diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,36416 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 5198,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0,
+      "loss": 1.8036,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.392632117637794,
+      "learning_rate": 1.282051282051282e-07,
+      "loss": 1.7593,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.392632117637794,
+      "learning_rate": 1.282051282051282e-07,
+      "loss": 1.8204,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.649435863911009,
+      "learning_rate": 2.564102564102564e-07,
+      "loss": 1.8638,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.086129248867127,
+      "learning_rate": 3.846153846153847e-07,
+      "loss": 1.8184,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.24871721951395,
+      "learning_rate": 5.128205128205128e-07,
+      "loss": 1.771,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.887621321675178,
+      "learning_rate": 6.41025641025641e-07,
+      "loss": 1.8939,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.888589115502417,
+      "learning_rate": 7.692307692307694e-07,
+      "loss": 1.9989,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.408495719365018,
+      "learning_rate": 8.974358974358975e-07,
+      "loss": 1.6954,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.870216323377528,
+      "learning_rate": 1.0256410256410257e-06,
+      "loss": 1.8055,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.853155052345261,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 1.825,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.162987016864364,
+      "learning_rate": 1.282051282051282e-06,
+      "loss": 1.8265,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.625806785087965,
+      "learning_rate": 1.4102564102564104e-06,
+      "loss": 1.6569,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.587919126994953,
+      "learning_rate": 1.5384615384615387e-06,
+      "loss": 1.7908,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.189136690238893,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 1.7856,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.991483500639154,
+      "learning_rate": 1.794871794871795e-06,
+      "loss": 1.6377,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.819566975645913,
+      "learning_rate": 1.9230769230769234e-06,
+      "loss": 1.6548,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.254911819966893,
+      "learning_rate": 2.0512820512820513e-06,
+      "loss": 1.6108,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.271208291299759,
+      "learning_rate": 2.1794871794871797e-06,
+      "loss": 1.675,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3761289097990117,
+      "learning_rate": 2.307692307692308e-06,
+      "loss": 1.6194,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.392120597949798,
+      "learning_rate": 2.435897435897436e-06,
+      "loss": 1.6155,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3071051931899933,
+      "learning_rate": 2.564102564102564e-06,
+      "loss": 1.5842,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.464482995196971,
+      "learning_rate": 2.6923076923076923e-06,
+      "loss": 1.5303,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.356072527526895,
+      "learning_rate": 2.8205128205128207e-06,
+      "loss": 1.5632,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.117952884918975,
+      "learning_rate": 2.948717948717949e-06,
+      "loss": 1.5642,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9260794729972779,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 1.5416,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8923272038888044,
+      "learning_rate": 3.205128205128206e-06,
+      "loss": 1.5448,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9476334022014774,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.5299,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8511959784688226,
+      "learning_rate": 3.4615384615384617e-06,
+      "loss": 1.4343,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7782558852451371,
+      "learning_rate": 3.58974358974359e-06,
+      "loss": 1.5903,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.591936532062632,
+      "learning_rate": 3.7179487179487184e-06,
+      "loss": 1.4834,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.682663196566898,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 1.4287,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.507533033406422,
+      "learning_rate": 3.974358974358974e-06,
+      "loss": 1.5118,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9608394066978427,
+      "learning_rate": 4.102564102564103e-06,
+      "loss": 1.4604,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6890143330700709,
+      "learning_rate": 4.230769230769231e-06,
+      "loss": 1.4364,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6231151984068477,
+      "learning_rate": 4.358974358974359e-06,
+      "loss": 1.3834,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3258683133142453,
+      "learning_rate": 4.487179487179488e-06,
+      "loss": 1.3807,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2389338676237673,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 1.4197,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4364220982013363,
+      "learning_rate": 4.743589743589744e-06,
+      "loss": 1.5278,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5467532507383799,
+      "learning_rate": 4.871794871794872e-06,
+      "loss": 1.4191,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.548864549274519,
+      "learning_rate": 5e-06,
+      "loss": 1.5151,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1047128719493817,
+      "learning_rate": 5.128205128205128e-06,
+      "loss": 1.4207,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1778439290004632,
+      "learning_rate": 5.256410256410257e-06,
+      "loss": 1.4088,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5550935591156625,
+      "learning_rate": 5.384615384615385e-06,
+      "loss": 1.4093,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0911598201074248,
+      "learning_rate": 5.512820512820514e-06,
+      "loss": 1.4497,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.161670719374125,
+      "learning_rate": 5.641025641025641e-06,
+      "loss": 1.4124,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1265749693149578,
+      "learning_rate": 5.769230769230769e-06,
+      "loss": 1.4402,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0045875115311311,
+      "learning_rate": 5.897435897435898e-06,
+      "loss": 1.4442,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8938530512123973,
+      "learning_rate": 6.025641025641026e-06,
+      "loss": 1.4418,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9974034017410497,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 1.3333,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0866341261229504,
+      "learning_rate": 6.282051282051282e-06,
+      "loss": 1.4038,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.953202134020128,
+      "learning_rate": 6.410256410256412e-06,
+      "loss": 1.4377,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.118372896539707,
+      "learning_rate": 6.538461538461539e-06,
+      "loss": 1.473,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8979148735223598,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.3492,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1065040040075123,
+      "learning_rate": 6.794871794871796e-06,
+      "loss": 1.4365,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9924651655197421,
+      "learning_rate": 6.923076923076923e-06,
+      "loss": 1.343,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9936693935885011,
+      "learning_rate": 7.051282051282053e-06,
+      "loss": 1.3716,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.883728006242286,
+      "learning_rate": 7.17948717948718e-06,
+      "loss": 1.4478,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0845504788036984,
+      "learning_rate": 7.307692307692308e-06,
+      "loss": 1.3412,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8907007939289703,
+      "learning_rate": 7.435897435897437e-06,
+      "loss": 1.3586,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2540170813643672,
+      "learning_rate": 7.564102564102564e-06,
+      "loss": 1.3403,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2284057885919346,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 1.4152,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8644102022826152,
+      "learning_rate": 7.820512820512822e-06,
+      "loss": 1.3452,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9345386122706648,
+      "learning_rate": 7.948717948717949e-06,
+      "loss": 1.4404,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8094065650266457,
+      "learning_rate": 8.076923076923077e-06,
+      "loss": 1.3616,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8502959132384741,
+      "learning_rate": 8.205128205128205e-06,
+      "loss": 1.4445,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8920061575107928,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 1.2911,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7894563058946519,
+      "learning_rate": 8.461538461538462e-06,
+      "loss": 1.3326,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8618721142947509,
+      "learning_rate": 8.58974358974359e-06,
+      "loss": 1.2913,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8148074065058423,
+      "learning_rate": 8.717948717948719e-06,
+      "loss": 1.3225,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8114878210576331,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 1.2667,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8923562588995999,
+      "learning_rate": 8.974358974358976e-06,
+      "loss": 1.37,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7467571567018628,
+      "learning_rate": 9.102564102564104e-06,
+      "loss": 1.3062,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7736401079482755,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 1.3433,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7678834158456634,
+      "learning_rate": 9.358974358974359e-06,
+      "loss": 1.3524,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7450372382866174,
+      "learning_rate": 9.487179487179487e-06,
+      "loss": 1.298,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8565417171377238,
+      "learning_rate": 9.615384615384616e-06,
+      "loss": 1.2999,
+      "step": 77
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8741172119253061,
+      "learning_rate": 9.743589743589744e-06,
+      "loss": 1.3307,
+      "step": 78
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1109967546045887,
+      "learning_rate": 9.871794871794872e-06,
+      "loss": 1.3141,
+      "step": 79
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8673511577095718,
+      "learning_rate": 1e-05,
+      "loss": 1.3545,
+      "step": 80
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8113255846723173,
+      "learning_rate": 1.012820512820513e-05,
+      "loss": 1.2726,
+      "step": 81
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7879781005347615,
+      "learning_rate": 1.0256410256410256e-05,
+      "loss": 1.379,
+      "step": 82
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.061745230833992,
+      "learning_rate": 1.0384615384615386e-05,
+      "loss": 1.3396,
+      "step": 83
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8376250946438392,
+      "learning_rate": 1.0512820512820514e-05,
+      "loss": 1.2905,
+      "step": 84
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8210798111201055,
+      "learning_rate": 1.0641025641025643e-05,
+      "loss": 1.3507,
+      "step": 85
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7573730432356858,
+      "learning_rate": 1.076923076923077e-05,
+      "loss": 1.3165,
+      "step": 86
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.750980892422162,
+      "learning_rate": 1.0897435897435898e-05,
+      "loss": 1.2302,
+      "step": 87
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7357559217292094,
+      "learning_rate": 1.1025641025641028e-05,
+      "loss": 1.2958,
+      "step": 88
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7888931316316014,
+      "learning_rate": 1.1153846153846154e-05,
+      "loss": 1.3323,
+      "step": 89
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7679080409646744,
+      "learning_rate": 1.1282051282051283e-05,
+      "loss": 1.299,
+      "step": 90
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7440536119418517,
+      "learning_rate": 1.1410256410256411e-05,
+      "loss": 1.3096,
+      "step": 91
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.764797926799834,
+      "learning_rate": 1.1538461538461538e-05,
+      "loss": 1.3615,
+      "step": 92
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7147948825933413,
+      "learning_rate": 1.1666666666666668e-05,
+      "loss": 1.3026,
+      "step": 93
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7543525847373825,
+      "learning_rate": 1.1794871794871796e-05,
+      "loss": 1.3009,
+      "step": 94
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7551551362216752,
+      "learning_rate": 1.1923076923076925e-05,
+      "loss": 1.3153,
+      "step": 95
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7323640838524181,
+      "learning_rate": 1.2051282051282051e-05,
+      "loss": 1.2806,
+      "step": 96
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7819925139903149,
+      "learning_rate": 1.217948717948718e-05,
+      "loss": 1.3156,
+      "step": 97
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.979401149145299,
+      "learning_rate": 1.230769230769231e-05,
+      "loss": 1.3483,
+      "step": 98
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7918866821791015,
+      "learning_rate": 1.2435897435897436e-05,
+      "loss": 1.2346,
+      "step": 99
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8269555436429149,
+      "learning_rate": 1.2564102564102565e-05,
+      "loss": 1.2404,
+      "step": 100
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7413270678495784,
+      "learning_rate": 1.2692307692307693e-05,
+      "loss": 1.2769,
+      "step": 101
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7619207984405567,
+      "learning_rate": 1.2820512820512823e-05,
+      "loss": 1.2905,
+      "step": 102
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.066423653016647,
+      "learning_rate": 1.294871794871795e-05,
+      "loss": 1.3712,
+      "step": 103
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8040758666191158,
+      "learning_rate": 1.3076923076923078e-05,
+      "loss": 1.2491,
+      "step": 104
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7814647420321116,
+      "learning_rate": 1.3205128205128207e-05,
+      "loss": 1.2191,
+      "step": 105
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7268917337772135,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 1.2878,
+      "step": 106
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7013953519789007,
+      "learning_rate": 1.3461538461538463e-05,
+      "loss": 1.3138,
+      "step": 107
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.730636573872388,
+      "learning_rate": 1.3589743589743592e-05,
+      "loss": 1.3207,
+      "step": 108
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.778195268043436,
+      "learning_rate": 1.3717948717948718e-05,
+      "loss": 1.2947,
+      "step": 109
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7752347681518034,
+      "learning_rate": 1.3846153846153847e-05,
+      "loss": 1.2654,
+      "step": 110
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.867770740895625,
+      "learning_rate": 1.3974358974358975e-05,
+      "loss": 1.2203,
+      "step": 111
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7050419082141126,
+      "learning_rate": 1.4102564102564105e-05,
+      "loss": 1.2705,
+      "step": 112
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7790199122970742,
+      "learning_rate": 1.4230769230769232e-05,
+      "loss": 1.3134,
+      "step": 113
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8446288329052468,
+      "learning_rate": 1.435897435897436e-05,
+      "loss": 1.2047,
+      "step": 114
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.954799629865856,
+      "learning_rate": 1.4487179487179489e-05,
+      "loss": 1.184,
+      "step": 115
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7947305747567648,
+      "learning_rate": 1.4615384615384615e-05,
+      "loss": 1.2075,
+      "step": 116
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8979452346243084,
+      "learning_rate": 1.4743589743589745e-05,
+      "loss": 1.2588,
+      "step": 117
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7250684463145183,
+      "learning_rate": 1.4871794871794874e-05,
+      "loss": 1.2527,
+      "step": 118
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7297089323629333,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 1.1801,
+      "step": 119
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7610131686142391,
+      "learning_rate": 1.5128205128205129e-05,
+      "loss": 1.2615,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7240051490639147,
+      "learning_rate": 1.5256410256410257e-05,
+      "loss": 1.2349,
+      "step": 121
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7026448647486977,
+      "learning_rate": 1.5384615384615387e-05,
+      "loss": 1.2419,
+      "step": 122
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7862025032444973,
+      "learning_rate": 1.5512820512820516e-05,
+      "loss": 1.3191,
+      "step": 123
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7096327769711908,
+      "learning_rate": 1.5641025641025644e-05,
+      "loss": 1.3048,
+      "step": 124
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6855700129187143,
+      "learning_rate": 1.576923076923077e-05,
+      "loss": 1.2421,
+      "step": 125
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7854763720430596,
+      "learning_rate": 1.5897435897435897e-05,
+      "loss": 1.3193,
+      "step": 126
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7566311814034028,
+      "learning_rate": 1.602564102564103e-05,
+      "loss": 1.2201,
+      "step": 127
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7055593157105556,
+      "learning_rate": 1.6153846153846154e-05,
+      "loss": 1.2637,
+      "step": 128
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8053454545168159,
+      "learning_rate": 1.6282051282051282e-05,
+      "loss": 1.2671,
+      "step": 129
