diff --git "a/checkpoint-1526387/trainer_state.json" "b/checkpoint-1526387/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1526387/trainer_state.json"
@@ -0,0 +1,18071 @@
+{
+  "best_metric": 3.995372772216797,
+  "best_model_checkpoint": "/mmfs1/gscratch/stf/abhinavp/corpus-filtering/outputs/full/lstm/3/checkpoints/checkpoint-1526387",
+  "epoch": 0.025000606015738065,
+  "eval_steps": 10,
+  "global_step": 1526387,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999998362119627e-05,
+      "loss": 10.8183,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999161405248948e-05,
+      "loss": 7.5661,
+      "step": 512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998322810497896e-05,
+      "loss": 7.0563,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997484215746844e-05,
+      "loss": 6.985,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996645620995792e-05,
+      "loss": 6.956,
+      "step": 2048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99580702624474e-05,
+      "loss": 6.8543,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994970069374061e-05,
+      "loss": 6.7253,
+      "step": 3072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994131474623009e-05,
+      "loss": 6.6189,
+      "step": 3584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993292879871958e-05,
+      "loss": 6.5339,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.992454285120906e-05,
+      "loss": 6.4422,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.991615690369854e-05,
+      "loss": 6.3916,
+      "step": 5120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.990777095618801e-05,
+      "loss": 6.3209,
+      "step": 5632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989938500867749e-05,
+      "loss": 6.253,
+      "step": 6144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.989099906116697e-05,
+      "loss": 6.1866,
+      "step": 6656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.988261311365645e-05,
+      "loss": 6.1383,
+      "step": 7168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.987422716614593e-05,
+      "loss": 6.0974,
+      "step": 7680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.986584121863541e-05,
+      "loss": 6.0514,
+      "step": 8192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.985745527112489e-05,
+      "loss": 5.9997,
+      "step": 8704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984906932361437e-05,
+      "loss": 5.9645,
+      "step": 9216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.984068337610385e-05,
+      "loss": 5.9268,
+      "step": 9728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.983229742859333e-05,
+      "loss": 5.8802,
+      "step": 10240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.982391148108281e-05,
+      "loss": 5.8432,
+      "step": 10752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.981552553357229e-05,
+      "loss": 5.8191,
+      "step": 11264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.980713958606178e-05,
+      "loss": 5.7811,
+      "step": 11776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9798770017354986e-05,
+      "loss": 5.7597,
+      "step": 12288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9790384069844466e-05,
+      "loss": 5.7314,
+      "step": 12800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9781998122333946e-05,
+      "loss": 5.7082,
+      "step": 13312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9773612174823426e-05,
+      "loss": 5.6786,
+      "step": 13824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.97652262273129e-05,
+      "loss": 5.6449,
+      "step": 14336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9756856658606115e-05,
+      "loss": 5.638,
+      "step": 14848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9748470711095595e-05,
+      "loss": 5.6037,
+      "step": 15360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.974008476358507e-05,
+      "loss": 5.5886,
+      "step": 15872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.973169881607455e-05,
+      "loss": 5.5626,
+      "step": 16384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.972331286856403e-05,
+      "loss": 5.5594,
+      "step": 16896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9714943299857244e-05,
+      "loss": 5.5281,
+      "step": 17408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9706557352346724e-05,
+      "loss": 5.5185,
+      "step": 17920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.969818778363994e-05,
+      "loss": 5.5051,
+      "step": 18432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.968980183612942e-05,
+      "loss": 5.4855,
+      "step": 18944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96814158886189e-05,
+      "loss": 5.475,
+      "step": 19456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.967302994110837e-05,
+      "loss": 5.4524,
+      "step": 19968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.966464399359785e-05,
+      "loss": 5.4292,
+      "step": 20480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.965625804608733e-05,
+      "loss": 5.4232,
+      "step": 20992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.964787209857681e-05,
+      "loss": 5.3952,
+      "step": 21504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.963948615106629e-05,
+      "loss": 5.408,
+      "step": 22016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.96311165823595e-05,
+      "loss": 5.3798,
+      "step": 22528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.962273063484898e-05,
+      "loss": 5.3763,
+      "step": 23040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.961434468733847e-05,
+      "loss": 5.3602,
+      "step": 23552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.960597511863168e-05,
+      "loss": 5.358,
+      "step": 24064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.959758917112116e-05,
+      "loss": 5.3509,
+      "step": 24576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958920322361064e-05,
+      "loss": 5.3221,
+      "step": 25088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.958081727610012e-05,
+      "loss": 5.3108,
+      "step": 25600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.95724313285896e-05,
+      "loss": 5.3022,
+      "step": 26112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.956404538107908e-05,
+      "loss": 5.3086,
+      "step": 26624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.955565943356856e-05,
+      "loss": 5.2857,
+      "step": 27136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.954727348605804e-05,
+      "loss": 5.2924,
+      "step": 27648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.953890391735125e-05,
+      "loss": 5.263,
+      "step": 28160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9530517969840727e-05,
+      "loss": 5.26,
+      "step": 28672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9522132022330207e-05,
+      "loss": 5.2459,
+      "step": 29184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9513746074819686e-05,
+      "loss": 5.241,
+      "step": 29696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9505360127309166e-05,
+      "loss": 5.2379,
+      "step": 30208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9496974179798646e-05,
+      "loss": 5.2178,
+      "step": 30720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948858823228813e-05,
+      "loss": 5.2112,
+      "step": 31232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.948021866358134e-05,
+      "loss": 5.2158,
+      "step": 31744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.947183271607082e-05,
+      "loss": 5.1943,
+      "step": 32256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.94634467685603e-05,
+      "loss": 5.1926,
+      "step": 32768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.945506082104978e-05,
+      "loss": 5.1823,
+      "step": 33280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9446674873539255e-05,
+      "loss": 5.1801,
+      "step": 33792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9438288926028735e-05,
+      "loss": 5.1659,
+      "step": 34304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9429902978518215e-05,
+      "loss": 5.1684,
+      "step": 34816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.942153340981143e-05,
+      "loss": 5.1552,
+      "step": 35328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9413147462300904e-05,
+      "loss": 5.146,
+      "step": 35840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9404761514790384e-05,
+      "loss": 5.148,
+      "step": 36352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.939637556727987e-05,
+      "loss": 5.134,
+      "step": 36864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.938800599857309e-05,
+      "loss": 5.1416,
+      "step": 37376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.937962005106256e-05,
+      "loss": 5.1388,
+      "step": 37888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.937123410355204e-05,
+      "loss": 5.129,
+      "step": 38400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9362864534845256e-05,
+      "loss": 5.1101,
+      "step": 38912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.935447858733473e-05,
+      "loss": 5.0947,
+      "step": 39424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.934609263982421e-05,
+      "loss": 5.0881,
+      "step": 39936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9337723071117425e-05,
+      "loss": 5.0927,
+      "step": 40448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9329337123606905e-05,
+      "loss": 5.0953,
+      "step": 40960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.932095117609638e-05,
+      "loss": 5.0905,
+      "step": 41472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.931256522858586e-05,
+      "loss": 5.0644,
+      "step": 41984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.930417928107534e-05,
+      "loss": 5.0589,
+      "step": 42496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9295793333564825e-05,
+      "loss": 5.0672,
+      "step": 43008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9287407386054305e-05,
+      "loss": 5.0535,
+      "step": 43520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9279021438543785e-05,
+      "loss": 5.0529,
+      "step": 44032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9270651869836994e-05,
+      "loss": 5.0443,
+      "step": 44544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9262265922326474e-05,
+      "loss": 5.0457,
+      "step": 45056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9253879974815954e-05,
+      "loss": 5.035,
+      "step": 45568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9245494027305433e-05,
+      "loss": 5.0266,
+      "step": 46080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9237108079794913e-05,
+      "loss": 5.0134,
+      "step": 46592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922872213228439e-05,
+      "loss": 5.0257,
+      "step": 47104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.922033618477387e-05,
+      "loss": 5.0113,
+      "step": 47616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.921195023726335e-05,
+      "loss": 4.9951,
+      "step": 48128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.920356428975283e-05,
+      "loss": 4.995,
+      "step": 48640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.919517834224231e-05,
+      "loss": 5.007,
+      "step": 49152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.918680877353552e-05,
+      "loss": 4.9967,
+      "step": 49664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917842282602501e-05,
+      "loss": 4.9875,
+      "step": 50176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.917003687851449e-05,
+      "loss": 4.9758,
+      "step": 50688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.916165093100397e-05,
+      "loss": 4.9797,
+      "step": 51200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.915328136229718e-05,
+      "loss": 4.9595,
+      "step": 51712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.914489541478666e-05,
+      "loss": 4.9742,
+      "step": 52224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.913650946727614e-05,
+      "loss": 4.9756,
+      "step": 52736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.912812351976562e-05,
+      "loss": 4.9512,
+      "step": 53248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911973757225509e-05,
+      "loss": 4.9452,
+      "step": 53760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.911136800354831e-05,
+      "loss": 4.922,
+      "step": 54272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.910298205603779e-05,
+      "loss": 4.9337,
+      "step": 54784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9094612487330996e-05,
+      "loss": 4.9349,
+      "step": 55296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.908624291862421e-05,
+      "loss": 4.9289,
+      "step": 55808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.907785697111369e-05,
+      "loss": 4.9306,
+      "step": 56320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906947102360317e-05,
+      "loss": 4.9136,
+      "step": 56832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.906108507609265e-05,
+      "loss": 4.9153,
+      "step": 57344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.905269912858213e-05,
+      "loss": 4.917,
+      "step": 57856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.904431318107161e-05,
+      "loss": 4.9185,
+      "step": 58368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.903592723356109e-05,
+      "loss": 4.9047,
+      "step": 58880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9027541286050565e-05,
+      "loss": 4.8931,
+      "step": 59392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9019155338540045e-05,
+      "loss": 4.8959,
+      "step": 59904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9010769391029525e-05,
+      "loss": 4.8983,
+      "step": 60416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9002383443519005e-05,
+      "loss": 4.8944,
+      "step": 60928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8994013874812214e-05,
+      "loss": 4.8881,
+      "step": 61440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.89856279273017e-05,
+      "loss": 4.8778,
+      "step": 61952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.897724197979118e-05,
+      "loss": 4.8712,
+      "step": 62464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896885603228066e-05,
+      "loss": 4.8642,
+      "step": 62976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.896047008477014e-05,
+      "loss": 4.8712,
+      "step": 63488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.895208413725962e-05,
+      "loss": 4.8685,
+      "step": 64000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.89436981897491e-05,
+      "loss": 4.8707,
+      "step": 64512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.893531224223858e-05,
+      "loss": 4.8451,
+      "step": 65024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.892692629472806e-05,
+      "loss": 4.8582,
+      "step": 65536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891854034721754e-05,
+      "loss": 4.8587,
+      "step": 66048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.891017077851075e-05,
+      "loss": 4.8442,
+      "step": 66560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.890178483100023e-05,
+      "loss": 4.8594,
+      "step": 67072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.889339888348971e-05,
+      "loss": 4.8397,
+      "step": 67584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.888502931478292e-05,
+      "loss": 4.8539,
+      "step": 68096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.88766433672724e-05,
+      "loss": 4.8418,
+      "step": 68608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.886825741976188e-05,
+      "loss": 4.8217,
+      "step": 69120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8859871472251365e-05,
+      "loss": 4.8207,
+      "step": 69632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8851485524740845e-05,
+      "loss": 4.8202,
+      "step": 70144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8843099577230325e-05,
+      "loss": 4.823,
+      "step": 70656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8834713629719805e-05,
+      "loss": 4.8275,
+      "step": 71168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8826344061013014e-05,
+      "loss": 4.8246,
+      "step": 71680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8817958113502494e-05,
+      "loss": 4.8196,
+      "step": 72192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8809572165991974e-05,
+      "loss": 4.8071,
+      "step": 72704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.8801186218481454e-05,
+      "loss": 4.7964,
+      "step": 73216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.879280027097093e-05,
+      "loss": 4.7952,
+      "step": 73728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.878441432346041e-05,
+      "loss": 4.805,
+      "step": 74240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.877604475475362e-05,
+      "loss": 4.8102,
+      "step": 74752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.87676588072431e-05,
+      "loss": 4.8008,
+      "step": 75264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875927285973258e-05,
+      "loss": 4.7837,
+      "step": 75776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.875088691222206e-05,
+      "loss": 4.7832,
+      "step": 76288
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.748494625091553,
+      "eval_runtime": 552.6156,
+      "eval_samples_per_second": 690.518,
+      "eval_steps_per_second": 21.579,
+      "step": 76319
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.874250096471154e-05,
+      "loss": 4.7924,
+      "step": 76800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.873411501720102e-05,
+      "loss": 4.7847,
+      "step": 77312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.87257290696905e-05,
+      "loss": 4.7762,
+      "step": 77824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.871734312217998e-05,
+      "loss": 4.7744,
+      "step": 78336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.870895717466946e-05,
+      "loss": 4.7644,
+      "step": 78848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.870057122715894e-05,
+      "loss": 4.7505,
+      "step": 79360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.869218527964842e-05,
+      "loss": 4.7583,
+      "step": 79872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.86837993321379e-05,
+      "loss": 4.7679,
+      "step": 80384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.867541338462738e-05,
+      "loss": 4.75,
+      "step": 80896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.866702743711686e-05,
+      "loss": 4.7578,
+      "step": 81408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.865864148960634e-05,
+      "loss": 4.7612,
+      "step": 81920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8650255542095815e-05,
+      "loss": 4.761,
+      "step": 82432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.86418695945853e-05,
+      "loss": 4.7497,
+      "step": 82944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.863348364707478e-05,
+      "loss": 4.7469,
+      "step": 83456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.862509769956426e-05,
+      "loss": 4.7446,
+      "step": 83968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.861671175205374e-05,
+      "loss": 4.7335,
+      "step": 84480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.860832580454322e-05,
+      "loss": 4.7369,
+      "step": 84992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.85999398570327e-05,
+      "loss": 4.7371,
+      "step": 85504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.859155390952218e-05,
+      "loss": 4.7361,
+      "step": 86016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.858320071961912e-05,
+      "loss": 4.7337,
+      "step": 86528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.85748147721086e-05,
+      "loss": 4.7221,
+      "step": 87040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.856642882459808e-05,
+      "loss": 4.7341,
+      "step": 87552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.855804287708756e-05,
+      "loss": 4.7204,
+      "step": 88064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.854967330838077e-05,
+      "loss": 4.7285,
+      "step": 88576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8541287360870256e-05,
+      "loss": 4.7185,
+      "step": 89088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8532901413359736e-05,
+      "loss": 4.7171,
+      "step": 89600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8524515465849216e-05,
+      "loss": 4.7203,
+      "step": 90112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8516145897142425e-05,
+      "loss": 4.7019,
+      "step": 90624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.850777632843564e-05,
+      "loss": 4.7161,
+      "step": 91136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8499390380925114e-05,
+      "loss": 4.6952,
+      "step": 91648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8491004433414594e-05,
+      "loss": 4.7013,
+      "step": 92160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8482618485904074e-05,
+      "loss": 4.698,
+      "step": 92672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8474232538393554e-05,
+      "loss": 4.7058,
+      "step": 93184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8465846590883034e-05,
+      "loss": 4.6932,
+      "step": 93696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8457460643372514e-05,
+      "loss": 4.6863,
+      "step": 94208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8449074695861993e-05,
+      "loss": 4.7029,
+      "step": 94720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8440688748351473e-05,
+      "loss": 4.6903,
+      "step": 95232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.843230280084095e-05,
+      "loss": 4.6971,
+      "step": 95744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.842393323213417e-05,
+      "loss": 4.6849,
+      "step": 96256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.841554728462365e-05,
+      "loss": 4.6705,
+      "step": 96768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.840716133711313e-05,
+      "loss": 4.6734,
+      "step": 97280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.839877538960261e-05,
+      "loss": 4.6661,
+      "step": 97792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.839038944209209e-05,
+      "loss": 4.6811,
+      "step": 98304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.83820198733853e-05,
+      "loss": 4.6713,
+      "step": 98816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.837363392587478e-05,
+      "loss": 4.6767,
+      "step": 99328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.836524797836426e-05,
+      "loss": 4.6682,
+      "step": 99840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.835686203085374e-05,
+      "loss": 4.6741,
+      "step": 100352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.834847608334322e-05,
+      "loss": 4.6696,
+      "step": 100864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.83400901358327e-05,
+      "loss": 4.6476,
+      "step": 101376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.833170418832218e-05,
+      "loss": 4.651,
+      "step": 101888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.832331824081166e-05,
+      "loss": 4.6494,
+      "step": 102400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8314948672104874e-05,
+      "loss": 4.6627,
+      "step": 102912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8306562724594354e-05,
+      "loss": 4.6403,
+      "step": 103424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.829819315588756e-05,
+      "loss": 4.6644,
+      "step": 103936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.828980720837704e-05,
+      "loss": 4.6381,
+      "step": 104448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.828142126086652e-05,
+      "loss": 4.6386,
+      "step": 104960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8273035313356e-05,
+      "loss": 4.6402,
+      "step": 105472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8264649365845476e-05,
+      "loss": 4.6369,
+      "step": 105984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8256263418334956e-05,
+      "loss": 4.639,
+      "step": 106496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8247877470824436e-05,
+      "loss": 4.6302,
+      "step": 107008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8239491523313916e-05,
+      "loss": 4.6264,
+      "step": 107520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8231121954607125e-05,
+      "loss": 4.6386,
+      "step": 108032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.822273600709661e-05,
+      "loss": 4.624,
+      "step": 108544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.821435005958609e-05,
+      "loss": 4.6222,
+      "step": 109056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.820596411207557e-05,
+      "loss": 4.6226,
+      "step": 109568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.819759454336878e-05,
+      "loss": 4.6268,
+      "step": 110080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.818920859585826e-05,
+      "loss": 4.6113,
+      "step": 110592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.818082264834774e-05,
+      "loss": 4.6301,
+      "step": 111104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.817243670083722e-05,
+      "loss": 4.6165,
+      "step": 111616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.816406713213043e-05,
+      "loss": 4.6144,
+      "step": 112128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.815568118461991e-05,
+      "loss": 4.6209,
+      "step": 112640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.814729523710939e-05,
+      "loss": 4.6083,
+      "step": 113152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.813890928959887e-05,
+      "loss": 4.6265,
+      "step": 113664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.813053972089208e-05,
+      "loss": 4.6297,
+      "step": 114176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8122153773381565e-05,
+      "loss": 4.623,
+      "step": 114688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8113767825871045e-05,
+      "loss": 4.604,
+      "step": 115200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8105381878360525e-05,
+      "loss": 4.5988,
+      "step": 115712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8096995930850005e-05,
+      "loss": 4.5934,
+      "step": 116224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8088609983339485e-05,
+      "loss": 4.6041,
+      "step": 116736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8080240414632694e-05,
+      "loss": 4.6161,
+      "step": 117248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8071854467122174e-05,
+      "loss": 4.6076,
+      "step": 117760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8063468519611654e-05,
+      "loss": 4.5948,
+      "step": 118272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8055082572101134e-05,
+      "loss": 4.5829,
+      "step": 118784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8046696624590614e-05,
+      "loss": 4.6014,
+      "step": 119296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8038310677080094e-05,
+      "loss": 4.5865,
+      "step": 119808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8029924729569574e-05,
+      "loss": 4.5891,
+      "step": 120320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.8021538782059054e-05,
+      "loss": 4.5909,
+      "step": 120832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.801316921335226e-05,
+      "loss": 4.5991,
+      "step": 121344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.800479964464548e-05,
+      "loss": 4.587,
+      "step": 121856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.799641369713496e-05,
+      "loss": 4.5844,
+      "step": 122368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.798802774962444e-05,
+      "loss": 4.5752,
+      "step": 122880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.797964180211392e-05,
+      "loss": 4.5873,
+      "step": 123392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.797127223340713e-05,
+      "loss": 4.5863,
+      "step": 123904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.796288628589661e-05,
+      "loss": 4.5621,
+      "step": 124416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.795450033838609e-05,
+      "loss": 4.5711,
+      "step": 124928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.794611439087557e-05,
+      "loss": 4.5864,
+      "step": 125440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.793772844336505e-05,
+      "loss": 4.5867,
+      "step": 125952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.792934249585453e-05,
+      "loss": 4.5735,
+      "step": 126464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.792095654834401e-05,
+      "loss": 4.5675,
+      "step": 126976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.791257060083349e-05,
+      "loss": 4.5785,
+      "step": 127488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7904201032126697e-05,
+      "loss": 4.5568,
+      "step": 128000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.789581508461618e-05,
+      "loss": 4.5744,
+      "step": 128512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.788744551590939e-05,
+      "loss": 4.5775,
+      "step": 129024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.787905956839887e-05,
+      "loss": 4.5649,
+      "step": 129536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.787067362088835e-05,
+      "loss": 4.5643,
+      "step": 130048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.786228767337783e-05,
+      "loss": 4.5412,
+      "step": 130560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.785390172586731e-05,
+      "loss": 4.5464,
+      "step": 131072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7845515778356785e-05,
+      "loss": 4.5531,
+      "step": 131584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7837129830846265e-05,
+      "loss": 4.5553,
+      "step": 132096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.782876026213948e-05,
+      "loss": 4.5584,
+      "step": 132608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.782037431462896e-05,
+      "loss": 4.5463,
+      "step": 133120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7811988367118434e-05,
+      "loss": 4.5451,
+      "step": 133632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.780360241960792e-05,
+      "loss": 4.5519,
+      "step": 134144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.77952164720974e-05,
+      "loss": 4.558,
+      "step": 134656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.778683052458688e-05,
+      "loss": 4.546,
+      "step": 135168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.777844457707636e-05,
+      "loss": 4.5458,
+      "step": 135680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.777005862956584e-05,
+      "loss": 4.5433,
+      "step": 136192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.776168906085905e-05,
+      "loss": 4.5465,
+      "step": 136704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.775330311334853e-05,
+      "loss": 4.547,
+      "step": 137216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.774491716583801e-05,
+      "loss": 4.5395,
+      "step": 137728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.773654759713122e-05,
+      "loss": 4.5382,
+      "step": 138240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.77281616496207e-05,
+      "loss": 4.5322,
+      "step": 138752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.771977570211018e-05,
+      "loss": 4.5274,
+      "step": 139264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.771138975459966e-05,
+      "loss": 4.5335,
+      "step": 139776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.770300380708914e-05,
+      "loss": 4.5363,
+      "step": 140288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.769461785957862e-05,
+      "loss": 4.5385,
+      "step": 140800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7686231912068105e-05,
+      "loss": 4.5224,
+      "step": 141312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7677862343361315e-05,
+      "loss": 4.5285,
+      "step": 141824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7669476395850795e-05,
+      "loss": 4.5403,
+      "step": 142336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7661090448340274e-05,
+      "loss": 4.5251,
+      "step": 142848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7652704500829754e-05,
+      "loss": 4.5384,
+      "step": 143360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7644318553319234e-05,
+      "loss": 4.523,
+      "step": 143872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7635932605808714e-05,
+      "loss": 4.541,
+      "step": 144384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7627546658298194e-05,
+      "loss": 4.5275,
+      "step": 144896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.7619160710787674e-05,
+      "loss": 4.5116,
+      "step": 145408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.761079114208088e-05,
+      "loss": 4.5152,
+      "step": 145920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.760242157337409e-05,
+      "loss": 4.5125,
+      "step": 146432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.759403562586357e-05,
+      "loss": 4.5231,
+      "step": 146944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.758566605715679e-05,
+      "loss": 4.5229,
+      "step": 147456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.757728010964627e-05,
+      "loss": 4.5252,
+      "step": 147968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.756889416213575e-05,
+      "loss": 4.52,
+      "step": 148480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.756050821462523e-05,
+      "loss": 4.5111,
+      "step": 148992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.755212226711471e-05,
+      "loss": 4.506,
+      "step": 149504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.754373631960419e-05,
+      "loss": 4.5029,
+      "step": 150016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.753535037209367e-05,
+      "loss": 4.5134,
+      "step": 150528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.752696442458315e-05,
+      "loss": 4.5265,
+      "step": 151040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.751857847707262e-05,
+      "loss": 4.5096,
+      "step": 151552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.751020890836584e-05,
+      "loss": 4.5027,
+      "step": 152064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.750182296085532e-05,
+      "loss": 4.5027,
+      "step": 152576
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.47078800201416,
+      "eval_runtime": 551.3077,
+      "eval_samples_per_second": 692.156,
+      "eval_steps_per_second": 21.63,
+      "step": 152638
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74934370133448e-05,
+      "loss": 4.5138,
+      "step": 153088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.748505106583428e-05,
+      "loss": 4.5086,
+      "step": 153600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.747666511832376e-05,
+      "loss": 4.4956,
+      "step": 154112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.746827917081324e-05,
+      "loss": 4.4975,
+      "step": 154624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.745989322330272e-05,
+      "loss": 4.4939,
+      "step": 155136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.74515072757922e-05,
+      "loss": 4.479,
+      "step": 155648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.744312132828168e-05,
+      "loss": 4.4885,
+      "step": 156160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.743473538077116e-05,
+      "loss": 4.5019,
+      "step": 156672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7426349433260637e-05,
+      "loss": 4.4809,
+      "step": 157184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7417963485750117e-05,
+      "loss": 4.4904,
+      "step": 157696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7409577538239597e-05,
+      "loss": 4.5005,
+      "step": 158208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7401191590729076e-05,
+      "loss": 4.4998,
+      "step": 158720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7392805643218556e-05,
+      "loss": 4.4867,
+      "step": 159232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7384419695708036e-05,
+      "loss": 4.4868,
+      "step": 159744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.737603374819751e-05,
+      "loss": 4.4857,
+      "step": 160256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.736764780068699e-05,
+      "loss": 4.4789,
+      "step": 160768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7359261853176476e-05,
+      "loss": 4.4781,
+      "step": 161280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7350892284469685e-05,
+      "loss": 4.4845,
+      "step": 161792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7342506336959165e-05,
+      "loss": 4.4875,
+      "step": 162304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7334120389448645e-05,
+      "loss": 4.4839,
+      "step": 162816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7325734441938125e-05,
+      "loss": 4.4755,
+      "step": 163328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7317348494427605e-05,
+      "loss": 4.4866,
+      "step": 163840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7308978925720814e-05,
+      "loss": 4.4764,
+      "step": 164352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7300592978210294e-05,
+      "loss": 4.4821,
+      "step": 164864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7292207030699774e-05,
+      "loss": 4.4766,
+      "step": 165376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7283821083189254e-05,
+      "loss": 4.4747,
+      "step": 165888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7275435135678734e-05,
+      "loss": 4.4804,
+      "step": 166400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7267049188168214e-05,
+      "loss": 4.4637,
+      "step": 166912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.7258663240657694e-05,
+      "loss": 4.4781,
+      "step": 167424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7250277293147174e-05,
+      "loss": 4.4627,
+      "step": 167936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.724190772444039e-05,
+      "loss": 4.4659,
+      "step": 168448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.72335381557336e-05,
+      "loss": 4.4669,
+      "step": 168960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.722515220822308e-05,
+      "loss": 4.4697,
+      "step": 169472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.721676626071256e-05,
+      "loss": 4.4634,
+      "step": 169984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.720838031320204e-05,
+      "loss": 4.4557,
+      "step": 170496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.719999436569152e-05,
+      "loss": 4.4744,
+      "step": 171008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7191608418181e-05,
+      "loss": 4.4625,
+      "step": 171520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.718322247067048e-05,
+      "loss": 4.4759,
+      "step": 172032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.717483652315996e-05,
+      "loss": 4.4619,
+      "step": 172544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.716646695445317e-05,
+      "loss": 4.4455,
+      "step": 173056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.715808100694265e-05,
+      "loss": 4.4461,
+      "step": 173568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.714969505943213e-05,
+      "loss": 4.4495,
+      "step": 174080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7141309111921614e-05,
+      "loss": 4.4664,
+      "step": 174592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7132939543214823e-05,
+      "loss": 4.4457,
+      "step": 175104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.7124553595704303e-05,
+      "loss": 4.459,
+      "step": 175616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.711616764819378e-05,
+      "loss": 4.4541,
+      "step": 176128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.710778170068326e-05,
+      "loss": 4.4534,
+      "step": 176640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709941213197647e-05,
+      "loss": 4.4595,
+      "step": 177152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.709104256326968e-05,
+      "loss": 4.4342,
+      "step": 177664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.708265661575916e-05,
+      "loss": 4.4396,
+      "step": 178176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.707427066824864e-05,
+      "loss": 4.4334,
+      "step": 178688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.706588472073812e-05,
+      "loss": 4.4548,
+      "step": 179200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.70574987732276e-05,
+      "loss": 4.431,
+      "step": 179712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704911282571708e-05,
+      "loss": 4.4573,
+      "step": 180224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.704072687820657e-05,
+      "loss": 4.4324,
+      "step": 180736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.703234093069605e-05,
+      "loss": 4.4326,
+      "step": 181248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.702397136198926e-05,
+      "loss": 4.4335,
+      "step": 181760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.701558541447874e-05,
+      "loss": 4.4301,
+      "step": 182272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.700719946696822e-05,
+      "loss": 4.4377,
+      "step": 182784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.69988135194577e-05,
+      "loss": 4.4316,
+      "step": 183296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6990443950750906e-05,
+      "loss": 4.4208,
+      "step": 183808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6982058003240386e-05,
+      "loss": 4.4367,
+      "step": 184320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6973672055729866e-05,
+      "loss": 4.4247,
+      "step": 184832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.696528610821934e-05,
+      "loss": 4.4277,
+      "step": 185344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6956916539512555e-05,
+      "loss": 4.4245,
+      "step": 185856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6948530592002035e-05,
+      "loss": 4.432,
+      "step": 186368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6940144644491515e-05,
+      "loss": 4.411,
+      "step": 186880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6931758696980995e-05,
+      "loss": 4.437,
+      "step": 187392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6923372749470475e-05,
+      "loss": 4.4218,
+      "step": 187904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6914986801959955e-05,
+      "loss": 4.4249,
+      "step": 188416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6906600854449435e-05,
+      "loss": 4.4279,
+      "step": 188928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6898231285742644e-05,
+      "loss": 4.4174,
+      "step": 189440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6889845338232124e-05,
+      "loss": 4.4357,
+      "step": 189952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6881459390721604e-05,
+      "loss": 4.438,
+      "step": 190464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6873073443211084e-05,
+      "loss": 4.4371,
+      "step": 190976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6864687495700564e-05,
+      "loss": 4.4216,
+      "step": 191488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.6856301548190044e-05,
