learning_rate": 5.512992005846096e-09,
+      "loss": 1.4969,
+      "step": 11470
+    },
+    {
+      "epoch": 1.9789528163546968,
+      "grad_norm": 0.5390625,
+      "learning_rate": 5.422993620892713e-09,
+      "loss": 1.4239,
+      "step": 11471
+    },
+    {
+      "epoch": 1.9791253342534287,
+      "grad_norm": 0.62890625,
+      "learning_rate": 5.333735696653186e-09,
+      "loss": 1.442,
+      "step": 11472
+    },
+    {
+      "epoch": 1.9792978521521607,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.245218239740002e-09,
+      "loss": 1.4307,
+      "step": 11473
+    },
+    {
+      "epoch": 1.9794703700508927,
+      "grad_norm": 0.578125,
+      "learning_rate": 5.157441256710138e-09,
+      "loss": 1.4257,
+      "step": 11474
+    },
+    {
+      "epoch": 1.9796428879496246,
+      "grad_norm": 0.6484375,
+      "learning_rate": 5.070404754068392e-09,
+      "loss": 1.5279,
+      "step": 11475
+    },
+    {
+      "epoch": 1.9798154058483568,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.984108738261828e-09,
+      "loss": 1.4082,
+      "step": 11476
+    },
+    {
+      "epoch": 1.9799879237470888,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.898553215685331e-09,
+      "loss": 1.4421,
+      "step": 11477
+    },
+    {
+      "epoch": 1.9801604416458207,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.813738192676054e-09,
+      "loss": 1.338,
+      "step": 11478
+    },
+    {
+      "epoch": 1.980332959544553,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.729663675516749e-09,
+      "loss": 1.5036,
+      "step": 11479
+    },
+    {
+      "epoch": 1.9805054774432849,
+      "grad_norm": 0.5859375,
+      "learning_rate": 4.6463296704379876e-09,
+      "loss": 1.402,
+      "step": 11480
+    },
+    {
+      "epoch": 1.9806779953420168,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.5637361836126106e-09,
+      "loss": 1.39,
+      "step": 11481
+    },
+    {
+      "epoch": 1.9808505132407488,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.481883221160166e-09,
+      "loss": 1.3499,
+      "step": 11482
+    },
+    {
+      "epoch": 1.9810230311394808,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.400770789145803e-09,
+      "loss": 1.5186,
+      "step": 11483
+    },
+    {
+      "epoch": 1.9811955490382127,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.320398893576938e-09,
+      "loss": 1.4402,
+      "step": 11484
+    },
+    {
+      "epoch": 1.9813680669369447,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.240767540407698e-09,
+      "loss": 1.5271,
+      "step": 11485
+    },
+    {
+      "epoch": 1.9815405848356766,
+      "grad_norm": 0.5546875,
+      "learning_rate": 4.161876735540027e-09,
+      "loss": 1.4659,
+      "step": 11486
+    },
+    {
+      "epoch": 1.9817131027344086,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.08372648481703e-09,
+      "loss": 1.4919,
+      "step": 11487
+    },
+    {
+      "epoch": 1.9818856206331406,
+      "grad_norm": 0.57421875,
+      "learning_rate": 4.00631679402963e-09,
+      "loss": 1.3573,
+      "step": 11488
+    },
+    {
+      "epoch": 1.9820581385318725,
+      "grad_norm": 0.66015625,
+      "learning_rate": 3.9296476689110185e-09,
+      "loss": 1.469,
+      "step": 11489
+    },
+    {
+      "epoch": 1.9822306564306047,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.853719115143317e-09,
+      "loss": 1.4017,
+      "step": 11490
+    },
+    {
+      "epoch": 1.9824031743293367,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.778531138350916e-09,
+      "loss": 1.4678,
+      "step": 11491
+    },
+    {
+      "epoch": 1.9825756922280686,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.7040837441038035e-09,
+      "loss": 1.4091,
+      "step": 11492
+    },
+    {
+      "epoch": 1.9827482101268008,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.630376937917568e-09,
+      "loss": 1.4825,
+      "step": 11493
+    },
+    {
+      "epoch": 1.9829207280255328,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.5574107252533963e-09,
+      "loss": 1.3433,
+      "step": 11494
+    },
+    {
+      "epoch": 1.9830932459242647,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.4851851115180745e-09,