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7814231285688767,
+      "learning_rate": 1.641025641025641e-05,
+      "loss": 1.2604,
+      "step": 130
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.741329100230582,
+      "learning_rate": 1.653846153846154e-05,
+      "loss": 1.2826,
+      "step": 131
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7559199812641729,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 1.2083,
+      "step": 132
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8343400220465166,
+      "learning_rate": 1.6794871794871796e-05,
+      "loss": 1.3123,
+      "step": 133
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7169475575542021,
+      "learning_rate": 1.6923076923076924e-05,
+      "loss": 1.291,
+      "step": 134
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9865752767766852,
+      "learning_rate": 1.7051282051282053e-05,
+      "loss": 1.3612,
+      "step": 135
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6496584196878319,
+      "learning_rate": 1.717948717948718e-05,
+      "loss": 1.2257,
+      "step": 136
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7797368911371989,
+      "learning_rate": 1.730769230769231e-05,
+      "loss": 1.2192,
+      "step": 137
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6713736196984944,
+      "learning_rate": 1.7435897435897438e-05,
+      "loss": 1.3132,
+      "step": 138
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7908596487865275,
+      "learning_rate": 1.7564102564102566e-05,
+      "loss": 1.2689,
+      "step": 139
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7135178535757124,
+      "learning_rate": 1.7692307692307694e-05,
+      "loss": 1.2047,
+      "step": 140
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.730404013421838,
+      "learning_rate": 1.7820512820512823e-05,
+      "loss": 1.1742,
+      "step": 141
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7710372682313935,
+      "learning_rate": 1.794871794871795e-05,
+      "loss": 1.254,
+      "step": 142
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7745959333758601,
+      "learning_rate": 1.807692307692308e-05,
+      "loss": 1.2217,
+      "step": 143
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8444769237633165,
+      "learning_rate": 1.8205128205128208e-05,
+      "loss": 1.2832,
+      "step": 144
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7846918965773385,
+      "learning_rate": 1.8333333333333333e-05,
+      "loss": 1.2391,
+      "step": 145
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6858099131488958,
+      "learning_rate": 1.8461538461538465e-05,
+      "loss": 1.2028,
+      "step": 146
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.814145037178926,
+      "learning_rate": 1.8589743589743593e-05,
+      "loss": 1.2848,
+      "step": 147
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8754182460098853,
+      "learning_rate": 1.8717948717948718e-05,
+      "loss": 1.2909,
+      "step": 148
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7626279613890875,
+      "learning_rate": 1.8846153846153846e-05,
+      "loss": 1.224,
+      "step": 149
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6905978311186907,
+      "learning_rate": 1.8974358974358975e-05,
+      "loss": 1.1689,
+      "step": 150
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6714671524741169,
+      "learning_rate": 1.9102564102564106e-05,
+      "loss": 1.2957,
+      "step": 151
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.756793672372782,
+      "learning_rate": 1.923076923076923e-05,
+      "loss": 1.2202,
+      "step": 152
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6713305199897481,
+      "learning_rate": 1.935897435897436e-05,
+      "loss": 1.2585,
+      "step": 153
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.645599708830294,
+      "learning_rate": 1.9487179487179488e-05,
+      "loss": 1.2324,
+      "step": 154
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7424644475712061,
+      "learning_rate": 1.9615384615384617e-05,
+      "loss": 1.2565,
+      "step": 155
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6752387229057167,
+      "learning_rate": 1.9743589743589745e-05,
+      "loss": 1.2125,
+      "step": 156
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.700089565088466,
+      "learning_rate": 1.9871794871794873e-05,
+      "loss": 1.1938,
+      "step": 157
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6642726197482461,
+      "learning_rate": 2e-05,
+      "loss": 1.2301,
+      "step": 158
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7370502922633804,
+      "learning_rate": 1.9999998058827844e-05,
+      "loss": 1.2205,
+      "step": 159
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6859721928870679,
+      "learning_rate": 1.9999992235312136e-05,
+      "loss": 1.2699,
+      "step": 160
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8891526593144146,
+      "learning_rate": 1.9999982529455127e-05,
+      "loss": 1.1755,
+      "step": 161
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6892927129913222,
+      "learning_rate": 1.9999968941260596e-05,
+      "loss": 1.1803,
+      "step": 162
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8478642312087844,
+      "learning_rate": 1.9999951470733808e-05,
+      "loss": 1.2758,
+      "step": 163
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6506191756150801,
+      "learning_rate": 1.9999930117881548e-05,
+      "loss": 1.1569,
+      "step": 164
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7477363646606103,
+      "learning_rate": 1.9999904882712115e-05,
+      "loss": 1.2815,
+      "step": 165
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7127913521161391,
+      "learning_rate": 1.99998757652353e-05,
+      "loss": 1.1693,
+      "step": 166
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6805508212140985,
+      "learning_rate": 1.9999842765462403e-05,
+      "loss": 1.3187,
+      "step": 167
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6980814861621865,
+      "learning_rate": 1.999980588340624e-05,
+      "loss": 1.1271,
+      "step": 168
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7714102805044136,
+      "learning_rate": 1.9999765119081132e-05,
+      "loss": 1.2744,
+      "step": 169
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6793688585706885,
+      "learning_rate": 1.9999720472502902e-05,
+      "loss": 1.2821,
+      "step": 170
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.871486093739056,
+      "learning_rate": 1.9999671943688885e-05,
+      "loss": 1.1753,
+      "step": 171
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.690461789346987,
+      "learning_rate": 1.9999619532657915e-05,
+      "loss": 1.1723,
+      "step": 172
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.68052669459998,
+      "learning_rate": 1.9999563239430352e-05,
+      "loss": 1.1697,
+      "step": 173
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6915714483254263,
+      "learning_rate": 1.9999503064028043e-05,
+      "loss": 1.249,
+      "step": 174
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7136085398610883,
+      "learning_rate": 1.999943900647435e-05,
+      "loss": 1.1135,
+      "step": 175
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.710119816759835,
+      "learning_rate": 1.9999371066794146e-05,
+      "loss": 1.1963,
+      "step": 176
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6721909530421784,
+      "learning_rate": 1.9999299245013805e-05,
+      "loss": 1.1716,
+      "step": 177
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6959600888666073,
+      "learning_rate": 1.999922354116121e-05,
+      "loss": 1.2227,
+      "step": 178
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6743206504406951,
+      "learning_rate": 1.999914395526575e-05,
+      "loss": 1.2166,
+      "step": 179
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.853829057059371,
+      "learning_rate": 1.9999060487358333e-05,
+      "loss": 1.186,
+      "step": 180
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7406829643680521,
+      "learning_rate": 1.9998973137471352e-05,
+      "loss": 1.2484,
+      "step": 181
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7354042906863061,
+      "learning_rate": 1.9998881905638727e-05,
+      "loss": 1.2434,
+      "step": 182
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7133928076063463,
+      "learning_rate": 1.9998786791895874e-05,
+      "loss": 1.2068,
+      "step": 183
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7459052872397416,
+      "learning_rate": 1.999868779627972e-05,
+      "loss": 1.1803,
+      "step": 184
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7057199235248272,
+      "learning_rate": 1.9998584918828695e-05,
+      "loss": 1.1224,
+      "step": 185
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8213884374103616,
+      "learning_rate": 1.9998478159582747e-05,
+      "loss": 1.2457,
+      "step": 186
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6282740338281271,
+      "learning_rate": 1.999836751858332e-05,
+      "loss": 1.1707,
+      "step": 187
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6978415192078632,
+      "learning_rate": 1.9998252995873367e-05,
+      "loss": 1.1529,
+      "step": 188
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7369911706210684,
+      "learning_rate": 1.999813459149735e-05,
+      "loss": 1.2622,
+      "step": 189
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7786460576104434,
+      "learning_rate": 1.9998012305501243e-05,
+      "loss": 1.1775,
+      "step": 190
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7659256067025416,
+      "learning_rate": 1.999788613793251e-05,
+      "loss": 1.2398,
+      "step": 191
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7710427786758622,
+      "learning_rate": 1.999775608884015e-05,
+      "loss": 1.1267,
+      "step": 192
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7437023558722431,
+      "learning_rate": 1.9997622158274635e-05,
+      "loss": 1.1303,
+      "step": 193
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7315790747570348,
+      "learning_rate": 1.9997484346287973e-05,
+      "loss": 1.1978,
+      "step": 194
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7246964509265909,
+      "learning_rate": 1.9997342652933668e-05,
+      "loss": 1.2642,
+      "step": 195
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7143988340935427,
+      "learning_rate": 1.9997197078266723e-05,
+      "loss": 1.2175,
+      "step": 196
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8072803092892678,
+      "learning_rate": 1.999704762234366e-05,
+      "loss": 1.2521,
+      "step": 197
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6941514258036726,
+      "learning_rate": 1.99968942852225e-05,
+      "loss": 1.1501,
+      "step": 198
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7016824795423219,
+      "learning_rate": 1.9996737066962778e-05,
+      "loss": 1.2968,
+      "step": 199
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7300353960156142,
+      "learning_rate": 1.9996575967625525e-05,
+      "loss": 1.224,
+      "step": 200
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6821488900483668,
+      "learning_rate": 1.999641098727329e-05,
+      "loss": 1.2386,
+      "step": 201
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7220656798621135,
+      "learning_rate": 1.999624212597013e-05,
+      "loss": 1.1949,
+      "step": 202
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6869106645201545,
+      "learning_rate": 1.9996069383781587e-05,
+      "loss": 1.2356,
+      "step": 203
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7604053438822989,
+      "learning_rate": 1.9995892760774738e-05,
+      "loss": 1.2035,
+      "step": 204
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7204540634075955,
+      "learning_rate": 1.9995712257018153e-05,
+      "loss": 1.1339,
+      "step": 205
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7155233124955801,
+      "learning_rate": 1.9995527872581903e-05,
+      "loss": 1.2004,
+      "step": 206
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7753384952147434,
+      "learning_rate": 1.9995339607537578e-05,
+      "loss": 1.1678,
+      "step": 207
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7215213442297274,
+      "learning_rate": 1.9995147461958267e-05,
+      "loss": 1.2694,
+      "step": 208
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7705937954195716,
+      "learning_rate": 1.999495143591857e-05,
+      "loss": 1.1617,
+      "step": 209
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7789664823939354,
+      "learning_rate": 1.999475152949459e-05,
+      "loss": 1.2484,
+      "step": 210
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7729181587807843,
+      "learning_rate": 1.9994547742763935e-05,
+      "loss": 1.1477,
+      "step": 211
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7647157078359609,
+      "learning_rate": 1.9994340075805724e-05,
+      "loss": 1.2527,
+      "step": 212
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6883587899866555,
+      "learning_rate": 1.9994128528700583e-05,
+      "loss": 1.1757,
+      "step": 213
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7662769816208213,
+      "learning_rate": 1.9993913101530635e-05,
+      "loss": 1.17,
+      "step": 214
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7081457551068928,
+      "learning_rate": 1.9993693794379525e-05,
+      "loss": 1.2499,
+      "step": 215
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7286388328851662,
+      "learning_rate": 1.9993470607332387e-05,
+      "loss": 1.1976,
+      "step": 216
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6492713363455858,
+      "learning_rate": 1.999324354047588e-05,
+      "loss": 1.2051,
+      "step": 217
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7761902031336291,
+      "learning_rate": 1.9993012593898146e-05,
+      "loss": 1.1805,
+      "step": 218
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6282895307083679,
+      "learning_rate": 1.9992777767688857e-05,
+      "loss": 1.1892,
+      "step": 219
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8168633767724819,
+      "learning_rate": 1.9992539061939175e-05,
+      "loss": 1.1434,
+      "step": 220
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7145128931165112,
+      "learning_rate": 1.999229647674178e-05,
+      "loss": 1.1517,
+      "step": 221
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6560225533529509,
+      "learning_rate": 1.9992050012190845e-05,
+      "loss": 1.1882,
+      "step": 222
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6613060272763377,
+      "learning_rate": 1.9991799668382058e-05,
+      "loss": 1.234,
+      "step": 223
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7161225087604862,
+      "learning_rate": 1.9991545445412614e-05,
+      "loss": 1.2162,
+      "step": 224
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7572782737933926,
+      "learning_rate": 1.9991287343381208e-05,
+      "loss": 1.2009,
+      "step": 225
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6582667110044204,
+      "learning_rate": 1.9991025362388044e-05,
+      "loss": 1.163,
+      "step": 226
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7115419239856433,
+      "learning_rate": 1.9990759502534835e-05,
+      "loss": 1.1224,
+      "step": 227
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7079904093116176,
+      "learning_rate": 1.9990489763924796e-05,
+      "loss": 1.1836,
+      "step": 228
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7235245676856912,
+      "learning_rate": 1.9990216146662648e-05,
+      "loss": 1.2078,
+      "step": 229
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.821798044591755,
+      "learning_rate": 1.9989938650854618e-05,
+      "loss": 1.173,
+      "step": 230
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6677184769235851,
+      "learning_rate": 1.998965727660844e-05,
+      "loss": 1.162,
+      "step": 231
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6365796693976542,
+      "learning_rate": 1.9989372024033352e-05,
+      "loss": 1.2238,
+      "step": 232
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6858128693130722,
+      "learning_rate": 1.99890828932401e-05,
+      "loss": 1.2118,
+      "step": 233
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6580103080140122,
+      "learning_rate": 1.9988789884340938e-05,
+      "loss": 1.2281,
+      "step": 234
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.681544445814685,
+      "learning_rate": 1.9988492997449615e-05,
+      "loss": 1.2181,
+      "step": 235
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.748514455616942,
+      "learning_rate": 1.9988192232681398e-05,
+      "loss": 1.1319,
+      "step": 236
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7715745988739108,
+      "learning_rate": 1.9987887590153055e-05,
+      "loss": 1.2235,
+      "step": 237
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6498407864340681,
+      "learning_rate": 1.9987579069982856e-05,
+      "loss": 1.2255,
+      "step": 238
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7099212688354762,
+      "learning_rate": 1.9987266672290577e-05,
+      "loss": 1.1592,
+      "step": 239
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8178535035778582,
+      "learning_rate": 1.9986950397197503e-05,
+      "loss": 1.1446,
+      "step": 240
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7702385039971258,
+      "learning_rate": 1.9986630244826425e-05,
+      "loss": 1.2681,
+      "step": 241
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.692742270606639,
+      "learning_rate": 1.998630621530164e-05,
+      "loss": 1.1414,
+      "step": 242
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.701136932305704,
+      "learning_rate": 1.998597830874894e-05,
+      "loss": 1.1112,
+      "step": 243
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.67653977951436,
+      "learning_rate": 1.9985646525295634e-05,
+      "loss": 1.1248,
+      "step": 244
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.723032834790449,
+      "learning_rate": 1.998531086507053e-05,
+      "loss": 1.0797,
+      "step": 245
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7406615138724238,
+      "learning_rate": 1.9984971328203945e-05,
+      "loss": 1.1958,
+      "step": 246
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6813067565499344,
+      "learning_rate": 1.9984627914827698e-05,
+      "loss": 1.0928,
+      "step": 247
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7576433618035561,
+      "learning_rate": 1.9984280625075115e-05,