+      "loss": 4.4115,
+      "step": 192000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.684793197948325e-05,
+      "loss": 4.4046,
+      "step": 192512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683954603197274e-05,
+      "loss": 4.4228,
+      "step": 193024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.683116008446222e-05,
+      "loss": 4.43,
+      "step": 193536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.68227741369517e-05,
+      "loss": 4.4261,
+      "step": 194048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.681438818944118e-05,
+      "loss": 4.4129,
+      "step": 194560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680600224193066e-05,
+      "loss": 4.4032,
+      "step": 195072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.679761629442014e-05,
+      "loss": 4.4162,
+      "step": 195584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678923034690962e-05,
+      "loss": 4.4031,
+      "step": 196096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.678086077820283e-05,
+      "loss": 4.4083,
+      "step": 196608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.677247483069231e-05,
+      "loss": 4.4168,
+      "step": 197120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.676408888318179e-05,
+      "loss": 4.4165,
+      "step": 197632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.675570293567127e-05,
+      "loss": 4.4094,
+      "step": 198144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.674731698816075e-05,
+      "loss": 4.4124,
+      "step": 198656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673893104065023e-05,
+      "loss": 4.3982,
+      "step": 199168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.673054509313971e-05,
+      "loss": 4.4121,
+      "step": 199680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.672215914562919e-05,
+      "loss": 4.4082,
+      "step": 200192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.671380595572613e-05,
+      "loss": 4.3928,
+      "step": 200704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.670542000821561e-05,
+      "loss": 4.3989,
+      "step": 201216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.669703406070509e-05,
+      "loss": 4.4133,
+      "step": 201728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668864811319457e-05,
+      "loss": 4.4144,
+      "step": 202240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.668026216568405e-05,
+      "loss": 4.4001,
+      "step": 202752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.667187621817353e-05,
+      "loss": 4.4019,
+      "step": 203264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6663490270663006e-05,
+      "loss": 4.4075,
+      "step": 203776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6655104323152486e-05,
+      "loss": 4.3892,
+      "step": 204288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.66467347544457e-05,
+      "loss": 4.4063,
+      "step": 204800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6638348806935175e-05,
+      "loss": 4.4113,
+      "step": 205312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662996285942466e-05,
+      "loss": 4.3958,
+      "step": 205824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.662157691191414e-05,
+      "loss": 4.4024,
+      "step": 206336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.661320734320736e-05,
+      "loss": 4.3777,
+      "step": 206848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.660482139569683e-05,
+      "loss": 4.3787,
+      "step": 207360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.659643544818631e-05,
+      "loss": 4.3912,
+      "step": 207872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.658804950067579e-05,
+      "loss": 4.3924,
+      "step": 208384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6579679931969006e-05,
+      "loss": 4.3987,
+      "step": 208896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.657129398445848e-05,
+      "loss": 4.3834,
+      "step": 209408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6562924415751695e-05,
+      "loss": 4.3815,
+      "step": 209920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6554538468241175e-05,
+      "loss": 4.3877,
+      "step": 210432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.654615252073065e-05,
+      "loss": 4.3984,
+      "step": 210944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.653776657322013e-05,
+      "loss": 4.3905,
+      "step": 211456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6529380625709615e-05,
+      "loss": 4.3934,
+      "step": 211968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6520994678199095e-05,
+      "loss": 4.3828,
+      "step": 212480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6512608730688575e-05,
+      "loss": 4.384,
+      "step": 212992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6504222783178055e-05,
+      "loss": 4.3954,
+      "step": 213504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6495853214471264e-05,
+      "loss": 4.3812,
+      "step": 214016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6487467266960744e-05,
+      "loss": 4.382,
+      "step": 214528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.6479081319450224e-05,
+      "loss": 4.3794,
+      "step": 215040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.647071175074343e-05,
+      "loss": 4.3716,
+      "step": 215552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.646232580323291e-05,
+      "loss": 4.3803,
+      "step": 216064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.645393985572239e-05,
+      "loss": 4.38,
+      "step": 216576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.644555390821187e-05,
+      "loss": 4.3893,
+      "step": 217088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.643718433950508e-05,
+      "loss": 4.37,
+      "step": 217600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642879839199457e-05,
+      "loss": 4.3743,
+      "step": 218112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.642041244448405e-05,
+      "loss": 4.3919,
+      "step": 218624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.641202649697353e-05,
+      "loss": 4.374,
+      "step": 219136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.640365692826674e-05,
+      "loss": 4.3887,
+      "step": 219648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.639527098075622e-05,
+      "loss": 4.3764,
+      "step": 220160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.63868850332457e-05,
+      "loss": 4.3921,
+      "step": 220672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637849908573518e-05,
+      "loss": 4.3781,
+      "step": 221184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.637011313822466e-05,
+      "loss": 4.3683,
+      "step": 221696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.636174356951787e-05,
+      "loss": 4.3674,
+      "step": 222208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.635335762200735e-05,
+      "loss": 4.3647,
+      "step": 222720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.634497167449683e-05,
+      "loss": 4.376,
+      "step": 223232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.633658572698631e-05,
+      "loss": 4.3844,
+      "step": 223744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.632819977947579e-05,
+      "loss": 4.3763,
+      "step": 224256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631981383196527e-05,
+      "loss": 4.3735,
+      "step": 224768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.631144426325848e-05,
+      "loss": 4.3668,
+      "step": 225280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.630305831574796e-05,
+      "loss": 4.3644,
+      "step": 225792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.629467236823744e-05,
+      "loss": 4.3598,
+      "step": 226304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.628628642072692e-05,
+      "loss": 4.3698,
+      "step": 226816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.62779004732164e-05,
+      "loss": 4.3816,
+      "step": 227328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626951452570588e-05,
+      "loss": 4.3672,
+      "step": 227840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.626114495699909e-05,
+      "loss": 4.3635,
+      "step": 228352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.625275900948857e-05,
+      "loss": 4.3608,
+      "step": 228864
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.33704948425293,
+      "eval_runtime": 541.6671,
+      "eval_samples_per_second": 704.475,
+      "eval_steps_per_second": 22.015,
+      "step": 228957
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.624437306197805e-05,
+      "loss": 4.3708,
+      "step": 229376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.623598711446753e-05,
+      "loss": 4.3725,
+      "step": 229888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.622760116695701e-05,
+      "loss": 4.3573,
+      "step": 230400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6219215219446484e-05,
+      "loss": 4.3573,
+      "step": 230912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.621082927193597e-05,
+      "loss": 4.3587,
+      "step": 231424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.620245970322919e-05,
+      "loss": 4.3399,
+      "step": 231936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.619407375571867e-05,
+      "loss": 4.3501,
+      "step": 232448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.618568780820814e-05,
+      "loss": 4.3619,
+      "step": 232960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.617730186069762e-05,
+      "loss": 4.3494,
+      "step": 233472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6168932291990836e-05,
+      "loss": 4.3574,
+      "step": 233984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.616054634448031e-05,
+      "loss": 4.3637,
+      "step": 234496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.615216039696979e-05,
+      "loss": 4.3618,
+      "step": 235008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6143790828263005e-05,
+      "loss": 4.3534,
+      "step": 235520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6135404880752485e-05,
+      "loss": 4.3565,
+      "step": 236032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.612701893324196e-05,
+      "loss": 4.3497,
+      "step": 236544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.611863298573144e-05,
+      "loss": 4.3485,
+      "step": 237056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6110247038220925e-05,
+      "loss": 4.3438,
+      "step": 237568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6101861090710405e-05,
+      "loss": 4.3517,
+      "step": 238080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6093475143199885e-05,
+      "loss": 4.3533,
+      "step": 238592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6085089195689365e-05,
+      "loss": 4.3535,
+      "step": 239104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6076703248178845e-05,
+      "loss": 4.3463,
+      "step": 239616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6068317300668325e-05,
+      "loss": 4.3576,
+      "step": 240128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6059931353157805e-05,
+      "loss": 4.3474,
+      "step": 240640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.6051561784451014e-05,
+      "loss": 4.3479,
+      "step": 241152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.604319221574422e-05,
+      "loss": 4.3496,
+      "step": 241664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.60348062682337e-05,
+      "loss": 4.3446,
+      "step": 242176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.602642032072318e-05,
+      "loss": 4.3515,
+      "step": 242688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.601803437321266e-05,
+      "loss": 4.3369,
+      "step": 243200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.600964842570214e-05,
+      "loss": 4.3521,
+      "step": 243712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.600126247819162e-05,
+      "loss": 4.336,
+      "step": 244224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.599289290948484e-05,
+      "loss": 4.337,
+      "step": 244736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.598452334077805e-05,
+      "loss": 4.3422,
+      "step": 245248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.597613739326753e-05,
+      "loss": 4.3479,
+      "step": 245760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.596775144575701e-05,
+      "loss": 4.3352,
+      "step": 246272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595936549824649e-05,
+      "loss": 4.3385,
+      "step": 246784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.595097955073597e-05,
+      "loss": 4.3432,
+      "step": 247296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.594259360322545e-05,
+      "loss": 4.3386,
+      "step": 247808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.593420765571493e-05,
+      "loss": 4.3554,
+      "step": 248320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.592582170820441e-05,
+      "loss": 4.3385,
+      "step": 248832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5917452139497616e-05,
+      "loss": 4.3244,
+      "step": 249344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5909066191987096e-05,
+      "loss": 4.3271,
+      "step": 249856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5900680244476576e-05,
+      "loss": 4.3235,
+      "step": 250368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.589229429696606e-05,
+      "loss": 4.3433,
+      "step": 250880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.588392472825927e-05,
+      "loss": 4.3276,
+      "step": 251392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.587553878074875e-05,
+      "loss": 4.3371,
+      "step": 251904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.586715283323823e-05,
+      "loss": 4.3328,
+      "step": 252416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585876688572771e-05,
+      "loss": 4.3343,
+      "step": 252928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.585038093821719e-05,
+      "loss": 4.3395,
+      "step": 253440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.584199499070667e-05,
+      "loss": 4.3162,
+      "step": 253952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.583362542199988e-05,
+      "loss": 4.317,
+      "step": 254464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.582523947448936e-05,
+      "loss": 4.323,
+      "step": 254976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.581686990578257e-05,
+      "loss": 4.3355,
+      "step": 255488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580848395827205e-05,
+      "loss": 4.3152,
+      "step": 256000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.580009801076153e-05,
+      "loss": 4.3373,
+      "step": 256512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5791712063251017e-05,
+      "loss": 4.3165,
+      "step": 257024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5783326115740497e-05,
+      "loss": 4.3129,
+      "step": 257536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.577494016822997e-05,
+      "loss": 4.3199,
+      "step": 258048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.576655422071945e-05,
+      "loss": 4.3089,
+      "step": 258560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.575816827320893e-05,
+      "loss": 4.3271,
+      "step": 259072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574978232569841e-05,
+      "loss": 4.3188,
+      "step": 259584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.574139637818789e-05,
+      "loss": 4.3067,
+      "step": 260096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.573301043067737e-05,
+      "loss": 4.3216,
+      "step": 260608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.572462448316685e-05,
+      "loss": 4.3102,
+      "step": 261120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.571623853565633e-05,
+      "loss": 4.3164,
+      "step": 261632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.570786896694954e-05,
+      "loss": 4.3086,
+      "step": 262144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.569948301943902e-05,
+      "loss": 4.3209,
+      "step": 262656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56910970719285e-05,
+      "loss": 4.3007,
+      "step": 263168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.568271112441798e-05,
+      "loss": 4.3223,
+      "step": 263680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5674341555711194e-05,
+      "loss": 4.3083,
+      "step": 264192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.56659719870044e-05,
+      "loss": 4.3101,
+      "step": 264704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.565758603949388e-05,
+      "loss": 4.3174,
+      "step": 265216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564920009198336e-05,
+      "loss": 4.3099,
+      "step": 265728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.564081414447284e-05,
+      "loss": 4.3186,
+      "step": 266240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.563242819696232e-05,
+      "loss": 4.3304,
+      "step": 266752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.562405862825553e-05,
+      "loss": 4.325,
+      "step": 267264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.561567268074501e-05,
+      "loss": 4.3113,
+      "step": 267776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.560728673323449e-05,
+      "loss": 4.3054,
+      "step": 268288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559890078572397e-05,
+      "loss": 4.2969,
+      "step": 268800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.559051483821345e-05,
+      "loss": 4.3133,
+      "step": 269312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.558214526950667e-05,
+      "loss": 4.3215,
+      "step": 269824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.557375932199615e-05,
+      "loss": 4.3182,
+      "step": 270336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.556537337448563e-05,
+      "loss": 4.3029,
+      "step": 270848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.555698742697511e-05,
+      "loss": 4.2951,
+      "step": 271360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.554861785826832e-05,
+      "loss": 4.3139,
+      "step": 271872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.55402319107578e-05,
+      "loss": 4.2908,
+      "step": 272384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.553184596324728e-05,
+      "loss": 4.3045,
+      "step": 272896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.552346001573676e-05,
+      "loss": 4.3069,
+      "step": 273408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.551507406822624e-05,
+      "loss": 4.3111,
+      "step": 273920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.5506704499519446e-05,
+      "loss": 4.3042,
+      "step": 274432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5498318552008926e-05,
+      "loss": 4.3097,
+      "step": 274944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5489932604498406e-05,
+      "loss": 4.2914,
+      "step": 275456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5481546656987886e-05,
+      "loss": 4.3032,
+      "step": 275968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.547316070947737e-05,
+      "loss": 4.302,
+      "step": 276480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.546477476196685e-05,
+      "loss": 4.2921,
+      "step": 276992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.545640519326006e-05,
+      "loss": 4.294,
+      "step": 277504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.544801924574954e-05,
+      "loss": 4.3107,
+      "step": 278016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543964967704275e-05,
+      "loss": 4.3074,
+      "step": 278528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.543126372953223e-05,
+      "loss": 4.3038,
+      "step": 279040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.542287778202171e-05,
+      "loss": 4.2965,
+      "step": 279552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.541449183451119e-05,
+      "loss": 4.306,
+      "step": 280064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.540610588700067e-05,
+      "loss": 4.2881,
+      "step": 280576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.539771993949015e-05,
+      "loss": 4.303,
+      "step": 281088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538933399197963e-05,
+      "loss": 4.3092,
+      "step": 281600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.538094804446911e-05,
+      "loss": 4.2914,
+      "step": 282112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.537256209695859e-05,
+      "loss": 4.3076,
+      "step": 282624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.536417614944807e-05,
+      "loss": 4.2723,
+      "step": 283136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.535580658074128e-05,
+      "loss": 4.28,
+      "step": 283648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.534742063323076e-05,
+      "loss": 4.293,
+      "step": 284160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533903468572024e-05,
+      "loss": 4.2929,
+      "step": 284672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.533064873820972e-05,
+      "loss": 4.3006,
+      "step": 285184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.53222627906992e-05,
+      "loss": 4.2832,
+      "step": 285696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.531387684318868e-05,
+      "loss": 4.2822,
+      "step": 286208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.530549089567816e-05,
+      "loss": 4.2893,
+      "step": 286720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.529710494816764e-05,
+      "loss": 4.299,
+      "step": 287232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528873537946085e-05,
+      "loss": 4.2952,
+      "step": 287744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.528034943195033e-05,
+      "loss": 4.2951,
+      "step": 288256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.527196348443981e-05,
+      "loss": 4.2831,
+      "step": 288768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5263577536929295e-05,
+      "loss": 4.2857,
+      "step": 289280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5255207968222504e-05,
+      "loss": 4.3017,
+      "step": 289792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5246822020711984e-05,
+      "loss": 4.2842,
+      "step": 290304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5238436073201464e-05,
+      "loss": 4.2826,
+      "step": 290816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5230050125690944e-05,
+      "loss": 4.2856,
+      "step": 291328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.522168055698415e-05,
+      "loss": 4.2725,
+      "step": 291840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.521329460947363e-05,
+      "loss": 4.2871,
+      "step": 292352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.520490866196311e-05,
+      "loss": 4.284,
+      "step": 292864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.519652271445259e-05,
+      "loss": 4.2948,
+      "step": 293376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.51881531457458e-05,
+      "loss": 4.2747,
+      "step": 293888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517976719823528e-05,
+      "loss": 4.2773,
+      "step": 294400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.517138125072476e-05,
+      "loss": 4.2998,
+      "step": 294912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.516299530321425e-05,
+      "loss": 4.2789,
+      "step": 295424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.515460935570373e-05,
+      "loss": 4.2894,
+      "step": 295936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.514622340819321e-05,
+      "loss": 4.2862,
+      "step": 296448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.513783746068269e-05,
+      "loss": 4.2955,
+      "step": 296960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512945151317217e-05,
+      "loss": 4.2871,
+      "step": 297472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.512108194446538e-05,
+      "loss": 4.2773,
+      "step": 297984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.511269599695486e-05,
+      "loss": 4.2719,
+      "step": 298496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.510431004944434e-05,
+      "loss": 4.2721,
+      "step": 299008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.509592410193382e-05,
+      "loss": 4.2824,
+      "step": 299520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5087554533227026e-05,
+      "loss": 4.2928,
+      "step": 300032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5079168585716506e-05,
+      "loss": 4.2843,
+      "step": 300544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5070782638205986e-05,
+      "loss": 4.2851,
+      "step": 301056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5062413069499195e-05,
+      "loss": 4.2736,
+      "step": 301568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.505402712198868e-05,
+      "loss": 4.2742,
+      "step": 302080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.504564117447816e-05,
+      "loss": 4.2667,
+      "step": 302592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.503725522696764e-05,
+      "loss": 4.282,
+      "step": 303104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5028869279457115e-05,
+      "loss": 4.2848,
+      "step": 303616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5020483331946595e-05,
+      "loss": 4.2852,
+      "step": 304128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5012097384436075e-05,
+      "loss": 4.2705,
+      "step": 304640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.5003711436925555e-05,
+      "loss": 4.2693,
+      "step": 305152
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.255746364593506,
+      "eval_runtime": 538.6451,
+      "eval_samples_per_second": 708.427,
+      "eval_steps_per_second": 22.139,
+      "step": 305276
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4995325489415035e-05,
+      "loss": 4.283,
+      "step": 305664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4986939541904515e-05,
+      "loss": 4.279,
+      "step": 306176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4978553594393995e-05,
+      "loss": 4.2683,
+      "step": 306688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4970167646883475e-05,
+      "loss": 4.2678,
+      "step": 307200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4961781699372955e-05,
+      "loss": 4.2693,
+      "step": 307712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4953395751862435e-05,
+      "loss": 4.2563,
+      "step": 308224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4945009804351915e-05,
+      "loss": 4.2615,
+      "step": 308736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4936623856841395e-05,
+      "loss": 4.2674,
+      "step": 309248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.492823790933088e-05,
+      "loss": 4.2666,
+      "step": 309760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4919851961820354e-05,
+      "loss": 4.2718,
+      "step": 310272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.491148239311357e-05,
+      "loss": 4.2751,
+      "step": 310784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.490309644560305e-05,
+      "loss": 4.2703,
+      "step": 311296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.489471049809253e-05,
+      "loss": 4.268,
+      "step": 311808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4886324550582003e-05,
+      "loss": 4.2718,
+      "step": 312320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4877938603071483e-05,
+      "loss": 4.2652,
+      "step": 312832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486955265556096e-05,
+      "loss": 4.2591,
+      "step": 313344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.486116670805044e-05,
+      "loss": 4.2572,
+      "step": 313856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.485278076053992e-05,
+      "loss": 4.2695,
+      "step": 314368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.48443948130294e-05,
+      "loss": 4.2645,
+      "step": 314880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.483600886551888e-05,
+      "loss": 4.2718,
+      "step": 315392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.482762291800836e-05,
+      "loss": 4.2607,
+      "step": 315904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481923697049785e-05,
+      "loss": 4.2708,
+      "step": 316416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.481085102298733e-05,
+      "loss": 4.2647,
+      "step": 316928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.480248145428054e-05,
+      "loss": 4.2652,
+      "step": 317440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.479409550677002e-05,
+      "loss": 4.2622,
+      "step": 317952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.47857095592595e-05,
+      "loss": 4.2599,
+      "step": 318464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.477732361174898e-05,
+      "loss": 4.267,
+      "step": 318976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476893766423846e-05,
+      "loss": 4.2516,
+      "step": 319488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.476055171672794e-05,
+      "loss": 4.2673,
+      "step": 320000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.475218214802115e-05,
+      "loss": 4.2533,
+      "step": 320512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.474379620051063e-05,
+      "loss": 4.2498,
+      "step": 321024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.473541025300011e-05,
+      "loss": 4.2611,
+      "step": 321536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.472702430548959e-05,
+      "loss": 4.263,
+      "step": 322048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471863835797907e-05,
+      "loss": 4.2565,
+      "step": 322560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.471025241046855e-05,
+      "loss": 4.2537,
+      "step": 323072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4701882841761763e-05,
+      "loss": 4.2588,
+      "step": 323584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4693496894251243e-05,
+      "loss": 4.2619,
+      "step": 324096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4685110946740717e-05,
+      "loss": 4.2737,
+      "step": 324608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4676724999230197e-05,
+      "loss": 4.2544,
+      "step": 325120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4668339051719676e-05,
+      "loss": 4.2459,
+      "step": 325632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.465996948301289e-05,
+      "loss": 4.244,
+      "step": 326144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4651583535502366e-05,
+      "loss": 4.2456,
+      "step": 326656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4643197587991846e-05,
+      "loss": 4.2607,
+      "step": 327168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4634811640481325e-05,
+      "loss": 4.2479,
+      "step": 327680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4626425692970805e-05,
+      "loss": 4.2593,
+      "step": 328192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4618039745460285e-05,
+      "loss": 4.2521,
+      "step": 328704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4609653797949765e-05,
+      "loss": 4.2501,
+      "step": 329216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.460126785043925e-05,
+      "loss": 4.2614,
+      "step": 329728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.459289828173246e-05,
+      "loss": 4.2392,
+      "step": 330240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.458451233422194e-05,
+      "loss": 4.2341,
+      "step": 330752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.457614276551515e-05,
+      "loss": 4.2462,
+      "step": 331264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.456775681800463e-05,
+      "loss": 4.2525,
+      "step": 331776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455937087049411e-05,
+      "loss": 4.2341,
+      "step": 332288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.455098492298359e-05,
+      "loss": 4.2575,
+      "step": 332800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.45426153542768e-05,
+      "loss": 4.238,
+      "step": 333312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.453422940676628e-05,
+      "loss": 4.2333,
+      "step": 333824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.452584345925576e-05,
+      "loss": 4.2412,
+      "step": 334336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.451745751174524e-05,
+      "loss": 4.2308,
+      "step": 334848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4509087943038455e-05,
+      "loss": 4.2519,
+      "step": 335360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4500701995527935e-05,
+      "loss": 4.2431,
+      "step": 335872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4492316048017415e-05,
+      "loss": 4.2274,
+      "step": 336384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4483930100506895e-05,
+      "loss": 4.2403,
+      "step": 336896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4475544152996375e-05,
+      "loss": 4.234,
+      "step": 337408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4467158205485855e-05,
+      "loss": 4.2453,
+      "step": 337920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4458772257975335e-05,
+      "loss": 4.2298,
+      "step": 338432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4450402689268544e-05,
+      "loss": 4.2425,
+      "step": 338944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4442016741758024e-05,
+      "loss": 4.2228,
+      "step": 339456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4433630794247504e-05,
+      "loss": 4.2429,
+      "step": 339968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4425244846736984e-05,
+      "loss": 4.2331,
+      "step": 340480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4416858899226464e-05,
+      "loss": 4.2316,
+      "step": 340992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4408472951715944e-05,
+      "loss": 4.2441,
+      "step": 341504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4400087004205423e-05,
+      "loss": 4.2323,
+      "step": 342016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4391701056694903e-05,
+      "loss": 4.2429,
+      "step": 342528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.438333148798812e-05,
+      "loss": 4.2509,
+      "step": 343040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43749455404776e-05,
+      "loss": 4.2514,
+      "step": 343552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.436655959296708e-05,
+      "loss": 4.2359,
+      "step": 344064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.435817364545655e-05,
+      "loss": 4.2322,
+      "step": 344576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434978769794603e-05,
+      "loss": 4.2222,
+      "step": 345088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.434143450804298e-05,
+      "loss": 4.2376,
+      "step": 345600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.433304856053246e-05,
+      "loss": 4.2518,
+      "step": 346112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.432466261302194e-05,
+      "loss": 4.2415,
+      "step": 346624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.431627666551142e-05,
+      "loss": 4.2316,
+      "step": 347136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.43078907180009e-05,
+      "loss": 4.2181,
+      "step": 347648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429950477049038e-05,
+      "loss": 4.2443,
+      "step": 348160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.429111882297986e-05,
+      "loss": 4.219,
+      "step": 348672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.428274925427307e-05,
+      "loss": 4.2282,
+      "step": 349184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.427436330676255e-05,
+      "loss": 4.2303,
+      "step": 349696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4265977359252026e-05,
+      "loss": 4.2356,
+      "step": 350208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.4257591411741506e-05,
+      "loss": 4.2354,
+      "step": 350720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.424922184303472e-05,
+      "loss": 4.2372,
+      "step": 351232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.42408358955242e-05,
+      "loss": 4.2187,
+      "step": 351744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4232449948013675e-05,
+      "loss": 4.2311,
+      "step": 352256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4224064000503155e-05,
+      "loss": 4.2296,
+      "step": 352768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4215678052992635e-05,
+      "loss": 4.2212,
+      "step": 353280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.420730848428585e-05,
+      "loss": 4.22,
+      "step": 353792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419892253677533e-05,
+      "loss": 4.2401,
+      "step": 354304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.419053658926481e-05,
+      "loss": 4.2338,
+      "step": 354816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.418215064175429e-05,
+      "loss": 4.2326,
+      "step": 355328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.417376469424377e-05,
+      "loss": 4.2232,
+      "step": 355840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.416539512553698e-05,
+      "loss": 4.2312,
+      "step": 356352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.415700917802646e-05,
+      "loss": 4.2188,
+      "step": 356864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414862323051594e-05,
+      "loss": 4.2333,
+      "step": 357376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.414023728300542e-05,
+      "loss": 4.2348,
+      "step": 357888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.413186771429863e-05,
+      "loss": 4.2217,
+      "step": 358400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.412348176678811e-05,
+      "loss": 4.2369,
+      "step": 358912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.411509581927759e-05,
+      "loss": 4.2046,
+      "step": 359424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.410670987176707e-05,
+      "loss": 4.2112,
+      "step": 359936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.409832392425655e-05,
+      "loss": 4.2196,
+      "step": 360448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.408993797674603e-05,
+      "loss": 4.221,
+      "step": 360960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4081568408039244e-05,
+      "loss": 4.233,
+      "step": 361472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4073182460528724e-05,
+      "loss": 4.2182,
+      "step": 361984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4064796513018204e-05,
+      "loss": 4.2097,
+      "step": 362496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4056410565507684e-05,
+      "loss": 4.2171,
+      "step": 363008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4048024617997164e-05,
+      "loss": 4.2303,
+      "step": 363520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4039638670486644e-05,
+      "loss": 4.2273,
+      "step": 364032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.4031252722976124e-05,
+      "loss": 4.2286,
+      "step": 364544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.402288315426933e-05,
+      "loss": 4.2148,
+      "step": 365056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.401449720675881e-05,
+      "loss": 4.2162,
+      "step": 365568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.400611125924829e-05,
+      "loss": 4.2313,
+      "step": 366080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.399772531173777e-05,
+      "loss": 4.2188,
+      "step": 366592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398935574303098e-05,
+      "loss": 4.2144,
+      "step": 367104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.398096979552047e-05,
+      "loss": 4.2179,
+      "step": 367616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.397258384800995e-05,
+      "loss": 4.2059,
+      "step": 368128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.396419790049943e-05,
+      "loss": 4.2195,
+      "step": 368640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.395581195298891e-05,
+      "loss": 4.2192,
+      "step": 369152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.394744238428212e-05,
+      "loss": 4.2258,
+      "step": 369664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.39390564367716e-05,
+      "loss": 4.2087,
+      "step": 370176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.393067048926108e-05,
+      "loss": 4.2113,
+      "step": 370688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.392228454175056e-05,
+      "loss": 4.2281,
+      "step": 371200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.391391497304377e-05,
+      "loss": 4.2163,
+      "step": 371712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.390552902553325e-05,
+      "loss": 4.2238,
+      "step": 372224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.389714307802273e-05,
+      "loss": 4.2222,
+      "step": 372736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388875713051221e-05,
+      "loss": 4.2252,
+      "step": 373248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.388037118300169e-05,
+      "loss": 4.2241,
+      "step": 373760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.387198523549117e-05,
+      "loss": 4.2081,
+      "step": 374272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3863599287980647e-05,
+      "loss": 4.2067,
+      "step": 374784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3855213340470127e-05,
+      "loss": 4.209,
+      "step": 375296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3846827392959606e-05,
+      "loss": 4.2157,
+      "step": 375808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3838441445449086e-05,
+      "loss": 4.2226,
+      "step": 376320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3830071876742296e-05,
+      "loss": 4.2206,
+      "step": 376832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3821702308035505e-05,
+      "loss": 4.2198,
+      "step": 377344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3813316360524985e-05,
+      "loss": 4.2111,
+      "step": 377856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3804930413014465e-05,
+      "loss": 4.2111,
+      "step": 378368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3796544465503944e-05,
+      "loss": 4.2003,
+      "step": 378880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3788158517993424e-05,
+      "loss": 4.2183,
+      "step": 379392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.3779772570482904e-05,
+      "loss": 4.2208,
+      "step": 379904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.377138662297239e-05,
+      "loss": 4.2219,
+      "step": 380416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.376300067546187e-05,
+      "loss": 4.2039,
+      "step": 380928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.375463110675508e-05,
+      "loss": 4.2052,
+      "step": 381440
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.200170993804932,
+      "eval_runtime": 541.6701,
+      "eval_samples_per_second": 704.471,
+      "eval_steps_per_second": 22.015,
+      "step": 381595
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.374626153804829e-05,
+      "loss": 4.2216,
+      "step": 381952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.373787559053777e-05,
+      "loss": 4.213,
+      "step": 382464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372948964302725e-05,
+      "loss": 4.2051,
+      "step": 382976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.372110369551673e-05,