+      "loss": 1.4208,
+      "step": 11495
+    },
+    {
+      "epoch": 1.9832657638229967,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.4137001020595473e-09,
+      "loss": 1.5352,
+      "step": 11496
+    },
+    {
+      "epoch": 1.9834382817217286,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.3429557021769087e-09,
+      "loss": 1.3472,
+      "step": 11497
+    },
+    {
+      "epoch": 1.9836107996204606,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.2729519171093018e-09,
+      "loss": 1.4102,
+      "step": 11498
+    },
+    {
+      "epoch": 1.9837833175191926,
+      "grad_norm": 0.578125,
+      "learning_rate": 3.203688752044798e-09,
+      "loss": 1.4648,
+      "step": 11499
+    },
+    {
+      "epoch": 1.9839558354179245,
+      "grad_norm": 0.66015625,
+      "learning_rate": 3.1351662121137384e-09,
+      "loss": 1.2301,
+      "step": 11500
+    },
+    {
+      "epoch": 1.9839558354179245,
+      "eval_loss": 1.4070264101028442,
+      "eval_runtime": 10.8116,
+      "eval_samples_per_second": 94.713,
+      "eval_steps_per_second": 23.678,
+      "step": 11500
+    },
+    {
+      "epoch": 1.9841283533166565,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.0673843023920623e-09,
+      "loss": 1.3121,
+      "step": 11501
+    },
+    {
+      "epoch": 1.9843008712153885,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.0003430279024193e-09,
+      "loss": 1.3848,
+      "step": 11502
+    },
+    {
+      "epoch": 1.9844733891141206,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.9340423936119466e-09,
+      "loss": 1.4146,
+      "step": 11503
+    },
+    {
+      "epoch": 1.9846459070128526,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.868482404432271e-09,
+      "loss": 1.4652,
+      "step": 11504
+    },
+    {
+      "epoch": 1.9848184249115846,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.8036630652206187e-09,
+      "loss": 1.4103,
+      "step": 11505
+    },
+    {
+      "epoch": 1.9849909428103165,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.7395843807775934e-09,
+      "loss": 1.3924,
+      "step": 11506
+    },
+    {
+      "epoch": 1.9851634607090487,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.67624635585384e-09,
+      "loss": 1.4179,
+      "step": 11507
+    },
+    {
+      "epoch": 1.9853359786077807,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.6136489951378295e-09,
+      "loss": 1.3898,
+      "step": 11508
+    },
+    {
+      "epoch": 1.9855084965065126,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.5517923032714053e-09,
+      "loss": 1.457,
+      "step": 11509
+    },
+    {
+      "epoch": 1.9856810144052446,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.490676284833127e-09,
+      "loss": 1.3667,
+      "step": 11510
+    },
+    {
+      "epoch": 1.9858535323039765,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.430300944353814e-09,
+      "loss": 1.4318,
+      "step": 11511
+    },
+    {
+      "epoch": 1.9860260502027085,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.3706662863054452e-09,
+      "loss": 1.3655,
+      "step": 11512
+    },
+    {
+      "epoch": 1.9861985681014405,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.311772315106708e-09,
+      "loss": 1.4294,
+      "step": 11513
+    },
+    {
+      "epoch": 1.9863710860001724,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.2536190351196697e-09,
+      "loss": 1.4547,
+      "step": 11514
+    },
+    {
+      "epoch": 1.9865436038989044,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.1962064506542146e-09,
+      "loss": 1.4139,
+      "step": 11515
+    },
+    {
+      "epoch": 1.9867161217976363,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.1395345659613877e-09,
+      "loss": 1.4734,
+      "step": 11516
+    },
+    {
+      "epoch": 1.9868886396963685,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.0836033852422723e-09,
+      "loss": 1.3815,
+      "step": 11517
+    },
+    {
+      "epoch": 1.9870611575951005,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.0284129126402207e-09,
+      "loss": 1.4186,
+      "step": 11518
+    },
+    {