+      "loss": 1.1886,
+      "step": 248
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6563244564114679,
+      "learning_rate": 1.9983929459081022e-05,
+      "loss": 1.2374,
+      "step": 249
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7605876030332414,
+      "learning_rate": 1.998357441698176e-05,
+      "loss": 1.1909,
+      "step": 250
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.706008998457151,
+      "learning_rate": 1.998321549891516e-05,
+      "loss": 1.1677,
+      "step": 251
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7207914177557863,
+      "learning_rate": 1.9982852705020572e-05,
+      "loss": 1.164,
+      "step": 252
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7564655675583722,
+      "learning_rate": 1.9982486035438848e-05,
+      "loss": 1.0911,
+      "step": 253
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6573805482963057,
+      "learning_rate": 1.9982115490312334e-05,
+      "loss": 1.2018,
+      "step": 254
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7900249260828748,
+      "learning_rate": 1.9981741069784894e-05,
+      "loss": 1.1039,
+      "step": 255
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7237434624411251,
+      "learning_rate": 1.9981362774001886e-05,
+      "loss": 1.2234,
+      "step": 256
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7582247503260461,
+      "learning_rate": 1.9980980603110185e-05,
+      "loss": 1.2018,
+      "step": 257
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6917318666968727,
+      "learning_rate": 1.9980594557258158e-05,
+      "loss": 1.2086,
+      "step": 258
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6679995301812832,
+      "learning_rate": 1.9980204636595682e-05,
+      "loss": 1.1147,
+      "step": 259
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7545934478219363,
+      "learning_rate": 1.9979810841274135e-05,
+      "loss": 1.1741,
+      "step": 260
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7284143306371138,
+      "learning_rate": 1.9979413171446403e-05,
+      "loss": 1.1599,
+      "step": 261
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7018468020059422,
+      "learning_rate": 1.9979011627266884e-05,
+      "loss": 1.1636,
+      "step": 262
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7419960650668772,
+      "learning_rate": 1.997860620889146e-05,
+      "loss": 1.2019,
+      "step": 263
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6608260403716455,
+      "learning_rate": 1.997819691647753e-05,
+      "loss": 1.159,
+      "step": 264
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6060815180215597,
+      "learning_rate": 1.9977783750184e-05,
+      "loss": 1.1296,
+      "step": 265
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.728851505660618,
+      "learning_rate": 1.9977366710171274e-05,
+      "loss": 1.1426,
+      "step": 266
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7899422965567231,
+      "learning_rate": 1.9976945796601258e-05,
+      "loss": 1.2188,
+      "step": 267
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6872852959757944,
+      "learning_rate": 1.9976521009637366e-05,
+      "loss": 1.2087,
+      "step": 268
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.701389638177654,
+      "learning_rate": 1.997609234944452e-05,
+      "loss": 1.1259,
+      "step": 269
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7056200574811612,
+      "learning_rate": 1.9975659816189137e-05,
+      "loss": 1.1941,
+      "step": 270
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.703995853960454,
+      "learning_rate": 1.997522341003914e-05,
+      "loss": 1.2438,
+      "step": 271
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7533987443751103,
+      "learning_rate": 1.9974783131163957e-05,
+      "loss": 1.2301,
+      "step": 272
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7062555116854766,
+      "learning_rate": 1.9974338979734523e-05,
+      "loss": 1.1454,
+      "step": 273
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7937607074344634,
+      "learning_rate": 1.997389095592327e-05,
+      "loss": 1.1583,
+      "step": 274
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7636593317765026,
+      "learning_rate": 1.9973439059904133e-05,
+      "loss": 1.1661,
+      "step": 275
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6618638766329515,
+      "learning_rate": 1.9972983291852565e-05,
+      "loss": 1.0809,
+      "step": 276
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8175282418981842,
+      "learning_rate": 1.9972523651945496e-05,
+      "loss": 1.246,
+      "step": 277
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.724663138598895,
+      "learning_rate": 1.9972060140361384e-05,
+      "loss": 1.1664,
+      "step": 278
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5612179145039993,
+      "learning_rate": 1.997159275728018e-05,
+      "loss": 1.0908,
+      "step": 279
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6698761783211568,
+      "learning_rate": 1.9971121502883332e-05,
+      "loss": 1.1544,
+      "step": 280
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6315557091036534,
+      "learning_rate": 1.9970646377353802e-05,
+      "loss": 1.0829,
+      "step": 281
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7053753256021735,
+      "learning_rate": 1.997016738087605e-05,
+      "loss": 1.1466,
+      "step": 282
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6829457209058323,
+      "learning_rate": 1.9969684513636035e-05,
+      "loss": 1.1674,
+      "step": 283
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6969464304927909,
+      "learning_rate": 1.9969197775821227e-05,
+      "loss": 1.0485,
+      "step": 284
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7051885716241267,
+      "learning_rate": 1.9968707167620593e-05,
+      "loss": 1.1182,
+      "step": 285
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6522539962627258,
+      "learning_rate": 1.9968212689224603e-05,
+      "loss": 1.1867,
+      "step": 286
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6765534697739571,
+      "learning_rate": 1.996771434082523e-05,
+      "loss": 1.1091,
+      "step": 287
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5706749389518032,
+      "learning_rate": 1.9967212122615958e-05,
+      "loss": 1.1864,
+      "step": 288
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7268896485102951,
+      "learning_rate": 1.9966706034791752e-05,
+      "loss": 1.2059,
+      "step": 289
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7744392527768145,
+      "learning_rate": 1.9966196077549106e-05,
+      "loss": 1.2156,
+      "step": 290
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7266793688444242,
+      "learning_rate": 1.996568225108599e-05,
+      "loss": 1.1332,
+      "step": 291
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7471484211913241,
+      "learning_rate": 1.99651645556019e-05,
+      "loss": 1.1467,
+      "step": 292
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6985378242044223,
+      "learning_rate": 1.9964642991297817e-05,
+      "loss": 1.1083,
+      "step": 293
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7098112274475468,
+      "learning_rate": 1.996411755837623e-05,
+      "loss": 1.1264,
+      "step": 294
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7951394116878303,
+      "learning_rate": 1.9963588257041137e-05,
+      "loss": 1.1279,
+      "step": 295
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7507906865111684,
+      "learning_rate": 1.996305508749802e-05,
+      "loss": 1.1411,
+      "step": 296
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.792229090269127,
+      "learning_rate": 1.9962518049953887e-05,
+      "loss": 1.1725,
+      "step": 297
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.734081504424788,
+      "learning_rate": 1.9961977144617225e-05,
+      "loss": 1.1357,
+      "step": 298
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.771196333869013,
+      "learning_rate": 1.996143237169803e-05,
+      "loss": 1.1812,
+      "step": 299
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.732511481756585,
+      "learning_rate": 1.996088373140781e-05,
+      "loss": 1.1909,
+      "step": 300
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6500774104980205,
+      "learning_rate": 1.9960331223959564e-05,
+      "loss": 1.2191,
+      "step": 301
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7158503171274536,
+      "learning_rate": 1.995977484956779e-05,
+      "loss": 1.1335,
+      "step": 302
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7409931442686917,
+      "learning_rate": 1.9959214608448495e-05,
+      "loss": 1.1699,
+      "step": 303
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7196721561745003,
+      "learning_rate": 1.9958650500819183e-05,
+      "loss": 1.1425,
+      "step": 304
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7112436517324864,
+      "learning_rate": 1.995808252689886e-05,
+      "loss": 1.1202,
+      "step": 305
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6555594710476855,
+      "learning_rate": 1.9957510686908034e-05,
+      "loss": 1.1567,
+      "step": 306
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8021677920925815,
+      "learning_rate": 1.9956934981068713e-05,
+      "loss": 1.1919,
+      "step": 307
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6800557034015636,
+      "learning_rate": 1.9956355409604402e-05,
+      "loss": 1.2415,
+      "step": 308
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7174667486093853,
+      "learning_rate": 1.9955771972740118e-05,
+      "loss": 1.2272,
+      "step": 309
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7101413206443971,
+      "learning_rate": 1.9955184670702363e-05,
+      "loss": 1.2037,
+      "step": 310
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7448893972574331,
+      "learning_rate": 1.995459350371915e-05,
+      "loss": 1.143,
+      "step": 311
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6947034321765927,
+      "learning_rate": 1.9953998472019996e-05,
+      "loss": 1.2297,
+      "step": 312
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6296893289532655,
+      "learning_rate": 1.995339957583591e-05,
+      "loss": 1.1611,
+      "step": 313
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6555284277189896,
+      "learning_rate": 1.9952796815399403e-05,
+      "loss": 1.1549,
+      "step": 314
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6431787381675065,
+      "learning_rate": 1.9952190190944484e-05,
+      "loss": 1.1393,
+      "step": 315
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6760904280328096,
+      "learning_rate": 1.9951579702706668e-05,
+      "loss": 1.1833,
+      "step": 316
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7062710262140488,
+      "learning_rate": 1.9950965350922975e-05,
+      "loss": 1.1336,
+      "step": 317
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6850089405731257,
+      "learning_rate": 1.9950347135831907e-05,
+      "loss": 1.1676,
+      "step": 318
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7639235879994501,
+      "learning_rate": 1.994972505767348e-05,
+      "loss": 1.1372,
+      "step": 319
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6444207366947412,
+      "learning_rate": 1.994909911668921e-05,
+      "loss": 1.1096,
+      "step": 320
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6855784572104565,
+      "learning_rate": 1.99484693131221e-05,
+      "loss": 1.1664,
+      "step": 321
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7603643029953564,
+      "learning_rate": 1.994783564721667e-05,
+      "loss": 1.1519,
+      "step": 322
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6004679404820837,
+      "learning_rate": 1.9947198119218924e-05,
+      "loss": 1.2156,
+      "step": 323
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7774863273344531,
+      "learning_rate": 1.994655672937638e-05,
+      "loss": 1.1842,
+      "step": 324
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7065190607325077,
+      "learning_rate": 1.9945911477938044e-05,
+      "loss": 1.1509,
+      "step": 325
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6737084222655518,
+      "learning_rate": 1.994526236515442e-05,
+      "loss": 1.1566,
+      "step": 326
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5645822458836974,
+      "learning_rate": 1.994460939127753e-05,
+      "loss": 1.1941,
+      "step": 327
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7611269006969976,
+      "learning_rate": 1.9943952556560863e-05,
+      "loss": 1.1905,
+      "step": 328
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6943557609962231,
+      "learning_rate": 1.9943291861259433e-05,
+      "loss": 1.1882,
+      "step": 329
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6661499841892283,
+      "learning_rate": 1.9942627305629747e-05,
+      "loss": 1.2821,
+      "step": 330
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8150571073053675,
+      "learning_rate": 1.9941958889929808e-05,
+      "loss": 1.1707,
+      "step": 331
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6836125058070764,
+      "learning_rate": 1.9941286614419113e-05,
+      "loss": 1.1392,
+      "step": 332
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7533726062145043,
+      "learning_rate": 1.994061047935867e-05,
+      "loss": 1.0944,
+      "step": 333
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7249258778760096,
+      "learning_rate": 1.9939930485010968e-05,
+      "loss": 1.2095,
+      "step": 334
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6285649640268047,
+      "learning_rate": 1.9939246631640014e-05,
+      "loss": 1.1486,
+      "step": 335
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7434753843446424,
+      "learning_rate": 1.99385589195113e-05,
+      "loss": 1.1519,
+      "step": 336
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6621478269719245,
+      "learning_rate": 1.9937867348891815e-05,
+      "loss": 1.2002,
+      "step": 337
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7281341117541069,
+      "learning_rate": 1.9937171920050057e-05,
+      "loss": 1.0711,
+      "step": 338
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6966974305985796,
+      "learning_rate": 1.9936472633256012e-05,
+      "loss": 1.1698,
+      "step": 339
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6630907472453598,
+      "learning_rate": 1.9935769488781167e-05,
+      "loss": 1.202,
+      "step": 340
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.693595116555087,
+      "learning_rate": 1.993506248689851e-05,
+      "loss": 1.1941,
+      "step": 341
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.75849887113895,
+      "learning_rate": 1.993435162788252e-05,
+      "loss": 1.1681,
+      "step": 342
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6952617245860016,
+      "learning_rate": 1.993363691200918e-05,
+      "loss": 1.1432,
+      "step": 343
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7578059897854379,
+      "learning_rate": 1.9932918339555965e-05,
+      "loss": 1.1266,
+      "step": 344
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7104511666315011,
+      "learning_rate": 1.9932195910801848e-05,
+      "loss": 1.1624,
+      "step": 345
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7608252152187127,
+      "learning_rate": 1.9931469626027305e-05,
+      "loss": 1.1174,
+      "step": 346
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7211042462397494,
+      "learning_rate": 1.9930739485514304e-05,
+      "loss": 1.1448,
+      "step": 347
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7536185330260695,
+      "learning_rate": 1.9930005489546308e-05,
+      "loss": 1.0814,
+      "step": 348
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6630763971536932,
+      "learning_rate": 1.9929267638408277e-05,
+      "loss": 1.1646,
+      "step": 349
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.701009246870605,
+      "learning_rate": 1.9928525932386678e-05,
+      "loss": 1.1683,
+      "step": 350
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5932931612721024,
+      "learning_rate": 1.9927780371769463e-05,
+      "loss": 1.1653,
+      "step": 351
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7836580804086299,
+      "learning_rate": 1.9927030956846083e-05,
+      "loss": 1.248,
+      "step": 352
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6026764952563511,
+      "learning_rate": 1.992627768790749e-05,
+      "loss": 1.1768,
+      "step": 353
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6377169470282621,
+      "learning_rate": 1.9925520565246125e-05,
+      "loss": 1.2096,
+      "step": 354
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7355299501184754,
+      "learning_rate": 1.9924759589155932e-05,
+      "loss": 1.2345,
+      "step": 355
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6946051862832243,
+      "learning_rate": 1.9923994759932344e-05,
+      "loss": 1.1459,
+      "step": 356
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5831649546301736,
+      "learning_rate": 1.9923226077872296e-05,
+      "loss": 1.208,
+      "step": 357
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7798892487673164,
+      "learning_rate": 1.9922453543274223e-05,
+      "loss": 1.1807,
+      "step": 358
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7504500740879361,
+      "learning_rate": 1.9921677156438044e-05,
+      "loss": 1.1556,
+      "step": 359
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7527702592154902,
+      "learning_rate": 1.9920896917665178e-05,
+      "loss": 1.1118,
+      "step": 360
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6938216611884355,
+      "learning_rate": 1.992011282725854e-05,
+      "loss": 1.1277,
+      "step": 361
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6460077759549928,
+      "learning_rate": 1.9919324885522548e-05,
+      "loss": 1.1871,
+      "step": 362
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6453992675502472,
+      "learning_rate": 1.99185330927631e-05,
+      "loss": 1.1196,
+      "step": 363
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7043000661104767,
+      "learning_rate": 1.99177374492876e-05,
+      "loss": 1.1453,
+      "step": 364
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8115823882874854,
+      "learning_rate": 1.991693795540494e-05,
+      "loss": 1.217,
+      "step": 365
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6186990740488456,
+      "learning_rate": 1.9916134611425522e-05,
+      "loss": 1.1633,
+      "step": 366
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7613499206122579,