+      "loss": 4.2004,
+      "step": 383488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.371271774800621e-05,
+      "loss": 4.2047,
+      "step": 384000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.370433180049569e-05,
+      "loss": 4.1946,
+      "step": 384512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.369594585298517e-05,
+      "loss": 4.1998,
+      "step": 385024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.368755990547465e-05,
+      "loss": 4.1979,
+      "step": 385536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.367919033676786e-05,
+      "loss": 4.204,
+      "step": 386048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3670820768061074e-05,
+      "loss": 4.2137,
+      "step": 386560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3662434820550554e-05,
+      "loss": 4.2068,
+      "step": 387072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3654048873040034e-05,
+      "loss": 4.2054,
+      "step": 387584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3645662925529514e-05,
+      "loss": 4.2093,
+      "step": 388096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.3637276978018994e-05,
+      "loss": 4.2074,
+      "step": 388608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.36289074093122e-05,
+      "loss": 4.2009,
+      "step": 389120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.362052146180168e-05,
+      "loss": 4.1961,
+      "step": 389632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.361215189309489e-05,
+      "loss": 4.1956,
+      "step": 390144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.360376594558437e-05,
+      "loss": 4.2041,
+      "step": 390656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.359537999807385e-05,
+      "loss": 4.2044,
+      "step": 391168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.358699405056333e-05,
+      "loss": 4.2128,
+      "step": 391680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.357860810305281e-05,
+      "loss": 4.1991,
+      "step": 392192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.35702221555423e-05,
+      "loss": 4.2097,
+      "step": 392704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.356183620803178e-05,
+      "loss": 4.2066,
+      "step": 393216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.355345026052126e-05,
+      "loss": 4.2026,
+      "step": 393728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.354506431301074e-05,
+      "loss": 4.2003,
+      "step": 394240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.353667836550022e-05,
+      "loss": 4.197,
+      "step": 394752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.352830879679343e-05,
+      "loss": 4.2077,
+      "step": 395264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351992284928291e-05,
+      "loss": 4.1882,
+      "step": 395776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.351153690177239e-05,
+      "loss": 4.2097,
+      "step": 396288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.350315095426187e-05,
+      "loss": 4.192,
+      "step": 396800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.349476500675134e-05,
+      "loss": 4.187,
+      "step": 397312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.348637905924082e-05,
+      "loss": 4.2042,
+      "step": 397824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.34779931117303e-05,
+      "loss": 4.2004,
+      "step": 398336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3469623543023516e-05,
+      "loss": 4.2019,
+      "step": 398848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3461237595512996e-05,
+      "loss": 4.1937,
+      "step": 399360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3452851648002476e-05,
+      "loss": 4.1936,
+      "step": 399872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3444465700491956e-05,
+      "loss": 4.1992,
+      "step": 400384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3436079752981436e-05,
+      "loss": 4.2142,
+      "step": 400896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3427693805470916e-05,
+      "loss": 4.1959,
+      "step": 401408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3419307857960396e-05,
+      "loss": 4.1863,
+      "step": 401920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3410921910449876e-05,
+      "loss": 4.1866,
+      "step": 402432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3402552341743085e-05,
+      "loss": 4.1835,
+      "step": 402944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3394166394232565e-05,
+      "loss": 4.198,
+      "step": 403456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3385780446722045e-05,
+      "loss": 4.1932,
+      "step": 403968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3377410878015254e-05,
+      "loss": 4.1956,
+      "step": 404480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3369024930504734e-05,
+      "loss": 4.1968,
+      "step": 404992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3360638982994214e-05,
+      "loss": 4.1906,
+      "step": 405504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33522530354837e-05,
+      "loss": 4.202,
+      "step": 406016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.334386708797318e-05,
+      "loss": 4.1815,
+      "step": 406528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.333548114046266e-05,
+      "loss": 4.1742,
+      "step": 407040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.332709519295214e-05,
+      "loss": 4.1916,
+      "step": 407552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.331870924544162e-05,
+      "loss": 4.1871,
+      "step": 408064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.33103232979311e-05,
+      "loss": 4.1797,
+      "step": 408576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.330195372922431e-05,
+      "loss": 4.1973,
+      "step": 409088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.329358416051752e-05,
+      "loss": 4.1787,
+      "step": 409600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3285198213007e-05,
+      "loss": 4.1784,
+      "step": 410112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.327681226549648e-05,
+      "loss": 4.1798,
+      "step": 410624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326842631798596e-05,
+      "loss": 4.1744,
+      "step": 411136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.326004037047544e-05,
+      "loss": 4.1932,
+      "step": 411648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.325165442296492e-05,
+      "loss": 4.1829,
+      "step": 412160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.32432684754544e-05,
+      "loss": 4.1794,
+      "step": 412672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3234882527943885e-05,
+      "loss": 4.1782,
+      "step": 413184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3226512959237094e-05,
+      "loss": 4.1714,
+      "step": 413696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3218127011726574e-05,
+      "loss": 4.1922,
+      "step": 414208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3209741064216054e-05,
+      "loss": 4.1742,
+      "step": 414720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.320137149550926e-05,
+      "loss": 4.1907,
+      "step": 415232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.319298554799874e-05,
+      "loss": 4.1648,
+      "step": 415744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.318459960048822e-05,
+      "loss": 4.1837,
+      "step": 416256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.31762136529777e-05,
+      "loss": 4.1783,
+      "step": 416768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3167827705467176e-05,
+      "loss": 4.1731,
+      "step": 417280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3159441757956656e-05,
+      "loss": 4.1865,
+      "step": 417792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3151055810446136e-05,
+      "loss": 4.1789,
+      "step": 418304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.314266986293562e-05,
+      "loss": 4.184,
+      "step": 418816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.313430029422883e-05,
+      "loss": 4.1931,
+      "step": 419328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.312593072552205e-05,
+      "loss": 4.1943,
+      "step": 419840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.311754477801153e-05,
+      "loss": 4.1846,
+      "step": 420352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3109158830501e-05,
+      "loss": 4.177,
+      "step": 420864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.310077288299048e-05,
+      "loss": 4.1689,
+      "step": 421376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.30924033142837e-05,
+      "loss": 4.18,
+      "step": 421888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.308401736677318e-05,
+      "loss": 4.1922,
+      "step": 422400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.307563141926265e-05,
+      "loss": 4.1859,
+      "step": 422912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.306724547175213e-05,
+      "loss": 4.1785,
+      "step": 423424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3058875903045346e-05,
+      "loss": 4.161,
+      "step": 423936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3050489955534826e-05,
+      "loss": 4.1916,
+      "step": 424448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3042104008024306e-05,
+      "loss": 4.1595,
+      "step": 424960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3033718060513786e-05,
+      "loss": 4.1765,
+      "step": 425472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3025332113003266e-05,
+      "loss": 4.1761,
+      "step": 425984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3016946165492746e-05,
+      "loss": 4.181,
+      "step": 426496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 4.3008560217982226e-05,
+      "loss": 4.1773,
+      "step": 427008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.3000174270471705e-05,
+      "loss": 4.1871,
+      "step": 427520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2991788322961185e-05,
+      "loss": 4.1618,
+      "step": 428032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2983402375450665e-05,
+      "loss": 4.1742,
+      "step": 428544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2975016427940145e-05,
+      "loss": 4.1784,
+      "step": 429056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2966630480429625e-05,
+      "loss": 4.1686,
+      "step": 429568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2958260911722834e-05,
+      "loss": 4.166,
+      "step": 430080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2949874964212314e-05,
+      "loss": 4.1849,
+      "step": 430592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.294150539550553e-05,
+      "loss": 4.1785,
+      "step": 431104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.293311944799501e-05,
+      "loss": 4.1857,
+      "step": 431616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.292473350048449e-05,
+      "loss": 4.1702,
+      "step": 432128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.291634755297397e-05,
+      "loss": 4.1773,
+      "step": 432640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.290796160546345e-05,
+      "loss": 4.1637,
+      "step": 433152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289957565795293e-05,
+      "loss": 4.179,
+      "step": 433664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.289118971044241e-05,
+      "loss": 4.1845,
+      "step": 434176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.288280376293189e-05,
+      "loss": 4.167,
+      "step": 434688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.287441781542136e-05,
+      "loss": 4.1833,
+      "step": 435200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.286604824671458e-05,
+      "loss": 4.1502,
+      "step": 435712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.285766229920406e-05,
+      "loss": 4.1622,
+      "step": 436224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284929273049727e-05,
+      "loss": 4.1626,
+      "step": 436736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.284090678298675e-05,
+      "loss": 4.1689,
+      "step": 437248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.283252083547623e-05,
+      "loss": 4.1801,
+      "step": 437760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.282413488796571e-05,
+      "loss": 4.1709,
+      "step": 438272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2815765319258924e-05,
+      "loss": 4.1481,
+      "step": 438784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2807379371748404e-05,
+      "loss": 4.1699,
+      "step": 439296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2798993424237884e-05,
+      "loss": 4.1746,
+      "step": 439808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2790607476727364e-05,
+      "loss": 4.1763,
+      "step": 440320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.278222152921684e-05,
+      "loss": 4.1738,
+      "step": 440832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.277383558170632e-05,
+      "loss": 4.1639,
+      "step": 441344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.27654496341958e-05,
+      "loss": 4.163,
+      "step": 441856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.275708006548901e-05,
+      "loss": 4.1829,
+      "step": 442368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.274871049678222e-05,
+      "loss": 4.1647,
+      "step": 442880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.27403245492717e-05,
+      "loss": 4.1648,
+      "step": 443392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.273193860176118e-05,
+      "loss": 4.1642,
+      "step": 443904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.272355265425066e-05,
+      "loss": 4.1559,
+      "step": 444416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.271516670674014e-05,
+      "loss": 4.1661,
+      "step": 444928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.270678075922962e-05,
+      "loss": 4.1671,
+      "step": 445440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26983948117191e-05,
+      "loss": 4.1742,
+      "step": 445952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.269000886420858e-05,
+      "loss": 4.1603,
+      "step": 446464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.268162291669806e-05,
+      "loss": 4.159,
+      "step": 446976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.267323696918754e-05,
+      "loss": 4.1761,
+      "step": 447488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.266485102167702e-05,
+      "loss": 4.1622,
+      "step": 448000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.26564650741665e-05,
+      "loss": 4.1696,
+      "step": 448512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.264807912665598e-05,
+      "loss": 4.1723,
+      "step": 449024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263970955794919e-05,
+      "loss": 4.1783,
+      "step": 449536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.263132361043867e-05,
+      "loss": 4.1719,
+      "step": 450048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.262293766292815e-05,
+      "loss": 4.1595,
+      "step": 450560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.261455171541763e-05,
+      "loss": 4.1511,
+      "step": 451072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2606182146710846e-05,
+      "loss": 4.1603,
+      "step": 451584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2597796199200326e-05,
+      "loss": 4.1668,
+      "step": 452096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2589410251689806e-05,
+      "loss": 4.1748,
+      "step": 452608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2581024304179286e-05,
+      "loss": 4.1674,
+      "step": 453120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2572638356668766e-05,
+      "loss": 4.1647,
+      "step": 453632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2564268787961975e-05,
+      "loss": 4.166,
+      "step": 454144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2555882840451455e-05,
+      "loss": 4.1629,
+      "step": 454656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2547496892940935e-05,
+      "loss": 4.1522,
+      "step": 455168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2539110945430415e-05,
+      "loss": 4.1609,
+      "step": 455680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2530741376723624e-05,
+      "loss": 4.1731,
+      "step": 456192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2522355429213104e-05,
+      "loss": 4.1733,
+      "step": 456704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.2513969481702584e-05,
+      "loss": 4.1536,
+      "step": 457216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 4.250558353419207e-05,
+      "loss": 4.1539,
+      "step": 457728
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.158684730529785,
+      "eval_runtime": 538.9879,
+      "eval_samples_per_second": 707.977,
+      "eval_steps_per_second": 22.125,
+      "step": 457914
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.249719758668155e-05,
+      "loss": 4.1794,
+      "step": 458240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2488811639171024e-05,
+      "loss": 4.163,
+      "step": 458752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2480425691660504e-05,
+      "loss": 4.1558,
+      "step": 459264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2472039744149984e-05,
+      "loss": 4.1562,
+      "step": 459776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2463653796639464e-05,
+      "loss": 4.1534,
+      "step": 460288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2455267849128943e-05,
+      "loss": 4.1507,
+      "step": 460800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2446881901618423e-05,
+      "loss": 4.148,
+      "step": 461312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.24384959541079e-05,
+      "loss": 4.1524,
+      "step": 461824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.243011000659738e-05,
+      "loss": 4.1561,
+      "step": 462336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.242172405908686e-05,
+      "loss": 4.163,
+      "step": 462848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.241333811157634e-05,
+      "loss": 4.1618,
+      "step": 463360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.240495216406582e-05,
+      "loss": 4.1513,
+      "step": 463872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.239658259535903e-05,
+      "loss": 4.167,
+      "step": 464384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.238819664784852e-05,
+      "loss": 4.1564,
+      "step": 464896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2379810700338e-05,
+      "loss": 4.1567,
+      "step": 465408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.237142475282748e-05,
+      "loss": 4.1439,
+      "step": 465920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.236303880531696e-05,
+      "loss": 4.1453,
+      "step": 466432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.235465285780644e-05,
+      "loss": 4.159,
+      "step": 466944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.234626691029591e-05,
+      "loss": 4.1564,
+      "step": 467456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.233788096278539e-05,
+      "loss": 4.1635,
+      "step": 467968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232949501527487e-05,
+      "loss": 4.1521,
+      "step": 468480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.232110906776435e-05,
+      "loss": 4.1624,
+      "step": 468992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.231273949905756e-05,
+      "loss": 4.1585,
+      "step": 469504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.230436993035078e-05,
+      "loss": 4.1557,
+      "step": 470016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.229598398284026e-05,
+      "loss": 4.1535,
+      "step": 470528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.228759803532974e-05,
+      "loss": 4.1507,
+      "step": 471040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.227921208781922e-05,
+      "loss": 4.1597,
+      "step": 471552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.22708261403087e-05,
+      "loss": 4.1426,
+      "step": 472064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.226244019279818e-05,
+      "loss": 4.1613,
+      "step": 472576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2254070624091386e-05,
+      "loss": 4.1492,
+      "step": 473088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2245684676580866e-05,
+      "loss": 4.1406,
+      "step": 473600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2237298729070346e-05,
+      "loss": 4.1552,
+      "step": 474112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2228912781559826e-05,
+      "loss": 4.1549,
+      "step": 474624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2220526834049306e-05,
+      "loss": 4.1597,
+      "step": 475136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2212140886538786e-05,
+      "loss": 4.1451,
+      "step": 475648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2203754939028265e-05,
+      "loss": 4.15,
+      "step": 476160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2195368991517745e-05,
+      "loss": 4.1519,
+      "step": 476672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2186983044007225e-05,
+      "loss": 4.1647,
+      "step": 477184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2178597096496705e-05,
+      "loss": 4.1522,
+      "step": 477696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2170211148986185e-05,
+      "loss": 4.1393,
+      "step": 478208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.216182520147567e-05,
+      "loss": 4.1378,
+      "step": 478720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.215345563276888e-05,
+      "loss": 4.1369,
+      "step": 479232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.214506968525836e-05,
+      "loss": 4.1561,
+      "step": 479744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.213668373774784e-05,
+      "loss": 4.1479,
+      "step": 480256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.212829779023732e-05,
+      "loss": 4.1465,
+      "step": 480768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.211992822153053e-05,
+      "loss": 4.1554,
+      "step": 481280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.211154227402001e-05,
+      "loss": 4.1456,
+      "step": 481792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.210315632650949e-05,
+      "loss": 4.1565,
+      "step": 482304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.209477037899897e-05,
+      "loss": 4.1378,
+      "step": 482816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.208638443148845e-05,
+      "loss": 4.131,
+      "step": 483328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.207801486278166e-05,
+      "loss": 4.1431,
+      "step": 483840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.206962891527114e-05,
+      "loss": 4.139,
+      "step": 484352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2061259346564355e-05,
+      "loss": 4.1392,
+      "step": 484864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2052873399053835e-05,
+      "loss": 4.1479,
+      "step": 485376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2044503830347044e-05,
+      "loss": 4.1388,
+      "step": 485888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2036117882836524e-05,
+      "loss": 4.134,
+      "step": 486400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2027731935326004e-05,
+      "loss": 4.1335,
+      "step": 486912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2019345987815484e-05,
+      "loss": 4.1345,
+      "step": 487424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2010960040304964e-05,
+      "loss": 4.1449,
+      "step": 487936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.2002574092794444e-05,
+      "loss": 4.1376,
+      "step": 488448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1994188145283924e-05,
+      "loss": 4.132,
+      "step": 488960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.19858021977734e-05,
+      "loss": 4.1363,
+      "step": 489472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.197741625026288e-05,
+      "loss": 4.123,
+      "step": 489984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1969030302752363e-05,
+      "loss": 4.1498,
+      "step": 490496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1960644355241843e-05,
+      "loss": 4.1323,
+      "step": 491008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.195227478653505e-05,
+      "loss": 4.1421,
+      "step": 491520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.194388883902453e-05,
+      "loss": 4.1285,
+      "step": 492032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.193550289151401e-05,
+      "loss": 4.1348,
+      "step": 492544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.192711694400349e-05,
+      "loss": 4.1374,
+      "step": 493056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191873099649297e-05,
+      "loss": 4.1301,
+      "step": 493568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.191034504898245e-05,
+      "loss": 4.1433,
+      "step": 494080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.190195910147193e-05,
+      "loss": 4.1335,
+      "step": 494592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.189357315396141e-05,
+      "loss": 4.1417,
+      "step": 495104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.188520358525462e-05,
+      "loss": 4.1487,
+      "step": 495616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18768176377441e-05,
+      "loss": 4.147,
+      "step": 496128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186843169023358e-05,
+      "loss": 4.1439,
+      "step": 496640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.186004574272306e-05,
+      "loss": 4.1343,
+      "step": 497152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.185165979521254e-05,
+      "loss": 4.1267,
+      "step": 497664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.184329022650576e-05,
+      "loss": 4.1347,
+      "step": 498176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.183490427899524e-05,
+      "loss": 4.1467,
+      "step": 498688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.182651833148472e-05,
+      "loss": 4.1429,
+      "step": 499200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.18181323839742e-05,
+      "loss": 4.1341,
+      "step": 499712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.180974643646368e-05,
+      "loss": 4.1188,
+      "step": 500224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1801376867756886e-05,
+      "loss": 4.1493,
+      "step": 500736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1792990920246366e-05,
+      "loss": 4.1158,
+      "step": 501248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1784604972735846e-05,
+      "loss": 4.134,
+      "step": 501760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1776219025225326e-05,
+      "loss": 4.1364,
+      "step": 502272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1767849456518535e-05,
+      "loss": 4.134,
+      "step": 502784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.1759463509008015e-05,
+      "loss": 4.1382,
+      "step": 503296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1751077561497495e-05,
+      "loss": 4.1448,
+      "step": 503808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.174269161398698e-05,
+      "loss": 4.1195,
+      "step": 504320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.173432204528019e-05,
+      "loss": 4.1343,
+      "step": 504832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.172593609776967e-05,
+      "loss": 4.1365,
+      "step": 505344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.171755015025915e-05,
+      "loss": 4.1258,
+      "step": 505856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170916420274863e-05,
+      "loss": 4.1228,
+      "step": 506368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.170077825523811e-05,
+      "loss": 4.1434,
+      "step": 506880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.169240868653132e-05,
+      "loss": 4.1355,
+      "step": 507392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.16840227390208e-05,
+      "loss": 4.1461,
+      "step": 507904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.167563679151028e-05,
+      "loss": 4.1283,
+      "step": 508416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.166725084399975e-05,
+      "loss": 4.1356,
+      "step": 508928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165888127529297e-05,
+      "loss": 4.1213,
+      "step": 509440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.165049532778245e-05,
+      "loss": 4.1366,
+      "step": 509952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1642109380271935e-05,
+      "loss": 4.1418,
+      "step": 510464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.163372343276141e-05,
+      "loss": 4.1328,
+      "step": 510976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1625353864054624e-05,
+      "loss": 4.1398,
+      "step": 511488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1616967916544104e-05,
+      "loss": 4.1122,
+      "step": 512000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1608581969033584e-05,
+      "loss": 4.1167,
+      "step": 512512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.160019602152306e-05,
+      "loss": 4.1186,
+      "step": 513024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.159181007401254e-05,
+      "loss": 4.1274,
+      "step": 513536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.158342412650202e-05,
+      "loss": 4.1383,
+      "step": 514048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.15750381789915e-05,
+      "loss": 4.1349,
+      "step": 514560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.156665223148098e-05,
+      "loss": 4.1077,
+      "step": 515072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1558282662774186e-05,
+      "loss": 4.1276,
+      "step": 515584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154989671526367e-05,
+      "loss": 4.1357,
+      "step": 516096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.154151076775315e-05,
+      "loss": 4.1371,
+      "step": 516608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.153312482024263e-05,
+      "loss": 4.1334,
+      "step": 517120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.152473887273211e-05,
+      "loss": 4.1245,
+      "step": 517632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.151635292522159e-05,
+      "loss": 4.1259,
+      "step": 518144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.150799973531853e-05,
+      "loss": 4.137,
+      "step": 518656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149961378780801e-05,
+      "loss": 4.1279,
+      "step": 519168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.149122784029749e-05,
+      "loss": 4.1244,
+      "step": 519680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.148284189278697e-05,
+      "loss": 4.1202,
+      "step": 520192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.147445594527645e-05,
+      "loss": 4.1156,
+      "step": 520704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.146606999776593e-05,
+      "loss": 4.1284,
+      "step": 521216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.145770042905914e-05,
+      "loss": 4.1285,
+      "step": 521728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144931448154863e-05,
+      "loss": 4.1305,
+      "step": 522240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.144092853403811e-05,
+      "loss": 4.1233,
+      "step": 522752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1432542586527587e-05,
+      "loss": 4.1184,
+      "step": 523264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1424156639017067e-05,
+      "loss": 4.1366,
+      "step": 523776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1415770691506546e-05,
+      "loss": 4.1184,
+      "step": 524288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1407384743996026e-05,
+      "loss": 4.1324,
+      "step": 524800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1398998796485506e-05,
+      "loss": 4.1363,
+      "step": 525312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1390629227778716e-05,
+      "loss": 4.1322,
+      "step": 525824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1382243280268195e-05,
+      "loss": 4.1357,
+      "step": 526336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1373873711561405e-05,
+      "loss": 4.1239,
+      "step": 526848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1365487764050885e-05,
+      "loss": 4.1106,
+      "step": 527360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.1357101816540364e-05,
+      "loss": 4.1188,
+      "step": 527872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134873224783358e-05,
+      "loss": 4.1247,
+      "step": 528384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.134034630032306e-05,
+      "loss": 4.1391,
+      "step": 528896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.133196035281254e-05,
+      "loss": 4.1271,
+      "step": 529408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.132357440530202e-05,
+      "loss": 4.1303,
+      "step": 529920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.13151884577915e-05,
+      "loss": 4.1221,
+      "step": 530432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.130680251028098e-05,
+      "loss": 4.1244,
+      "step": 530944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129841656277046e-05,
+      "loss": 4.1104,
+      "step": 531456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.129003061525994e-05,
+      "loss": 4.1234,
+      "step": 531968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.128164466774942e-05,
+      "loss": 4.1354,
+      "step": 532480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.127325872023889e-05,
+      "loss": 4.1353,
+      "step": 532992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.126488915153211e-05,
+      "loss": 4.1134,
+      "step": 533504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.125650320402159e-05,
+      "loss": 4.1188,
+      "step": 534016
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.127756595611572,
+      "eval_runtime": 538.0808,
+      "eval_samples_per_second": 709.17,
+      "eval_steps_per_second": 22.162,
+      "step": 534233
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.124811725651106e-05,
+      "loss": 4.1415,
+      "step": 534528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.123973130900054e-05,
+      "loss": 4.1235,
+      "step": 535040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.123134536149003e-05,
+      "loss": 4.1167,
+      "step": 535552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.122295941397951e-05,
+      "loss": 4.121,
+      "step": 536064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.121457346646899e-05,
+      "loss": 4.1109,
+      "step": 536576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.120618751895847e-05,
+      "loss": 4.1157,
+      "step": 537088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.119780157144795e-05,
+      "loss": 4.1043,
+      "step": 537600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.118941562393743e-05,
+      "loss": 4.1174,
+      "step": 538112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.118102967642691e-05,
+      "loss": 4.1139,
+      "step": 538624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.117264372891639e-05,
+      "loss": 4.1246,
+      "step": 539136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.116425778140587e-05,
+      "loss": 4.1211,
+      "step": 539648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.115587183389535e-05,
+      "loss": 4.1172,
+      "step": 540160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.114748588638483e-05,
+      "loss": 4.1285,
+      "step": 540672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.11390999388743e-05,
+      "loss": 4.1159,
+      "step": 541184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.113071399136378e-05,
+      "loss": 4.1238,
+      "step": 541696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1122344422657e-05,
+      "loss": 4.1035,
+      "step": 542208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.111395847514648e-05,
+      "loss": 4.1095,
+      "step": 542720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.110557252763596e-05,
+      "loss": 4.12,
+      "step": 543232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.109720295892917e-05,
+      "loss": 4.1142,
+      "step": 543744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108881701141865e-05,
+      "loss": 4.1268,
+      "step": 544256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.108043106390813e-05,
+      "loss": 4.116,
+      "step": 544768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1072045116397606e-05,
+      "loss": 4.1224,
+      "step": 545280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1063659168887086e-05,
+      "loss": 4.1206,
+      "step": 545792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1055273221376566e-05,
+      "loss": 4.1126,
+      "step": 546304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1046887273866046e-05,
+      "loss": 4.1199,
+      "step": 546816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1038501326355526e-05,
+      "loss": 4.1099,
+      "step": 547328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1030115378845006e-05,
+      "loss": 4.123,
+      "step": 547840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1021729431334486e-05,
+      "loss": 4.1124,
+      "step": 548352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1013343483823966e-05,
+      "loss": 4.1221,
+      "step": 548864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.1004957536313446e-05,
+      "loss": 4.109,
+      "step": 549376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0996571588802926e-05,
+      "loss": 4.1063,
+      "step": 549888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0988185641292406e-05,
+      "loss": 4.119,
+      "step": 550400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.097979969378189e-05,
+      "loss": 4.1162,
+      "step": 550912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.09714301250751e-05,
+      "loss": 4.1193,
+      "step": 551424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.096304417756458e-05,
+      "loss": 4.115,
+      "step": 551936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.095467460885779e-05,
+      "loss": 4.1082,
+      "step": 552448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.094628866134727e-05,
+      "loss": 4.1172,
+      "step": 552960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.093790271383675e-05,
+      "loss": 4.1261,
+      "step": 553472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.092951676632623e-05,
+      "loss": 4.1153,
+      "step": 553984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.092113081881571e-05,
+      "loss": 4.1061,
+      "step": 554496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.091274487130519e-05,
+      "loss": 4.1015,
+      "step": 555008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0904358923794664e-05,
+      "loss": 4.1021,
+      "step": 555520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.089597297628415e-05,
+      "loss": 4.1198,
+      "step": 556032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.088760340757736e-05,
+      "loss": 4.1111,
+      "step": 556544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0879217460066846e-05,
+      "loss": 4.1056,
+      "step": 557056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0870847891360055e-05,
+      "loss": 4.1248,
+      "step": 557568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0862461943849535e-05,
+      "loss": 4.107,
+      "step": 558080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0854075996339015e-05,
+      "loss": 4.1202,
+      "step": 558592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0845690048828495e-05,
+      "loss": 4.1037,
+      "step": 559104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0837320480121704e-05,
+      "loss": 4.0968,
+      "step": 559616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0828934532611184e-05,
+      "loss": 4.105,
+      "step": 560128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0820548585100664e-05,
+      "loss": 4.1063,
+      "step": 560640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.081216263759014e-05,
+      "loss": 4.1005,
+      "step": 561152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.080377669007962e-05,
+      "loss": 4.1152,
+      "step": 561664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.07953907425691e-05,
+      "loss": 4.1042,
+      "step": 562176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.078702117386231e-05,
+      "loss": 4.095,
+      "step": 562688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077863522635179e-05,
+      "loss": 4.1035,
+      "step": 563200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.077024927884127e-05,
+      "loss": 4.0931,
+      "step": 563712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.076186333133075e-05,
+      "loss": 4.1086,
+      "step": 564224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.075347738382023e-05,
+      "loss": 4.1029,
+      "step": 564736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.074509143630971e-05,
+      "loss": 4.0999,
+      "step": 565248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.073672186760292e-05,
+      "loss": 4.0979,
+      "step": 565760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.07283359200924e-05,
+      "loss": 4.0911,
+      "step": 566272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.071994997258188e-05,
+      "loss": 4.114,
+      "step": 566784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.071156402507136e-05,
+      "loss": 4.0952,
+      "step": 567296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.070317807756084e-05,
+      "loss": 4.1079,
+      "step": 567808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.069479213005032e-05,
+      "loss": 4.0938,
+      "step": 568320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.06864061825398e-05,
+      "loss": 4.1006,
+      "step": 568832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.067802023502928e-05,
+      "loss": 4.0979,
+      "step": 569344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.06696506663225e-05,
+      "loss": 4.1013,
+      "step": 569856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.066126471881198e-05,
+      "loss": 4.1069,
+      "step": 570368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.065287877130146e-05,
+      "loss": 4.099,
+      "step": 570880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.064449282379094e-05,
+      "loss": 4.1055,
+      "step": 571392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0636123255084147e-05,
+      "loss": 4.1159,
+      "step": 571904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0627737307573627e-05,
+      "loss": 4.1157,
+      "step": 572416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0619351360063106e-05,
+      "loss": 4.106,
+      "step": 572928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0610965412552586e-05,
+      "loss": 4.1047,
+      "step": 573440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0602579465042066e-05,
+      "loss": 4.0864,
+      "step": 573952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.0594226275139005e-05,
+      "loss": 4.1023,
+      "step": 574464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.058584032762849e-05,
+      "loss": 4.1146,
+      "step": 574976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.057745438011797e-05,
+      "loss": 4.108,
+      "step": 575488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056906843260745e-05,