+      "epoch": 1.9872336754938325,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.9739631522430746e-09,
+      "loss": 1.3891,
+      "step": 11519
+    },
+    {
+      "epoch": 1.9874061933925646,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.9202541080853844e-09,
+      "loss": 1.489,
+      "step": 11520
+    },
+    {
+      "epoch": 1.9875787112912966,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.867285784146189e-09,
+      "loss": 1.4045,
+      "step": 11521
+    },
+    {
+      "epoch": 1.9877512291900286,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.8150581843490167e-09,
+      "loss": 1.3625,
+      "step": 11522
+    },
+    {
+      "epoch": 1.9879237470887605,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.7635713125641052e-09,
+      "loss": 1.473,
+      "step": 11523
+    },
+    {
+      "epoch": 1.9880962649874925,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.7128251726061805e-09,
+      "loss": 1.4697,
+      "step": 11524
+    },
+    {
+      "epoch": 1.9882687828862244,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.6628197682344583e-09,
+      "loss": 1.5217,
+      "step": 11525
+    },
+    {
+      "epoch": 1.9884413007849564,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.613555103152642e-09,
+      "loss": 1.5008,
+      "step": 11526
+    },
+    {
+      "epoch": 1.9886138186836884,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.5650311810122555e-09,
+      "loss": 1.3161,
+      "step": 11527
+    },
+    {
+      "epoch": 1.9887863365824203,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.5172480054070903e-09,
+      "loss": 1.4783,
+      "step": 11528
+    },
+    {
+      "epoch": 1.9889588544811523,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.4702055798776482e-09,
+      "loss": 1.5052,
+      "step": 11529
+    },
+    {
+      "epoch": 1.9891313723798845,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.423903907908919e-09,
+      "loss": 1.4766,
+      "step": 11530
+    },
+    {
+      "epoch": 1.9893038902786164,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.3783429929314918e-09,
+      "loss": 1.3936,
+      "step": 11531
+    },
+    {
+      "epoch": 1.9894764081773484,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.3335228383215548e-09,
+      "loss": 1.4857,
+      "step": 11532
+    },
+    {
+      "epoch": 1.9896489260760803,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.2894434473975648e-09,
+      "loss": 1.435,
+      "step": 11533
+    },
+    {
+      "epoch": 1.9898214439748125,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.246104823426908e-09,
+      "loss": 1.4502,
+      "step": 11534
+    },
+    {
+      "epoch": 1.9899939618735445,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.2035069696203494e-09,
+      "loss": 1.3413,
+      "step": 11535
+    },
+    {
+      "epoch": 1.9901664797722765,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.161649889133143e-09,
+      "loss": 1.4472,
+      "step": 11536
+    },
+    {
+      "epoch": 1.9903389976710084,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.1205335850661414e-09,
+      "loss": 1.5178,
+      "step": 11537
+    },
+    {
+      "epoch": 1.9905115155697404,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.080158060465797e-09,
+      "loss": 1.3904,
+      "step": 11538
+    },
+    {
+      "epoch": 1.9906840334684723,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.0405233183241604e-09,
+      "loss": 1.4666,
+      "step": 11539
+    },
+    {
+      "epoch": 1.9908565513672043,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.0016293615766615e-09,
+      "loss": 1.4624,
+      "step": 11540
+    },
+    {
+      "epoch": 1.9910290692659363,
+      "grad_norm": 0.6484375,
+      "learning_rate": 9.63476193104329e-10,
+      "loss": 1.3808,
+      "step": 11541
+    },
+    {
+      "epoch": 1.9912015871646682,
+      "grad_norm": 0.609375,
+      "learning_rate": 9.260638157360113e-10,
+      "loss": 1.4517,
+      "step": 11542
+    },
+    {
+      "epoch": 1.9913741050634002,
+      "grad_norm": 0.578125,
+      "learning_rate": 8.893922322406046e-10,
+      "loss": 1.4662,
+      "step": 11543
+    },
+    {
+      "epoch": 1.9915466229621324,