+      "learning_rate": 1.9915327417661226e-05,
+      "loss": 1.1942,
+      "step": 367
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6308340901722499,
+      "learning_rate": 1.991451637442543e-05,
+      "loss": 1.1,
+      "step": 368
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7195062828947171,
+      "learning_rate": 1.9913701482033008e-05,
+      "loss": 1.1957,
+      "step": 369
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7079424809072743,
+      "learning_rate": 1.9912882740800336e-05,
+      "loss": 1.1006,
+      "step": 370
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6050198992458018,
+      "learning_rate": 1.9912060151045273e-05,
+      "loss": 1.1743,
+      "step": 371
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.632424300044297,
+      "learning_rate": 1.9911233713087172e-05,
+      "loss": 1.1339,
+      "step": 372
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7201958818522626,
+      "learning_rate": 1.9910403427246895e-05,
+      "loss": 1.22,
+      "step": 373
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6581886998101526,
+      "learning_rate": 1.990956929384678e-05,
+      "loss": 1.1865,
+      "step": 374
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.654239325594963,
+      "learning_rate": 1.990873131321067e-05,
+      "loss": 1.0664,
+      "step": 375
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7812675071521543,
+      "learning_rate": 1.9907889485663897e-05,
+      "loss": 1.1519,
+      "step": 376
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6380208462073286,
+      "learning_rate": 1.9907043811533283e-05,
+      "loss": 1.1168,
+      "step": 377
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6921321664132063,
+      "learning_rate": 1.9906194291147155e-05,
+      "loss": 1.0933,
+      "step": 378
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7691984572504157,
+      "learning_rate": 1.9905340924835322e-05,
+      "loss": 1.1135,
+      "step": 379
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5965775105632853,
+      "learning_rate": 1.9904483712929094e-05,
+      "loss": 1.1964,
+      "step": 380
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7528672025766631,
+      "learning_rate": 1.9903622655761267e-05,
+      "loss": 1.1317,
+      "step": 381
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6981650019059337,
+      "learning_rate": 1.990275775366613e-05,
+      "loss": 1.1241,
+      "step": 382
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7756789519518531,
+      "learning_rate": 1.9901889006979473e-05,
+      "loss": 1.1835,
+      "step": 383
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7291827561569454,
+      "learning_rate": 1.990101641603857e-05,
+      "loss": 1.1093,
+      "step": 384
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7200713107159433,
+      "learning_rate": 1.9900139981182193e-05,
+      "loss": 1.1945,
+      "step": 385
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7339200853057544,
+      "learning_rate": 1.9899259702750604e-05,
+      "loss": 1.1102,
+      "step": 386
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6625210174674538,
+      "learning_rate": 1.9898375581085555e-05,
+      "loss": 1.1392,
+      "step": 387
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7337238543303566,
+      "learning_rate": 1.9897487616530296e-05,
+      "loss": 1.0106,
+      "step": 388
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7591475058523958,
+      "learning_rate": 1.9896595809429565e-05,
+      "loss": 1.1512,
+      "step": 389
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6822435887606652,
+      "learning_rate": 1.9895700160129593e-05,
+      "loss": 1.1286,
+      "step": 390
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7107625112526386,
+      "learning_rate": 1.9894800668978095e-05,
+      "loss": 1.1234,
+      "step": 391
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7413871021382318,
+      "learning_rate": 1.9893897336324292e-05,
+      "loss": 1.1611,
+      "step": 392
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7235572128798108,
+      "learning_rate": 1.9892990162518884e-05,
+      "loss": 1.2103,
+      "step": 393
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7359379685951377,
+      "learning_rate": 1.9892079147914072e-05,
+      "loss": 1.1847,
+      "step": 394
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5858887478826892,
+      "learning_rate": 1.9891164292863537e-05,
+      "loss": 1.1196,
+      "step": 395
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7695392961644535,
+      "learning_rate": 1.9890245597722465e-05,
+      "loss": 1.1407,
+      "step": 396
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7478919123266071,
+      "learning_rate": 1.9889323062847516e-05,
+      "loss": 1.1239,
+      "step": 397
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6964712680510038,
+      "learning_rate": 1.988839668859686e-05,
+      "loss": 1.1238,
+      "step": 398
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.732406215367089,
+      "learning_rate": 1.988746647533014e-05,
+      "loss": 1.1488,
+      "step": 399
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6572441896384323,
+      "learning_rate": 1.9886532423408495e-05,
+      "loss": 1.1447,
+      "step": 400
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7317552804639093,
+      "learning_rate": 1.9885594533194564e-05,
+      "loss": 1.1331,
+      "step": 401
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6201624019475236,
+      "learning_rate": 1.9884652805052465e-05,
+      "loss": 1.1635,
+      "step": 402
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6815944488273845,
+      "learning_rate": 1.9883707239347804e-05,
+      "loss": 1.1164,
+      "step": 403
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6699065279241674,
+      "learning_rate": 1.988275783644769e-05,
+      "loss": 1.1052,
+      "step": 404
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6722949685831223,
+      "learning_rate": 1.988180459672071e-05,
+      "loss": 1.1603,
+      "step": 405
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7309428445655229,
+      "learning_rate": 1.988084752053695e-05,
+      "loss": 1.0835,
+      "step": 406
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7429203375173432,
+      "learning_rate": 1.9879886608267967e-05,
+      "loss": 1.1367,
+      "step": 407
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6775378548146198,
+      "learning_rate": 1.9878921860286832e-05,
+      "loss": 1.1076,
+      "step": 408
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6741685638432184,
+      "learning_rate": 1.9877953276968088e-05,
+      "loss": 1.1425,
+      "step": 409
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7763363846247273,
+      "learning_rate": 1.9876980858687777e-05,
+      "loss": 1.0738,
+      "step": 410
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7072725146908377,
+      "learning_rate": 1.9876004605823417e-05,
+      "loss": 1.0813,
+      "step": 411
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6891818397970018,
+      "learning_rate": 1.987502451875403e-05,
+      "loss": 1.1614,
+      "step": 412
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6849457407812337,
+      "learning_rate": 1.987404059786012e-05,
+      "loss": 1.1582,
+      "step": 413
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7028989105224268,
+      "learning_rate": 1.9873052843523676e-05,
+      "loss": 1.1434,
+      "step": 414
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7002202817154949,
+      "learning_rate": 1.987206125612818e-05,
+      "loss": 1.1858,
+      "step": 415
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6939610168252043,
+      "learning_rate": 1.98710658360586e-05,
+      "loss": 1.1321,
+      "step": 416
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7433780910831888,
+      "learning_rate": 1.987006658370139e-05,
+      "loss": 1.1655,
+      "step": 417
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5471559186769006,
+      "learning_rate": 1.9869063499444495e-05,
+      "loss": 1.1326,
+      "step": 418
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7284660998135325,
+      "learning_rate": 1.9868056583677346e-05,
+      "loss": 1.0908,
+      "step": 419
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6459977688027898,
+      "learning_rate": 1.9867045836790867e-05,
+      "loss": 1.1036,
+      "step": 420
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5860294407597697,
+      "learning_rate": 1.9866031259177463e-05,
+      "loss": 1.1156,
+      "step": 421
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6889727740260306,
+      "learning_rate": 1.9865012851231022e-05,
+      "loss": 1.1053,
+      "step": 422
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.771253424705983,
+      "learning_rate": 1.9863990613346936e-05,
+      "loss": 1.1164,
+      "step": 423
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7106264472883955,
+      "learning_rate": 1.986296454592206e-05,
+      "loss": 1.1523,
+      "step": 424
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7507993834481723,
+      "learning_rate": 1.9861934649354763e-05,
+      "loss": 1.1506,
+      "step": 425
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5774479910185982,
+      "learning_rate": 1.9860900924044873e-05,
+      "loss": 1.1516,
+      "step": 426
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8082168908299288,
+      "learning_rate": 1.9859863370393726e-05,
+      "loss": 1.1287,
+      "step": 427
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7131940724780981,
+      "learning_rate": 1.9858821988804132e-05,
+      "loss": 1.1392,
+      "step": 428
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7401522377668833,
+      "learning_rate": 1.9857776779680393e-05,
+      "loss": 1.1505,
+      "step": 429
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7150290784837048,
+      "learning_rate": 1.98567277434283e-05,
+      "loss": 1.2405,
+      "step": 430
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7650974740269331,
+      "learning_rate": 1.9855674880455115e-05,
+      "loss": 1.1351,
+      "step": 431
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7112920325841698,
+      "learning_rate": 1.98546181911696e-05,
+      "loss": 1.1379,
+      "step": 432
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7216007596354014,
+      "learning_rate": 1.9853557675982e-05,
+      "loss": 1.1976,
+      "step": 433
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5669317150859281,
+      "learning_rate": 1.985249333530404e-05,
+      "loss": 1.1581,
+      "step": 434
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.69458092671465,
+      "learning_rate": 1.9851425169548938e-05,
+      "loss": 1.1209,
+      "step": 435
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6591220712182309,
+      "learning_rate": 1.9850353179131392e-05,
+      "loss": 1.1695,
+      "step": 436
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.669735137192875,
+      "learning_rate": 1.9849277364467585e-05,
+      "loss": 1.1489,
+      "step": 437
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7341086797531258,
+      "learning_rate": 1.984819772597518e-05,
+      "loss": 1.0864,
+      "step": 438
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6686247427888973,
+      "learning_rate": 1.9847114264073336e-05,
+      "loss": 1.1826,
+      "step": 439
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7015525695786888,
+      "learning_rate": 1.984602697918269e-05,
+      "loss": 1.1613,
+      "step": 440
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6778358832474367,
+      "learning_rate": 1.9844935871725363e-05,
+      "loss": 1.1262,
+      "step": 441
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.752705215147602,
+      "learning_rate": 1.9843840942124956e-05,
+      "loss": 1.184,
+      "step": 442
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7123277606573524,
+      "learning_rate": 1.9842742190806566e-05,
+      "loss": 1.0511,
+      "step": 443
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6742477799707172,
+      "learning_rate": 1.984163961819676e-05,
+      "loss": 1.1129,
+      "step": 444
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7247842957980725,
+      "learning_rate": 1.9840533224723595e-05,
+      "loss": 1.1437,
+      "step": 445
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.76726992451268,
+      "learning_rate": 1.9839423010816616e-05,
+      "loss": 1.2389,
+      "step": 446
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7259801095952174,
+      "learning_rate": 1.983830897690684e-05,
+      "loss": 1.0571,
+      "step": 447
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7305123076947123,
+      "learning_rate": 1.9837191123426777e-05,
+      "loss": 1.1467,
+      "step": 448
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6734355857409862,
+      "learning_rate": 1.983606945081042e-05,
+      "loss": 1.178,
+      "step": 449
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8513538071350008,
+      "learning_rate": 1.983494395949323e-05,
+      "loss": 1.1507,
+      "step": 450
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7348298424549189,
+      "learning_rate": 1.983381464991217e-05,
+      "loss": 1.1246,
+      "step": 451
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.786106496321788,
+      "learning_rate": 1.9832681522505676e-05,
+      "loss": 1.1401,
+      "step": 452
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.768712057349468,
+      "learning_rate": 1.9831544577713663e-05,
+      "loss": 1.1033,
+      "step": 453
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6093441701497824,
+      "learning_rate": 1.983040381597754e-05,
+      "loss": 1.1685,
+      "step": 454
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5841358022733592,
+      "learning_rate": 1.982925923774018e-05,
+      "loss": 1.081,
+      "step": 455
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6440596435826569,
+      "learning_rate": 1.9828110843445954e-05,
+      "loss": 1.0866,
+      "step": 456
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6019074518103203,
+      "learning_rate": 1.982695863354071e-05,
+      "loss": 1.2335,
+      "step": 457
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7682643917197928,
+      "learning_rate": 1.9825802608471767e-05,
+      "loss": 1.1547,
+      "step": 458
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7633766474748785,
+      "learning_rate": 1.982464276868794e-05,
+      "loss": 1.0967,
+      "step": 459
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6752798619743369,
+      "learning_rate": 1.982347911463952e-05,
+      "loss": 1.1229,
+      "step": 460
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6157197566064779,
+      "learning_rate": 1.9822311646778277e-05,
+      "loss": 1.1537,
+      "step": 461
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7277681744663871,
+      "learning_rate": 1.982114036555746e-05,
+      "loss": 1.122,
+      "step": 462
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7324965007763199,
+      "learning_rate": 1.9819965271431797e-05,
+      "loss": 1.1062,
+      "step": 463
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7249389237454766,
+      "learning_rate": 1.9818786364857506e-05,
+      "loss": 1.1638,
+      "step": 464
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7348294053317046,
+      "learning_rate": 1.9817603646292278e-05,
+      "loss": 1.0905,
+      "step": 465
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7239540577242838,
+      "learning_rate": 1.9816417116195287e-05,
+      "loss": 1.1161,
+      "step": 466
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7774151246132808,
+      "learning_rate": 1.9815226775027182e-05,
+      "loss": 1.1317,
+      "step": 467
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7273065921537067,
+      "learning_rate": 1.9814032623250093e-05,
+      "loss": 1.1575,
+      "step": 468
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6982860164268229,
+      "learning_rate": 1.9812834661327632e-05,
+      "loss": 1.1534,
+      "step": 469
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.678311823787742,
+      "learning_rate": 1.9811632889724888e-05,
+      "loss": 1.1168,
+      "step": 470
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.687607555495862,
+      "learning_rate": 1.9810427308908437e-05,
+      "loss": 1.1105,
+      "step": 471
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7012137582405243,
+      "learning_rate": 1.9809217919346318e-05,
+      "loss": 1.1519,
+      "step": 472
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7737320899985661,
+      "learning_rate": 1.980800472150806e-05,
+      "loss": 1.1313,
+      "step": 473
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7731578304398685,
+      "learning_rate": 1.9806787715864674e-05,
+      "loss": 1.1111,
+      "step": 474
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8007838291046915,
+      "learning_rate": 1.9805566902888637e-05,
+      "loss": 1.1513,
+      "step": 475
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6974199699730655,
+      "learning_rate": 1.9804342283053916e-05,
+      "loss": 1.1467,
+      "step": 476
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6947901933551158,
+      "learning_rate": 1.980311385683594e-05,
+      "loss": 1.1185,
+      "step": 477
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.610636696735961,
+      "learning_rate": 1.980188162471164e-05,
+      "loss": 1.1592,
+      "step": 478
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5842198780662904,
+      "learning_rate": 1.98006455871594e-05,
+      "loss": 1.1709,
+      "step": 479
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6816494470818093,
+      "learning_rate": 1.97994057446591e-05,
+      "loss": 1.1125,
+      "step": 480
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6161398876899591,
+      "learning_rate": 1.979816209769209e-05,
+      "loss": 1.1318,
+      "step": 481
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.704214148871699,
+      "learning_rate": 1.9796914646741187e-05,
+      "loss": 1.1028,
+      "step": 482
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7398150492665445,
+      "learning_rate": 1.9795663392290702e-05,
+      "loss": 1.0935,
+      "step": 483
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7461060245373187,
+      "learning_rate": 1.9794408334826415e-05,
+      "loss": 1.1519,
+      "step": 484
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6980055191118132,
+      "learning_rate": 1.979314947483558e-05,