+      "loss": 4.1055,
+      "step": 576000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.056069886390066e-05,
+      "loss": 4.0833,
+      "step": 576512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.055231291639014e-05,
+      "loss": 4.1099,
+      "step": 577024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.054392696887962e-05,
+      "loss": 4.0856,
+      "step": 577536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.05355410213691e-05,
+      "loss": 4.0988,
+      "step": 578048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.052715507385858e-05,
+      "loss": 4.0973,
+      "step": 578560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.051876912634806e-05,
+      "loss": 4.1039,
+      "step": 579072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.051038317883754e-05,
+      "loss": 4.1054,
+      "step": 579584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.050201361013075e-05,
+      "loss": 4.1059,
+      "step": 580096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.049362766262023e-05,
+      "loss": 4.0892,
+      "step": 580608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.048524171510971e-05,
+      "loss": 4.1047,
+      "step": 581120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.047685576759919e-05,
+      "loss": 4.0977,
+      "step": 581632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0468486198892405e-05,
+      "loss": 4.0936,
+      "step": 582144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0460100251381885e-05,
+      "loss": 4.0866,
+      "step": 582656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0451714303871365e-05,
+      "loss": 4.1069,
+      "step": 583168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0443328356360845e-05,
+      "loss": 4.1024,
+      "step": 583680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0434942408850325e-05,
+      "loss": 4.1122,
+      "step": 584192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0426572840143534e-05,
+      "loss": 4.0977,
+      "step": 584704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0418186892633014e-05,
+      "loss": 4.1021,
+      "step": 585216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0409800945122494e-05,
+      "loss": 4.0864,
+      "step": 585728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0401414997611974e-05,
+      "loss": 4.1079,
+      "step": 586240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.039304542890518e-05,
+      "loss": 4.1066,
+      "step": 586752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.038465948139466e-05,
+      "loss": 4.1005,
+      "step": 587264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.037627353388414e-05,
+      "loss": 4.1012,
+      "step": 587776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.036788758637363e-05,
+      "loss": 4.0831,
+      "step": 588288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.03595016388631e-05,
+      "loss": 4.082,
+      "step": 588800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.035111569135258e-05,
+      "loss": 4.0872,
+      "step": 589312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.034272974384206e-05,
+      "loss": 4.0947,
+      "step": 589824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.033434379633154e-05,
+      "loss": 4.1061,
+      "step": 590336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.032597422762475e-05,
+      "loss": 4.0993,
+      "step": 590848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.031758828011423e-05,
+      "loss": 4.0725,
+      "step": 591360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.030920233260371e-05,
+      "loss": 4.098,
+      "step": 591872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.030083276389692e-05,
+      "loss": 4.1029,
+      "step": 592384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.02924468163864e-05,
+      "loss": 4.1006,
+      "step": 592896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.028406086887588e-05,
+      "loss": 4.1009,
+      "step": 593408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.027567492136537e-05,
+      "loss": 4.0941,
+      "step": 593920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.026728897385485e-05,
+      "loss": 4.0883,
+      "step": 594432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.025890302634433e-05,
+      "loss": 4.105,
+      "step": 594944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.025051707883381e-05,
+      "loss": 4.097,
+      "step": 595456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.024213113132329e-05,
+      "loss": 4.0904,
+      "step": 595968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.023374518381277e-05,
+      "loss": 4.0857,
+      "step": 596480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.022535923630225e-05,
+      "loss": 4.0865,
+      "step": 596992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.021697328879173e-05,
+      "loss": 4.0938,
+      "step": 597504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.020858734128121e-05,
+      "loss": 4.0928,
+      "step": 598016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.020020139377069e-05,
+      "loss": 4.1017,
+      "step": 598528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0191831825063896e-05,
+      "loss": 4.0891,
+      "step": 599040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0183445877553376e-05,
+      "loss": 4.0874,
+      "step": 599552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0175059930042856e-05,
+      "loss": 4.1058,
+      "step": 600064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0166673982532336e-05,
+      "loss": 4.0845,
+      "step": 600576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0158304413825545e-05,
+      "loss": 4.1024,
+      "step": 601088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.014991846631503e-05,
+      "loss": 4.1038,
+      "step": 601600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.014153251880451e-05,
+      "loss": 4.0984,
+      "step": 602112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.013314657129399e-05,
+      "loss": 4.1025,
+      "step": 602624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.01247770025872e-05,
+      "loss": 4.0918,
+      "step": 603136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.011640743388041e-05,
+      "loss": 4.0819,
+      "step": 603648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.010802148636989e-05,
+      "loss": 4.0864,
+      "step": 604160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.009963553885937e-05,
+      "loss": 4.097,
+      "step": 604672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.009124959134885e-05,
+      "loss": 4.1008,
+      "step": 605184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.008286364383833e-05,
+      "loss": 4.0991,
+      "step": 605696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.007447769632781e-05,
+      "loss": 4.0964,
+      "step": 606208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.006610812762102e-05,
+      "loss": 4.0894,
+      "step": 606720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.00577221801105e-05,
+      "loss": 4.0928,
+      "step": 607232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0049352611403714e-05,
+      "loss": 4.0795,
+      "step": 607744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0040966663893194e-05,
+      "loss": 4.0913,
+      "step": 608256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0032580716382674e-05,
+      "loss": 4.1014,
+      "step": 608768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0024194768872154e-05,
+      "loss": 4.1046,
+      "step": 609280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0015808821361634e-05,
+      "loss": 4.0798,
+      "step": 609792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.000742287385111e-05,
+      "loss": 4.0903,
+      "step": 610304
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.104250431060791,
+      "eval_runtime": 537.2975,
+      "eval_samples_per_second": 710.204,
+      "eval_steps_per_second": 22.194,
+      "step": 610552
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.999903692634059e-05,
+      "loss": 4.0991,
+      "step": 610816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.999065097883007e-05,
+      "loss": 4.0971,
+      "step": 611328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.998226503131955e-05,
+      "loss": 4.0845,
+      "step": 611840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.997387908380903e-05,
+      "loss": 4.089,
+      "step": 612352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.996549313629851e-05,
+      "loss": 4.0783,
+      "step": 612864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.995710718878799e-05,
+      "loss": 4.0832,
+      "step": 613376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.994872124127747e-05,
+      "loss": 4.0791,
+      "step": 613888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.994035167257068e-05,
+      "loss": 4.0805,
+      "step": 614400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.993196572506016e-05,
+      "loss": 4.0834,
+      "step": 614912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.992357977754964e-05,
+      "loss": 4.0942,
+      "step": 615424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.991522658764658e-05,
+      "loss": 4.0893,
+      "step": 615936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.990684064013606e-05,
+      "loss": 4.0876,
+      "step": 616448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.989845469262554e-05,
+      "loss": 4.0974,
+      "step": 616960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.989006874511502e-05,
+      "loss": 4.0806,
+      "step": 617472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.98816827976045e-05,
+      "loss": 4.0948,
+      "step": 617984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.987329685009398e-05,
+      "loss": 4.0734,
+      "step": 618496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.986491090258346e-05,
+      "loss": 4.0785,
+      "step": 619008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.985654133387668e-05,
+      "loss": 4.0893,
+      "step": 619520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.984815538636616e-05,
+      "loss": 4.087,
+      "step": 620032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.983976943885564e-05,
+      "loss": 4.0963,
+      "step": 620544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.983138349134512e-05,
+      "loss": 4.0873,
+      "step": 621056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9822997543834597e-05,
+      "loss": 4.0891,
+      "step": 621568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9814611596324077e-05,
+      "loss": 4.0918,
+      "step": 622080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9806225648813557e-05,
+      "loss": 4.0823,
+      "step": 622592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9797839701303036e-05,
+      "loss": 4.0899,
+      "step": 623104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9789453753792516e-05,
+      "loss": 4.0799,
+      "step": 623616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.9781067806281996e-05,
+      "loss": 4.0913,
+      "step": 624128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.977268185877147e-05,
+      "loss": 4.0796,
+      "step": 624640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.976429591126095e-05,
+      "loss": 4.0952,
+      "step": 625152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.975590996375043e-05,
+      "loss": 4.0757,
+      "step": 625664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.974752401623991e-05,
+      "loss": 4.0759,
+      "step": 626176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9739154447533125e-05,
+      "loss": 4.0885,
+      "step": 626688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9730768500022605e-05,
+      "loss": 4.0858,
+      "step": 627200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9722382552512085e-05,
+      "loss": 4.0902,
+      "step": 627712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9713996605001565e-05,
+      "loss": 4.0831,
+      "step": 628224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9705627036294774e-05,
+      "loss": 4.0836,
+      "step": 628736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9697241088784254e-05,
+      "loss": 4.084,
+      "step": 629248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9688855141273734e-05,
+      "loss": 4.093,
+      "step": 629760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9680469193763214e-05,
+      "loss": 4.0876,
+      "step": 630272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9672083246252694e-05,
+      "loss": 4.0772,
+      "step": 630784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.96637136775459e-05,
+      "loss": 4.0725,
+      "step": 631296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.965532773003538e-05,
+      "loss": 4.0712,
+      "step": 631808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.964694178252486e-05,
+      "loss": 4.0888,
+      "step": 632320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.963855583501434e-05,
+      "loss": 4.084,
+      "step": 632832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.963018626630756e-05,
+      "loss": 4.0723,
+      "step": 633344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.962180031879704e-05,
+      "loss": 4.0947,
+      "step": 633856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.961343075009025e-05,
+      "loss": 4.075,
+      "step": 634368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.960504480257973e-05,
+      "loss": 4.0889,
+      "step": 634880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.959665885506921e-05,
+      "loss": 4.0753,
+      "step": 635392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.958827290755869e-05,
+      "loss": 4.0703,
+      "step": 635904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.957988696004817e-05,
+      "loss": 4.0749,
+      "step": 636416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.957150101253765e-05,
+      "loss": 4.0738,
+      "step": 636928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.956311506502713e-05,
+      "loss": 4.077,
+      "step": 637440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.955472911751661e-05,
+      "loss": 4.0772,
+      "step": 637952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9546375927613546e-05,
+      "loss": 4.0771,
+      "step": 638464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.953800635890676e-05,
+      "loss": 4.0637,
+      "step": 638976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.952962041139624e-05,
+      "loss": 4.0746,
+      "step": 639488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.952123446388572e-05,
+      "loss": 4.065,
+      "step": 640000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.95128485163752e-05,
+      "loss": 4.0768,
+      "step": 640512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.950446256886468e-05,
+      "loss": 4.0685,
+      "step": 641024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.949607662135416e-05,
+      "loss": 4.0749,
+      "step": 641536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.948769067384364e-05,
+      "loss": 4.0715,
+      "step": 642048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.947930472633312e-05,
+      "loss": 4.062,
+      "step": 642560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.94709187788226e-05,
+      "loss": 4.0811,
+      "step": 643072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.946253283131208e-05,
+      "loss": 4.0667,
+      "step": 643584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.945414688380156e-05,
+      "loss": 4.079,
+      "step": 644096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.944576093629104e-05,
+      "loss": 4.0678,
+      "step": 644608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.943739136758425e-05,
+      "loss": 4.0687,
+      "step": 645120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.942900542007373e-05,
+      "loss": 4.07,
+      "step": 645632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.942061947256322e-05,
+      "loss": 4.0703,
+      "step": 646144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.94122335250527e-05,
+      "loss": 4.0759,
+      "step": 646656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9403863956345906e-05,
+      "loss": 4.0732,
+      "step": 647168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9395478008835386e-05,
+      "loss": 4.0756,
+      "step": 647680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9387092061324866e-05,
+      "loss": 4.092,
+      "step": 648192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9378706113814346e-05,
+      "loss": 4.0827,
+      "step": 648704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9370336545107555e-05,
+      "loss": 4.0749,
+      "step": 649216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9361966976400764e-05,
+      "loss": 4.076,
+      "step": 649728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9353581028890244e-05,
+      "loss": 4.0598,
+      "step": 650240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9345195081379724e-05,
+      "loss": 4.0711,
+      "step": 650752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9336809133869204e-05,
+      "loss": 4.0849,
+      "step": 651264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.9328423186358684e-05,
+      "loss": 4.0844,
+      "step": 651776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.932003723884817e-05,
+      "loss": 4.0725,
+      "step": 652288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.931166767014138e-05,
+      "loss": 4.0584,
+      "step": 652800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.930328172263086e-05,
+      "loss": 4.0794,
+      "step": 653312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.929489577512034e-05,
+      "loss": 4.054,
+      "step": 653824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.928650982760982e-05,
+      "loss": 4.0706,
+      "step": 654336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.92781238800993e-05,
+      "loss": 4.071,
+      "step": 654848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.926975431139251e-05,
+      "loss": 4.0711,
+      "step": 655360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.926136836388199e-05,
+      "loss": 4.0794,
+      "step": 655872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.925298241637147e-05,
+      "loss": 4.0778,
+      "step": 656384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.924459646886095e-05,
+      "loss": 4.064,
+      "step": 656896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.923622690015416e-05,
+      "loss": 4.0687,
+      "step": 657408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.922784095264364e-05,
+      "loss": 4.0702,
+      "step": 657920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9219455005133124e-05,
+      "loss": 4.0637,
+      "step": 658432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9211069057622604e-05,
+      "loss": 4.0617,
+      "step": 658944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.920268311011208e-05,
+      "loss": 4.0731,
+      "step": 659456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.919429716260156e-05,
+      "loss": 4.0788,
+      "step": 659968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.918592759389477e-05,
+      "loss": 4.0876,
+      "step": 660480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.917754164638425e-05,
+      "loss": 4.0677,
+      "step": 660992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9169155698873727e-05,
+      "loss": 4.0726,
+      "step": 661504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9160769751363206e-05,
+      "loss": 4.0647,
+      "step": 662016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9152383803852686e-05,
+      "loss": 4.0736,
+      "step": 662528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9143997856342166e-05,
+      "loss": 4.0809,
+      "step": 663040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9135628287635375e-05,
+      "loss": 4.0734,
+      "step": 663552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.912724234012486e-05,
+      "loss": 4.076,
+      "step": 664064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.911885639261434e-05,
+      "loss": 4.0532,
+      "step": 664576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.911047044510382e-05,
+      "loss": 4.0555,
+      "step": 665088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.91020844975933e-05,
+      "loss": 4.0608,
+      "step": 665600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.909371492888651e-05,
+      "loss": 4.0678,
+      "step": 666112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.908532898137599e-05,
+      "loss": 4.0727,
+      "step": 666624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.90769594126692e-05,
+      "loss": 4.0735,
+      "step": 667136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.906857346515868e-05,
+      "loss": 4.0449,
+      "step": 667648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.906018751764816e-05,
+      "loss": 4.0711,
+      "step": 668160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.905180157013764e-05,
+      "loss": 4.0764,
+      "step": 668672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.904341562262712e-05,
+      "loss": 4.07,
+      "step": 669184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.903504605392033e-05,
+      "loss": 4.0717,
+      "step": 669696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9026660106409816e-05,
+      "loss": 4.0699,
+      "step": 670208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9018274158899296e-05,
+      "loss": 4.0567,
+      "step": 670720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9009888211388776e-05,
+      "loss": 4.0763,
+      "step": 671232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.9001502263878256e-05,
+      "loss": 4.0753,
+      "step": 671744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8993116316367736e-05,
+      "loss": 4.059,
+      "step": 672256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8984730368857216e-05,
+      "loss": 4.0585,
+      "step": 672768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8976344421346696e-05,
+      "loss": 4.0586,
+      "step": 673280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8967958473836176e-05,
+      "loss": 4.064,
+      "step": 673792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8959572526325656e-05,
+      "loss": 4.0636,
+      "step": 674304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8951186578815135e-05,
+      "loss": 4.0722,
+      "step": 674816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8942817010108345e-05,
+      "loss": 4.0671,
+      "step": 675328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8934447441401554e-05,
+      "loss": 4.0562,
+      "step": 675840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8926061493891034e-05,
+      "loss": 4.0787,
+      "step": 676352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8917675546380514e-05,
+      "loss": 4.058,
+      "step": 676864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.8909289598869994e-05,
+      "loss": 4.0737,
+      "step": 677376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.890090365135948e-05,
+      "loss": 4.0782,
+      "step": 677888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.889251770384896e-05,
+      "loss": 4.0712,
+      "step": 678400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.888413175633844e-05,
+      "loss": 4.0767,
+      "step": 678912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.887574580882791e-05,
+      "loss": 4.0658,
+      "step": 679424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.886737624012113e-05,
+      "loss": 4.0545,
+      "step": 679936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.885899029261061e-05,
+      "loss": 4.0617,
+      "step": 680448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.885060434510009e-05,
+      "loss": 4.0636,
+      "step": 680960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.884221839758956e-05,
+      "loss": 4.0782,
+      "step": 681472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.883383245007904e-05,
+      "loss": 4.0715,
+      "step": 681984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.882544650256852e-05,
+      "loss": 4.0685,
+      "step": 682496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.881707693386173e-05,
+      "loss": 4.0685,
+      "step": 683008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.880869098635122e-05,
+      "loss": 4.0619,
+      "step": 683520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.88003050388407e-05,
+      "loss": 4.0525,
+      "step": 684032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.879191909133018e-05,
+      "loss": 4.0639,
+      "step": 684544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.878353314381966e-05,
+      "loss": 4.0753,
+      "step": 685056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.877514719630914e-05,
+      "loss": 4.0794,
+      "step": 685568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.876676124879862e-05,
+      "loss": 4.0565,
+      "step": 686080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.87583753012881e-05,
+      "loss": 4.0598,
+      "step": 686592
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.08475399017334,
+      "eval_runtime": 537.5317,
+      "eval_samples_per_second": 709.895,
+      "eval_steps_per_second": 22.185,
+      "step": 686871
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.875000573258131e-05,
+      "loss": 4.0824,
+      "step": 687104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.874161978507079e-05,
+      "loss": 4.0696,
+      "step": 687616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.873323383756027e-05,
+      "loss": 4.0604,
+      "step": 688128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.872484789004975e-05,
+      "loss": 4.0609,
+      "step": 688640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.871646194253923e-05,
+      "loss": 4.0554,
+      "step": 689152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.870807599502871e-05,
+      "loss": 4.0533,
+      "step": 689664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.869969004751819e-05,
+      "loss": 4.0513,
+      "step": 690176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8691304100007667e-05,
+      "loss": 4.0547,
+      "step": 690688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8682918152497147e-05,
+      "loss": 4.0545,
+      "step": 691200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8674532204986626e-05,
+      "loss": 4.0691,
+      "step": 691712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8666146257476106e-05,
+      "loss": 4.0623,
+      "step": 692224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8657760309965586e-05,
+      "loss": 4.0616,
+      "step": 692736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8649374362455066e-05,
+      "loss": 4.0723,
+      "step": 693248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8640988414944546e-05,
+      "loss": 4.0572,
+      "step": 693760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8632602467434026e-05,
+      "loss": 4.0648,
+      "step": 694272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8624232898727235e-05,
+      "loss": 4.047,
+      "step": 694784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8615846951216715e-05,
+      "loss": 4.0513,
+      "step": 695296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8607461003706195e-05,
+      "loss": 4.0643,
+      "step": 695808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8599091434999404e-05,
+      "loss": 4.059,
+      "step": 696320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.8590705487488884e-05,
+      "loss": 4.0711,
+      "step": 696832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.858231953997837e-05,
+      "loss": 4.0563,
+      "step": 697344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.857393359246785e-05,
+      "loss": 4.0723,
+      "step": 697856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.856554764495733e-05,
+      "loss": 4.0582,
+      "step": 698368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.855716169744681e-05,
+      "loss": 4.0567,
+      "step": 698880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.854877574993629e-05,
+      "loss": 4.0623,
+      "step": 699392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.854038980242577e-05,
+      "loss": 4.0595,
+      "step": 699904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.853200385491525e-05,
+      "loss": 4.063,
+      "step": 700416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.852361790740473e-05,
+      "loss": 4.0519,
+      "step": 700928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.851523195989421e-05,
+      "loss": 4.0697,
+      "step": 701440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.850684601238369e-05,
+      "loss": 4.0522,
+      "step": 701952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8498460064873164e-05,
+      "loss": 4.0452,
+      "step": 702464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8490074117362644e-05,
+      "loss": 4.0662,
+      "step": 702976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8481688169852124e-05,
+      "loss": 4.0574,
+      "step": 703488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.847331860114533e-05,
+      "loss": 4.0651,
+      "step": 704000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.846493265363482e-05,
+      "loss": 4.0586,
+      "step": 704512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.84565467061243e-05,
+      "loss": 4.0545,
+      "step": 705024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.844816075861378e-05,
+      "loss": 4.0609,
+      "step": 705536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.843977481110326e-05,
+      "loss": 4.0639,
+      "step": 706048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.843138886359274e-05,
+      "loss": 4.0675,
+      "step": 706560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.842300291608222e-05,
+      "loss": 4.0448,
+      "step": 707072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.84146169685717e-05,
+      "loss": 4.0515,
+      "step": 707584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.840624739986491e-05,
+      "loss": 4.0495,
+      "step": 708096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.839787783115812e-05,
+      "loss": 4.0555,
+      "step": 708608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.83894918836476e-05,
+      "loss": 4.0633,
+      "step": 709120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.838110593613708e-05,
+      "loss": 4.0472,
+      "step": 709632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.837271998862656e-05,
+      "loss": 4.0672,
+      "step": 710144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.836433404111604e-05,
+      "loss": 4.0495,
+      "step": 710656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.835594809360552e-05,
+      "loss": 4.0662,
+      "step": 711168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8347562146095004e-05,
+      "loss": 4.0516,
+      "step": 711680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.833919257738821e-05,
+      "loss": 4.042,
+      "step": 712192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.833080662987769e-05,
+      "loss": 4.0526,
+      "step": 712704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.832242068236717e-05,
+      "loss": 4.0427,
+      "step": 713216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.831403473485665e-05,
+      "loss": 4.056,
+      "step": 713728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.830564878734613e-05,
+      "loss": 4.0495,
+      "step": 714240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.829726283983561e-05,
+      "loss": 4.0547,
+      "step": 714752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.828887689232509e-05,
+      "loss": 4.0346,
+      "step": 715264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.828049094481457e-05,
+      "loss": 4.0506,
+      "step": 715776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.827212137610778e-05,
+      "loss": 4.0379,
+      "step": 716288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.826373542859726e-05,
+      "loss": 4.0559,
+      "step": 716800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.825534948108674e-05,
+      "loss": 4.0447,
+      "step": 717312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.824697991237996e-05,
+      "loss": 4.049,
+      "step": 717824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.823859396486944e-05,
+      "loss": 4.0463,
+      "step": 718336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.823020801735892e-05,
+      "loss": 4.0369,
+      "step": 718848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.82218220698484e-05,
+      "loss": 4.0529,
+      "step": 719360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.821343612233788e-05,
+      "loss": 4.0442,
+      "step": 719872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8205066553631087e-05,
+      "loss": 4.0564,
+      "step": 720384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8196680606120567e-05,
+      "loss": 4.039,
+      "step": 720896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8188294658610047e-05,
+      "loss": 4.0469,
+      "step": 721408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.817990871109952e-05,
+      "loss": 4.0422,
+      "step": 721920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8171539142392736e-05,
+      "loss": 4.0485,
+      "step": 722432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8163153194882216e-05,
+      "loss": 4.0472,
+      "step": 722944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8154767247371695e-05,
+      "loss": 4.0473,
+      "step": 723456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8146381299861175e-05,
+      "loss": 4.0503,
+      "step": 723968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8137995352350655e-05,
+      "loss": 4.0675,
+      "step": 724480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8129609404840135e-05,
+      "loss": 4.0545,
+      "step": 724992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.812123983613335e-05,
+      "loss": 4.0533,
+      "step": 725504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8112853888622824e-05,
+      "loss": 4.0507,
+      "step": 726016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8104467941112304e-05,
+      "loss": 4.0388,
+      "step": 726528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8096081993601784e-05,
+      "loss": 4.0417,
+      "step": 727040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8087696046091264e-05,
+      "loss": 4.0562,
+      "step": 727552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8079310098580744e-05,
+      "loss": 4.0631,
+      "step": 728064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8070924151070224e-05,
+      "loss": 4.0531,
+      "step": 728576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8062538203559704e-05,
+      "loss": 4.0327,
+      "step": 729088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.805416863485291e-05,
+      "loss": 4.0493,
+      "step": 729600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.804578268734239e-05,
+      "loss": 4.0336,
+      "step": 730112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.803739673983187e-05,
+      "loss": 4.0442,
+      "step": 730624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.802901079232136e-05,
+      "loss": 4.0462,
+      "step": 731136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.802062484481084e-05,
+      "loss": 4.0458,
+      "step": 731648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.801223889730032e-05,
+      "loss": 4.0561,
+      "step": 732160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.800386932859353e-05,
+      "loss": 4.0496,
+      "step": 732672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.799548338108301e-05,
+      "loss": 4.0412,
+      "step": 733184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.798711381237622e-05,
+      "loss": 4.0445,
+      "step": 733696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.79787278648657e-05,
+      "loss": 4.0489,
+      "step": 734208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.797035829615891e-05,
+      "loss": 4.0378,
+      "step": 734720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.796197234864839e-05,
+      "loss": 4.0405,
+      "step": 735232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.795358640113787e-05,
+      "loss": 4.0462,
+      "step": 735744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.794520045362735e-05,
+      "loss": 4.0536,
+      "step": 736256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.793681450611683e-05,
+      "loss": 4.0608,
+      "step": 736768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7928428558606314e-05,
+      "loss": 4.0414,
+      "step": 737280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7920042611095793e-05,
+      "loss": 4.0525,
+      "step": 737792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7911656663585273e-05,
+      "loss": 4.0423,
+      "step": 738304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7903270716074753e-05,
+      "loss": 4.0485,
+      "step": 738816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.789488476856423e-05,
+      "loss": 4.0487,
+      "step": 739328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.788651519985744e-05,
+      "loss": 4.0511,
+      "step": 739840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.787812925234692e-05,
+      "loss": 4.0537,
+      "step": 740352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.78697433048364e-05,
+      "loss": 4.0293,
+      "step": 740864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.786135735732588e-05,
+      "loss": 4.032,
+      "step": 741376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7852971409815355e-05,
+      "loss": 4.0344,
+      "step": 741888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.784460184110857e-05,
+      "loss": 4.0392,
+      "step": 742400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.783621589359805e-05,
+      "loss": 4.0476,
+      "step": 742912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.782782994608753e-05,
+      "loss": 4.0484,
+      "step": 743424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.781944399857701e-05,
+      "loss": 4.0242,
+      "step": 743936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.781105805106649e-05,
+      "loss": 4.0439,
+      "step": 744448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.780268848235971e-05,
+      "loss": 4.0496,
+      "step": 744960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.779430253484919e-05,
+      "loss": 4.0456,
+      "step": 745472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.778591658733866e-05,
+      "loss": 4.0503,
+      "step": 745984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.777753063982814e-05,
+      "loss": 4.0418,
+      "step": 746496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.776914469231762e-05,
+      "loss": 4.0356,
+      "step": 747008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.776077512361083e-05,
+      "loss": 4.0471,
+      "step": 747520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.775238917610031e-05,
+      "loss": 4.05,
+      "step": 748032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.774400322858979e-05,
+      "loss": 4.0416,
+      "step": 748544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.773561728107927e-05,
+      "loss": 4.0291,
+      "step": 749056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.772723133356875e-05,
+      "loss": 4.0361,
+      "step": 749568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7718845386058236e-05,
+      "loss": 4.0371,
+      "step": 750080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7710459438547716e-05,
+      "loss": 4.044,
+      "step": 750592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7702073491037196e-05,
+      "loss": 4.04,
+      "step": 751104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7693720301134134e-05,
+      "loss": 4.0473,
+      "step": 751616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7685334353623614e-05,
+      "loss": 4.0308,
+      "step": 752128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7676948406113094e-05,
+      "loss": 4.0531,
+      "step": 752640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7668562458602574e-05,
+      "loss": 4.0347,
+      "step": 753152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7660176511092054e-05,
+      "loss": 4.0466,
+      "step": 753664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7651790563581534e-05,
+      "loss": 4.0515,
+      "step": 754176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7643404616071014e-05,
+      "loss": 4.0455,
+      "step": 754688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.7635018668560494e-05,
+      "loss": 4.0516,
+      "step": 755200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.76266490998537e-05,
+      "loss": 4.0427,
+      "step": 755712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.761827953114692e-05,
+      "loss": 4.0304,
+      "step": 756224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.76098935836364e-05,
+      "loss": 4.0365,
+      "step": 756736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.760150763612588e-05,
+      "loss": 4.036,
+      "step": 757248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.759312168861536e-05,
+      "loss": 4.0523,
+      "step": 757760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.758473574110484e-05,
+      "loss": 4.0484,
+      "step": 758272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.757634979359432e-05,
+      "loss": 4.0443,
+      "step": 758784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.75679638460838e-05,
+      "loss": 4.0459,
+      "step": 759296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.755957789857328e-05,
+      "loss": 4.0362,
+      "step": 759808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.755120832986649e-05,
+      "loss": 4.027,
+      "step": 760320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.754282238235597e-05,
+      "loss": 4.0399,
+      "step": 760832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.753443643484545e-05,
+      "loss": 4.0524,
+      "step": 761344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.752605048733493e-05,
+      "loss": 4.0509,
+      "step": 761856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.751766453982441e-05,
+      "loss": 4.0301,
+      "step": 762368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.750929497111762e-05,
+      "loss": 4.036,
+      "step": 762880
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.067467212677002,
+      "eval_runtime": 559.811,
+      "eval_samples_per_second": 681.643,
+      "eval_steps_per_second": 21.302,
+      "step": 763190
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.750092540241083e-05,
+      "loss": 4.0527,
+      "step": 763392
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.749253945490031e-05,
+      "loss": 4.0399,
+      "step": 763904
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.748415350738979e-05,
+      "loss": 4.041,
+      "step": 764416
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.747576755987927e-05,
+      "loss": 4.0321,
+      "step": 764928
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.746738161236875e-05,
+      "loss": 4.0377,
+      "step": 765440
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.745901204366196e-05,
+      "loss": 4.028,
+      "step": 765952
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.745062609615144e-05,
+      "loss": 4.0301,
+      "step": 766464
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.744224014864092e-05,
+      "loss": 4.0261,
+      "step": 766976