+      "grad_norm": 0.59375,
+      "learning_rate": 8.534614453370449e-10,
+      "loss": 1.4781,
+      "step": 11544
+    },
+    {
+      "epoch": 1.9917191408608643,
+      "grad_norm": 0.5859375,
+      "learning_rate": 8.182714576865369e-10,
+      "loss": 1.3584,
+      "step": 11545
+    },
+    {
+      "epoch": 1.9918916587595963,
+      "grad_norm": 0.6484375,
+      "learning_rate": 7.838222718958844e-10,
+      "loss": 1.4533,
+      "step": 11546
+    },
+    {
+      "epoch": 1.9920641766583285,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.501138905186e-10,
+      "loss": 1.4187,
+      "step": 11547
+    },
+    {
+      "epoch": 1.9922366945570604,
+      "grad_norm": 0.56640625,
+      "learning_rate": 7.171463160504655e-10,
+      "loss": 1.5579,
+      "step": 11548
+    },
+    {
+      "epoch": 1.9924092124557924,
+      "grad_norm": 0.609375,
+      "learning_rate": 6.849195509339712e-10,
+      "loss": 1.5175,
+      "step": 11549
+    },
+    {
+      "epoch": 1.9925817303545243,
+      "grad_norm": 0.6015625,
+      "learning_rate": 6.534335975583173e-10,
+      "loss": 1.3491,
+      "step": 11550
+    },
+    {
+      "epoch": 1.9927542482532563,
+      "grad_norm": 0.55859375,
+      "learning_rate": 6.226884582538618e-10,
+      "loss": 1.3489,
+      "step": 11551
+    },
+    {
+      "epoch": 1.9929267661519883,
+      "grad_norm": 0.5703125,
+      "learning_rate": 5.926841353010026e-10,
+      "loss": 1.4231,
+      "step": 11552
+    },
+    {
+      "epoch": 1.9930992840507202,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.634206309201861e-10,
+      "loss": 1.4619,
+      "step": 11553
+    },
+    {
+      "epoch": 1.9932718019494522,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.34897947280788e-10,
+      "loss": 1.4616,
+      "step": 11554
+    },
+    {
+      "epoch": 1.9934443198481842,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.071160864966729e-10,
+      "loss": 1.3633,
+      "step": 11555
+    },
+    {
+      "epoch": 1.9936168377469161,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.800750506239737e-10,
+      "loss": 1.4285,
+      "step": 11556
+    },
+    {
+      "epoch": 1.993789355645648,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.537748416677534e-10,
+      "loss": 1.432,
+      "step": 11557
+    },
+    {
+      "epoch": 1.9939618735443803,
+      "grad_norm": 0.609375,
+      "learning_rate": 4.2821546157534313e-10,
+      "loss": 1.3494,
+      "step": 11558
+    },
+    {
+      "epoch": 1.9941343914431122,
+      "grad_norm": 0.5703125,
+      "learning_rate": 4.0339691224189346e-10,
+      "loss": 1.457,
+      "step": 11559
+    },
+    {
+      "epoch": 1.9943069093418442,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.7931919550482364e-10,
+      "loss": 1.4494,
+      "step": 11560
+    },
+    {
+      "epoch": 1.9944794272405764,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.559823131471518e-10,
+      "loss": 1.345,
+      "step": 11561
+    },
+    {
+      "epoch": 1.9946519451393083,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.3338626690082587e-10,
+      "loss": 1.4648,
+      "step": 11562
+    },
+    {
+      "epoch": 1.9948244630380403,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.115310584367315e-10,
+      "loss": 1.3445,
+      "step": 11563
+    },
+    {
+      "epoch": 1.9949969809367722,
+      "grad_norm": 0.59375,
+      "learning_rate": 2.9041668937579426e-10,
+      "loss": 1.3534,
+      "step": 11564
+    },
+    {
+      "epoch": 1.9951694988355042,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.7004316128231844e-10,
+      "loss": 1.428,
+      "step": 11565
+    },
+    {
+      "epoch": 1.9953420167342362,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.504104756639869e-10,
+      "loss": 1.4076,
+      "step": 11566
+    },
+    {
+      "epoch": 1.9955145346329681,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.3151863397741225e-10,
+      "loss": 1.406,
+      "step": 11567
+    },
+    {
+      "epoch": 1.9956870525317,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.133676376214755e-10,
+      "loss": 1.4402,
+      "step": 11568
+    },
+    {