+      "loss": 1.1154,
+      "step": 485
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6872851999955822,
+      "learning_rate": 1.9791886812806932e-05,
+      "loss": 1.0994,
+      "step": 486
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6346923721337314,
+      "learning_rate": 1.9790620349230676e-05,
+      "loss": 1.1434,
+      "step": 487
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.719750445131268,
+      "learning_rate": 1.9789350084598504e-05,
+      "loss": 1.1653,
+      "step": 488
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6872436757603463,
+      "learning_rate": 1.9788076019403565e-05,
+      "loss": 1.1472,
+      "step": 489
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.657694424549737,
+      "learning_rate": 1.9786798154140507e-05,
+      "loss": 1.1287,
+      "step": 490
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7021351352661943,
+      "learning_rate": 1.9785516489305437e-05,
+      "loss": 1.1359,
+      "step": 491
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7667795089000105,
+      "learning_rate": 1.9784231025395936e-05,
+      "loss": 1.1265,
+      "step": 492
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.637550185285362,
+      "learning_rate": 1.9782941762911075e-05,
+      "loss": 1.0856,
+      "step": 493
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6893137311258615,
+      "learning_rate": 1.9781648702351383e-05,
+      "loss": 1.1643,
+      "step": 494
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7081567085601528,
+      "learning_rate": 1.9780351844218874e-05,
+      "loss": 1.1362,
+      "step": 495
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8127836958045928,
+      "learning_rate": 1.977905118901703e-05,
+      "loss": 1.1086,
+      "step": 496
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7814658997487776,
+      "learning_rate": 1.977774673725081e-05,
+      "loss": 1.1526,
+      "step": 497
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6894208282829798,
+      "learning_rate": 1.977643848942665e-05,
+      "loss": 1.0508,
+      "step": 498
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7178684340790373,
+      "learning_rate": 1.977512644605246e-05,
+      "loss": 1.1294,
+      "step": 499
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7240833374822012,
+      "learning_rate": 1.9773810607637612e-05,
+      "loss": 1.1287,
+      "step": 500
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7306397222158821,
+      "learning_rate": 1.9772490974692962e-05,
+      "loss": 1.2028,
+      "step": 501
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7969935258894385,
+      "learning_rate": 1.9771167547730844e-05,
+      "loss": 1.112,
+      "step": 502
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7709965359953949,
+      "learning_rate": 1.976984032726505e-05,
+      "loss": 1.1382,
+      "step": 503
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7061583914218047,
+      "learning_rate": 1.976850931381086e-05,
+      "loss": 1.1009,
+      "step": 504
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7145470511984596,
+      "learning_rate": 1.976717450788501e-05,
+      "loss": 1.0891,
+      "step": 505
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6523449704665889,
+      "learning_rate": 1.9765835910005726e-05,
+      "loss": 1.1567,
+      "step": 506
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7418698680626513,
+      "learning_rate": 1.9764493520692685e-05,
+      "loss": 1.1042,
+      "step": 507
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7531819735290837,
+      "learning_rate": 1.9763147340467067e-05,
+      "loss": 1.1484,
+      "step": 508
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7435252982578849,
+      "learning_rate": 1.9761797369851498e-05,
+      "loss": 1.0214,
+      "step": 509
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6950152267330575,
+      "learning_rate": 1.9760443609370074e-05,
+      "loss": 1.1791,
+      "step": 510
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.729423119223485,
+      "learning_rate": 1.975908605954838e-05,
+      "loss": 1.1389,
+      "step": 511
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7271597285278848,
+      "learning_rate": 1.9757724720913466e-05,
+      "loss": 1.1936,
+      "step": 512
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7963930155327531,
+      "learning_rate": 1.9756359593993845e-05,
+      "loss": 1.144,
+      "step": 513
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7066396361685061,
+      "learning_rate": 1.975499067931951e-05,
+      "loss": 1.1142,
+      "step": 514
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7602180316835385,
+      "learning_rate": 1.975361797742192e-05,
+      "loss": 1.1415,
+      "step": 515
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7042310111116885,
+      "learning_rate": 1.9752241488834002e-05,
+      "loss": 1.1492,
+      "step": 516
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7394507934333411,
+      "learning_rate": 1.975086121409016e-05,
+      "loss": 1.1572,
+      "step": 517
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7762928506030143,
+      "learning_rate": 1.974947715372626e-05,
+      "loss": 1.0834,
+      "step": 518
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7220468277476689,
+      "learning_rate": 1.974808930827965e-05,
+      "loss": 0.9861,
+      "step": 519
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5294407858880192,
+      "learning_rate": 1.9746697678289128e-05,
+      "loss": 1.1405,
+      "step": 520
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.681277241738237,
+      "learning_rate": 1.9745302264294982e-05,
+      "loss": 1.1008,
+      "step": 521
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6896641466766222,
+      "learning_rate": 1.9743903066838954e-05,
+      "loss": 1.1256,
+      "step": 522
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5907556415688429,
+      "learning_rate": 1.9742500086464266e-05,
+      "loss": 1.1134,
+      "step": 523
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7823120459633898,
+      "learning_rate": 1.9741093323715597e-05,
+      "loss": 1.0978,
+      "step": 524
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7897306456447875,
+      "learning_rate": 1.9739682779139107e-05,
+      "loss": 1.1584,
+      "step": 525
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5374482243772916,
+      "learning_rate": 1.9738268453282414e-05,
+      "loss": 1.1343,
+      "step": 526
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7292620809747712,
+      "learning_rate": 1.9736850346694608e-05,
+      "loss": 1.1726,
+      "step": 527
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8075469183292824,
+      "learning_rate": 1.973542845992625e-05,
+      "loss": 1.113,
+      "step": 528
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5805338705175567,
+      "learning_rate": 1.9734002793529362e-05,
+      "loss": 1.1439,
+      "step": 529
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7574434112376857,
+      "learning_rate": 1.9732573348057437e-05,
+      "loss": 1.1754,
+      "step": 530
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.741199091101836,
+      "learning_rate": 1.973114012406544e-05,
+      "loss": 1.1372,
+      "step": 531
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6377154429262023,
+      "learning_rate": 1.9729703122109788e-05,
+      "loss": 1.0811,
+      "step": 532
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7124956899617472,
+      "learning_rate": 1.9728262342748384e-05,
+      "loss": 1.1537,
+      "step": 533
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7527200527000097,
+      "learning_rate": 1.9726817786540584e-05,
+      "loss": 1.0874,
+      "step": 534
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7153708542284745,
+      "learning_rate": 1.9725369454047215e-05,
+      "loss": 1.0657,
+      "step": 535
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7992604170428712,
+      "learning_rate": 1.9723917345830568e-05,
+      "loss": 1.1299,
+      "step": 536
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7043905180090546,
+      "learning_rate": 1.9722461462454405e-05,
+      "loss": 1.0576,
+      "step": 537
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7211319977465496,
+      "learning_rate": 1.9721001804483947e-05,
+      "loss": 1.1185,
+      "step": 538
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6885772599685217,
+      "learning_rate": 1.9719538372485887e-05,
+      "loss": 1.0951,
+      "step": 539
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6915100337529689,
+      "learning_rate": 1.9718071167028376e-05,
+      "loss": 1.1407,
+      "step": 540
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7060509459794483,
+      "learning_rate": 1.9716600188681038e-05,
+      "loss": 1.0939,
+      "step": 541
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6974224443071825,
+      "learning_rate": 1.971512543801495e-05,
+      "loss": 1.084,
+      "step": 542
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7068019893099365,
+      "learning_rate": 1.9713646915602663e-05,
+      "loss": 1.1404,
+      "step": 543
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7226995334383631,
+      "learning_rate": 1.9712164622018197e-05,
+      "loss": 1.0974,
+      "step": 544
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7238295371726832,
+      "learning_rate": 1.9710678557837024e-05,
+      "loss": 1.1033,
+      "step": 545
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6855921326457342,
+      "learning_rate": 1.9709188723636088e-05,
+      "loss": 1.1191,
+      "step": 546
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7049597046835516,
+      "learning_rate": 1.970769511999379e-05,
+      "loss": 1.1074,
+      "step": 547
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7703434742654682,
+      "learning_rate": 1.9706197747490004e-05,
+      "loss": 1.2134,
+      "step": 548
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7401071406585429,
+      "learning_rate": 1.9704696606706055e-05,
+      "loss": 1.1317,
+      "step": 549
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7668337783646898,
+      "learning_rate": 1.9703191698224742e-05,
+      "loss": 1.1201,
+      "step": 550
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6976211464264989,
+      "learning_rate": 1.9701683022630323e-05,
+      "loss": 1.1816,
+      "step": 551
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6326324429905372,
+      "learning_rate": 1.9700170580508514e-05,
+      "loss": 1.1251,
+      "step": 552
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6474842024436397,
+      "learning_rate": 1.9698654372446495e-05,
+      "loss": 1.1737,
+      "step": 553
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7253419161643988,
+      "learning_rate": 1.969713439903292e-05,
+      "loss": 1.0333,
+      "step": 554
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6413459108638907,
+      "learning_rate": 1.9695610660857886e-05,
+      "loss": 1.1168,
+      "step": 555
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6855484951641235,
+      "learning_rate": 1.9694083158512965e-05,
+      "loss": 1.0861,
+      "step": 556
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7142705775408277,
+      "learning_rate": 1.9692551892591185e-05,
+      "loss": 1.1328,
+      "step": 557
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6947278198777586,
+      "learning_rate": 1.9691016863687037e-05,
+      "loss": 1.1056,
+      "step": 558
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.569853365755238,
+      "learning_rate": 1.968947807239647e-05,
+      "loss": 1.1092,
+      "step": 559
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6842433266188208,
+      "learning_rate": 1.9687935519316897e-05,
+      "loss": 1.1113,
+      "step": 560
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7274338260426947,
+      "learning_rate": 1.9686389205047186e-05,
+      "loss": 0.9991,
+      "step": 561
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7574191575385314,
+      "learning_rate": 1.9684839130187678e-05,
+      "loss": 1.0677,
+      "step": 562
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6902460799484037,
+      "learning_rate": 1.968328529534016e-05,
+      "loss": 1.1089,
+      "step": 563
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7796522881773431,
+      "learning_rate": 1.9681727701107885e-05,
+      "loss": 1.1247,
+      "step": 564
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7884927039430832,
+      "learning_rate": 1.9680166348095568e-05,
+      "loss": 1.1322,
+      "step": 565
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6662994672533354,
+      "learning_rate": 1.967860123690937e-05,
+      "loss": 1.0663,
+      "step": 566
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7255884627055011,
+      "learning_rate": 1.9677032368156934e-05,
+      "loss": 1.1373,
+      "step": 567
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6861194785961084,
+      "learning_rate": 1.967545974244734e-05,
+      "loss": 1.1597,
+      "step": 568
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8048638513937136,
+      "learning_rate": 1.9673883360391138e-05,
+      "loss": 1.1144,
+      "step": 569
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7444793596563625,
+      "learning_rate": 1.9672303222600333e-05,
+      "loss": 1.1187,
+      "step": 570
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6250666577940501,
+      "learning_rate": 1.967071932968839e-05,
+      "loss": 1.0804,
+      "step": 571
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7187895708831898,
+      "learning_rate": 1.9669131682270232e-05,
+      "loss": 1.1754,
+      "step": 572
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7202242195199098,
+      "learning_rate": 1.9667540280962235e-05,
+      "loss": 1.0855,
+      "step": 573
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7854862910927118,
+      "learning_rate": 1.966594512638224e-05,
+      "loss": 1.1418,
+      "step": 574
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5656439861473376,
+      "learning_rate": 1.9664346219149538e-05,
+      "loss": 1.1217,
+      "step": 575
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7281332992760278,
+      "learning_rate": 1.966274355988488e-05,
+      "loss": 1.2053,
+      "step": 576
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6753822853310782,
+      "learning_rate": 1.9661137149210473e-05,
+      "loss": 1.1467,
+      "step": 577
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7468125417348132,
+      "learning_rate": 1.9659526987749987e-05,
+      "loss": 1.1412,
+      "step": 578
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5869171265838955,
+      "learning_rate": 1.9657913076128532e-05,
+      "loss": 1.043,
+      "step": 579
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7412231647217881,
+      "learning_rate": 1.965629541497269e-05,
+      "loss": 1.1188,
+      "step": 580
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7289672157398351,
+      "learning_rate": 1.9654674004910493e-05,
+      "loss": 1.1855,
+      "step": 581
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6814713220097868,
+      "learning_rate": 1.9653048846571427e-05,
+      "loss": 1.1246,
+      "step": 582
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7677936673405868,
+      "learning_rate": 1.9651419940586437e-05,
+      "loss": 1.0798,
+      "step": 583
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6932797082253758,
+      "learning_rate": 1.964978728758791e-05,
+      "loss": 1.061,
+      "step": 584
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7157404353373106,
+      "learning_rate": 1.9648150888209715e-05,
+      "loss": 1.1234,
+      "step": 585
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7426791207960447,
+      "learning_rate": 1.9646510743087144e-05,
+      "loss": 1.1257,
+      "step": 586
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7399688742690367,
+      "learning_rate": 1.964486685285697e-05,
+      "loss": 1.0236,
+      "step": 587
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7221903329565911,
+      "learning_rate": 1.9643219218157395e-05,
+      "loss": 1.1176,
+      "step": 588
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6752101733310502,
+      "learning_rate": 1.9641567839628092e-05,
+      "loss": 1.0879,
+      "step": 589
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7105884631723548,
+      "learning_rate": 1.963991271791019e-05,
+      "loss": 1.1584,
+      "step": 590
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6823227282269847,
+      "learning_rate": 1.9638253853646255e-05,
+      "loss": 1.0784,
+      "step": 591
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5654313499841065,
+      "learning_rate": 1.9636591247480323e-05,
+      "loss": 1.1154,
+      "step": 592
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7083344588993327,
+      "learning_rate": 1.9634924900057867e-05,
+      "loss": 1.0872,
+      "step": 593
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5978924643100868,
+      "learning_rate": 1.963325481202583e-05,
+      "loss": 1.1403,
+      "step": 594
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7698601716372853,
+      "learning_rate": 1.963158098403259e-05,
+      "loss": 1.0759,
+      "step": 595
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7321266071758208,
+      "learning_rate": 1.9629903416727987e-05,
+      "loss": 1.1276,
+      "step": 596
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6265122167018551,
+      "learning_rate": 1.962822211076331e-05,
+      "loss": 1.0969,
+      "step": 597
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7272280090212183,
+      "learning_rate": 1.96265370667913e-05,
+      "loss": 1.0829,
+      "step": 598
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6729141029334618,
+      "learning_rate": 1.9624848285466146e-05,
+      "loss": 1.1243,
+      "step": 599
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6576792193592125,
+      "learning_rate": 1.9623155767443498e-05,
+      "loss": 1.1941,
+      "step": 600
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7091594519412999,
+      "learning_rate": 1.9621459513380445e-05,
+      "loss": 1.1884,
+      "step": 601
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7531463061681734,
+      "learning_rate": 1.9619759523935532e-05,
+      "loss": 1.188,
+      "step": 602
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7588726526810049,
+      "learning_rate": 1.9618055799768757e-05,
+      "loss": 1.0797,
+      "step": 603
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6953823374993325,