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.74338542011304e-05,
+      "loss": 4.0343,
+      "step": 767488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.742546825361988e-05,
+      "loss": 4.0424,
+      "step": 768000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7417115063716826e-05,
+      "loss": 4.0413,
+      "step": 768512
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7408729116206306e-05,
+      "loss": 4.0392,
+      "step": 769024
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7400343168695786e-05,
+      "loss": 4.0441,
+      "step": 769536
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7391957221185266e-05,
+      "loss": 4.0322,
+      "step": 770048
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7383571273674746e-05,
+      "loss": 4.044,
+      "step": 770560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7375185326164226e-05,
+      "loss": 4.0221,
+      "step": 771072
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7366799378653706e-05,
+      "loss": 4.0236,
+      "step": 771584
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7358413431143186e-05,
+      "loss": 4.0444,
+      "step": 772096
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7350027483632666e-05,
+      "loss": 4.0358,
+      "step": 772608
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.734164153612214e-05,
+      "loss": 4.0475,
+      "step": 773120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7333271967415355e-05,
+      "loss": 4.0295,
+      "step": 773632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7324886019904835e-05,
+      "loss": 4.0494,
+      "step": 774144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7316500072394315e-05,
+      "loss": 4.037,
+      "step": 774656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7308114124883794e-05,
+      "loss": 4.036,
+      "step": 775168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.729974455617701e-05,
+      "loss": 4.0368,
+      "step": 775680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.729135860866649e-05,
+      "loss": 4.0366,
+      "step": 776192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7282972661155963e-05,
+      "loss": 4.0397,
+      "step": 776704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.7274586713645443e-05,
+      "loss": 4.0265,
+      "step": 777216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.726620076613492e-05,
+      "loss": 4.0447,
+      "step": 777728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.72578148186244e-05,
+      "loss": 4.0347,
+      "step": 778240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.724942887111388e-05,
+      "loss": 4.0205,
+      "step": 778752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.724104292360336e-05,
+      "loss": 4.04,
+      "step": 779264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.723267335489657e-05,
+      "loss": 4.0352,
+      "step": 779776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.722430378618979e-05,
+      "loss": 4.0409,
+      "step": 780288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.721591783867927e-05,
+      "loss": 4.0306,
+      "step": 780800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.720753189116875e-05,
+      "loss": 4.0359,
+      "step": 781312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.719914594365823e-05,
+      "loss": 4.037,
+      "step": 781824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.719075999614771e-05,
+      "loss": 4.0398,
+      "step": 782336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.718237404863719e-05,
+      "loss": 4.0439,
+      "step": 782848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.717398810112667e-05,
+      "loss": 4.0225,
+      "step": 783360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.716560215361615e-05,
+      "loss": 4.028,
+      "step": 783872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.715723258490936e-05,
+      "loss": 4.0261,
+      "step": 784384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.714884663739884e-05,
+      "loss": 4.0363,
+      "step": 784896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.714046068988832e-05,
+      "loss": 4.0378,
+      "step": 785408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.71320747423778e-05,
+      "loss": 4.0245,
+      "step": 785920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.712368879486728e-05,
+      "loss": 4.0445,
+      "step": 786432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7115319226160486e-05,
+      "loss": 4.0271,
+      "step": 786944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.7106933278649966e-05,
+      "loss": 4.0424,
+      "step": 787456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.709856370994318e-05,
+      "loss": 4.0346,
+      "step": 787968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.709017776243266e-05,
+      "loss": 4.018,
+      "step": 788480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.708179181492214e-05,
+      "loss": 4.0293,
+      "step": 788992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.707340586741162e-05,
+      "loss": 4.0205,
+      "step": 789504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.70650199199011e-05,
+      "loss": 4.0351,
+      "step": 790016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.705663397239058e-05,
+      "loss": 4.0272,
+      "step": 790528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.704824802488006e-05,
+      "loss": 4.0291,
+      "step": 791040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.703987845617327e-05,
+      "loss": 4.0114,
+      "step": 791552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.703149250866275e-05,
+      "loss": 4.0289,
+      "step": 792064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.702310656115223e-05,
+      "loss": 4.0168,
+      "step": 792576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.701472061364171e-05,
+      "loss": 4.0338,
+      "step": 793088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.700633466613119e-05,
+      "loss": 4.0232,
+      "step": 793600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.699794871862067e-05,
+      "loss": 4.0261,
+      "step": 794112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.698956277111015e-05,
+      "loss": 4.0265,
+      "step": 794624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.698117682359963e-05,
+      "loss": 4.0133,
+      "step": 795136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6972807254892846e-05,
+      "loss": 4.0292,
+      "step": 795648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6964421307382326e-05,
+      "loss": 4.0242,
+      "step": 796160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.69560353598718e-05,
+      "loss": 4.0352,
+      "step": 796672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.694764941236128e-05,
+      "loss": 4.0172,
+      "step": 797184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6939279843654495e-05,
+      "loss": 4.0227,
+      "step": 797696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6930893896143975e-05,
+      "loss": 4.0202,
+      "step": 798208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.692250794863345e-05,
+      "loss": 4.024,
+      "step": 798720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.691412200112293e-05,
+      "loss": 4.0237,
+      "step": 799232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6905752432416144e-05,
+      "loss": 4.0288,
+      "step": 799744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6897366484905624e-05,
+      "loss": 4.0241,
+      "step": 800256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6888980537395104e-05,
+      "loss": 4.0478,
+      "step": 800768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6880594589884584e-05,
+      "loss": 4.0352,
+      "step": 801280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.68722250211778e-05,
+      "loss": 4.0286,
+      "step": 801792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.686383907366727e-05,
+      "loss": 4.0316,
+      "step": 802304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.685545312615675e-05,
+      "loss": 4.017,
+      "step": 802816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.684706717864623e-05,
+      "loss": 4.0261,
+      "step": 803328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.683869760993945e-05,
+      "loss": 4.0278,
+      "step": 803840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.683031166242892e-05,
+      "loss": 4.0393,
+      "step": 804352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.682194209372214e-05,
+      "loss": 4.0321,
+      "step": 804864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.681355614621162e-05,
+      "loss": 4.0143,
+      "step": 805376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.68051701987011e-05,
+      "loss": 4.0269,
+      "step": 805888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.679678425119058e-05,
+      "loss": 4.0108,
+      "step": 806400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6788414682483794e-05,
+      "loss": 4.0252,
+      "step": 806912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.6780028734973274e-05,
+      "loss": 4.0215,
+      "step": 807424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.677164278746275e-05,
+      "loss": 4.0288,
+      "step": 807936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.676325683995223e-05,
+      "loss": 4.0348,
+      "step": 808448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.675487089244171e-05,
+      "loss": 4.0286,
+      "step": 808960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6746484944931187e-05,
+      "loss": 4.0162,
+      "step": 809472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6738098997420667e-05,
+      "loss": 4.024,
+      "step": 809984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6729713049910146e-05,
+      "loss": 4.0262,
+      "step": 810496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6721327102399626e-05,
+      "loss": 4.0172,
+      "step": 811008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6712957533692836e-05,
+      "loss": 4.0179,
+      "step": 811520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6704571586182315e-05,
+      "loss": 4.0265,
+      "step": 812032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6696185638671795e-05,
+      "loss": 4.0355,
+      "step": 812544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6687799691161275e-05,
+      "loss": 4.0375,
+      "step": 813056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.667941374365076e-05,
+      "loss": 4.023,
+      "step": 813568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.667102779614024e-05,
+      "loss": 4.03,
+      "step": 814080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.666264184862972e-05,
+      "loss": 4.0218,
+      "step": 814592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.66542559011192e-05,
+      "loss": 4.028,
+      "step": 815104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.664586995360868e-05,
+      "loss": 4.026,
+      "step": 815616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.663748400609816e-05,
+      "loss": 4.0354,
+      "step": 816128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.662911443739137e-05,
+      "loss": 4.0291,
+      "step": 816640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.662072848988085e-05,
+      "loss": 4.0081,
+      "step": 817152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.661234254237033e-05,
+      "loss": 4.0121,
+      "step": 817664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.660397297366354e-05,
+      "loss": 4.0153,
+      "step": 818176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.659560340495675e-05,
+      "loss": 4.02,
+      "step": 818688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.658721745744623e-05,
+      "loss": 4.0271,
+      "step": 819200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6578831509935716e-05,
+      "loss": 4.0311,
+      "step": 819712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6570445562425196e-05,
+      "loss": 4.0058,
+      "step": 820224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6562059614914676e-05,
+      "loss": 4.0212,
+      "step": 820736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6553673667404156e-05,
+      "loss": 4.0297,
+      "step": 821248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6545287719893636e-05,
+      "loss": 4.0259,
+      "step": 821760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.653690177238311e-05,
+      "loss": 4.031,
+      "step": 822272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6528532203676325e-05,
+      "loss": 4.0231,
+      "step": 822784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6520146256165805e-05,
+      "loss": 4.0114,
+      "step": 823296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6511760308655285e-05,
+      "loss": 4.0271,
+      "step": 823808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.650337436114476e-05,
+      "loss": 4.0334,
+      "step": 824320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.649498841363424e-05,
+      "loss": 4.0195,
+      "step": 824832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.648660246612372e-05,
+      "loss": 4.0114,
+      "step": 825344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.64782165186132e-05,
+      "loss": 4.0169,
+      "step": 825856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6469830571102684e-05,
+      "loss": 4.0162,
+      "step": 826368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6461461002395893e-05,
+      "loss": 4.0245,
+      "step": 826880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6453075054885373e-05,
+      "loss": 4.0182,
+      "step": 827392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.644468910737485e-05,
+      "loss": 4.032,
+      "step": 827904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.643630315986433e-05,
+      "loss": 4.0092,
+      "step": 828416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.642794996996128e-05,
+      "loss": 4.031,
+      "step": 828928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.641956402245075e-05,
+      "loss": 4.0194,
+      "step": 829440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.641117807494023e-05,
+      "loss": 4.0267,
+      "step": 829952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.640279212742971e-05,
+      "loss": 4.0379,
+      "step": 830464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.639440617991919e-05,
+      "loss": 4.0219,
+      "step": 830976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.638603661121241e-05,
+      "loss": 4.0296,
+      "step": 831488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.637765066370189e-05,
+      "loss": 4.0282,
+      "step": 832000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.636926471619137e-05,
+      "loss": 4.0114,
+      "step": 832512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.636087876868085e-05,
+      "loss": 4.0146,
+      "step": 833024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6352509199974056e-05,
+      "loss": 4.0183,
+      "step": 833536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6344123252463536e-05,
+      "loss": 4.0287,
+      "step": 834048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6335737304953016e-05,
+      "loss": 4.0299,
+      "step": 834560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6327351357442496e-05,
+      "loss": 4.0266,
+      "step": 835072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6318965409931976e-05,
+      "loss": 4.0251,
+      "step": 835584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6310579462421456e-05,
+      "loss": 4.0175,
+      "step": 836096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6302193514910936e-05,
+      "loss": 4.0101,
+      "step": 836608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6293807567400416e-05,
+      "loss": 4.0218,
+      "step": 837120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6285437998693625e-05,
+      "loss": 4.0296,
+      "step": 837632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.6277052051183105e-05,
+      "loss": 4.0316,
+      "step": 838144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.626866610367259e-05,
+      "loss": 4.0145,
+      "step": 838656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.626028015616207e-05,
+      "loss": 4.0172,
+      "step": 839168
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.05504846572876,
+      "eval_runtime": 563.1319,
+      "eval_samples_per_second": 677.623,
+      "eval_steps_per_second": 21.176,
+      "step": 839509
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.625189420865155e-05,
+      "loss": 4.0478,
+      "step": 839680
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.624350826114103e-05,
+      "loss": 4.0246,
+      "step": 840192
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.623512231363051e-05,
+      "loss": 4.0185,
+      "step": 840704
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.622673636611999e-05,
+      "loss": 4.017,
+      "step": 841216
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.621835041860947e-05,
+      "loss": 4.0199,
+      "step": 841728
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6209964471098945e-05,
+      "loss": 4.0111,
+      "step": 842240
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6201578523588425e-05,
+      "loss": 4.0085,
+      "step": 842752
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6193192576077905e-05,
+      "loss": 4.0092,
+      "step": 843264
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6184806628567384e-05,
+      "loss": 4.019,
+      "step": 843776
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6176420681056864e-05,
+      "loss": 4.0201,
+      "step": 844288
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6168034733546344e-05,
+      "loss": 4.0163,
+      "step": 844800
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6159648786035824e-05,
+      "loss": 4.022,
+      "step": 845312
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6151262838525304e-05,
+      "loss": 4.0251,
+      "step": 845824
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6142876891014784e-05,
+      "loss": 4.0154,
+      "step": 846336
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.613449094350427e-05,
+      "loss": 4.024,
+      "step": 846848
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.612610499599375e-05,
+      "loss": 4.0058,
+      "step": 847360
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.611771904848323e-05,
+      "loss": 4.0049,
+      "step": 847872
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.610933310097271e-05,
+      "loss": 4.0251,
+      "step": 848384
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6100947153462184e-05,
+      "loss": 4.0178,
+      "step": 848896
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.60925775847554e-05,
+      "loss": 4.024,
+      "step": 849408
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.608419163724488e-05,
+      "loss": 4.0138,
+      "step": 849920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.607580568973436e-05,
+      "loss": 4.0287,
+      "step": 850432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.606741974222383e-05,
+      "loss": 4.0179,
+      "step": 850944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.605903379471331e-05,
+      "loss": 4.0192,
+      "step": 851456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.605066422600653e-05,
+      "loss": 4.0175,
+      "step": 851968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.604227827849601e-05,
+      "loss": 4.0189,
+      "step": 852480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.603389233098549e-05,
+      "loss": 4.0172,
+      "step": 852992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.602550638347497e-05,
+      "loss": 4.0089,
+      "step": 853504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.601712043596445e-05,
+      "loss": 4.0245,
+      "step": 854016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.600875086725766e-05,
+      "loss": 4.014,
+      "step": 854528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.600036491974714e-05,
+      "loss": 4.007,
+      "step": 855040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.599197897223662e-05,
+      "loss": 4.0179,
+      "step": 855552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.59835930247261e-05,
+      "loss": 4.0182,
+      "step": 856064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.597520707721558e-05,
+      "loss": 4.0251,
+      "step": 856576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.596682112970506e-05,
+      "loss": 4.0116,
+      "step": 857088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.595843518219454e-05,
+      "loss": 4.0165,
+      "step": 857600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5950065613487747e-05,
+      "loss": 4.0206,
+      "step": 858112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5941679665977227e-05,
+      "loss": 4.0196,
+      "step": 858624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5933293718466706e-05,
+      "loss": 4.0273,
+      "step": 859136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.592490777095619e-05,
+      "loss": 4.0053,
+      "step": 859648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.591652182344567e-05,
+      "loss": 4.0121,
+      "step": 860160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.590813587593515e-05,
+      "loss": 4.008,
+      "step": 860672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.589974992842463e-05,
+      "loss": 4.012,
+      "step": 861184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.589136398091411e-05,
+      "loss": 4.0191,
+      "step": 861696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.588297803340359e-05,
+      "loss": 4.0085,
+      "step": 862208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.58746084646968e-05,
+      "loss": 4.0232,
+      "step": 862720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.586622251718628e-05,
+      "loss": 4.0114,
+      "step": 863232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.585783656967576e-05,
+      "loss": 4.0259,
+      "step": 863744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.584945062216524e-05,
+      "loss": 4.016,
+      "step": 864256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.584108105345845e-05,
+      "loss": 3.9963,
+      "step": 864768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.583269510594793e-05,
+      "loss": 4.0089,
+      "step": 865280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.582432553724114e-05,
+      "loss": 4.0059,
+      "step": 865792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.581593958973063e-05,
+      "loss": 4.015,
+      "step": 866304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.580755364222011e-05,
+      "loss": 4.0091,
+      "step": 866816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.579916769470959e-05,
+      "loss": 4.012,
+      "step": 867328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.579078174719907e-05,
+      "loss": 3.9951,
+      "step": 867840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5782412178492276e-05,
+      "loss": 4.0122,
+      "step": 868352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5774026230981756e-05,
+      "loss": 4.0019,
+      "step": 868864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5765640283471236e-05,
+      "loss": 4.0148,
+      "step": 869376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5757254335960716e-05,
+      "loss": 3.9976,
+      "step": 869888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.574886838845019e-05,
+      "loss": 4.0115,
+      "step": 870400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.574048244093967e-05,
+      "loss": 4.0058,
+      "step": 870912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.573209649342915e-05,
+      "loss": 3.9987,
+      "step": 871424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5723726924722365e-05,
+      "loss": 4.0106,
+      "step": 871936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5715340977211845e-05,
+      "loss": 4.0084,
+      "step": 872448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5706955029701325e-05,
+      "loss": 4.0165,
+      "step": 872960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5698569082190804e-05,
+      "loss": 3.9982,
+      "step": 873472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5690183134680284e-05,
+      "loss": 4.0088,
+      "step": 873984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5681813565973494e-05,
+      "loss": 4.0019,
+      "step": 874496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5673427618462974e-05,
+      "loss": 4.0065,
+      "step": 875008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5665041670952453e-05,
+      "loss": 4.0045,
+      "step": 875520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5656655723441933e-05,
+      "loss": 4.0138,
+      "step": 876032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.564828615473514e-05,
+      "loss": 4.005,
+      "step": 876544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.563990020722462e-05,
+      "loss": 4.0267,
+      "step": 877056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.56315142597141e-05,
+      "loss": 4.0227,
+      "step": 877568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.562312831220358e-05,
+      "loss": 4.0123,
+      "step": 878080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.561474236469306e-05,
+      "loss": 4.0113,
+      "step": 878592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.560635641718255e-05,
+      "loss": 4.0044,
+      "step": 879104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.559797046967203e-05,
+      "loss": 4.0043,
+      "step": 879616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.558958452216151e-05,
+      "loss": 4.0123,
+      "step": 880128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.558121495345472e-05,
+      "loss": 4.0179,
+      "step": 880640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.55728290059442e-05,
+      "loss": 4.0205,
+      "step": 881152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.556444305843368e-05,
+      "loss": 3.9944,
+      "step": 881664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.555605711092316e-05,
+      "loss": 4.0064,
+      "step": 882176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.554768754221637e-05,
+      "loss": 3.9938,
+      "step": 882688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5539317973509576e-05,
+      "loss": 4.0123,
+      "step": 883200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5530932025999056e-05,
+      "loss": 4.0016,
+      "step": 883712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5522546078488536e-05,
+      "loss": 4.0101,
+      "step": 884224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5514160130978016e-05,
+      "loss": 4.0165,
+      "step": 884736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.55057741834675e-05,
+      "loss": 4.0146,
+      "step": 885248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.549740461476071e-05,
+      "loss": 3.9984,
+      "step": 885760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.548901866725019e-05,
+      "loss": 4.0031,
+      "step": 886272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.548063271973967e-05,
+      "loss": 4.0108,
+      "step": 886784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.547224677222915e-05,
+      "loss": 3.9983,
+      "step": 887296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.546386082471863e-05,
+      "loss": 3.9973,
+      "step": 887808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.545547487720811e-05,
+      "loss": 4.0106,
+      "step": 888320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.544708892969759e-05,
+      "loss": 4.0166,
+      "step": 888832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.543870298218707e-05,
+      "loss": 4.0209,
+      "step": 889344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.543031703467655e-05,
+      "loss": 4.0102,
+      "step": 889856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.542194746596976e-05,
+      "loss": 4.0112,
+      "step": 890368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.541357789726297e-05,
+      "loss": 3.9998,
+      "step": 890880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5405191949752456e-05,
+      "loss": 4.0113,
+      "step": 891392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5396806002241936e-05,
+      "loss": 4.0082,
+      "step": 891904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5388420054731416e-05,
+      "loss": 4.0179,
+      "step": 892416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5380034107220896e-05,
+      "loss": 4.0086,
+      "step": 892928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5371664538514105e-05,
+      "loss": 3.9937,
+      "step": 893440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5363278591003585e-05,
+      "loss": 3.9947,
+      "step": 893952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5354909022296794e-05,
+      "loss": 3.9931,
+      "step": 894464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5346523074786274e-05,
+      "loss": 4.0056,
+      "step": 894976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5338137127275754e-05,
+      "loss": 4.0119,
+      "step": 895488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5329751179765234e-05,
+      "loss": 4.0097,
+      "step": 896000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5321365232254714e-05,
+      "loss": 3.9933,
+      "step": 896512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5312979284744194e-05,
+      "loss": 3.9992,
+      "step": 897024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5304593337233674e-05,
+      "loss": 4.012,
+      "step": 897536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5296207389723154e-05,
+      "loss": 4.0117,
+      "step": 898048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5287821442212634e-05,
+      "loss": 4.0157,
+      "step": 898560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5279435494702114e-05,
+      "loss": 4.0039,
+      "step": 899072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5271049547191594e-05,
+      "loss": 3.9967,
+      "step": 899584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5262663599681074e-05,
+      "loss": 4.0089,
+      "step": 900096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.525429403097428e-05,
+      "loss": 4.0148,
+      "step": 900608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.52459244622675e-05,
+      "loss": 4.0014,
+      "step": 901120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.523753851475697e-05,
+      "loss": 3.9936,
+      "step": 901632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.522915256724645e-05,
+      "loss": 3.9979,
+      "step": 902144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.522076661973593e-05,
+      "loss": 4.0002,
+      "step": 902656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.521238067222541e-05,
+      "loss": 4.0063,
+      "step": 903168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.520399472471489e-05,
+      "loss": 4.002,
+      "step": 903680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.519560877720437e-05,
+      "loss": 4.0135,
+      "step": 904192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.518722282969386e-05,
+      "loss": 3.9939,
+      "step": 904704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.517885326098707e-05,
+      "loss": 4.014,
+      "step": 905216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.517046731347655e-05,
+      "loss": 4.0008,
+      "step": 905728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.516208136596603e-05,
+      "loss": 4.011,
+      "step": 906240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.515369541845551e-05,
+      "loss": 4.0184,
+      "step": 906752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.514530947094499e-05,
+      "loss": 4.0103,
+      "step": 907264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5136939902238197e-05,
+      "loss": 4.0074,
+      "step": 907776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5128553954727677e-05,
+      "loss": 4.0117,
+      "step": 908288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5120168007217157e-05,
+      "loss": 3.9921,
+      "step": 908800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5111782059706636e-05,
+      "loss": 4.0012,
+      "step": 909312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5103412490999846e-05,
+      "loss": 3.9994,
+      "step": 909824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.5095026543489326e-05,
+      "loss": 4.0161,
+      "step": 910336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.508665697478254e-05,
+      "loss": 4.0126,
+      "step": 910848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.507827102727202e-05,
+      "loss": 4.0127,
+      "step": 911360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.50698850797615e-05,
+      "loss": 4.0067,
+      "step": 911872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.506149913225098e-05,
+      "loss": 3.9989,
+      "step": 912384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.505311318474046e-05,
+      "loss": 3.993,
+      "step": 912896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.504472723722994e-05,
+      "loss": 4.0066,
+      "step": 913408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.503635766852315e-05,
+      "loss": 4.0121,
+      "step": 913920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.502797172101263e-05,
+      "loss": 4.018,
+      "step": 914432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.501958577350211e-05,
+      "loss": 3.9954,
+      "step": 914944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.501119982599159e-05,
+      "loss": 4.001,
+      "step": 915456
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.044651985168457,
+      "eval_runtime": 606.1911,
+      "eval_samples_per_second": 629.49,
+      "eval_steps_per_second": 19.672,
+      "step": 915828
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.500281387848107e-05,
+      "loss": 4.0273,
+      "step": 915968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.499442793097055e-05,
+      "loss": 4.0084,
+      "step": 916480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.498604198346003e-05,
+      "loss": 4.0009,
+      "step": 916992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.497765603594951e-05,
+      "loss": 3.9996,
+      "step": 917504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496927008843899e-05,
+      "loss": 4.0025,
+      "step": 918016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.496088414092847e-05,
+      "loss": 3.9942,
+      "step": 918528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.495249819341795e-05,
+      "loss": 3.9952,
+      "step": 919040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.494411224590743e-05,
+      "loss": 3.988,
+      "step": 919552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.493572629839691e-05,
+      "loss": 4.008,
+      "step": 920064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.492734035088639e-05,
+      "loss": 3.9991,
+      "step": 920576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.491895440337587e-05,
+      "loss": 4.0023,
+      "step": 921088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.491056845586535e-05,
+      "loss": 4.0055,
+      "step": 921600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.490218250835483e-05,
+      "loss": 4.0037,
+      "step": 922112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.489379656084431e-05,
+      "loss": 4.0012,
+      "step": 922624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.488541061333379e-05,
+      "loss": 4.0086,
+      "step": 923136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.487702466582327e-05,
+      "loss": 3.9912,
+      "step": 923648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.486865509711648e-05,
+      "loss": 3.9876,
+      "step": 924160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4860269149605965e-05,
+      "loss": 4.0068,
+      "step": 924672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4851899580899174e-05,
+      "loss": 4.0004,
+      "step": 925184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4843513633388654e-05,
+      "loss": 4.0095,
+      "step": 925696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4835127685878134e-05,
+      "loss": 3.9953,
+      "step": 926208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4826741738367614e-05,
+      "loss": 4.013,
+      "step": 926720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4818355790857094e-05,
+      "loss": 4.0029,
+      "step": 927232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4809969843346574e-05,
+      "loss": 3.9984,
+      "step": 927744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.480158389583605e-05,
+      "loss": 4.002,
+      "step": 928256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.479319794832553e-05,
+      "loss": 4.0006,
+      "step": 928768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.478481200081501e-05,
+      "loss": 4.0056,
+      "step": 929280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.477642605330449e-05,
+      "loss": 3.9951,
+      "step": 929792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.476804010579397e-05,
+      "loss": 4.001,
+      "step": 930304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.475965415828345e-05,
+      "loss": 4.0051,
+      "step": 930816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.475128458957666e-05,
+      "loss": 3.9892,
+      "step": 931328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.474291502086988e-05,
+      "loss": 3.9975,
+      "step": 931840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.473452907335935e-05,
+      "loss": 4.0012,
+      "step": 932352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.472614312584883e-05,
+      "loss": 4.0089,
+      "step": 932864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.471775717833831e-05,
+      "loss": 3.9936,
+      "step": 933376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.470938760963152e-05,
+      "loss": 3.9976,
+      "step": 933888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4701001662121e-05,
+      "loss": 4.0088,
+      "step": 934400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.469261571461048e-05,
+      "loss": 3.998,
+      "step": 934912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.468422976709996e-05,
+      "loss": 4.012,
+      "step": 935424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.467586019839317e-05,
+      "loss": 3.9877,
+      "step": 935936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.466747425088266e-05,
+      "loss": 3.9955,
+      "step": 936448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.465908830337214e-05,
+      "loss": 3.9878,
+      "step": 936960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.465070235586162e-05,
+      "loss": 3.9977,
+      "step": 937472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4642316408351097e-05,
+      "loss": 4.0029,
+      "step": 937984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4633930460840577e-05,
+      "loss": 3.9946,
+      "step": 938496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4625544513330056e-05,
+      "loss": 4.0041,
+      "step": 939008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4617158565819536e-05,
+      "loss": 3.9955,
+      "step": 939520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4608772618309016e-05,
+      "loss": 4.0102,
+      "step": 940032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4600403049602225e-05,
+      "loss": 3.9996,
+      "step": 940544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4592017102091705e-05,
+      "loss": 3.976,
+      "step": 941056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4583631154581185e-05,
+      "loss": 3.9935,
+      "step": 941568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4575245207070665e-05,
+      "loss": 3.9877,
+      "step": 942080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4566875638363874e-05,
+      "loss": 4.0007,
+      "step": 942592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4558489690853354e-05,
+      "loss": 3.9912,
+      "step": 943104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4550103743342834e-05,
+      "loss": 3.9976,
+      "step": 943616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.454173417463605e-05,
+      "loss": 3.9776,
+      "step": 944128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.453334822712553e-05,
+      "loss": 3.997,
+      "step": 944640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.452497865841874e-05,
+      "loss": 3.9846,
+      "step": 945152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.451659271090822e-05,
+      "loss": 3.9993,
+      "step": 945664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.45082067633977e-05,
+      "loss": 3.9816,
+      "step": 946176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449982081588718e-05,
+      "loss": 3.9899,
+      "step": 946688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.449143486837666e-05,
+      "loss": 3.9915,
+      "step": 947200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.448304892086614e-05,
+      "loss": 3.9848,
+      "step": 947712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.447466297335562e-05,
+      "loss": 3.9892,
+      "step": 948224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.44662770258451e-05,
+      "loss": 3.9948,
+      "step": 948736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.445789107833458e-05,
+      "loss": 4.0,
+      "step": 949248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.444950513082406e-05,
+      "loss": 3.984,
+      "step": 949760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.444111918331354e-05,
+      "loss": 3.9879,
+      "step": 950272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.443273323580302e-05,
+      "loss": 3.9869,
+      "step": 950784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4424363667096235e-05,
+      "loss": 3.9889,
+      "step": 951296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.441597771958571e-05,
+      "loss": 3.9877,
+      "step": 951808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.440759177207519e-05,
+      "loss": 4.0008,
+      "step": 952320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.439920582456467e-05,
+      "loss": 3.987,
+      "step": 952832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.439081987705415e-05,
+      "loss": 4.0106,
+      "step": 953344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.438243392954363e-05,
+      "loss": 4.005,
+      "step": 953856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.437404798203311e-05,
+      "loss": 3.9983,
+      "step": 954368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.436566203452259e-05,
+      "loss": 3.9949,
+      "step": 954880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.43572924658158e-05,
+      "loss": 3.9866,
+      "step": 955392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.434892289710901e-05,
+      "loss": 3.9838,
+      "step": 955904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.434053694959849e-05,
+      "loss": 3.9982,
+      "step": 956416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.433215100208797e-05,
+      "loss": 4.003,
+      "step": 956928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.432376505457745e-05,
+      "loss": 4.0006,
+      "step": 957440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.431537910706693e-05,
+      "loss": 3.9812,
+      "step": 957952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.430699315955641e-05,
+      "loss": 3.9903,
+      "step": 958464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.429860721204589e-05,
+      "loss": 3.9792,
+      "step": 958976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.42902376433391e-05,
+      "loss": 3.9946,