+      "epoch": 1.995859570430432,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.9595748794065673e-10,
+      "loss": 1.5273,
+      "step": 11569
+    },
+    {
+      "epoch": 1.996032088329164,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.792881862250351e-10,
+      "loss": 1.457,
+      "step": 11570
+    },
+    {
+      "epoch": 1.9962046062278962,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.6335973370917858e-10,
+      "loss": 1.3983,
+      "step": 11571
+    },
+    {
+      "epoch": 1.9963771241266282,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.481721315743645e-10,
+      "loss": 1.4884,
+      "step": 11572
+    },
+    {
+      "epoch": 1.9965496420253601,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.3372538094413856e-10,
+      "loss": 1.4246,
+      "step": 11573
+    },
+    {
+      "epoch": 1.996722159924092,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.2001948288986598e-10,
+      "loss": 1.4212,
+      "step": 11574
+    },
+    {
+      "epoch": 1.9968946778228243,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.070544384262906e-10,
+      "loss": 1.3712,
+      "step": 11575
+    },
+    {
+      "epoch": 1.9970671957215562,
+      "grad_norm": 0.59375,
+      "learning_rate": 9.483024851486556e-11,
+      "loss": 1.3957,
+      "step": 11576
+    },
+    {
+      "epoch": 1.9972397136202882,
+      "grad_norm": 0.56640625,
+      "learning_rate": 8.334691406042261e-11,
+      "loss": 1.4156,
+      "step": 11577
+    },
+    {
+      "epoch": 1.9974122315190201,
+      "grad_norm": 0.53515625,
+      "learning_rate": 7.260443591450283e-11,
+      "loss": 1.4135,
+      "step": 11578
+    },
+    {
+      "epoch": 1.997584749417752,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.26028148720259e-11,
+      "loss": 1.3832,
+      "step": 11579
+    },
+    {
+      "epoch": 1.997757267316484,
+      "grad_norm": 0.59765625,
+      "learning_rate": 5.334205167462081e-11,
+      "loss": 1.4101,
+      "step": 11580
+    },
+    {
+      "epoch": 1.997929785215216,
+      "grad_norm": 0.5625,
+      "learning_rate": 4.482214700729515e-11,
+      "loss": 1.394,
+      "step": 11581
+    },
+    {
+      "epoch": 1.998102303113948,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.704310150287604e-11,
+      "loss": 1.4096,
+      "step": 11582
+    },
+    {
+      "epoch": 1.99827482101268,
+      "grad_norm": 0.57421875,
+      "learning_rate": 3.000491573756925e-11,
+      "loss": 1.4367,
+      "step": 11583
+    },
+    {
+      "epoch": 1.998447338911412,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.3707590230959144e-11,
+      "loss": 1.4118,
+      "step": 11584
+    },
+    {
+      "epoch": 1.998619856810144,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.8151125451559838e-11,
+      "loss": 1.3415,
+      "step": 11585
+    },
+    {
+      "epoch": 1.998792374708876,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.3335521810153851e-11,
+      "loss": 1.4252,
+      "step": 11586
+    },
+    {
+      "epoch": 1.998964892607608,
+      "grad_norm": 0.58203125,
+      "learning_rate": 9.260779664233e-12,
+      "loss": 1.4393,
+      "step": 11587
+    },
+    {
+      "epoch": 1.9991374105063402,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.926899314667723e-12,
+      "loss": 1.4364,
+      "step": 11588
+    },
+    {
+      "epoch": 1.9993099284050722,
+      "grad_norm": 0.5625,
+      "learning_rate": 3.3338810079275307e-12,
+      "loss": 1.4977,
+      "step": 11589
+    },
+    {
+      "epoch": 1.9994824463038041,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.4817249383014543e-12,
+      "loss": 1.3664,
+      "step": 11590
+    },
+    {
+      "epoch": 1.999654964202536,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.704312412367017e-13,
+      "loss": 1.3706,
+      "step": 11591
+    },
+    {
+      "epoch": 1.999827482101268,
+      "grad_norm": 0.55078125,
+      "learning_rate": 0.0,
+      "loss": 1.4666,
+      "step": 11592
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 11592,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 3000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.613498406381342e+19,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}