+      "learning_rate": 1.961634834154156e-05,
+      "loss": 1.1376,
+      "step": 604
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6656585174926234,
+      "learning_rate": 1.9614637149916834e-05,
+      "loss": 1.0294,
+      "step": 605
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6634267305705428,
+      "learning_rate": 1.9612922225558924e-05,
+      "loss": 1.0972,
+      "step": 606
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6683857439835209,
+      "learning_rate": 1.961120356913363e-05,
+      "loss": 1.1437,
+      "step": 607
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.697168967343409,
+      "learning_rate": 1.960948118130818e-05,
+      "loss": 1.1158,
+      "step": 608
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6927532387932568,
+      "learning_rate": 1.9607755062751273e-05,
+      "loss": 1.151,
+      "step": 609
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5089744931056635,
+      "learning_rate": 1.9606025214133046e-05,
+      "loss": 1.119,
+      "step": 610
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6337467606711766,
+      "learning_rate": 1.9604291636125084e-05,
+      "loss": 1.0306,
+      "step": 611
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6678035077952152,
+      "learning_rate": 1.960255432940043e-05,
+      "loss": 1.0707,
+      "step": 612
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7251175591334869,
+      "learning_rate": 1.9600813294633552e-05,
+      "loss": 1.1453,
+      "step": 613
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.711726584580558,
+      "learning_rate": 1.9599068532500394e-05,
+      "loss": 1.0789,
+      "step": 614
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7193280365415088,
+      "learning_rate": 1.9597320043678322e-05,
+      "loss": 1.1929,
+      "step": 615
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.67590590822157,
+      "learning_rate": 1.9595567828846166e-05,
+      "loss": 1.1967,
+      "step": 616
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7626294901612286,
+      "learning_rate": 1.9593811888684192e-05,
+      "loss": 1.0857,
+      "step": 617
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7678409750269887,
+      "learning_rate": 1.9592052223874115e-05,
+      "loss": 1.0839,
+      "step": 618
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6551463012090313,
+      "learning_rate": 1.959028883509911e-05,
+      "loss": 1.0898,
+      "step": 619
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7734192350054508,
+      "learning_rate": 1.9588521723043764e-05,
+      "loss": 1.0946,
+      "step": 620
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6229464412966554,
+      "learning_rate": 1.958675088839415e-05,
+      "loss": 1.1367,
+      "step": 621
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5522594626747575,
+      "learning_rate": 1.9584976331837758e-05,
+      "loss": 1.0774,
+      "step": 622
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6950990062741964,
+      "learning_rate": 1.9583198054063535e-05,
+      "loss": 1.1151,
+      "step": 623
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7181454851884006,
+      "learning_rate": 1.9581416055761865e-05,
+      "loss": 1.0569,
+      "step": 624
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7512217159607729,
+      "learning_rate": 1.9579630337624585e-05,
+      "loss": 1.1404,
+      "step": 625
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.620520563297347,
+      "learning_rate": 1.9577840900344974e-05,
+      "loss": 1.0964,
+      "step": 626
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5219332881154474,
+      "learning_rate": 1.9576047744617752e-05,
+      "loss": 1.1063,
+      "step": 627
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.724055413139711,
+      "learning_rate": 1.957425087113908e-05,
+      "loss": 1.1414,
+      "step": 628
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7647545422077406,
+      "learning_rate": 1.9572450280606568e-05,
+      "loss": 1.0764,
+      "step": 629
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.74579942767317,
+      "learning_rate": 1.9570645973719273e-05,
+      "loss": 1.1306,
+      "step": 630
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7079405351706866,
+      "learning_rate": 1.9568837951177677e-05,
+      "loss": 1.1006,
+      "step": 631
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6754812712234064,
+      "learning_rate": 1.9567026213683728e-05,
+      "loss": 1.1112,
+      "step": 632
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7146222347794403,
+      "learning_rate": 1.9565210761940798e-05,
+      "loss": 1.1447,
+      "step": 633
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7469205205997731,
+      "learning_rate": 1.956339159665371e-05,
+      "loss": 1.1444,
+      "step": 634
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7482448141880468,
+      "learning_rate": 1.956156871852873e-05,
+      "loss": 1.115,
+      "step": 635
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6719415882045036,
+      "learning_rate": 1.9559742128273558e-05,
+      "loss": 1.1102,
+      "step": 636
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6305993564523358,
+      "learning_rate": 1.9557911826597337e-05,
+      "loss": 1.0769,
+      "step": 637
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7645575333561179,
+      "learning_rate": 1.9556077814210662e-05,
+      "loss": 1.2139,
+      "step": 638
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6994283385530845,
+      "learning_rate": 1.955424009182555e-05,
+      "loss": 1.0994,
+      "step": 639
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5902216423513406,
+      "learning_rate": 1.955239866015547e-05,
+      "loss": 1.188,
+      "step": 640
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.696132586366459,
+      "learning_rate": 1.9550553519915335e-05,
+      "loss": 1.1334,
+      "step": 641
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6915753079085342,
+      "learning_rate": 1.954870467182149e-05,
+      "loss": 1.109,
+      "step": 642
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6149502541266277,
+      "learning_rate": 1.954685211659172e-05,
+      "loss": 1.1255,
+      "step": 643
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6613878836229872,
+      "learning_rate": 1.9544995854945248e-05,
+      "loss": 1.1644,
+      "step": 644
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7343095733713401,
+      "learning_rate": 1.954313588760274e-05,
+      "loss": 1.1219,
+      "step": 645
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6838683390910062,
+      "learning_rate": 1.9541272215286304e-05,
+      "loss": 1.1168,
+      "step": 646
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7618751941679186,
+      "learning_rate": 1.9539404838719477e-05,
+      "loss": 1.1351,
+      "step": 647
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7650311469918712,
+      "learning_rate": 1.9537533758627242e-05,
+      "loss": 1.0553,
+      "step": 648
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6981520752463816,
+      "learning_rate": 1.953565897573601e-05,
+      "loss": 1.1166,
+      "step": 649
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7783366003159811,
+      "learning_rate": 1.9533780490773645e-05,
+      "loss": 1.114,
+      "step": 650
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6946394847925227,
+      "learning_rate": 1.9531898304469435e-05,
+      "loss": 1.0367,
+      "step": 651
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5479991416936699,
+      "learning_rate": 1.953001241755411e-05,
+      "loss": 1.1582,
+      "step": 652
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6860162317773716,
+      "learning_rate": 1.952812283075984e-05,
+      "loss": 1.0999,
+      "step": 653
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7127001321204931,
+      "learning_rate": 1.952622954482022e-05,
+      "loss": 1.0854,
+      "step": 654
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.648589457864316,
+      "learning_rate": 1.9524332560470293e-05,
+      "loss": 1.064,
+      "step": 655
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6476008521616606,
+      "learning_rate": 1.9522431878446536e-05,
+      "loss": 1.1245,
+      "step": 656
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8380090276925275,
+      "learning_rate": 1.9520527499486856e-05,
+      "loss": 1.1614,
+      "step": 657
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7235755089311282,
+      "learning_rate": 1.95186194243306e-05,
+      "loss": 1.053,
+      "step": 658
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6669420976465522,
+      "learning_rate": 1.9516707653718546e-05,
+      "loss": 1.0898,
+      "step": 659
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6936339735693602,
+      "learning_rate": 1.9514792188392914e-05,
+      "loss": 1.1433,
+      "step": 660
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7180558058667452,
+      "learning_rate": 1.9512873029097347e-05,
+      "loss": 1.0657,
+      "step": 661
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7808862752513407,
+      "learning_rate": 1.9510950176576933e-05,
+      "loss": 1.1111,
+      "step": 662
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7538855156276504,
+      "learning_rate": 1.950902363157819e-05,
+      "loss": 1.1572,
+      "step": 663
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6559718042387737,
+      "learning_rate": 1.950709339484907e-05,
+      "loss": 1.1115,
+      "step": 664
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6472760768067235,
+      "learning_rate": 1.9505159467138954e-05,
+      "loss": 1.153,
+      "step": 665
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7755537150899483,
+      "learning_rate": 1.9503221849198655e-05,
+      "loss": 1.1265,
+      "step": 666
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6921364822462591,
+      "learning_rate": 1.9501280541780435e-05,
+      "loss": 1.1555,
+      "step": 667
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6917695772496943,
+      "learning_rate": 1.9499335545637968e-05,
+      "loss": 1.1523,
+      "step": 668
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6835634636741946,
+      "learning_rate": 1.949738686152637e-05,
+      "loss": 1.1244,
+      "step": 669
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6935756244887104,
+      "learning_rate": 1.9495434490202188e-05,
+      "loss": 1.1144,
+      "step": 670
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.749980312061298,
+      "learning_rate": 1.94934784324234e-05,
+      "loss": 1.0585,
+      "step": 671
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7384890397834906,
+      "learning_rate": 1.9491518688949417e-05,
+      "loss": 1.076,
+      "step": 672
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7070058735027431,
+      "learning_rate": 1.9489555260541074e-05,
+      "loss": 1.101,
+      "step": 673
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7419232319950377,
+      "learning_rate": 1.948758814796064e-05,
+      "loss": 1.1163,
+      "step": 674
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7351301374411214,
+      "learning_rate": 1.9485617351971827e-05,
+      "loss": 1.1044,
+      "step": 675
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6493337042206134,
+      "learning_rate": 1.9483642873339753e-05,
+      "loss": 1.1014,
+      "step": 676
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7576213959095417,
+      "learning_rate": 1.9481664712830987e-05,
+      "loss": 1.1321,
+      "step": 677
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.70861600766359,
+      "learning_rate": 1.9479682871213515e-05,
+      "loss": 1.0822,
+      "step": 678
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7349475507651863,
+      "learning_rate": 1.9477697349256756e-05,
+      "loss": 1.1431,
+      "step": 679
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6906857913709649,
+      "learning_rate": 1.947570814773156e-05,
+      "loss": 1.1095,
+      "step": 680
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7903054384520309,
+      "learning_rate": 1.9473715267410206e-05,
+      "loss": 1.2195,
+      "step": 681
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7197792542798533,
+      "learning_rate": 1.9471718709066392e-05,
+      "loss": 1.1091,
+      "step": 682
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5517662778285205,
+      "learning_rate": 1.9469718473475256e-05,
+      "loss": 1.1101,
+      "step": 683
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7194840038801243,
+      "learning_rate": 1.9467714561413358e-05,
+      "loss": 1.0956,
+      "step": 684
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6525571039310383,
+      "learning_rate": 1.9465706973658683e-05,
+      "loss": 1.0903,
+      "step": 685
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6795346141403105,
+      "learning_rate": 1.9463695710990648e-05,
+      "loss": 1.0953,
+      "step": 686
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7180476320919993,
+      "learning_rate": 1.946168077419009e-05,
+      "loss": 1.0184,
+      "step": 687
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.628846853489212,
+      "learning_rate": 1.9459662164039283e-05,
+      "loss": 1.1364,
+      "step": 688
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.800820578002895,
+      "learning_rate": 1.9457639881321917e-05,
+      "loss": 1.1295,
+      "step": 689
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7490529762897399,
+      "learning_rate": 1.9455613926823115e-05,
+      "loss": 1.1321,
+      "step": 690
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7001314693508524,
+      "learning_rate": 1.945358430132942e-05,
+      "loss": 1.0649,
+      "step": 691
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7297706728013142,
+      "learning_rate": 1.9451551005628803e-05,
+      "loss": 1.1199,
+      "step": 692
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7774220520862066,
+      "learning_rate": 1.9449514040510654e-05,
+      "loss": 1.1384,
+      "step": 693
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6281134528024044,
+      "learning_rate": 1.9447473406765803e-05,
+      "loss": 1.1171,
+      "step": 694
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.664072015308591,
+      "learning_rate": 1.9445429105186487e-05,
+      "loss": 1.1295,
+      "step": 695
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8069780603699435,
+      "learning_rate": 1.9443381136566382e-05,
+      "loss": 1.1127,
+      "step": 696
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7384646184951787,
+      "learning_rate": 1.9441329501700568e-05,
+      "loss": 1.1235,
+      "step": 697
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7568325474386091,
+      "learning_rate": 1.943927420138557e-05,
+      "loss": 1.0968,
+      "step": 698
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6864844515286681,
+      "learning_rate": 1.9437215236419322e-05,
+      "loss": 1.0284,
+      "step": 699
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.732517527702547,
+      "learning_rate": 1.9435152607601187e-05,
+      "loss": 1.144,
+      "step": 700
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6749884084764984,
+      "learning_rate": 1.943308631573195e-05,
+      "loss": 1.0518,
+      "step": 701
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6962998601912628,
+      "learning_rate": 1.9431016361613816e-05,
+      "loss": 1.1226,
+      "step": 702
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6947665873896774,
+      "learning_rate": 1.9428942746050406e-05,
+      "loss": 1.0843,
+      "step": 703
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6482554442661476,
+      "learning_rate": 1.9426865469846773e-05,
+      "loss": 1.0355,
+      "step": 704
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7147991481157909,
+      "learning_rate": 1.9424784533809393e-05,
+      "loss": 1.1235,
+      "step": 705
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7193392081322759,
+      "learning_rate": 1.942269993874615e-05,
+      "loss": 1.153,
+      "step": 706
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5084775175850217,
+      "learning_rate": 1.9420611685466358e-05,
+      "loss": 1.1199,
+      "step": 707
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7022888396434249,
+      "learning_rate": 1.9418519774780748e-05,
+      "loss": 1.1243,
+      "step": 708
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7139036910073672,
+      "learning_rate": 1.9416424207501474e-05,
+      "loss": 1.1124,
+      "step": 709
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6812019730615415,
+      "learning_rate": 1.9414324984442102e-05,
+      "loss": 1.0683,
+      "step": 710
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7575652645321478,
+      "learning_rate": 1.9412222106417632e-05,
+      "loss": 0.9924,
+      "step": 711
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7405585362734467,
+      "learning_rate": 1.9410115574244462e-05,
+      "loss": 1.0363,
+      "step": 712
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7723763172928233,
+      "learning_rate": 1.9408005388740433e-05,
+      "loss": 1.0016,
+      "step": 713
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6579982842497508,
+      "learning_rate": 1.9405891550724778e-05,
+      "loss": 1.1329,
+      "step": 714
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7010505413679019,
+      "learning_rate": 1.940377406101817e-05,
+      "loss": 0.9654,
+      "step": 715
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7375517693764438,
+      "learning_rate": 1.9401652920442694e-05,
+      "loss": 1.0288,
+      "step": 716
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.729134414971048,
+      "learning_rate": 1.9399528129821842e-05,
+      "loss": 1.1616,
+      "step": 717
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7775961282524412,
+      "learning_rate": 1.939739968998054e-05,
+      "loss": 1.0905,
+      "step": 718
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7143964957188201,
+      "learning_rate": 1.939526760174511e-05,
+      "loss": 1.0428,
+      "step": 719
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5546638368396869,
+      "learning_rate": 1.939313186594331e-05,
+      "loss": 1.203,
+      "step": 720
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7111667751336932,
+      "learning_rate": 1.9390992483404308e-05,
+      "loss": 1.1328,
+      "step": 721
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6513528623178632,
+      "learning_rate": 1.938884945495868e-05,