+      "step": 959488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.428185169582858e-05,
+      "loss": 3.9846,
+      "step": 960000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.427346574831806e-05,
+      "loss": 3.9907,
+      "step": 960512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.426507980080754e-05,
+      "loss": 4.0024,
+      "step": 961024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.425669385329702e-05,
+      "loss": 3.9953,
+      "step": 961536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.42483079057865e-05,
+      "loss": 3.9827,
+      "step": 962048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.423992195827598e-05,
+      "loss": 3.9906,
+      "step": 962560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.423153601076546e-05,
+      "loss": 3.9965,
+      "step": 963072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.422316644205868e-05,
+      "loss": 3.9828,
+      "step": 963584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.421478049454816e-05,
+      "loss": 3.9804,
+      "step": 964096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.420639454703764e-05,
+      "loss": 3.9925,
+      "step": 964608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.419800859952712e-05,
+      "loss": 3.9985,
+      "step": 965120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.41896226520166e-05,
+      "loss": 4.0043,
+      "step": 965632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.418123670450607e-05,
+      "loss": 3.9945,
+      "step": 966144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.417285075699555e-05,
+      "loss": 3.9979,
+      "step": 966656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.416446480948503e-05,
+      "loss": 3.9822,
+      "step": 967168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4156095240778246e-05,
+      "loss": 3.9925,
+      "step": 967680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4147725672071455e-05,
+      "loss": 3.9934,
+      "step": 968192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4139339724560935e-05,
+      "loss": 4.005,
+      "step": 968704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4130953777050415e-05,
+      "loss": 3.9909,
+      "step": 969216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4122567829539895e-05,
+      "loss": 3.984,
+      "step": 969728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4114181882029375e-05,
+      "loss": 3.9773,
+      "step": 970240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4105795934518855e-05,
+      "loss": 3.9762,
+      "step": 970752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4097409987008335e-05,
+      "loss": 3.9841,
+      "step": 971264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4089024039497815e-05,
+      "loss": 3.9986,
+      "step": 971776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4080654470791024e-05,
+      "loss": 3.993,
+      "step": 972288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4072268523280504e-05,
+      "loss": 3.977,
+      "step": 972800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4063882575769984e-05,
+      "loss": 3.9845,
+      "step": 973312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.4055496628259463e-05,
+      "loss": 3.9965,
+      "step": 973824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.404712705955267e-05,
+      "loss": 3.9965,
+      "step": 974336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.403874111204215e-05,
+      "loss": 3.9978,
+      "step": 974848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.403035516453163e-05,
+      "loss": 3.9896,
+      "step": 975360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.402196921702111e-05,
+      "loss": 3.9825,
+      "step": 975872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.40135832695106e-05,
+      "loss": 3.99,
+      "step": 976384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.400521370080381e-05,
+      "loss": 3.9988,
+      "step": 976896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.399682775329329e-05,
+      "loss": 3.9863,
+      "step": 977408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.398844180578277e-05,
+      "loss": 3.9756,
+      "step": 977920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.398005585827225e-05,
+      "loss": 3.9856,
+      "step": 978432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.397168628956546e-05,
+      "loss": 3.9827,
+      "step": 978944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.396330034205494e-05,
+      "loss": 3.9944,
+      "step": 979456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.395491439454442e-05,
+      "loss": 3.9871,
+      "step": 979968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.39465284470339e-05,
+      "loss": 3.9977,
+      "step": 980480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.393814249952338e-05,
+      "loss": 3.9801,
+      "step": 980992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.392975655201286e-05,
+      "loss": 3.9985,
+      "step": 981504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3921386983306066e-05,
+      "loss": 3.9848,
+      "step": 982016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.391300103579555e-05,
+      "loss": 3.9952,
+      "step": 982528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.390461508828503e-05,
+      "loss": 3.9983,
+      "step": 983040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.389622914077451e-05,
+      "loss": 3.9992,
+      "step": 983552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.388784319326399e-05,
+      "loss": 3.9928,
+      "step": 984064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.38794736245572e-05,
+      "loss": 3.9993,
+      "step": 984576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.387108767704668e-05,
+      "loss": 3.9768,
+      "step": 985088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.386270172953616e-05,
+      "loss": 3.9796,
+      "step": 985600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.385431578202564e-05,
+      "loss": 3.9873,
+      "step": 986112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.384592983451512e-05,
+      "loss": 3.9959,
+      "step": 986624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.38375438870046e-05,
+      "loss": 4.0016,
+      "step": 987136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.382915793949408e-05,
+      "loss": 3.9944,
+      "step": 987648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3820771991983555e-05,
+      "loss": 3.9941,
+      "step": 988160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3812386044473035e-05,
+      "loss": 3.9831,
+      "step": 988672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3804032854569986e-05,
+      "loss": 3.9779,
+      "step": 989184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3795646907059466e-05,
+      "loss": 3.9888,
+      "step": 989696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3787260959548946e-05,
+      "loss": 3.9979,
+      "step": 990208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3778875012038426e-05,
+      "loss": 4.0033,
+      "step": 990720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.3770489064527906e-05,
+      "loss": 3.9826,
+      "step": 991232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.376210311701738e-05,
+      "loss": 3.9809,
+      "step": 991744
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.035505771636963,
+      "eval_runtime": 544.8363,
+      "eval_samples_per_second": 700.377,
+      "eval_steps_per_second": 21.887,
+      "step": 992147
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.375371716950686e-05,
+      "loss": 3.9722,
+      "step": 992256
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.374533122199634e-05,
+      "loss": 4.0042,
+      "step": 992768
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.373694527448582e-05,
+      "loss": 3.984,
+      "step": 993280
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.37285593269753e-05,
+      "loss": 3.9699,
+      "step": 993792
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.372017337946478e-05,
+      "loss": 3.992,
+      "step": 994304
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.371178743195426e-05,
+      "loss": 3.9857,
+      "step": 994816
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.370340148444374e-05,
+      "loss": 3.9909,
+      "step": 995328
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.369501553693322e-05,
+      "loss": 3.9798,
+      "step": 995840
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.36866295894227e-05,
+      "loss": 3.9833,
+      "step": 996352
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3678243641912186e-05,
+      "loss": 3.9691,
+      "step": 996864
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3669857694401666e-05,
+      "loss": 3.9993,
+      "step": 997376
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3661488125694875e-05,
+      "loss": 3.9825,
+      "step": 997888
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3653102178184355e-05,
+      "loss": 3.9912,
+      "step": 998400
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3644716230673835e-05,
+      "loss": 3.9952,
+      "step": 998912
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3636330283163315e-05,
+      "loss": 3.9842,
+      "step": 999424
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3627944335652795e-05,
+      "loss": 3.9741,
+      "step": 999936
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.361955838814227e-05,
+      "loss": 3.994,
+      "step": 1000448
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.361117244063175e-05,
+      "loss": 3.9873,
+      "step": 1000960
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.360278649312123e-05,
+      "loss": 3.9811,
+      "step": 1001472
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.359440054561071e-05,
+      "loss": 3.9989,
+      "step": 1001984
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3586030976903924e-05,
+      "loss": 4.0081,
+      "step": 1002496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3577645029393404e-05,
+      "loss": 3.9884,
+      "step": 1003008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3569259081882884e-05,
+      "loss": 3.979,
+      "step": 1003520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3560873134372363e-05,
+      "loss": 3.9855,
+      "step": 1004032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.3552487186861843e-05,
+      "loss": 3.9839,
+      "step": 1004544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.354410123935132e-05,
+      "loss": 3.9793,
+      "step": 1005056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.35357152918408e-05,
+      "loss": 3.986,
+      "step": 1005568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.352732934433028e-05,
+      "loss": 3.9778,
+      "step": 1006080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.351894339681976e-05,
+      "loss": 3.9856,
+      "step": 1006592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.351055744930924e-05,
+      "loss": 3.9917,
+      "step": 1007104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.350218788060245e-05,
+      "loss": 3.9927,
+      "step": 1007616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.349380193309193e-05,
+      "loss": 3.987,
+      "step": 1008128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.348543236438514e-05,
+      "loss": 3.9882,
+      "step": 1008640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.347704641687462e-05,
+      "loss": 3.9979,
+      "step": 1009152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.346866046936411e-05,
+      "loss": 3.9774,
+      "step": 1009664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.346027452185359e-05,
+      "loss": 3.9848,
+      "step": 1010176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.345188857434307e-05,
+      "loss": 3.9802,
+      "step": 1010688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.344350262683255e-05,
+      "loss": 3.9948,
+      "step": 1011200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.343511667932203e-05,
+      "loss": 3.995,
+      "step": 1011712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.342673073181151e-05,
+      "loss": 3.9989,
+      "step": 1012224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.341834478430098e-05,
+      "loss": 3.9917,
+      "step": 1012736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.340995883679046e-05,
+      "loss": 3.9882,
+      "step": 1013248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.340157288927994e-05,
+      "loss": 3.9779,
+      "step": 1013760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.339320332057316e-05,
+      "loss": 3.9802,
+      "step": 1014272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3384833751866366e-05,
+      "loss": 3.9738,
+      "step": 1014784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3376447804355846e-05,
+      "loss": 3.9727,
+      "step": 1015296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3368061856845326e-05,
+      "loss": 3.9783,
+      "step": 1015808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3359675909334806e-05,
+      "loss": 3.983,
+      "step": 1016320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3351289961824286e-05,
+      "loss": 3.9826,
+      "step": 1016832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3342904014313766e-05,
+      "loss": 3.9842,
+      "step": 1017344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3334518066803246e-05,
+      "loss": 3.9691,
+      "step": 1017856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3326132119292726e-05,
+      "loss": 3.9829,
+      "step": 1018368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3317746171782206e-05,
+      "loss": 3.9829,
+      "step": 1018880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3309360224271685e-05,
+      "loss": 3.9659,
+      "step": 1019392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3300974276761165e-05,
+      "loss": 3.9711,
+      "step": 1019904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3292604708054375e-05,
+      "loss": 3.9792,
+      "step": 1020416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3284235139347584e-05,
+      "loss": 3.9924,
+      "step": 1020928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3275849191837064e-05,
+      "loss": 3.9635,
+      "step": 1021440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3267463244326544e-05,
+      "loss": 3.9835,
+      "step": 1021952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.325907729681603e-05,
+      "loss": 3.9862,
+      "step": 1022464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.325069134930551e-05,
+      "loss": 3.9713,
+      "step": 1022976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.324230540179499e-05,
+      "loss": 3.9678,
+      "step": 1023488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.323391945428447e-05,
+      "loss": 3.9663,
+      "step": 1024000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.322553350677395e-05,
+      "loss": 3.968,
+      "step": 1024512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.321714755926343e-05,
+      "loss": 3.9887,
+      "step": 1025024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.320876161175291e-05,
+      "loss": 3.9888,
+      "step": 1025536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.320039204304612e-05,
+      "loss": 3.9736,
+      "step": 1026048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.31920060955356e-05,
+      "loss": 3.9818,
+      "step": 1026560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.318362014802508e-05,
+      "loss": 3.9849,
+      "step": 1027072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.317523420051456e-05,
+      "loss": 3.9743,
+      "step": 1027584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.316686463180777e-05,
+      "loss": 3.9878,
+      "step": 1028096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.315847868429725e-05,
+      "loss": 3.9785,
+      "step": 1028608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.315009273678673e-05,
+      "loss": 3.979,
+      "step": 1029120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.314170678927621e-05,
+      "loss": 3.9742,
+      "step": 1029632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3133320841765695e-05,
+      "loss": 3.9899,
+      "step": 1030144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.312493489425517e-05,
+      "loss": 3.9766,
+      "step": 1030656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.311654894674465e-05,
+      "loss": 3.9876,
+      "step": 1031168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.310816299923413e-05,
+      "loss": 3.9818,
+      "step": 1031680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.309977705172361e-05,
+      "loss": 3.983,
+      "step": 1032192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.309142386182055e-05,
+      "loss": 3.9771,
+      "step": 1032704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.308303791431003e-05,
+      "loss": 3.9823,
+      "step": 1033216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.307465196679951e-05,
+      "loss": 3.975,
+      "step": 1033728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.306626601928899e-05,
+      "loss": 3.9741,
+      "step": 1034240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.3057880071778466e-05,
+      "loss": 3.9756,
+      "step": 1034752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.304951050307168e-05,
+      "loss": 3.9681,
+      "step": 1035264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.304112455556116e-05,
+      "loss": 3.9811,
+      "step": 1035776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.303273860805064e-05,
+      "loss": 3.9682,
+      "step": 1036288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.302435266054012e-05,
+      "loss": 3.9792,
+      "step": 1036800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.30159667130296e-05,
+      "loss": 3.9646,
+      "step": 1037312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.300758076551908e-05,
+      "loss": 3.9863,
+      "step": 1037824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.299919481800856e-05,
+      "loss": 3.9841,
+      "step": 1038336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.299082524930177e-05,
+      "loss": 3.9861,
+      "step": 1038848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.298243930179125e-05,
+      "loss": 3.983,
+      "step": 1039360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.297405335428073e-05,
+      "loss": 3.9705,
+      "step": 1039872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.296568378557394e-05,
+      "loss": 3.983,
+      "step": 1040384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.295729783806342e-05,
+      "loss": 3.9773,
+      "step": 1040896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.29489118905529e-05,
+      "loss": 3.9757,
+      "step": 1041408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2940525943042386e-05,
+      "loss": 3.9692,
+      "step": 1041920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2932139995531866e-05,
+      "loss": 3.9723,
+      "step": 1042432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2923770426825075e-05,
+      "loss": 3.9835,
+      "step": 1042944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2915384479314555e-05,
+      "loss": 3.9786,
+      "step": 1043456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2906998531804035e-05,
+      "loss": 3.9874,
+      "step": 1043968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2898612584293515e-05,
+      "loss": 3.9782,
+      "step": 1044480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2890226636782995e-05,
+      "loss": 3.9753,
+      "step": 1044992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2881840689272475e-05,
+      "loss": 3.9683,
+      "step": 1045504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2873454741761955e-05,
+      "loss": 3.9703,
+      "step": 1046016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2865068794251435e-05,
+      "loss": 3.9834,
+      "step": 1046528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2856682846740915e-05,
+      "loss": 3.9823,
+      "step": 1047040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2848313278034124e-05,
+      "loss": 3.971,
+      "step": 1047552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2839927330523604e-05,
+      "loss": 3.9638,
+      "step": 1048064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.283155776181682e-05,
+      "loss": 3.9787,
+      "step": 1048576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.28231718143063e-05,
+      "loss": 3.9818,
+      "step": 1049088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.281478586679578e-05,
+      "loss": 3.9653,
+      "step": 1049600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.280639991928526e-05,
+      "loss": 3.9782,
+      "step": 1050112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.279803035057847e-05,
+      "loss": 3.9693,
+      "step": 1050624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.278964440306795e-05,
+      "loss": 3.9724,
+      "step": 1051136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.278125845555743e-05,
+      "loss": 3.9684,
+      "step": 1051648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.277287250804691e-05,
+      "loss": 3.9785,
+      "step": 1052160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.276448656053639e-05,
+      "loss": 3.975,
+      "step": 1052672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.27561169918296e-05,
+      "loss": 3.9718,
+      "step": 1053184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.274773104431908e-05,
+      "loss": 3.9836,
+      "step": 1053696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.273934509680856e-05,
+      "loss": 3.9916,
+      "step": 1054208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.273095914929804e-05,
+      "loss": 3.9771,
+      "step": 1054720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2722573201787524e-05,
+      "loss": 3.9776,
+      "step": 1055232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2714187254277004e-05,
+      "loss": 3.9747,
+      "step": 1055744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.270580130676648e-05,
+      "loss": 3.9757,
+      "step": 1056256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.269741535925596e-05,
+      "loss": 3.976,
+      "step": 1056768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.268904579054917e-05,
+      "loss": 3.9704,
+      "step": 1057280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.268067622184238e-05,
+      "loss": 3.9785,
+      "step": 1057792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.267229027433186e-05,
+      "loss": 3.9704,
+      "step": 1058304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.266390432682134e-05,
+      "loss": 3.9898,
+      "step": 1058816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.265551837931082e-05,
+      "loss": 3.9728,
+      "step": 1059328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.264714881060403e-05,
+      "loss": 3.9889,
+      "step": 1059840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.263876286309351e-05,
+      "loss": 3.9763,
+      "step": 1060352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.263037691558299e-05,
+      "loss": 3.9813,
+      "step": 1060864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.262199096807247e-05,
+      "loss": 3.9743,
+      "step": 1061376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.261360502056195e-05,
+      "loss": 3.9807,
+      "step": 1061888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.260523545185517e-05,
+      "loss": 3.9861,
+      "step": 1062400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.259684950434465e-05,
+      "loss": 3.9748,
+      "step": 1062912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.258846355683413e-05,
+      "loss": 3.9778,
+      "step": 1063424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.25800776093236e-05,
+      "loss": 3.9823,
+      "step": 1063936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2571708040616816e-05,
+      "loss": 3.9721,
+      "step": 1064448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2563322093106296e-05,
+      "loss": 3.9751,
+      "step": 1064960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.255493614559577e-05,
+      "loss": 3.9825,
+      "step": 1065472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.254655019808525e-05,
+      "loss": 3.977,
+      "step": 1065984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.253816425057473e-05,
+      "loss": 3.9764,
+      "step": 1066496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2529794681867945e-05,
+      "loss": 3.9791,
+      "step": 1067008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2521408734357425e-05,
+      "loss": 3.9733,
+      "step": 1067520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.2513022786846905e-05,
+      "loss": 3.9667,
+      "step": 1068032
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.026306629180908,
+      "eval_runtime": 314.9074,
+      "eval_samples_per_second": 1211.756,
+      "eval_steps_per_second": 37.868,
+      "step": 1068467
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2504636839336385e-05,
+      "loss": 3.9746,
+      "step": 1068544
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2496250891825865e-05,
+      "loss": 3.9887,
+      "step": 1069056
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2487881323119074e-05,
+      "loss": 3.9733,
+      "step": 1069568
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2479495375608554e-05,
+      "loss": 3.9563,
+      "step": 1070080
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.247112580690177e-05,
+      "loss": 3.974,
+      "step": 1070592
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.246273985939124e-05,
+      "loss": 3.9732,
+      "step": 1071104
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.245435391188072e-05,
+      "loss": 3.9771,
+      "step": 1071616
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.24459679643702e-05,
+      "loss": 3.9658,
+      "step": 1072128
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.243758201685968e-05,
+      "loss": 3.9643,
+      "step": 1072640
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.242919606934916e-05,
+      "loss": 3.9613,
+      "step": 1073152
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.242081012183865e-05,
+      "loss": 3.9799,
+      "step": 1073664
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.241244055313186e-05,
+      "loss": 3.9658,
+      "step": 1074176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.240405460562134e-05,
+      "loss": 3.9781,
+      "step": 1074688
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.239566865811082e-05,
+      "loss": 3.978,
+      "step": 1075200
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.23872827106003e-05,
+      "loss": 3.9748,
+      "step": 1075712
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.237889676308978e-05,
+      "loss": 3.9591,
+      "step": 1076224
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.237051081557926e-05,
+      "loss": 3.9765,
+      "step": 1076736
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.236212486806874e-05,
+      "loss": 3.9746,
+      "step": 1077248
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.235373892055822e-05,
+      "loss": 3.9655,
+      "step": 1077760
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.23453529730477e-05,
+      "loss": 3.9849,
+      "step": 1078272
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.233696702553718e-05,
+      "loss": 3.9948,
+      "step": 1078784
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.232858107802666e-05,
+      "loss": 3.9794,
+      "step": 1079296
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.232019513051613e-05,
+      "loss": 3.9642,
+      "step": 1079808
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.231182556180935e-05,
+      "loss": 3.9712,
+      "step": 1080320
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2303439614298834e-05,
+      "loss": 3.9682,
+      "step": 1080832
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2295053666788314e-05,
+      "loss": 3.9657,
+      "step": 1081344
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.228666771927779e-05,
+      "loss": 3.972,
+      "step": 1081856
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2278298150571e-05,
+      "loss": 3.9636,
+      "step": 1082368
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.226991220306048e-05,
+      "loss": 3.9661,
+      "step": 1082880
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.2261526255549956e-05,
+      "loss": 3.9804,
+      "step": 1083392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2253140308039436e-05,
+      "loss": 3.975,
+      "step": 1083904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2244754360528916e-05,
+      "loss": 3.9761,
+      "step": 1084416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.223638479182213e-05,
+      "loss": 3.9723,
+      "step": 1084928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2227998844311605e-05,
+      "loss": 3.9812,
+      "step": 1085440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2219612896801085e-05,
+      "loss": 3.9604,
+      "step": 1085952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.221122694929057e-05,
+      "loss": 3.9728,
+      "step": 1086464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.220284100178005e-05,
+      "loss": 3.9635,
+      "step": 1086976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.219445505426953e-05,
+      "loss": 3.9863,
+      "step": 1087488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.218606910675901e-05,
+      "loss": 3.9748,
+      "step": 1088000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.217768315924849e-05,
+      "loss": 3.986,
+      "step": 1088512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.216929721173797e-05,
+      "loss": 3.9809,
+      "step": 1089024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.216092764303118e-05,
+      "loss": 3.9721,
+      "step": 1089536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.215254169552066e-05,
+      "loss": 3.9664,
+      "step": 1090048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.214415574801014e-05,
+      "loss": 3.9622,
+      "step": 1090560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.213576980049962e-05,
+      "loss": 3.9645,
+      "step": 1091072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.212740023179283e-05,
+      "loss": 3.9567,
+      "step": 1091584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.211901428428231e-05,
+      "loss": 3.9665,
+      "step": 1092096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.211062833677179e-05,
+      "loss": 3.9696,
+      "step": 1092608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.210224238926127e-05,
+      "loss": 3.9698,
+      "step": 1093120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.209385644175075e-05,
+      "loss": 3.9665,
+      "step": 1093632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2085470494240236e-05,
+      "loss": 3.9592,
+      "step": 1094144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2077084546729716e-05,
+      "loss": 3.9666,
+      "step": 1094656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2068714978022925e-05,
+      "loss": 3.9744,
+      "step": 1095168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2060329030512405e-05,
+      "loss": 3.9432,
+      "step": 1095680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2051943083001885e-05,
+      "loss": 3.9626,
+      "step": 1096192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2043557135491365e-05,
+      "loss": 3.9594,
+      "step": 1096704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2035171187980845e-05,
+      "loss": 3.981,
+      "step": 1097216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.202678524047032e-05,
+      "loss": 3.9489,
+      "step": 1097728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2018415671763534e-05,
+      "loss": 3.971,
+      "step": 1098240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2010029724253014e-05,
+      "loss": 3.9675,
+      "step": 1098752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.2001643776742494e-05,
+      "loss": 3.9617,
+      "step": 1099264
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1993257829231974e-05,
+      "loss": 3.9571,
+      "step": 1099776
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1984871881721454e-05,
+      "loss": 3.9497,
+      "step": 1100288
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1976485934210934e-05,
+      "loss": 3.9525,
+      "step": 1100800
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1968099986700414e-05,
+      "loss": 3.9757,
+      "step": 1101312
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1959714039189894e-05,
+      "loss": 3.9711,
+      "step": 1101824
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1951328091679373e-05,
+      "loss": 3.9613,
+      "step": 1102336
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.194295852297258e-05,
+      "loss": 3.9708,
+      "step": 1102848
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.193457257546206e-05,
+      "loss": 3.9681,
+      "step": 1103360
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.192618662795154e-05,
+      "loss": 3.9624,
+      "step": 1103872
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.191780068044102e-05,
+      "loss": 3.9721,
+      "step": 1104384
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.19094147329305e-05,
+      "loss": 3.9687,
+      "step": 1104896
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.190104516422371e-05,
+      "loss": 3.9618,
+      "step": 1105408
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.189265921671319e-05,
+      "loss": 3.9634,
+      "step": 1105920
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.188427326920267e-05,
+      "loss": 3.9751,
+      "step": 1106432
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.187588732169216e-05,
+      "loss": 3.9648,
+      "step": 1106944
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.186751775298537e-05,
+      "loss": 3.9757,
+      "step": 1107456
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1859148184278576e-05,
+      "loss": 3.9684,
+      "step": 1107968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1850762236768056e-05,
+      "loss": 3.9646,
+      "step": 1108480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1842392668061265e-05,
+      "loss": 3.9639,
+      "step": 1108992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1834006720550745e-05,
+      "loss": 3.969,
+      "step": 1109504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1825620773040225e-05,
+      "loss": 3.9624,
+      "step": 1110016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1817234825529705e-05,
+      "loss": 3.9608,
+      "step": 1110528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1808848878019185e-05,
+      "loss": 3.964,
+      "step": 1111040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1800462930508665e-05,
+      "loss": 3.9545,
+      "step": 1111552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1792076982998145e-05,
+      "loss": 3.9658,
+      "step": 1112064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.1783691035487625e-05,
+      "loss": 3.9573,
+      "step": 1112576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.177530508797711e-05,
+      "loss": 3.965,
+      "step": 1113088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.176691914046659e-05,
+      "loss": 3.9519,
+      "step": 1113600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 3.17585495717598e-05,
+      "loss": 3.969,
+      "step": 1114112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.175016362424928e-05,
+      "loss": 3.9722,
+      "step": 1114624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.174177767673876e-05,
+      "loss": 3.9707,
+      "step": 1115136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.173339172922824e-05,
+      "loss": 3.9673,
+      "step": 1115648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.172500578171772e-05,
+      "loss": 3.9608,
+      "step": 1116160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.17166198342072e-05,
+      "loss": 3.9621,
+      "step": 1116672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.170823388669668e-05,
+      "loss": 3.964,
+      "step": 1117184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1699847939186154e-05,
+      "loss": 3.97,
+      "step": 1117696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.169147837047937e-05,
+      "loss": 3.9525,
+      "step": 1118208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.168309242296885e-05,
+      "loss": 3.9577,
+      "step": 1118720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.167470647545833e-05,
+      "loss": 3.9735,
+      "step": 1119232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.166632052794781e-05,
+      "loss": 3.9621,
+      "step": 1119744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1657950959241025e-05,
+      "loss": 3.9776,
+      "step": 1120256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1649565011730505e-05,
+      "loss": 3.9601,
+      "step": 1120768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.164117906421998e-05,
+      "loss": 3.9678,
+      "step": 1121280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.163279311670946e-05,
+      "loss": 3.9514,
+      "step": 1121792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.162440716919894e-05,
+      "loss": 3.9568,
+      "step": 1122304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1616037600492154e-05,
+      "loss": 3.9665,
+      "step": 1122816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.160765165298163e-05,
+      "loss": 3.9686,
+      "step": 1123328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.159926570547111e-05,
+      "loss": 3.9604,
+      "step": 1123840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.159089613676432e-05,
+      "loss": 3.9475,
+      "step": 1124352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.15825101892538e-05,
+      "loss": 3.9643,
+      "step": 1124864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.157412424174328e-05,
+      "loss": 3.9678,
+      "step": 1125376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.156573829423276e-05,
+      "loss": 3.9579,
+      "step": 1125888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.155735234672224e-05,
+      "loss": 3.9606,
+      "step": 1126400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.154898277801545e-05,
+      "loss": 3.955,
+      "step": 1126912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.154059683050493e-05,
+      "loss": 3.9567,
+      "step": 1127424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.153221088299441e-05,
+      "loss": 3.9566,
+      "step": 1127936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.152382493548389e-05,
+      "loss": 3.9627,
+      "step": 1128448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.15154553667771e-05,
+      "loss": 3.9621,
+      "step": 1128960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.150706941926658e-05,
+      "loss": 3.9627,
+      "step": 1129472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.149868347175606e-05,
+      "loss": 3.9681,
+      "step": 1129984
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.149029752424554e-05,
+      "loss": 3.9817,
+      "step": 1130496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.148191157673502e-05,
+      "loss": 3.9593,
+      "step": 1131008
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.14735256292245e-05,
+      "loss": 3.9685,
+      "step": 1131520
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.146513968171398e-05,
+      "loss": 3.9637,
+      "step": 1132032
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.145675373420347e-05,
+      "loss": 3.9602,
+      "step": 1132544
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.144836778669295e-05,
+      "loss": 3.9616,
+      "step": 1133056
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.143998183918243e-05,
+      "loss": 3.9566,
+      "step": 1133568
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.143159589167191e-05,
+      "loss": 3.9682,
+      "step": 1134080
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.142320994416139e-05,
+      "loss": 3.9522,
+      "step": 1134592
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1414840375454597e-05,
+      "loss": 3.978,
+      "step": 1135104
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1406454427944077e-05,
+      "loss": 3.9615,
+      "step": 1135616
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1398068480433556e-05,
+      "loss": 3.9736,
+      "step": 1136128
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1389682532923036e-05,
+      "loss": 3.9621,
+      "step": 1136640
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1381296585412516e-05,
+      "loss": 3.9678,
+      "step": 1137152
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1372927016705726e-05,
+      "loss": 3.9605,
+      "step": 1137664
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1364541069195205e-05,
+      "loss": 3.9688,
+      "step": 1138176
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.135617150048842e-05,
+      "loss": 3.9743,
+      "step": 1138688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.13477855529779e-05,
+      "loss": 3.96,
+      "step": 1139200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.133939960546738e-05,
+      "loss": 3.966,
+      "step": 1139712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.133101365795686e-05,
+      "loss": 3.9688,
+      "step": 1140224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.132262771044634e-05,
+      "loss": 3.9626,
+      "step": 1140736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1314241762935814e-05,
+      "loss": 3.9557,
+      "step": 1141248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.1305855815425294e-05,
+      "loss": 3.9703,
+      "step": 1141760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.129748624671851e-05,
+      "loss": 3.9646,
+      "step": 1142272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.128910029920799e-05,
+      "loss": 3.962,
+      "step": 1142784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.128071435169746e-05,
+      "loss": 3.9621,
+      "step": 1143296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.127232840418694e-05,
+      "loss": 3.9652,
+      "step": 1143808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 3.126394245667642e-05,
+      "loss": 3.9546,
+      "step": 1144320
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.018762111663818,
+      "eval_runtime": 289.5623,
+      "eval_samples_per_second": 1317.82,
+      "eval_steps_per_second": 41.183,
+      "step": 1144787
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.12555565091659e-05,
+      "loss": 3.9267,
+      "step": 1144832
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.124717056165539e-05,
+      "loss": 3.9558,
+      "step": 1145344
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.123878461414487e-05,
+      "loss": 3.9504,
+      "step": 1145856
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.123039866663435e-05,
+      "loss": 3.9733,
+      "step": 1146368
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.122201271912383e-05,
+      "loss": 3.9632,
+      "step": 1146880
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.121362677161331e-05,
+      "loss": 3.9645,
+      "step": 1147392
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.120524082410279e-05,
+      "loss": 3.9571,
+      "step": 1147904
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.119685487659227e-05,
+      "loss": 3.9595,
+      "step": 1148416
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.118846892908175e-05,
+      "loss": 3.9502,
+      "step": 1148928
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.118008298157123e-05,