+      "loss": 1.1234,
+      "step": 722
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7436525970124912,
+      "learning_rate": 1.9386702781438425e-05,
+      "loss": 1.1124,
+      "step": 723
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7138708785664832,
+      "learning_rate": 1.938455246367696e-05,
+      "loss": 1.0346,
+      "step": 724
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7352796671423861,
+      "learning_rate": 1.9382398502509107e-05,
+      "loss": 1.1074,
+      "step": 725
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6947289245910957,
+      "learning_rate": 1.938024089877111e-05,
+      "loss": 1.0856,
+      "step": 726
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7242107685598967,
+      "learning_rate": 1.9378079653300624e-05,
+      "loss": 1.1471,
+      "step": 727
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7125257793332552,
+      "learning_rate": 1.9375914766936723e-05,
+      "loss": 1.1195,
+      "step": 728
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7565050447099276,
+      "learning_rate": 1.9373746240519884e-05,
+      "loss": 1.0796,
+      "step": 729
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6919766093472539,
+      "learning_rate": 1.937157407489201e-05,
+      "loss": 1.17,
+      "step": 730
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6857481846142615,
+      "learning_rate": 1.9369398270896403e-05,
+      "loss": 1.1129,
+      "step": 731
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7426539955599744,
+      "learning_rate": 1.936721882937779e-05,
+      "loss": 1.05,
+      "step": 732
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5831850550806623,
+      "learning_rate": 1.9365035751182307e-05,
+      "loss": 1.091,
+      "step": 733
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6596475157977179,
+      "learning_rate": 1.93628490371575e-05,
+      "loss": 1.1023,
+      "step": 734
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7058728126423313,
+      "learning_rate": 1.9360658688152322e-05,
+      "loss": 1.051,
+      "step": 735
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6598002272401847,
+      "learning_rate": 1.9358464705017143e-05,
+      "loss": 1.09,
+      "step": 736
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5080266086961561,
+      "learning_rate": 1.9356267088603745e-05,
+      "loss": 1.087,
+      "step": 737
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7710436677380652,
+      "learning_rate": 1.9354065839765316e-05,
+      "loss": 1.0458,
+      "step": 738
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.536752638641648,
+      "learning_rate": 1.9351860959356462e-05,
+      "loss": 1.0979,
+      "step": 739
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7162960298786131,
+      "learning_rate": 1.9349652448233187e-05,
+      "loss": 1.1077,
+      "step": 740
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6273700096836287,
+      "learning_rate": 1.934744030725291e-05,
+      "loss": 1.0853,
+      "step": 741
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6942901768441148,
+      "learning_rate": 1.934522453727447e-05,
+      "loss": 1.073,
+      "step": 742
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6690370182586646,
+      "learning_rate": 1.93430051391581e-05,
+      "loss": 1.0539,
+      "step": 743
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6555552437643969,
+      "learning_rate": 1.934078211376544e-05,
+      "loss": 1.1176,
+      "step": 744
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7365962869083227,
+      "learning_rate": 1.9338555461959554e-05,
+      "loss": 1.0634,
+      "step": 745
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7086621848546155,
+      "learning_rate": 1.93363251846049e-05,
+      "loss": 1.0762,
+      "step": 746
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5869715846110867,
+      "learning_rate": 1.9334091282567352e-05,
+      "loss": 1.1508,
+      "step": 747
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7762925744537899,
+      "learning_rate": 1.9331853756714185e-05,
+      "loss": 1.0656,
+      "step": 748
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6738563191125894,
+      "learning_rate": 1.9329612607914088e-05,
+      "loss": 1.1278,
+      "step": 749
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6960814013856869,
+      "learning_rate": 1.9327367837037142e-05,
+      "loss": 1.1516,
+      "step": 750
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7444336955305982,
+      "learning_rate": 1.9325119444954855e-05,
+      "loss": 1.0413,
+      "step": 751
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8052828748037388,
+      "learning_rate": 1.9322867432540126e-05,
+      "loss": 1.1588,
+      "step": 752
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6940260833353077,
+      "learning_rate": 1.9320611800667268e-05,
+      "loss": 1.1267,
+      "step": 753
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7234750885811341,
+      "learning_rate": 1.9318352550211986e-05,
+      "loss": 0.9792,
+      "step": 754
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5743491778569038,
+      "learning_rate": 1.9316089682051403e-05,
+      "loss": 1.1178,
+      "step": 755
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6602159273016988,
+      "learning_rate": 1.9313823197064042e-05,
+      "loss": 1.1033,
+      "step": 756
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6568501113779419,
+      "learning_rate": 1.9311553096129835e-05,
+      "loss": 1.1415,
+      "step": 757
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7268299407556794,
+      "learning_rate": 1.9309279380130112e-05,
+      "loss": 1.0605,
+      "step": 758
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6821611027198548,
+      "learning_rate": 1.93070020499476e-05,
+      "loss": 1.0984,
+      "step": 759
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8022415345988075,
+      "learning_rate": 1.930472110646645e-05,
+      "loss": 1.1716,
+      "step": 760
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7119383820648347,
+      "learning_rate": 1.9302436550572187e-05,
+      "loss": 1.0309,
+      "step": 761
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.728153246915069,
+      "learning_rate": 1.930014838315177e-05,
+      "loss": 1.0918,
+      "step": 762
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6442573053416989,
+      "learning_rate": 1.9297856605093534e-05,
+      "loss": 1.1296,
+      "step": 763
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6541708613287237,
+      "learning_rate": 1.9295561217287226e-05,
+      "loss": 1.1116,
+      "step": 764
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6570838037140911,
+      "learning_rate": 1.9293262220624002e-05,
+      "loss": 1.1343,
+      "step": 765
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8060798285828975,
+      "learning_rate": 1.9290959615996407e-05,
+      "loss": 1.0572,
+      "step": 766
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7111553626889839,
+      "learning_rate": 1.9288653404298392e-05,
+      "loss": 1.003,
+      "step": 767
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6516636426776734,
+      "learning_rate": 1.9286343586425307e-05,
+      "loss": 1.0883,
+      "step": 768
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6934124585105678,
+      "learning_rate": 1.9284030163273907e-05,
+      "loss": 1.1102,
+      "step": 769
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6237182089294938,
+      "learning_rate": 1.9281713135742333e-05,
+      "loss": 1.0953,
+      "step": 770
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6958289559193712,
+      "learning_rate": 1.9279392504730147e-05,
+      "loss": 1.0429,
+      "step": 771
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7208583743400162,
+      "learning_rate": 1.9277068271138287e-05,
+      "loss": 1.071,
+      "step": 772
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7026262460016143,
+      "learning_rate": 1.9274740435869107e-05,
+      "loss": 1.1653,
+      "step": 773
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6898555906395673,
+      "learning_rate": 1.927240899982635e-05,
+      "loss": 1.0729,
+      "step": 774
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6337193374957947,
+      "learning_rate": 1.9270073963915162e-05,
+      "loss": 1.1881,
+      "step": 775
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6952392430799793,
+      "learning_rate": 1.9267735329042086e-05,
+      "loss": 1.0377,
+      "step": 776
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7306503992033444,
+      "learning_rate": 1.9265393096115056e-05,
+      "loss": 1.1018,
+      "step": 777
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.739303361009525,
+      "learning_rate": 1.926304726604341e-05,
+      "loss": 1.1891,
+      "step": 778
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6110164214402608,
+      "learning_rate": 1.9260697839737875e-05,
+      "loss": 1.1405,
+      "step": 779
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6769249262484196,
+      "learning_rate": 1.925834481811059e-05,
+      "loss": 1.07,
+      "step": 780
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5667400652728766,
+      "learning_rate": 1.9255988202075065e-05,
+      "loss": 1.1277,
+      "step": 781
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6663430887185696,
+      "learning_rate": 1.925362799254623e-05,
+      "loss": 1.1191,
+      "step": 782
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6912208925150568,
+      "learning_rate": 1.9251264190440398e-05,
+      "loss": 1.0993,
+      "step": 783
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6299713532917003,
+      "learning_rate": 1.9248896796675277e-05,
+      "loss": 1.0619,
+      "step": 784
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6929227177217815,
+      "learning_rate": 1.924652581216997e-05,
+      "loss": 1.0303,
+      "step": 785
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7435534311582911,
+      "learning_rate": 1.9244151237844975e-05,
+      "loss": 1.1595,
+      "step": 786
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8020762000005973,
+      "learning_rate": 1.9241773074622182e-05,
+      "loss": 1.0729,
+      "step": 787
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6831383673696396,
+      "learning_rate": 1.923939132342488e-05,
+      "loss": 1.1357,
+      "step": 788
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6849160014119183,
+      "learning_rate": 1.923700598517775e-05,
+      "loss": 1.1274,
+      "step": 789
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6844095331536705,
+      "learning_rate": 1.923461706080685e-05,
+      "loss": 1.1227,
+      "step": 790
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7936628829475513,
+      "learning_rate": 1.923222455123965e-05,
+      "loss": 1.0551,
+      "step": 791
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6666651716024847,
+      "learning_rate": 1.9229828457405005e-05,
+      "loss": 1.1116,
+      "step": 792
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.671032760340139,
+      "learning_rate": 1.9227428780233162e-05,
+      "loss": 1.0084,
+      "step": 793
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.786318077857065,
+      "learning_rate": 1.922502552065576e-05,
+      "loss": 1.0816,
+      "step": 794
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6961698897249817,
+      "learning_rate": 1.922261867960582e-05,
+      "loss": 1.0817,
+      "step": 795
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7538799571905132,
+      "learning_rate": 1.9220208258017763e-05,
+      "loss": 1.0645,
+      "step": 796
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7923743027994927,
+      "learning_rate": 1.92177942568274e-05,
+      "loss": 1.0234,
+      "step": 797
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7201830637452646,
+      "learning_rate": 1.921537667697193e-05,
+      "loss": 1.1394,
+      "step": 798
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6563369007684183,
+      "learning_rate": 1.9212955519389938e-05,
+      "loss": 1.112,
+      "step": 799
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7214341519660327,
+      "learning_rate": 1.9210530785021405e-05,
+      "loss": 1.0948,
+      "step": 800
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6542585942740119,
+      "learning_rate": 1.9208102474807692e-05,
+      "loss": 1.1015,
+      "step": 801
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7513163323518253,
+      "learning_rate": 1.920567058969155e-05,
+      "loss": 1.0321,
+      "step": 802
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.661003185151641,
+      "learning_rate": 1.920323513061713e-05,
+      "loss": 1.071,
+      "step": 803
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6839810969302764,
+      "learning_rate": 1.9200796098529956e-05,
+      "loss": 1.0814,
+      "step": 804
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6786538821787851,
+      "learning_rate": 1.919835349437694e-05,
+      "loss": 1.0595,
+      "step": 805
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6332081843865542,
+      "learning_rate": 1.9195907319106394e-05,
+      "loss": 1.1359,
+      "step": 806
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7010597870115665,
+      "learning_rate": 1.9193457573667996e-05,
+      "loss": 1.0663,
+      "step": 807
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7485887331608561,
+      "learning_rate": 1.919100425901283e-05,
+      "loss": 1.1003,
+      "step": 808
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6047296477509498,
+      "learning_rate": 1.9188547376093355e-05,
+      "loss": 1.0944,
+      "step": 809
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6553523087361216,
+      "learning_rate": 1.918608692586342e-05,
+      "loss": 1.1059,
+      "step": 810
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.731672362951955,
+      "learning_rate": 1.918362290927825e-05,
+      "loss": 1.1605,
+      "step": 811
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7116843335430338,
+      "learning_rate": 1.9181155327294468e-05,
+      "loss": 1.1011,
+      "step": 812
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7098628619151549,
+      "learning_rate": 1.9178684180870072e-05,
+      "loss": 0.9943,
+      "step": 813
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6527229636406345,
+      "learning_rate": 1.9176209470964446e-05,
+      "loss": 1.0239,
+      "step": 814
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7866167314872999,
+      "learning_rate": 1.9173731198538354e-05,
+      "loss": 1.1093,
+      "step": 815
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7190057301149294,
+      "learning_rate": 1.9171249364553956e-05,
+      "loss": 1.1445,
+      "step": 816
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7455895558725282,
+      "learning_rate": 1.9168763969974773e-05,
+      "loss": 1.0953,
+      "step": 817
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6617000973893035,
+      "learning_rate": 1.916627501576573e-05,
+      "loss": 1.0523,
+      "step": 818
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6852206575747428,
+      "learning_rate": 1.916378250289312e-05,
+      "loss": 1.0776,
+      "step": 819
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7069310754545992,
+      "learning_rate": 1.9161286432324628e-05,
+      "loss": 1.0541,
+      "step": 820
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6842640685883785,
+      "learning_rate": 1.9158786805029307e-05,
+      "loss": 1.1193,
+      "step": 821
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.672902958578767,
+      "learning_rate": 1.9156283621977603e-05,
+      "loss": 1.1735,
+      "step": 822
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7266034939705762,
+      "learning_rate": 1.9153776884141336e-05,
+      "loss": 1.0014,
+      "step": 823
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6531064704098668,
+      "learning_rate": 1.915126659249371e-05,
+      "loss": 0.9772,
+      "step": 824
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7515549769775075,
+      "learning_rate": 1.9148752748009304e-05,
+      "loss": 1.0978,
+      "step": 825
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7454973210719319,
+      "learning_rate": 1.914623535166408e-05,
+      "loss": 1.0117,
+      "step": 826
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.605904186309522,
+      "learning_rate": 1.9143714404435382e-05,
+      "loss": 1.0452,
+      "step": 827
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6803835422182306,
+      "learning_rate": 1.9141189907301922e-05,
+      "loss": 1.0611,
+      "step": 828
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7336139379043911,
+      "learning_rate": 1.9138661861243802e-05,
+      "loss": 1.187,
+      "step": 829
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7465475738118239,
+      "learning_rate": 1.913613026724249e-05,
+      "loss": 1.0989,
+      "step": 830
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7225858181869839,
+      "learning_rate": 1.9133595126280848e-05,
+      "loss": 1.0371,
+      "step": 831
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8004399473525721,
+      "learning_rate": 1.9131056439343095e-05,
+      "loss": 1.1261,
+      "step": 832
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7978880754762557,
+      "learning_rate": 1.9128514207414838e-05,
+      "loss": 1.0939,
+      "step": 833
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.742862848656816,
+      "learning_rate": 1.9125968431483068e-05,
+      "loss": 1.1729,
+      "step": 834
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6322567622271626,
+      "learning_rate": 1.9123419112536132e-05,
+      "loss": 1.2133,
+      "step": 835
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6755404520658844,
+      "learning_rate": 1.912086625156377e-05,
+      "loss": 1.1443,
+      "step": 836
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5956049260086979,
+      "learning_rate": 1.911830984955709e-05,
+      "loss": 1.1262,
+      "step": 837
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7000930646403908,
+      "learning_rate": 1.911574990750857e-05,
+      "loss": 1.1306,
+      "step": 838
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6753926191153353,
+      "learning_rate": 1.9113186426412073e-05,
+      "loss": 1.106,
+      "step": 839
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7124364604255603,
+      "learning_rate": 1.9110619407262828e-05,
+      "loss": 1.0543,
+      "step": 840
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7485028515377586,