+      "loss": 3.9606,
+      "step": 1149440
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.11716970340607e-05,
+      "loss": 3.9608,
+      "step": 1149952
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.116331108655018e-05,
+      "loss": 3.9749,
+      "step": 1150464
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.11549415178434e-05,
+      "loss": 3.9708,
+      "step": 1150976
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.114655557033288e-05,
+      "loss": 3.9434,
+      "step": 1151488
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.113816962282236e-05,
+      "loss": 3.9594,
+      "step": 1152000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.112978367531184e-05,
+      "loss": 3.9507,
+      "step": 1152512
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.112139772780132e-05,
+      "loss": 3.9631,
+      "step": 1153024
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.11130117802908e-05,
+      "loss": 3.9553,
+      "step": 1153536
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.110462583278028e-05,
+      "loss": 3.9624,
+      "step": 1154048
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.109623988526976e-05,
+      "loss": 3.9574,
+      "step": 1154560
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.108787031656297e-05,
+      "loss": 3.9731,
+      "step": 1155072
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.107948436905245e-05,
+      "loss": 3.9616,
+      "step": 1155584
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.107109842154193e-05,
+      "loss": 3.9614,
+      "step": 1156096
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.106271247403141e-05,
+      "loss": 3.9562,
+      "step": 1156608
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.105432652652089e-05,
+      "loss": 3.9702,
+      "step": 1157120
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.104594057901037e-05,
+      "loss": 3.9578,
+      "step": 1157632
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.103755463149985e-05,
+      "loss": 3.9517,
+      "step": 1158144
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1029185062793056e-05,
+      "loss": 3.9604,
+      "step": 1158656
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.1020799115282536e-05,
+      "loss": 3.9537,
+      "step": 1159168
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.101241316777202e-05,
+      "loss": 3.9549,
+      "step": 1159680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.10040272202615e-05,
+      "loss": 3.9591,
+      "step": 1160192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.099564127275098e-05,
+      "loss": 3.9544,
+      "step": 1160704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.098727170404419e-05,
+      "loss": 3.9675,
+      "step": 1161216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.097888575653367e-05,
+      "loss": 3.9648,
+      "step": 1161728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.097049980902315e-05,
+      "loss": 3.9641,
+      "step": 1162240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.096213024031636e-05,
+      "loss": 3.9627,
+      "step": 1162752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.095374429280584e-05,
+      "loss": 3.9657,
+      "step": 1163264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.094535834529532e-05,
+      "loss": 3.9454,
+      "step": 1163776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.09369723977848e-05,
+      "loss": 3.9545,
+      "step": 1164288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.092858645027428e-05,
+      "loss": 3.9541,
+      "step": 1164800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.092020050276376e-05,
+      "loss": 3.9463,
+      "step": 1165312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.091181455525324e-05,
+      "loss": 3.959,
+      "step": 1165824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.090342860774272e-05,
+      "loss": 3.9577,
+      "step": 1166336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0895059039035936e-05,
+      "loss": 3.9501,
+      "step": 1166848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0886673091525416e-05,
+      "loss": 3.9622,
+      "step": 1167360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.087828714401489e-05,
+      "loss": 3.9625,
+      "step": 1167872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.086990119650437e-05,
+      "loss": 3.9616,
+      "step": 1168384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.086151524899385e-05,
+      "loss": 3.9589,
+      "step": 1168896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.085312930148333e-05,
+      "loss": 3.9408,
+      "step": 1169408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.084474335397281e-05,
+      "loss": 3.9579,
+      "step": 1169920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.083635740646229e-05,
+      "loss": 3.9634,
+      "step": 1170432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.08279878377555e-05,
+      "loss": 3.9586,
+      "step": 1170944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0819618269048714e-05,
+      "loss": 3.9528,
+      "step": 1171456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0811232321538194e-05,
+      "loss": 3.9507,
+      "step": 1171968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0802846374027674e-05,
+      "loss": 3.9462,
+      "step": 1172480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0794460426517154e-05,
+      "loss": 3.9464,
+      "step": 1172992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0786074479006634e-05,
+      "loss": 3.9568,
+      "step": 1173504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0777688531496114e-05,
+      "loss": 3.9403,
+      "step": 1174016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0769302583985594e-05,
+      "loss": 3.9594,
+      "step": 1174528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0760916636475074e-05,
+      "loss": 3.9592,
+      "step": 1175040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0752530688964554e-05,
+      "loss": 3.9426,
+      "step": 1175552
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.074416112025776e-05,
+      "loss": 3.965,
+      "step": 1176064
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.073577517274724e-05,
+      "loss": 3.9421,
+      "step": 1176576
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.072738922523672e-05,
+      "loss": 3.9469,
+      "step": 1177088
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.07190032777262e-05,
+      "loss": 3.9466,
+      "step": 1177600
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.071063370901941e-05,
+      "loss": 3.9578,
+      "step": 1178112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.07022477615089e-05,
+      "loss": 3.9432,
+      "step": 1178624
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.069386181399838e-05,
+      "loss": 3.9591,
+      "step": 1179136
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.068547586648786e-05,
+      "loss": 3.935,
+      "step": 1179648
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.067710629778107e-05,
+      "loss": 3.9466,
+      "step": 1180160
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.066873672907428e-05,
+      "loss": 3.9635,
+      "step": 1180672
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.066035078156376e-05,
+      "loss": 3.9537,
+      "step": 1181184
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.065196483405324e-05,
+      "loss": 3.9567,
+      "step": 1181696
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.064357888654272e-05,
+      "loss": 3.96,
+      "step": 1182208
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.06351929390322e-05,
+      "loss": 3.9701,
+      "step": 1182720
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.062680699152168e-05,
+      "loss": 3.9489,
+      "step": 1183232
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0618421044011157e-05,
+      "loss": 3.9671,
+      "step": 1183744
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0610035096500637e-05,
+      "loss": 3.9445,
+      "step": 1184256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.060166552779385e-05,
+      "loss": 3.9502,
+      "step": 1184768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.059327958028333e-05,
+      "loss": 3.9511,
+      "step": 1185280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.058489363277281e-05,
+      "loss": 3.9615,
+      "step": 1185792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.057650768526229e-05,
+      "loss": 3.9575,
+      "step": 1186304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.05681381165555e-05,
+      "loss": 3.9672,
+      "step": 1186816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.055975216904498e-05,
+      "loss": 3.9478,
+      "step": 1187328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.055136622153446e-05,
+      "loss": 3.9394,
+      "step": 1187840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.054298027402394e-05,
+      "loss": 3.9468,
+      "step": 1188352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.053461070531715e-05,
+      "loss": 3.9544,
+      "step": 1188864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.052622475780663e-05,
+      "loss": 3.95,
+      "step": 1189376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.051785518909984e-05,
+      "loss": 3.9492,
+      "step": 1189888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0509469241589323e-05,
+      "loss": 3.9552,
+      "step": 1190400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0501083294078803e-05,
+      "loss": 3.9466,
+      "step": 1190912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0492697346568283e-05,
+      "loss": 3.9498,
+      "step": 1191424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0484311399057763e-05,
+      "loss": 3.9497,
+      "step": 1191936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0475925451547243e-05,
+      "loss": 3.9608,
+      "step": 1192448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0467539504036723e-05,
+      "loss": 3.9674,
+      "step": 1192960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0459169935329935e-05,
+      "loss": 3.9601,
+      "step": 1193472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0450783987819415e-05,
+      "loss": 3.9393,
+      "step": 1193984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0442398040308895e-05,
+      "loss": 3.9514,
+      "step": 1194496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0434012092798375e-05,
+      "loss": 3.9479,
+      "step": 1195008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.042562614528785e-05,
+      "loss": 3.9563,
+      "step": 1195520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.041724019777733e-05,
+      "loss": 3.9589,
+      "step": 1196032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.040885425026681e-05,
+      "loss": 3.9452,
+      "step": 1196544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.040046830275629e-05,
+      "loss": 3.9548,
+      "step": 1197056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.039208235524577e-05,
+      "loss": 3.9552,
+      "step": 1197568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.038371278653898e-05,
+      "loss": 3.9514,
+      "step": 1198080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0375343217832196e-05,
+      "loss": 3.9475,
+      "step": 1198592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.036697364912541e-05,
+      "loss": 3.9476,
+      "step": 1199104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.035858770161489e-05,
+      "loss": 3.9411,
+      "step": 1199616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.035020175410437e-05,
+      "loss": 3.9523,
+      "step": 1200128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.034181580659385e-05,
+      "loss": 3.9452,
+      "step": 1200640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0333429859083322e-05,
+      "loss": 3.9446,
+      "step": 1201152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0325043911572805e-05,
+      "loss": 3.948,
+      "step": 1201664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0316657964062285e-05,
+      "loss": 3.955,
+      "step": 1202176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0308272016551765e-05,
+      "loss": 3.9446,
+      "step": 1202688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0299886069041245e-05,
+      "loss": 3.9503,
+      "step": 1203200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0291500121530725e-05,
+      "loss": 3.9641,
+      "step": 1203712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0283114174020205e-05,
+      "loss": 3.9597,
+      "step": 1204224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0274728226509685e-05,
+      "loss": 3.9597,
+      "step": 1204736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0266358657802897e-05,
+      "loss": 3.949,
+      "step": 1205248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0257972710292377e-05,
+      "loss": 3.9521,
+      "step": 1205760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0249586762781857e-05,
+      "loss": 3.9502,
+      "step": 1206272
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0241233572878795e-05,
+      "loss": 3.9398,
+      "step": 1206784
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0232847625368275e-05,
+      "loss": 3.9516,
+      "step": 1207296
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.022446167785776e-05,
+      "loss": 3.9445,
+      "step": 1207808
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.021607573034724e-05,
+      "loss": 3.9548,
+      "step": 1208320
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.020768978283672e-05,
+      "loss": 3.9468,
+      "step": 1208832
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.01993038353262e-05,
+      "loss": 3.9542,
+      "step": 1209344
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0190934266619408e-05,
+      "loss": 3.947,
+      "step": 1209856
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0182548319108888e-05,
+      "loss": 3.9448,
+      "step": 1210368
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0174162371598368e-05,
+      "loss": 3.9535,
+      "step": 1210880
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.016577642408785e-05,
+      "loss": 3.9556,
+      "step": 1211392
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.015739047657733e-05,
+      "loss": 3.9592,
+      "step": 1211904
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.014900452906681e-05,
+      "loss": 3.9501,
+      "step": 1212416
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.014061858155629e-05,
+      "loss": 3.9563,
+      "step": 1212928
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.013223263404577e-05,
+      "loss": 3.9553,
+      "step": 1213440
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.012384668653525e-05,
+      "loss": 3.9546,
+      "step": 1213952
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.011546073902473e-05,
+      "loss": 3.9544,
+      "step": 1214464
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.010707479151421e-05,
+      "loss": 3.9507,
+      "step": 1214976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0098705222807423e-05,
+      "loss": 3.9512,
+      "step": 1215488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0090319275296903e-05,
+      "loss": 3.9592,
+      "step": 1216000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0081933327786383e-05,
+      "loss": 3.9479,
+      "step": 1216512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0073547380275856e-05,
+      "loss": 3.9455,
+      "step": 1217024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0065161432765336e-05,
+      "loss": 3.9516,
+      "step": 1217536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0056775485254816e-05,
+      "loss": 3.9527,
+      "step": 1218048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0048405916548032e-05,
+      "loss": 3.9509,
+      "step": 1218560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.004001996903751e-05,
+      "loss": 3.9557,
+      "step": 1219072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.003163402152699e-05,
+      "loss": 3.948,
+      "step": 1219584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.002324807401647e-05,
+      "loss": 3.9613,
+      "step": 1220096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.0014878505309678e-05,
+      "loss": 3.9525,
+      "step": 1220608
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.0123610496521,
+      "eval_runtime": 289.9188,
+      "eval_samples_per_second": 1316.199,
+      "eval_steps_per_second": 41.132,
+      "step": 1221107
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.000649255779916e-05,
+      "loss": 3.9485,
+      "step": 1221120
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.999810661028864e-05,
+      "loss": 3.9417,
+      "step": 1221632
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.998972066277812e-05,
+      "loss": 3.9415,
+      "step": 1222144
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.99813347152676e-05,
+      "loss": 3.9584,
+      "step": 1222656
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.997294876775708e-05,
+      "loss": 3.9529,
+      "step": 1223168
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.996456282024656e-05,
+      "loss": 3.9524,
+      "step": 1223680
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.995617687273604e-05,
+      "loss": 3.9465,
+      "step": 1224192
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9947807304029253e-05,
+      "loss": 3.9483,
+      "step": 1224704
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9939421356518733e-05,
+      "loss": 3.9383,
+      "step": 1225216
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9931035409008213e-05,
+      "loss": 3.9494,
+      "step": 1225728
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9922649461497693e-05,
+      "loss": 3.949,
+      "step": 1226240
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9914279892790902e-05,
+      "loss": 3.9653,
+      "step": 1226752
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9905926702887847e-05,
+      "loss": 3.9597,
+      "step": 1227264
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.989754075537733e-05,
+      "loss": 3.9317,
+      "step": 1227776
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9889154807866804e-05,
+      "loss": 3.9456,
+      "step": 1228288
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9880768860356284e-05,
+      "loss": 3.9405,
+      "step": 1228800
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9872382912845764e-05,
+      "loss": 3.9445,
+      "step": 1229312
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9863996965335244e-05,
+      "loss": 3.9493,
+      "step": 1229824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9855611017824723e-05,
+      "loss": 3.9484,
+      "step": 1230336
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9847225070314207e-05,
+      "loss": 3.9442,
+      "step": 1230848
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9838839122803687e-05,
+      "loss": 3.9635,
+      "step": 1231360
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9830453175293167e-05,
+      "loss": 3.9511,
+      "step": 1231872
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9822067227782647e-05,
+      "loss": 3.947,
+      "step": 1232384
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9813681280272127e-05,
+      "loss": 3.9426,
+      "step": 1232896
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9805311711565336e-05,
+      "loss": 3.9625,
+      "step": 1233408
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9796925764054816e-05,
+      "loss": 3.9462,
+      "step": 1233920
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.97885398165443e-05,
+      "loss": 3.9401,
+      "step": 1234432
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.978015386903378e-05,
+      "loss": 3.9489,
+      "step": 1234944
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.977176792152326e-05,
+      "loss": 3.9421,
+      "step": 1235456
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.9763398352816468e-05,
+      "loss": 3.9441,
+      "step": 1235968
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9755012405305948e-05,
+      "loss": 3.9468,
+      "step": 1236480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9746626457795428e-05,
+      "loss": 3.9443,
+      "step": 1236992
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.973825688908864e-05,
+      "loss": 3.9553,
+      "step": 1237504
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.972987094157812e-05,
+      "loss": 3.9478,
+      "step": 1238016
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.97214849940676e-05,
+      "loss": 3.9559,
+      "step": 1238528
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.971311542536081e-05,
+      "loss": 3.9516,
+      "step": 1239040
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.970472947785029e-05,
+      "loss": 3.9548,
+      "step": 1239552
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.969634353033977e-05,
+      "loss": 3.9356,
+      "step": 1240064
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9687957582829253e-05,
+      "loss": 3.9449,
+      "step": 1240576
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9679571635318733e-05,
+      "loss": 3.9385,
+      "step": 1241088
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9671185687808213e-05,
+      "loss": 3.943,
+      "step": 1241600
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9662799740297693e-05,
+      "loss": 3.941,
+      "step": 1242112
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9654413792787166e-05,
+      "loss": 3.9488,
+      "step": 1242624
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9646027845276646e-05,
+      "loss": 3.9404,
+      "step": 1243136
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.963764189776613e-05,
+      "loss": 3.9508,
+      "step": 1243648
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.962925595025561e-05,
+      "loss": 3.9492,
+      "step": 1244160
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9620886381548818e-05,
+      "loss": 3.9538,
+      "step": 1244672
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9612500434038298e-05,
+      "loss": 3.947,
+      "step": 1245184
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9604114486527778e-05,
+      "loss": 3.93,
+      "step": 1245696
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9595728539017258e-05,
+      "loss": 3.9408,
+      "step": 1246208
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.958735897031047e-05,
+      "loss": 3.9559,
+      "step": 1246720
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.957897302279995e-05,
+      "loss": 3.9484,
+      "step": 1247232
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.957058707528943e-05,
+      "loss": 3.9398,
+      "step": 1247744
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.956220112777891e-05,
+      "loss": 3.9397,
+      "step": 1248256
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.955381518026839e-05,
+      "loss": 3.9356,
+      "step": 1248768
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.954542923275787e-05,
+      "loss": 3.9327,
+      "step": 1249280
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.953704328524735e-05,
+      "loss": 3.9546,
+      "step": 1249792
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.952865733773683e-05,
+      "loss": 3.9292,
+      "step": 1250304
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.952027139022631e-05,
+      "loss": 3.9484,
+      "step": 1250816
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9511901821519523e-05,
+      "loss": 3.9487,
+      "step": 1251328
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9503515874009003e-05,
+      "loss": 3.9313,
+      "step": 1251840
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9495129926498483e-05,
+      "loss": 3.9539,
+      "step": 1252352
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9486743978987962e-05,
+      "loss": 3.9344,
+      "step": 1252864
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9478358031477442e-05,
+      "loss": 3.9373,
+      "step": 1253376
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9469988462770655e-05,
+      "loss": 3.9293,
+      "step": 1253888
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9461602515260135e-05,
+      "loss": 3.951,
+      "step": 1254400
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9453216567749615e-05,
+      "loss": 3.9346,
+      "step": 1254912
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9444830620239095e-05,
+      "loss": 3.9413,
+      "step": 1255424
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9436461051532304e-05,
+      "loss": 3.9267,
+      "step": 1255936
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9428075104021784e-05,
+      "loss": 3.9352,
+      "step": 1256448
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9419689156511264e-05,
+      "loss": 3.9509,
+      "step": 1256960
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9411319587804476e-05,
+      "loss": 3.9496,
+      "step": 1257472
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9402933640293956e-05,
+      "loss": 3.9422,
+      "step": 1257984
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9394547692783436e-05,
+      "loss": 3.9465,
+      "step": 1258496
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9386161745272916e-05,
+      "loss": 3.9601,
+      "step": 1259008
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9377775797762396e-05,
+      "loss": 3.9374,
+      "step": 1259520
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9369389850251876e-05,
+      "loss": 3.9551,
+      "step": 1260032
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9361003902741353e-05,
+      "loss": 3.9338,
+      "step": 1260544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9352617955230833e-05,
+      "loss": 3.943,
+      "step": 1261056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.934424838652405e-05,
+      "loss": 3.9429,
+      "step": 1261568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.933586243901353e-05,
+      "loss": 3.9477,
+      "step": 1262080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9327476491503e-05,
+      "loss": 3.9462,
+      "step": 1262592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9319090543992485e-05,
+      "loss": 3.9564,
+      "step": 1263104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.93107209752857e-05,
+      "loss": 3.942,
+      "step": 1263616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.930235140657891e-05,
+      "loss": 3.9264,
+      "step": 1264128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.929396545906839e-05,
+      "loss": 3.9341,
+      "step": 1264640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.928557951155787e-05,
+      "loss": 3.9461,
+      "step": 1265152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.927719356404735e-05,
+      "loss": 3.9373,
+      "step": 1265664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9268807616536826e-05,
+      "loss": 3.9405,
+      "step": 1266176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.9260421669026306e-05,
+      "loss": 3.9458,
+      "step": 1266688
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9252052100319522e-05,
+      "loss": 3.935,
+      "step": 1267200
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9243666152809002e-05,
+      "loss": 3.9408,
+      "step": 1267712
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9235280205298475e-05,
+      "loss": 3.9368,
+      "step": 1268224
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9226894257787955e-05,
+      "loss": 3.9509,
+      "step": 1268736
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.921852468908117e-05,
+      "loss": 3.9533,
+      "step": 1269248
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9210138741570648e-05,
+      "loss": 3.9487,
+      "step": 1269760
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9201752794060128e-05,
+      "loss": 3.9299,
+      "step": 1270272
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9193366846549608e-05,
+      "loss": 3.9399,
+      "step": 1270784
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9184997277842823e-05,
+      "loss": 3.943,
+      "step": 1271296
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.91766113303323e-05,
+      "loss": 3.9448,
+      "step": 1271808
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.916822538282178e-05,
+      "loss": 3.9495,
+      "step": 1272320
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.915983943531126e-05,
+      "loss": 3.9343,
+      "step": 1272832
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.915145348780074e-05,
+      "loss": 3.9412,
+      "step": 1273344
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.914306754029022e-05,
+      "loss": 3.9448,
+      "step": 1273856
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.91346815927797e-05,
+      "loss": 3.9438,
+      "step": 1274368
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.912629564526918e-05,
+      "loss": 3.9318,
+      "step": 1274880
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9117926076562392e-05,
+      "loss": 3.9419,
+      "step": 1275392
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9109540129051872e-05,
+      "loss": 3.9317,
+      "step": 1275904
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9101154181541352e-05,
+      "loss": 3.936,
+      "step": 1276416
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9092768234030832e-05,
+      "loss": 3.938,
+      "step": 1276928
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9084382286520312e-05,
+      "loss": 3.9348,
+      "step": 1277440
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.907601271781352e-05,
+      "loss": 3.9343,
+      "step": 1277952
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9067626770303e-05,
+      "loss": 3.9428,
+      "step": 1278464
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9059240822792484e-05,
+      "loss": 3.9355,
+      "step": 1278976
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9050854875281964e-05,
+      "loss": 3.9393,
+      "step": 1279488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9042468927771444e-05,
+      "loss": 3.9507,
+      "step": 1280000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9034099359064653e-05,
+      "loss": 3.9525,
+      "step": 1280512
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9025713411554133e-05,
+      "loss": 3.947,
+      "step": 1281024
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9017327464043613e-05,
+      "loss": 3.9377,
+      "step": 1281536
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9008941516533093e-05,
+      "loss": 3.9404,
+      "step": 1282048
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.9000555569022573e-05,
+      "loss": 3.9382,
+      "step": 1282560
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8992169621512057e-05,
+      "loss": 3.9318,
+      "step": 1283072
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8983783674001537e-05,
+      "loss": 3.9381,
+      "step": 1283584
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.897539772649101e-05,
+      "loss": 3.9373,
+      "step": 1284096
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8967028157784226e-05,
+      "loss": 3.9435,
+      "step": 1284608
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8958642210273706e-05,
+      "loss": 3.9376,
+      "step": 1285120
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8950256262763186e-05,
+      "loss": 3.9442,
+      "step": 1285632
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8941886694056398e-05,
+      "loss": 3.9404,
+      "step": 1286144
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8933500746545878e-05,
+      "loss": 3.9367,
+      "step": 1286656
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8925114799035358e-05,
+      "loss": 3.9339,
+      "step": 1287168
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8916728851524838e-05,
+      "loss": 3.952,
+      "step": 1287680
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.890834290401431e-05,
+      "loss": 3.9445,
+      "step": 1288192
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8899956956503794e-05,
+      "loss": 3.9447,
+      "step": 1288704
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8891571008993274e-05,
+      "loss": 3.9438,
+      "step": 1289216
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8883185061482754e-05,
+      "loss": 3.9473,
+      "step": 1289728
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.887484825038343e-05,
+      "loss": 3.9464,
+      "step": 1290240
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.886646230287291e-05,
+      "loss": 3.9454,
+      "step": 1290752
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.885807635536239e-05,
+      "loss": 3.9401,
+      "step": 1291264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8849690407851872e-05,
+      "loss": 3.9428,
+      "step": 1291776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8841304460341352e-05,
+      "loss": 3.9497,
+      "step": 1292288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.883291851283083e-05,
+      "loss": 3.9371,
+      "step": 1292800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8824532565320305e-05,
+      "loss": 3.9348,
+      "step": 1293312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8816146617809785e-05,
+      "loss": 3.94,
+      "step": 1293824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8807760670299265e-05,
+      "loss": 3.9469,
+      "step": 1294336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8799374722788748e-05,
+      "loss": 3.9332,
+      "step": 1294848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8790988775278228e-05,
+      "loss": 3.9476,
+      "step": 1295360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8782602827767708e-05,
+      "loss": 3.9386,
+      "step": 1295872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8774249637864653e-05,
+      "loss": 3.95,
+      "step": 1296384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.8765863690354133e-05,
+      "loss": 3.9442,
+      "step": 1296896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.875747774284361e-05,
+      "loss": 3.9332,
+      "step": 1297408
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 4.00740385055542,
+      "eval_runtime": 289.9055,
+      "eval_samples_per_second": 1316.26,
+      "eval_steps_per_second": 41.134,
+      "step": 1297427
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.874909179533309e-05,
+      "loss": 3.9306,
+      "step": 1297920
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.874070584782257e-05,
+      "loss": 3.9307,
+      "step": 1298432
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.873231990031205e-05,
+      "loss": 3.9493,
+      "step": 1298944
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.872393395280153e-05,
+      "loss": 3.9407,
+      "step": 1299456
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.871554800529101e-05,
+      "loss": 3.9441,
+      "step": 1299968
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.870716205778049e-05,
+      "loss": 3.9398,
+      "step": 1300480
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.869877611026997e-05,
+      "loss": 3.9348,
+      "step": 1300992
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.869039016275945e-05,
+      "loss": 3.9275,
+      "step": 1301504
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8682004215248933e-05,
+      "loss": 3.9436,
+      "step": 1302016
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8673618267738412e-05,
+      "loss": 3.9326,
+      "step": 1302528
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8665232320227892e-05,
+      "loss": 3.9576,
+      "step": 1303040
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8656846372717372e-05,
+      "loss": 3.9465,
+      "step": 1303552
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.864847680401058e-05,
+      "loss": 3.9302,
+      "step": 1304064
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.864009085650006e-05,
+      "loss": 3.9322,
+      "step": 1304576
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.863170490898954e-05,
+      "loss": 3.9319,
+      "step": 1305088
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8623318961479025e-05,
+      "loss": 3.9317,
+      "step": 1305600
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8614933013968498e-05,
+      "loss": 3.9396,
+      "step": 1306112
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8606547066457978e-05,
+      "loss": 3.9422,
+      "step": 1306624
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8598161118947458e-05,
+      "loss": 3.9286,
+      "step": 1307136
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8589775171436938e-05,
+      "loss": 3.9582,
+      "step": 1307648
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8581389223926418e-05,
+      "loss": 3.9366,
+      "step": 1308160
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.85730032764159e-05,
+      "loss": 3.9408,
+      "step": 1308672
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.856461732890538e-05,
+      "loss": 3.9321,
+      "step": 1309184
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.855624776019859e-05,
+      "loss": 3.9519,
+      "step": 1309696
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.854786181268807e-05,
+      "loss": 3.9321,
+      "step": 1310208
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.853947586517755e-05,
+      "loss": 3.937,
+      "step": 1310720
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.853108991766703e-05,
+      "loss": 3.9335,
+      "step": 1311232
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.852270397015651e-05,
+      "loss": 3.9324,
+      "step": 1311744
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.851431802264599e-05,
+      "loss": 3.9333,
+      "step": 1312256
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8505932075135473e-05,
+      "loss": 3.9358,
+      "step": 1312768
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8497546127624953e-05,
+      "loss": 3.9371,
+      "step": 1313280
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8489176558918162e-05,
+      "loss": 3.9397,
+      "step": 1313792
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8480790611407642e-05,
+      "loss": 3.9422,
+      "step": 1314304
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8472404663897122e-05,
+      "loss": 3.9456,
+      "step": 1314816
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8464035095190335e-05,
+      "loss": 3.9406,
+      "step": 1315328
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8455649147679815e-05,
+      "loss": 3.9423,
+      "step": 1315840
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8447263200169295e-05,
+      "loss": 3.929,
+      "step": 1316352
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8438877252658775e-05,
+      "loss": 3.9352,
+      "step": 1316864
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8430507683951984e-05,
+      "loss": 3.9285,
+      "step": 1317376
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8422121736441464e-05,
+      "loss": 3.9355,
+      "step": 1317888
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8413735788930944e-05,
+      "loss": 3.9302,
+      "step": 1318400
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8405366220224156e-05,
+      "loss": 3.9376,
+      "step": 1318912
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8396980272713636e-05,
+      "loss": 3.9313,
+      "step": 1319424
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8388594325203116e-05,
+      "loss": 3.9424,
+      "step": 1319936
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8380208377692596e-05,
+      "loss": 3.9383,
+      "step": 1320448
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8371838808985805e-05,
+      "loss": 3.9434,
+      "step": 1320960
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.836345286147529e-05,
+      "loss": 3.9367,
+      "step": 1321472
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.835506691396477e-05,
+      "loss": 3.9224,
+      "step": 1321984
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8346680966454248e-05,
+      "loss": 3.9261,
+      "step": 1322496
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8338295018943728e-05,
+      "loss": 3.9477,
+      "step": 1323008
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8329909071433208e-05,
+      "loss": 3.9368,
+      "step": 1323520
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.832152312392268e-05,
+      "loss": 3.9302,
+      "step": 1324032
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8313137176412165e-05,
+      "loss": 3.9252,
+      "step": 1324544
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8304751228901645e-05,
+      "loss": 3.9329,
+      "step": 1325056
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8296365281391125e-05,
+      "loss": 3.9155,
+      "step": 1325568
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8287979333880605e-05,
+      "loss": 3.9468,
+      "step": 1326080
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8279593386370085e-05,
+      "loss": 3.9143,
+      "step": 1326592
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.827124019646703e-05,
+      "loss": 3.9376,
+      "step": 1327104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8262854248956506e-05,
+      "loss": 3.9386,
+      "step": 1327616
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8254468301445986e-05,
+      "loss": 3.9235,
+      "step": 1328128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8246082353935466e-05,
+      "loss": 3.9398,
+      "step": 1328640
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8237696406424946e-05,
+      "loss": 3.9265,
+      "step": 1329152
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8229310458914426e-05,
+      "loss": 3.9294,
+      "step": 1329664
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8220924511403906e-05,
+      "loss": 3.9176,
+      "step": 1330176
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8212538563893386e-05,
+      "loss": 3.9414,
+      "step": 1330688
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.82041689951866e-05,
+      "loss": 3.9259,
+      "step": 1331200
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.819578304767608e-05,
+      "loss": 3.9296,
+      "step": 1331712
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8187397100165558e-05,
+      "loss": 3.918,
+      "step": 1332224
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8179011152655038e-05,
+      "loss": 3.9265,
+      "step": 1332736
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8170641583948247e-05,
+      "loss": 3.936,
+      "step": 1333248
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8162255636437727e-05,
+      "loss": 3.9381,
+      "step": 1333760
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.815386968892721e-05,
+      "loss": 3.9337,
+      "step": 1334272
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.814548374141669e-05,
+      "loss": 3.9381,
+      "step": 1334784
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.813709779390617e-05,
+      "loss": 3.9431,
+      "step": 1335296
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.812871184639565e-05,
+      "loss": 3.9305,
+      "step": 1335808
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.812032589888513e-05,
+      "loss": 3.9437,
+      "step": 1336320
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.811193995137461e-05,
+      "loss": 3.9297,
+      "step": 1336832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.810357038266782e-05,
+      "loss": 3.9326,
+      "step": 1337344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8095200813961032e-05,
+      "loss": 3.9279,
+      "step": 1337856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8086814866450512e-05,
+      "loss": 3.9391,
+      "step": 1338368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8078428918939992e-05,
+      "loss": 3.9379,
+      "step": 1338880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8070042971429472e-05,
+      "loss": 3.9443,
+      "step": 1339392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.806167340272268e-05,
+      "loss": 3.9355,
+      "step": 1339904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8053287455212164e-05,
+      "loss": 3.9123,
+      "step": 1340416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8044901507701644e-05,
+      "loss": 3.9233,
+      "step": 1340928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8036515560191124e-05,
+      "loss": 3.9328,
+      "step": 1341440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8028129612680604e-05,
+      "loss": 3.9285,
+      "step": 1341952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8019743665170084e-05,
+      "loss": 3.9283,
+      "step": 1342464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8011357717659564e-05,
+      "loss": 3.9375,
+      "step": 1342976
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.8002971770149044e-05,
+      "loss": 3.9237,
+      "step": 1343488
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7994602201442257e-05,
+      "loss": 3.9294,
+      "step": 1344000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7986216253931736e-05,
+      "loss": 3.9307,
+      "step": 1344512
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7977846685224946e-05,
+      "loss": 3.9384,
+      "step": 1345024
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7969460737714426e-05,
+      "loss": 3.939,
+      "step": 1345536
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7961074790203905e-05,
+      "loss": 3.9457,
+      "step": 1346048
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7952688842693385e-05,
+      "loss": 3.9173,
+      "step": 1346560
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7944319273986598e-05,
+      "loss": 3.9338,
+      "step": 1347072
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7935933326476078e-05,
+      "loss": 3.9301,
+      "step": 1347584
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7927547378965558e-05,
+      "loss": 3.9344,
+      "step": 1348096
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7919161431455038e-05,
+      "loss": 3.9425,
+      "step": 1348608
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7910775483944518e-05,
+      "loss": 3.922,
+      "step": 1349120
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.790242229404146e-05,
+      "loss": 3.9303,
+      "step": 1349632
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.789403634653094e-05,
+      "loss": 3.9366,
+      "step": 1350144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.788565039902042e-05,
+      "loss": 3.9351,
+      "step": 1350656
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.78772644515099e-05,
+      "loss": 3.9208,
+      "step": 1351168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.786887850399938e-05,
+      "loss": 3.9321,
+      "step": 1351680
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.786049255648886e-05,
+      "loss": 3.9237,
+      "step": 1352192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.785210660897834e-05,
+      "loss": 3.924,
+      "step": 1352704
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7843720661467816e-05,
+      "loss": 3.9276,
+      "step": 1353216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7835334713957296e-05,
+      "loss": 3.9265,
+      "step": 1353728
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7826948766446776e-05,
+      "loss": 3.9243,
+      "step": 1354240
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7818562818936256e-05,
+      "loss": 3.9318,
+      "step": 1354752
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7810176871425735e-05,
+      "loss": 3.9249,
+      "step": 1355264
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7801790923915215e-05,
+      "loss": 3.9288,
+      "step": 1355776
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7793421355208428e-05,
+      "loss": 3.9448,
+      "step": 1356288
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7785035407697908e-05,
+      "loss": 3.9398,
+      "step": 1356800
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7776649460187388e-05,
+      "loss": 3.9348,
+      "step": 1357312
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7768263512676868e-05,
+      "loss": 3.9307,
+      "step": 1357824
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7759893943970077e-05,
+      "loss": 3.9304,
+      "step": 1358336
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7751507996459557e-05,
+      "loss": 3.9303,
+      "step": 1358848
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.774313842775277e-05,
+      "loss": 3.9197,
+      "step": 1359360
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.773475248024225e-05,
+      "loss": 3.9286,
+      "step": 1359872
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.772636653273173e-05,
+      "loss": 3.9276,
+      "step": 1360384
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.771798058522121e-05,
+      "loss": 3.932,
+      "step": 1360896
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.770959463771069e-05,
+      "loss": 3.9281,
+      "step": 1361408
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.770120869020017e-05,
+      "loss": 3.9333,
+      "step": 1361920
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.769282274268965e-05,
+      "loss": 3.9293,
+      "step": 1362432
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.768443679517913e-05,
+      "loss": 3.9277,
+      "step": 1362944
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.767606722647234e-05,
+      "loss": 3.9228,
+      "step": 1363456
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.766768127896182e-05,
+      "loss": 3.9417,
+      "step": 1363968
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.76592953314513e-05,
+      "loss": 3.9368,
+      "step": 1364480
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.765090938394078e-05,
+      "loss": 3.9322,
+      "step": 1364992
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.764253981523399e-05,
+      "loss": 3.9312,
+      "step": 1365504
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7634170246527203e-05,
+      "loss": 3.9362,
+      "step": 1366016
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7625784299016683e-05,
+      "loss": 3.9366,
+      "step": 1366528
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7617398351506163e-05,
+      "loss": 3.9405,
+      "step": 1367040
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7609012403995643e-05,
+      "loss": 3.9301,
+      "step": 1367552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7600642835288852e-05,
+      "loss": 3.9332,
+      "step": 1368064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7592256887778335e-05,
+      "loss": 3.9317,
+      "step": 1368576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7583870940267815e-05,
+      "loss": 3.9304,
+      "step": 1369088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7575484992757295e-05,
+      "loss": 3.9262,
+      "step": 1369600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7567099045246775e-05,
+      "loss": 3.9309,
+      "step": 1370112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7558713097736255e-05,
+      "loss": 3.9362,
+      "step": 1370624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7550327150225735e-05,
+      "loss": 3.9243,
+      "step": 1371136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7541941202715215e-05,
+      "loss": 3.9378,
+      "step": 1371648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7533571634008427e-05,
+      "loss": 3.9289,
+      "step": 1372160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7525185686497907e-05,
+      "loss": 3.9367,
+      "step": 1372672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7516799738987387e-05,
+      "loss": 3.9354,
+      "step": 1373184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.7508413791476867e-05,
+      "loss": 3.9251,
+      "step": 1373696
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 4.002753257751465,
+      "eval_runtime": 293.5279,
+      "eval_samples_per_second": 1300.016,
+      "eval_steps_per_second": 40.626,
+      "step": 1373747
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7500044222770076e-05,
+      "loss": 3.9226,
+      "step": 1374208
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7491658275259556e-05,
+      "loss": 3.9204,
+      "step": 1374720
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.748328870655277e-05,
+      "loss": 3.9423,
+      "step": 1375232
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.747490275904225e-05,
+      "loss": 3.9298,
+      "step": 1375744
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.746651681153173e-05,
+      "loss": 3.9374,
+      "step": 1376256
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.745813086402121e-05,
+      "loss": 3.9248,
+      "step": 1376768
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.744974491651069e-05,
+      "loss": 3.9278,
+      "step": 1377280
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.744135896900017e-05,
+      "loss": 3.9172,
+      "step": 1377792
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.743297302148965e-05,
+      "loss": 3.9312,
+      "step": 1378304
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7424587073979125e-05,
+      "loss": 3.9265,
+      "step": 1378816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.741621750527234e-05,
+      "loss": 3.9447,
+      "step": 1379328
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.740784793656555e-05,
+      "loss": 3.9358,
+      "step": 1379840
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.739947836785876e-05,
+      "loss": 3.9196,
+      "step": 1380352
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7391092420348243e-05,
+      "loss": 3.9217,
+      "step": 1380864
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7382706472837723e-05,
+      "loss": 3.9298,
+      "step": 1381376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7374320525327203e-05,
+      "loss": 3.9122,
+      "step": 1381888
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7365934577816682e-05,
+      "loss": 3.932,
+      "step": 1382400
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7357548630306162e-05,
+      "loss": 3.9296,
+      "step": 1382912
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7349162682795642e-05,
+      "loss": 3.924,
+      "step": 1383424
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7340776735285122e-05,
+      "loss": 3.95,
+      "step": 1383936
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.73323907877746e-05,
+      "loss": 3.9253,
+      "step": 1384448
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.732400484026408e-05,
+      "loss": 3.9322,
+      "step": 1384960
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.731561889275356e-05,
+      "loss": 3.9257,
+      "step": 1385472
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7307249324046768e-05,
+      "loss": 3.9412,
+      "step": 1385984
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7298863376536248e-05,
+      "loss": 3.9217,
+      "step": 1386496
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7290493807829464e-05,
+      "loss": 3.9285,
+      "step": 1387008
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.7282107860318944e-05,
+      "loss": 3.9185,
+      "step": 1387520
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.727372191280842e-05,
+      "loss": 3.9268,
+      "step": 1388032
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.72653359652979e-05,
+      "loss": 3.9266,
+      "step": 1388544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.725695001778738e-05,
+      "loss": 3.9236,
+      "step": 1389056
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.724856407027686e-05,
+      "loss": 3.9265,
+      "step": 1389568
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7240194501570073e-05,
+      "loss": 3.9325,
+      "step": 1390080
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.723182493286329e-05,
+      "loss": 3.9264,
+      "step": 1390592
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.722343898535277e-05,
+      "loss": 3.9405,
+      "step": 1391104
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7215069416645978e-05,
+      "loss": 3.9277,
+      "step": 1391616
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7206683469135457e-05,
+      "loss": 3.9307,
+      "step": 1392128
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7198297521624937e-05,
+      "loss": 3.9223,
+      "step": 1392640
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7189911574114417e-05,
+      "loss": 3.9259,
+      "step": 1393152
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7181525626603894e-05,
+      "loss": 3.918,
+      "step": 1393664
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7173139679093374e-05,
+      "loss": 3.927,
+      "step": 1394176
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7164753731582854e-05,
+      "loss": 3.92,
+      "step": 1394688
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7156367784072334e-05,
+      "loss": 3.9334,
+      "step": 1395200
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7147981836561814e-05,
+      "loss": 3.9188,
+      "step": 1395712
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7139595889051294e-05,
+      "loss": 3.9301,
+      "step": 1396224
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7131209941540774e-05,
+      "loss": 3.9332,
+      "step": 1396736
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7122823994030254e-05,
+      "loss": 3.9301,
+      "step": 1397248
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7114438046519737e-05,
+      "loss": 3.9293,
+      "step": 1397760
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7106052099009217e-05,
+      "loss": 3.9128,
+      "step": 1398272
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7097682530302426e-05,
+      "loss": 3.9187,
+      "step": 1398784
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7089296582791906e-05,
+      "loss": 3.935,
+      "step": 1399296
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7080910635281386e-05,
+      "loss": 3.9251,
+      "step": 1399808
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7072524687770866e-05,
+      "loss": 3.9253,
+      "step": 1400320
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7064138740260346e-05,
+      "loss": 3.9143,
+      "step": 1400832
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.705575279274983e-05,
+      "loss": 3.926,
+      "step": 1401344
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.704738322404304e-05,
+      "loss": 3.9033,
+      "step": 1401856
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7038997276532518e-05,
+      "loss": 3.94,
+      "step": 1402368
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7030611329021998e-05,
+      "loss": 3.9019,
+      "step": 1402880
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7022225381511478e-05,
+      "loss": 3.9344,
+      "step": 1403392
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7013839434000955e-05,
+      "loss": 3.9254,
+      "step": 1403904
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.7005453486490435e-05,
+      "loss": 3.9159,
+      "step": 1404416
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6997067538979915e-05,
+      "loss": 3.9322,
+      "step": 1404928
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6988681591469395e-05,
+      "loss": 3.9202,
+      "step": 1405440
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6980312022762604e-05,
+      "loss": 3.9171,
+      "step": 1405952
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6971926075252084e-05,
+      "loss": 3.9081,
+      "step": 1406464
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6963540127741567e-05,
+      "loss": 3.9325,
+      "step": 1406976
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6955154180231047e-05,
+      "loss": 3.915,
+      "step": 1407488
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6946768232720527e-05,
+      "loss": 3.9223,
+      "step": 1408000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6938398664013736e-05,
+      "loss": 3.9073,
+      "step": 1408512
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6930012716503216e-05,
+      "loss": 3.918,
+      "step": 1409024
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6921626768992696e-05,
+      "loss": 3.9261,
+      "step": 1409536
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6913240821482176e-05,
+      "loss": 3.9323,
+      "step": 1410048
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.690487125277539e-05,
+      "loss": 3.9214,
+      "step": 1410560
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.689648530526487e-05,
+      "loss": 3.9309,
+      "step": 1411072
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6888115736558077e-05,
+      "loss": 3.9325,
+      "step": 1411584
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6879729789047557e-05,
+      "loss": 3.9245,
+      "step": 1412096
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6871343841537037e-05,
+      "loss": 3.929,
+      "step": 1412608
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.686295789402652e-05,
+      "loss": 3.9242,
+      "step": 1413120
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6854571946516e-05,
+      "loss": 3.9236,
+      "step": 1413632
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.684620237780921e-05,
+      "loss": 3.9203,
+      "step": 1414144
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.683781643029869e-05,
+      "loss": 3.9292,
+      "step": 1414656
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.682943048278817e-05,
+      "loss": 3.9285,
+      "step": 1415168
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.682104453527765e-05,
+      "loss": 3.9363,
+      "step": 1415680
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.681265858776713e-05,
+      "loss": 3.9249,
+      "step": 1416192
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6804272640256613e-05,
+      "loss": 3.905,
+      "step": 1416704
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6795886692746093e-05,
+      "loss": 3.9126,
+      "step": 1417216
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6787500745235573e-05,
+      "loss": 3.9257,
+      "step": 1417728
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6779131176528782e-05,
+      "loss": 3.9207,
+      "step": 1418240
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6770745229018262e-05,
+      "loss": 3.919,
+      "step": 1418752
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 2.6762359281507742e-05,
+      "loss": 3.9272,
+      "step": 1419264
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6753989712800954e-05,
+      "loss": 3.9174,
+      "step": 1419776
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6745603765290434e-05,
+      "loss": 3.9152,
+      "step": 1420288
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6737217817779914e-05,
+      "loss": 3.9196,
+      "step": 1420800
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6728831870269394e-05,
+      "loss": 3.935,
+      "step": 1421312
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6720445922758874e-05,
+      "loss": 3.9291,
+      "step": 1421824
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6712076354052083e-05,
+      "loss": 3.9352,
+      "step": 1422336
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6703690406541567e-05,
+      "loss": 3.9128,
+      "step": 1422848
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6695304459031047e-05,
+      "loss": 3.9229,
+      "step": 1423360
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6686918511520526e-05,
+      "loss": 3.9217,
+      "step": 1423872
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6678532564010006e-05,
+      "loss": 3.9244,
+      "step": 1424384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6670146616499486e-05,
+      "loss": 3.9322,
+      "step": 1424896
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6661777047792695e-05,
+      "loss": 3.9169,
+      "step": 1425408
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6653391100282175e-05,
+      "loss": 3.9211,
+      "step": 1425920
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.664500515277166e-05,
+      "loss": 3.9263,
+      "step": 1426432
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.663661920526114e-05,
+      "loss": 3.927,
+      "step": 1426944
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.662823325775062e-05,
+      "loss": 3.91,
+      "step": 1427456
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6619847310240092e-05,
+      "loss": 3.9192,
+      "step": 1427968
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6611461362729572e-05,
+      "loss": 3.9198,
+      "step": 1428480
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6603075415219052e-05,
+      "loss": 3.9169,
+      "step": 1428992
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6594705846512264e-05,
+      "loss": 3.9159,
+      "step": 1429504
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6586319899001744e-05,
+      "loss": 3.916,
+      "step": 1430016
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6577933951491224e-05,
+      "loss": 3.9179,
+      "step": 1430528
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.656956438278444e-05,
+      "loss": 3.9245,
+      "step": 1431040
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6561178435273913e-05,
+      "loss": 3.9122,
+      "step": 1431552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6552792487763397e-05,
+      "loss": 3.9209,
+      "step": 1432064
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6544406540252877e-05,
+      "loss": 3.9342,
+      "step": 1432576
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6536020592742356e-05,
+      "loss": 3.9316,
+      "step": 1433088
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6527634645231836e-05,
+      "loss": 3.9256,
+      "step": 1433600
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6519265076525046e-05,
+      "loss": 3.9251,
+      "step": 1434112
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6510879129014525e-05,
+      "loss": 3.9199,
+      "step": 1434624
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6502493181504005e-05,
+      "loss": 3.919,
+      "step": 1435136
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6494107233993485e-05,
+      "loss": 3.9144,
+      "step": 1435648
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.648572128648297e-05,
+      "loss": 3.9155,
+      "step": 1436160
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.647733533897245e-05,
+      "loss": 3.9175,
+      "step": 1436672
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.646894939146193e-05,
+      "loss": 3.9279,
+      "step": 1437184
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.646056344395141e-05,
+      "loss": 3.9179,
+      "step": 1437696
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.645217749644089e-05,
+      "loss": 3.9232,
+      "step": 1438208
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6443807927734098e-05,
+      "loss": 3.9214,
+      "step": 1438720
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6435421980223578e-05,
+      "loss": 3.9165,
+      "step": 1439232
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.642705241151679e-05,
+      "loss": 3.9128,
+      "step": 1439744
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.641866646400627e-05,
+      "loss": 3.9352,
+      "step": 1440256
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.641028051649575e-05,
+      "loss": 3.9315,
+      "step": 1440768
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.640189456898523e-05,
+      "loss": 3.926,
+      "step": 1441280
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.639350862147471e-05,
+      "loss": 3.9198,
+      "step": 1441792
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6385139052767922e-05,
+      "loss": 3.9224,
+      "step": 1442304
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6376753105257402e-05,
+      "loss": 3.927,
+      "step": 1442816
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.636838353655061e-05,
+      "loss": 3.9357,
+      "step": 1443328
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.635999758904009e-05,
+      "loss": 3.9243,
+      "step": 1443840
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.635161164152957e-05,
+      "loss": 3.9196,
+      "step": 1444352
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.634322569401905e-05,
+      "loss": 3.9237,
+      "step": 1444864
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.633483974650853e-05,
+      "loss": 3.9227,
+      "step": 1445376
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6326453798998015e-05,
+      "loss": 3.9184,
+      "step": 1445888
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6318067851487495e-05,
+      "loss": 3.9234,
+      "step": 1446400
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6309681903976975e-05,
+      "loss": 3.9287,
+      "step": 1446912
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6301295956466448e-05,
+      "loss": 3.9148,
+      "step": 1447424
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6292910008955928e-05,
+      "loss": 3.9273,
+      "step": 1447936
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.6284524061445408e-05,
+      "loss": 3.9251,
+      "step": 1448448
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.627613811393489e-05,
+      "loss": 3.9216,
+      "step": 1448960
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.62677685452281e-05,
+      "loss": 3.9248,
+      "step": 1449472
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 2.625938259771758e-05,
+      "loss": 3.9148,
+      "step": 1449984
+    },
+    {
+      "epoch": 1.03,
+      "eval_loss": 3.998906373977661,
+      "eval_runtime": 419.5566,
+      "eval_samples_per_second": 909.51,
+      "eval_steps_per_second": 28.423,
+      "step": 1450067
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.625099665020706e-05,
+      "loss": 3.9119,
+      "step": 1450496
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.624261070269654e-05,
+      "loss": 3.9121,
+      "step": 1451008
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.623422475518602e-05,
+      "loss": 3.9354,
+      "step": 1451520
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.62258388076755e-05,
+      "loss": 3.9156,
+      "step": 1452032
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6217452860164983e-05,
+      "loss": 3.9319,
+      "step": 1452544
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6209066912654463e-05,
+      "loss": 3.9164,
+      "step": 1453056
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6200680965143943e-05,
+      "loss": 3.9194,
+      "step": 1453568
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6192295017633423e-05,
+      "loss": 3.9102,
+      "step": 1454080
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6183909070122903e-05,
+      "loss": 3.9241,
+      "step": 1454592
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6175523122612383e-05,
+      "loss": 3.915,
+      "step": 1455104
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6167137175101863e-05,
+      "loss": 3.9374,
+      "step": 1455616
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.615875122759134e-05,
+      "loss": 3.9256,
+      "step": 1456128
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6150381658884555e-05,
+      "loss": 3.9099,
+      "step": 1456640
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6141995711374035e-05,
+      "loss": 3.9147,
+      "step": 1457152
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6133609763863515e-05,
+      "loss": 3.9218,
+      "step": 1457664
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.612522381635299e-05,
+      "loss": 3.907,
+      "step": 1458176
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.611683786884247e-05,
+      "loss": 3.9206,
+      "step": 1458688
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6108451921331952e-05,
+      "loss": 3.9187,
+      "step": 1459200
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6100065973821432e-05,
+      "loss": 3.9149,
+      "step": 1459712
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.609168002631091e-05,
+      "loss": 3.9432,
+      "step": 1460224
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.608329407880039e-05,
+      "loss": 3.9188,
+      "step": 1460736
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.607490813128987e-05,
+      "loss": 3.9207,
+      "step": 1461248
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.606652218377935e-05,
+      "loss": 3.9164,
+      "step": 1461760
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.605813623626883e-05,
+      "loss": 3.9341,
+      "step": 1462272
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.604975028875831e-05,
+      "loss": 3.9116,
+      "step": 1462784
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6041380720051524e-05,
+      "loss": 3.9159,
+      "step": 1463296
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6032994772541004e-05,
+      "loss": 3.9135,
+      "step": 1463808
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6024608825030484e-05,
+      "loss": 3.9174,
+      "step": 1464320
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.6016222877519964e-05,
+      "loss": 3.9171,
+      "step": 1464832
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.6007853308813173e-05,
+      "loss": 3.9155,
+      "step": 1465344
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5999467361302653e-05,
+      "loss": 3.9207,
+      "step": 1465856
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5991081413792133e-05,
+      "loss": 3.9222,
+      "step": 1466368
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5982695466281616e-05,
+      "loss": 3.9203,
+      "step": 1466880
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5974325897574825e-05,
+      "loss": 3.9303,
+      "step": 1467392
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5965956328868034e-05,
+      "loss": 3.9202,
+      "step": 1467904
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5957570381357514e-05,
+      "loss": 3.9227,
+      "step": 1468416
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5949184433846994e-05,
+      "loss": 3.9114,
+      "step": 1468928
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5940798486336478e-05,
+      "loss": 3.9226,
+      "step": 1469440
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5932412538825958e-05,
+      "loss": 3.9078,
+      "step": 1469952
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5924042970119167e-05,
+      "loss": 3.9175,
+      "step": 1470464
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5915657022608647e-05,
+      "loss": 3.9125,
+      "step": 1470976
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5907271075098127e-05,
+      "loss": 3.9226,
+      "step": 1471488
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5898885127587607e-05,
+      "loss": 3.9087,
+      "step": 1472000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5890499180077086e-05,
+      "loss": 3.9194,
+      "step": 1472512
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.588211323256657e-05,
+      "loss": 3.927,
+      "step": 1473024
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.587372728505605e-05,
+      "loss": 3.9224,
+      "step": 1473536
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.586535771634926e-05,
+      "loss": 3.9249,
+      "step": 1474048
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.585697176883874e-05,
+      "loss": 3.9041,
+      "step": 1474560
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.584858582132822e-05,
+      "loss": 3.9087,
+      "step": 1475072
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.58401998738177e-05,
+      "loss": 3.9237,
+      "step": 1475584
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5831813926307175e-05,
+      "loss": 3.9172,
+      "step": 1476096
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5823427978796655e-05,
+      "loss": 3.9187,
+      "step": 1476608
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5815042031286135e-05,
+      "loss": 3.9056,
+      "step": 1477120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.580667246257935e-05,
+      "loss": 3.9201,
+      "step": 1477632
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5798286515068824e-05,
+      "loss": 3.8957,
+      "step": 1478144
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.578991694636204e-05,
+      "loss": 3.9232,
+      "step": 1478656
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5781530998851523e-05,
+      "loss": 3.9019,
+      "step": 1479168
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5773145051340997e-05,
+      "loss": 3.9173,
+      "step": 1479680
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5764759103830477e-05,
+      "loss": 3.9225,
+      "step": 1480192
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5756373156319957e-05,
+      "loss": 3.9068,
+      "step": 1480704
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5748003587613172e-05,
+      "loss": 3.9209,
+      "step": 1481216
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.573961764010265e-05,
+      "loss": 3.9134,
+      "step": 1481728
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.573123169259213e-05,
+      "loss": 3.9083,
+      "step": 1482240
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.572284574508161e-05,
+      "loss": 3.9025,
+      "step": 1482752
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.571445979757109e-05,
+      "loss": 3.9226,
+      "step": 1483264
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.570607385006057e-05,
+      "loss": 3.9077,
+      "step": 1483776
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.569768790255005e-05,
+      "loss": 3.9124,
+      "step": 1484288
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.568930195503953e-05,
+      "loss": 3.9017,
+      "step": 1484800
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.568093238633274e-05,
+      "loss": 3.9074,
+      "step": 1485312
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.567254643882222e-05,
+      "loss": 3.9173,
+      "step": 1485824
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.566417687011543e-05,
+      "loss": 3.9243,
+      "step": 1486336
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.565579092260491e-05,
+      "loss": 3.9108,
+      "step": 1486848
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.564740497509439e-05,
+      "loss": 3.9248,
+      "step": 1487360
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.563901902758387e-05,
+      "loss": 3.9248,
+      "step": 1487872
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5630649458877083e-05,
+      "loss": 3.9144,
+      "step": 1488384
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5622263511366563e-05,
+      "loss": 3.9195,
+      "step": 1488896
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5613877563856043e-05,
+      "loss": 3.9163,
+      "step": 1489408
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5605491616345522e-05,
+      "loss": 3.914,
+      "step": 1489920
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5597105668835002e-05,
+      "loss": 3.9128,
+      "step": 1490432
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5588719721324482e-05,
+      "loss": 3.9187,
+      "step": 1490944
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5580333773813962e-05,
+      "loss": 3.924,
+      "step": 1491456
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5571947826303442e-05,
+      "loss": 3.926,
+      "step": 1491968
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5563578257596655e-05,
+      "loss": 3.9173,
+      "step": 1492480
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5555192310086135e-05,
+      "loss": 3.896,
+      "step": 1492992
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5546806362575615e-05,
+      "loss": 3.9082,
+      "step": 1493504
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5538420415065095e-05,
+      "loss": 3.9128,
+      "step": 1494016
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5530034467554575e-05,
+      "loss": 3.9159,
+      "step": 1494528
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5521664898847787e-05,
+      "loss": 3.9078,
+      "step": 1495040
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5513311708944725e-05,
+      "loss": 3.9181,
+      "step": 1495552
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5504925761434205e-05,
+      "loss": 3.9084,
+      "step": 1496064
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5496539813923685e-05,
+      "loss": 3.9068,
+      "step": 1496576
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.548815386641317e-05,
+      "loss": 3.9167,
+      "step": 1497088
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.547976791890265e-05,
+      "loss": 3.9194,
+      "step": 1497600
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.547138197139213e-05,
+      "loss": 3.9169,
+      "step": 1498112
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.546299602388161e-05,
+      "loss": 3.9281,
+      "step": 1498624
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.545461007637109e-05,
+      "loss": 3.9064,
+      "step": 1499136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.544622412886057e-05,
+      "loss": 3.9114,
+      "step": 1499648
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.543783818135005e-05,
+      "loss": 3.9145,
+      "step": 1500160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5429452233839528e-05,
+      "loss": 3.9183,
+      "step": 1500672
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5421066286329008e-05,
+      "loss": 3.9232,
+      "step": 1501184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5412680338818485e-05,
+      "loss": 3.9079,
+      "step": 1501696
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.54043107701117e-05,
+      "loss": 3.914,
+      "step": 1502208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.539592482260118e-05,
+      "loss": 3.9163,
+      "step": 1502720
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.538753887509066e-05,
+      "loss": 3.9222,
+      "step": 1503232
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5379152927580134e-05,
+      "loss": 3.9024,
+      "step": 1503744
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5370799737677082e-05,
+      "loss": 3.9128,
+      "step": 1504256
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5362413790166562e-05,
+      "loss": 3.9123,
+      "step": 1504768
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5354027842656042e-05,
+      "loss": 3.9057,
+      "step": 1505280
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5345641895145522e-05,
+      "loss": 3.9076,
+      "step": 1505792
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5337255947635002e-05,
+      "loss": 3.9093,
+      "step": 1506304
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5328870000124482e-05,
+      "loss": 3.9139,
+      "step": 1506816
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.532048405261396e-05,
+      "loss": 3.9095,
+      "step": 1507328
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.531209810510344e-05,
+      "loss": 3.9049,
+      "step": 1507840
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5303728536396654e-05,
+      "loss": 3.912,
+      "step": 1508352
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5295342588886134e-05,
+      "loss": 3.9266,
+      "step": 1508864
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5286956641375607e-05,
+      "loss": 3.9246,
+      "step": 1509376
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5278570693865087e-05,
+      "loss": 3.9186,
+      "step": 1509888
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5270201125158303e-05,
+      "loss": 3.9137,
+      "step": 1510400
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.526181517764778e-05,
+      "loss": 3.9145,
+      "step": 1510912
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.525342923013726e-05,
+      "loss": 3.9147,
+      "step": 1511424
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.524504328262674e-05,
+      "loss": 3.9043,
+      "step": 1511936
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5236673713919956e-05,
+      "loss": 3.9074,
+      "step": 1512448
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5228287766409432e-05,
+      "loss": 3.9083,
+      "step": 1512960
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5219901818898912e-05,
+      "loss": 3.9209,
+      "step": 1513472
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5211515871388392e-05,
+      "loss": 3.9075,
+      "step": 1513984
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5203129923877872e-05,
+      "loss": 3.9188,
+      "step": 1514496
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5194743976367352e-05,
+      "loss": 3.91,
+      "step": 1515008
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5186358028856832e-05,
+      "loss": 3.9081,
+      "step": 1515520
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5177972081346312e-05,
+      "loss": 3.9061,
+      "step": 1516032
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5169586133835792e-05,
+      "loss": 3.9259,
+      "step": 1516544
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5161200186325272e-05,
+      "loss": 3.9264,
+      "step": 1517056
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5152814238814755e-05,
+      "loss": 3.9148,
+      "step": 1517568
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5144428291304235e-05,
+      "loss": 3.9153,
+      "step": 1518080
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5136058722597444e-05,
+      "loss": 3.9149,
+      "step": 1518592
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5127689153890653e-05,
+      "loss": 3.9159,
+      "step": 1519104
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5119303206380133e-05,
+      "loss": 3.9306,
+      "step": 1519616
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5110917258869617e-05,
+      "loss": 3.9124,
+      "step": 1520128
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5102531311359097e-05,
+      "loss": 3.9162,
+      "step": 1520640
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5094145363848577e-05,
+      "loss": 3.9104,
+      "step": 1521152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5085759416338057e-05,
+      "loss": 3.9169,
+      "step": 1521664
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5077389847631266e-05,
+      "loss": 3.9084,
+      "step": 1522176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5069003900120746e-05,
+      "loss": 3.9111,
+      "step": 1522688
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5060617952610226e-05,
+      "loss": 3.9195,
+      "step": 1523200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.505223200509971e-05,
+      "loss": 3.9097,
+      "step": 1523712
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.504384605758919e-05,
+      "loss": 3.9187,
+      "step": 1524224
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.503546011007867e-05,
+      "loss": 3.9221,
+      "step": 1524736
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5027074162568142e-05,
+      "loss": 3.9121,
+      "step": 1525248
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5018688215057622e-05,
+      "loss": 3.9165,
+      "step": 1525760
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 2.5010302267547102e-05,
+      "loss": 3.9065,
+      "step": 1526272
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 3.995372772216797,
+      "eval_runtime": 314.0386,
+      "eval_samples_per_second": 1215.109,
+      "eval_steps_per_second": 37.973,
+      "step": 1526387
+    }
+  ],
+  "logging_steps": 512,
+  "max_steps": 3052726,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 10,
+  "total_flos": 6.30189929319964e+17,
+  "trial_name": null,
+  "trial_params": null
+}