+      "learning_rate": 1.9108048851057447e-05,
+      "loss": 1.1301,
+      "step": 841
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.677210556504978,
+      "learning_rate": 1.9105474758793897e-05,
+      "loss": 1.0944,
+      "step": 842
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6837724692747091,
+      "learning_rate": 1.9102897131471536e-05,
+      "loss": 1.074,
+      "step": 843
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6640197237152033,
+      "learning_rate": 1.9100315970091088e-05,
+      "loss": 1.0551,
+      "step": 844
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.749902540474008,
+      "learning_rate": 1.9097731275654645e-05,
+      "loss": 1.16,
+      "step": 845
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6350813868085061,
+      "learning_rate": 1.909514304916568e-05,
+      "loss": 1.1505,
+      "step": 846
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.694173200182658,
+      "learning_rate": 1.9092551291629026e-05,
+      "loss": 1.1493,
+      "step": 847
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7529504440484347,
+      "learning_rate": 1.9089956004050893e-05,
+      "loss": 1.0676,
+      "step": 848
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7419006433113979,
+      "learning_rate": 1.908735718743887e-05,
+      "loss": 1.1017,
+      "step": 849
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7981752349097222,
+      "learning_rate": 1.908475484280189e-05,
+      "loss": 1.0565,
+      "step": 850
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7850410286828019,
+      "learning_rate": 1.908214897115029e-05,
+      "loss": 1.1178,
+      "step": 851
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6660531442740223,
+      "learning_rate": 1.907953957349575e-05,
+      "loss": 1.062,
+      "step": 852
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6501178088508476,
+      "learning_rate": 1.907692665085133e-05,
+      "loss": 1.0949,
+      "step": 853
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7204528745135828,
+      "learning_rate": 1.9074310204231457e-05,
+      "loss": 1.0598,
+      "step": 854
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6630470838991366,
+      "learning_rate": 1.9071690234651923e-05,
+      "loss": 1.0562,
+      "step": 855
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7717265164936367,
+      "learning_rate": 1.9069066743129893e-05,
+      "loss": 1.0692,
+      "step": 856
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7781642509906178,
+      "learning_rate": 1.90664397306839e-05,
+      "loss": 1.1164,
+      "step": 857
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7321409160429464,
+      "learning_rate": 1.9063809198333832e-05,
+      "loss": 1.0795,
+      "step": 858
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6889219228638164,
+      "learning_rate": 1.9061175147100957e-05,
+      "loss": 1.1177,
+      "step": 859
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6777699591387488,
+      "learning_rate": 1.905853757800791e-05,
+      "loss": 1.0887,
+      "step": 860
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7057673288841412,
+      "learning_rate": 1.9055896492078675e-05,
+      "loss": 1.0997,
+      "step": 861
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6947892843669026,
+      "learning_rate": 1.905325189033862e-05,
+      "loss": 1.0892,
+      "step": 862
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7598278870711114,
+      "learning_rate": 1.905060377381447e-05,
+      "loss": 1.1273,
+      "step": 863
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6990635122807672,
+      "learning_rate": 1.904795214353431e-05,
+      "loss": 1.0262,
+      "step": 864
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6901389836542378,
+      "learning_rate": 1.90452970005276e-05,
+      "loss": 1.0921,
+      "step": 865
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6443453022790674,
+      "learning_rate": 1.9042638345825155e-05,
+      "loss": 1.0366,
+      "step": 866
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6773693463575284,
+      "learning_rate": 1.9039976180459158e-05,
+      "loss": 1.0901,
+      "step": 867
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7134047642525744,
+      "learning_rate": 1.9037310505463153e-05,
+      "loss": 1.1041,
+      "step": 868
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6801616937252816,
+      "learning_rate": 1.9034641321872043e-05,
+      "loss": 1.1238,
+      "step": 869
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4144998869093647,
+      "learning_rate": 1.9031968630722104e-05,
+      "loss": 1.0295,
+      "step": 870
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7489632953794053,
+      "learning_rate": 1.902929243305096e-05,
+      "loss": 1.0649,
+      "step": 871
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7272635807063298,
+      "learning_rate": 1.902661272989761e-05,
+      "loss": 1.0967,
+      "step": 872
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7474370833400744,
+      "learning_rate": 1.9023929522302394e-05,
+      "loss": 1.1085,
+      "step": 873
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6431200749918418,
+      "learning_rate": 1.9021242811307044e-05,
+      "loss": 1.06,
+      "step": 874
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.700105010740571,
+      "learning_rate": 1.901855259795462e-05,
+      "loss": 1.1,
+      "step": 875
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.686798548650089,
+      "learning_rate": 1.9015858883289556e-05,
+      "loss": 1.021,
+      "step": 876
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7033669756891047,
+      "learning_rate": 1.9013161668357655e-05,
+      "loss": 1.1024,
+      "step": 877
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7441733936892874,
+      "learning_rate": 1.901046095420606e-05,
+      "loss": 1.1024,
+      "step": 878
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6381445803748899,
+      "learning_rate": 1.9007756741883284e-05,
+      "loss": 1.1157,
+      "step": 879
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6534669630108609,
+      "learning_rate": 1.9005049032439193e-05,
+      "loss": 1.0447,
+      "step": 880
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7419422497835401,
+      "learning_rate": 1.9002337826925012e-05,
+      "loss": 1.0881,
+      "step": 881
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6994434068826686,
+      "learning_rate": 1.899962312639333e-05,
+      "loss": 1.0323,
+      "step": 882
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5605110845559039,
+      "learning_rate": 1.8996904931898085e-05,
+      "loss": 1.1216,
+      "step": 883
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6686737024190771,
+      "learning_rate": 1.899418324449457e-05,
+      "loss": 1.0782,
+      "step": 884
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5429137318011835,
+      "learning_rate": 1.8991458065239444e-05,
+      "loss": 1.1125,
+      "step": 885
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5115266051608318,
+      "learning_rate": 1.8988729395190712e-05,
+      "loss": 1.0724,
+      "step": 886
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6519554824820927,
+      "learning_rate": 1.8985997235407735e-05,
+      "loss": 1.0646,
+      "step": 887
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7287748691971911,
+      "learning_rate": 1.898326158695124e-05,
+      "loss": 1.0847,
+      "step": 888
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6806006791675459,
+      "learning_rate": 1.8980522450883287e-05,
+      "loss": 1.016,
+      "step": 889
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7358781494611323,
+      "learning_rate": 1.8977779828267314e-05,
+      "loss": 1.0684,
+      "step": 890
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6701359043320417,
+      "learning_rate": 1.8975033720168094e-05,
+      "loss": 1.1558,
+      "step": 891
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6807105440265613,
+      "learning_rate": 1.897228412765177e-05,
+      "loss": 1.0864,
+      "step": 892
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7591972478540074,
+      "learning_rate": 1.896953105178582e-05,
+      "loss": 0.9972,
+      "step": 893
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6878937242866263,
+      "learning_rate": 1.8966774493639084e-05,
+      "loss": 1.0789,
+      "step": 894
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7078862713737236,
+      "learning_rate": 1.896401445428176e-05,
+      "loss": 1.1121,
+      "step": 895
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.722595178209212,
+      "learning_rate": 1.896125093478538e-05,
+      "loss": 1.0983,
+      "step": 896
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7595891558780771,
+      "learning_rate": 1.895848393622284e-05,
+      "loss": 1.0232,
+      "step": 897
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6670173651237444,
+      "learning_rate": 1.895571345966839e-05,
+      "loss": 1.0514,
+      "step": 898
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7456509511251023,
+      "learning_rate": 1.8952939506197622e-05,
+      "loss": 1.0738,
+      "step": 899
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.782500668067437,
+      "learning_rate": 1.8950162076887477e-05,
+      "loss": 1.0972,
+      "step": 900
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7698732421644873,
+      "learning_rate": 1.894738117281625e-05,
+      "loss": 1.1011,
+      "step": 901
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6593876645199647,
+      "learning_rate": 1.8944596795063584e-05,
+      "loss": 1.0668,
+      "step": 902
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7277187776303466,
+      "learning_rate": 1.894180894471047e-05,
+      "loss": 1.1081,
+      "step": 903
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.589038682010719,
+      "learning_rate": 1.8939017622839253e-05,
+      "loss": 1.0707,
+      "step": 904
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7248038319860728,
+      "learning_rate": 1.8936222830533613e-05,
+      "loss": 1.0884,
+      "step": 905
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6844859667468064,
+      "learning_rate": 1.8933424568878586e-05,
+      "loss": 1.1125,
+      "step": 906
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7246685708665126,
+      "learning_rate": 1.8930622838960555e-05,
+      "loss": 1.1201,
+      "step": 907
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6787605037916113,
+      "learning_rate": 1.8927817641867244e-05,
+      "loss": 1.1145,
+      "step": 908
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6810040675585096,
+      "learning_rate": 1.8925008978687737e-05,
+      "loss": 1.0923,
+      "step": 909
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6817106882295851,
+      "learning_rate": 1.8922196850512446e-05,
+      "loss": 1.1317,
+      "step": 910
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6628758292376109,
+      "learning_rate": 1.8919381258433135e-05,
+      "loss": 1.1534,
+      "step": 911
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.701952132584909,
+      "learning_rate": 1.8916562203542916e-05,
+      "loss": 1.0613,
+      "step": 912
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7223444217311599,
+      "learning_rate": 1.8913739686936244e-05,
+      "loss": 1.0635,
+      "step": 913
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7186138949937049,
+      "learning_rate": 1.8910913709708918e-05,
+      "loss": 1.0668,
+      "step": 914
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7181167671150747,
+      "learning_rate": 1.8908084272958077e-05,
+      "loss": 1.1103,
+      "step": 915
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7731825410756595,
+      "learning_rate": 1.8905251377782206e-05,
+      "loss": 1.1362,
+      "step": 916
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7181305918137796,
+      "learning_rate": 1.8902415025281136e-05,
+      "loss": 1.126,
+      "step": 917
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7141210481676854,
+      "learning_rate": 1.889957521655603e-05,
+      "loss": 1.0312,
+      "step": 918
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6616853970065444,
+      "learning_rate": 1.8896731952709408e-05,
+      "loss": 1.0325,
+      "step": 919
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6975922776453183,
+      "learning_rate": 1.8893885234845117e-05,
+      "loss": 1.1313,
+      "step": 920
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7235332070478706,
+      "learning_rate": 1.8891035064068354e-05,
+      "loss": 1.1925,
+      "step": 921
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7428816565811989,
+      "learning_rate": 1.888818144148565e-05,
+      "loss": 0.9901,
+      "step": 922
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7855548962773329,
+      "learning_rate": 1.888532436820488e-05,
+      "loss": 1.1185,
+      "step": 923
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6871316096383309,
+      "learning_rate": 1.8882463845335263e-05,
+      "loss": 1.1018,
+      "step": 924
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6235665957281616,
+      "learning_rate": 1.8879599873987343e-05,
+      "loss": 1.0365,
+      "step": 925
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6622781720588147,
+      "learning_rate": 1.8876732455273022e-05,
+      "loss": 1.0734,
+      "step": 926
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6582944479358258,
+      "learning_rate": 1.8873861590305527e-05,
+      "loss": 1.0739,
+      "step": 927
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7454199596089508,
+      "learning_rate": 1.8870987280199428e-05,
+      "loss": 1.0801,
+      "step": 928
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5712819415773454,
+      "learning_rate": 1.886810952607063e-05,
+      "loss": 1.0832,
+      "step": 929
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7240321173779786,
+      "learning_rate": 1.8865228329036372e-05,
+      "loss": 1.0713,
+      "step": 930
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6878914684693027,
+      "learning_rate": 1.886234369021524e-05,
+      "loss": 1.0672,
+      "step": 931
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.738762755347153,
+      "learning_rate": 1.885945561072715e-05,
+      "loss": 1.1235,
+      "step": 932
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7553447937249067,
+      "learning_rate": 1.885656409169335e-05,
+      "loss": 1.0107,
+      "step": 933
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5519204516395412,
+      "learning_rate": 1.885366913423643e-05,
+      "loss": 1.0765,
+      "step": 934
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6711457883711537,
+      "learning_rate": 1.8850770739480312e-05,
+      "loss": 1.0729,
+      "step": 935
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6781589383368517,
+      "learning_rate": 1.8847868908550252e-05,
+      "loss": 1.0819,
+      "step": 936
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6868838581291171,
+      "learning_rate": 1.8844963642572837e-05,
+      "loss": 1.0995,
+      "step": 937
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5872280069518147,
+      "learning_rate": 1.8842054942676e-05,
+      "loss": 1.0805,
+      "step": 938
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6571576421411348,
+      "learning_rate": 1.8839142809988987e-05,
+      "loss": 1.0986,
+      "step": 939
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5435289674993368,
+      "learning_rate": 1.88362272456424e-05,
+      "loss": 1.0638,
+      "step": 940
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5672937641319316,
+      "learning_rate": 1.8833308250768153e-05,
+      "loss": 1.1429,
+      "step": 941
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7002696575961184,
+      "learning_rate": 1.8830385826499507e-05,
+      "loss": 1.1088,
+      "step": 942
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.694210628524064,
+      "learning_rate": 1.882745997397104e-05,
+      "loss": 1.0713,
+      "step": 943
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.723352235069926,
+      "learning_rate": 1.8824530694318675e-05,
+      "loss": 1.0497,
+      "step": 944
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7307942394668275,
+      "learning_rate": 1.882159798867966e-05,
+      "loss": 1.063,
+      "step": 945
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7041603413031586,
+      "learning_rate": 1.8818661858192562e-05,
+      "loss": 1.1208,
+      "step": 946
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6626606704846895,
+      "learning_rate": 1.88157223039973e-05,
+      "loss": 1.0523,
+      "step": 947
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.554827781376577,
+      "learning_rate": 1.8812779327235106e-05,
+      "loss": 1.1309,
+      "step": 948
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6505330417865919,
+      "learning_rate": 1.880983292904854e-05,
+      "loss": 1.0959,
+      "step": 949
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7274270269378772,
+      "learning_rate": 1.88068831105815e-05,
+      "loss": 1.0873,
+      "step": 950
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6940592104050283,
+      "learning_rate": 1.8803929872979214e-05,
+      "loss": 1.1418,
+      "step": 951
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6738593248489014,
+      "learning_rate": 1.8800973217388215e-05,
+      "loss": 1.1626,
+      "step": 952
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6915574472340658,
+      "learning_rate": 1.879801314495639e-05,
+      "loss": 1.1382,
+      "step": 953
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7032109469373559,
+      "learning_rate": 1.879504965683294e-05,
+      "loss": 1.1019,
+      "step": 954
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6755709333673662,
+      "learning_rate": 1.8792082754168385e-05,
+      "loss": 1.0454,
+      "step": 955
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8460673597655551,
+      "learning_rate": 1.878911243811459e-05,
+      "loss": 1.1082,
+      "step": 956
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.695097530026736,
+      "learning_rate": 1.8786138709824726e-05,
+      "loss": 1.054,
+      "step": 957
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7144888882895353,
+      "learning_rate": 1.8783161570453295e-05,
+      "loss": 1.0386,
+      "step": 958
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6930868378316069,
+      "learning_rate": 1.878018102115614e-05,
+      "loss": 1.0539,
+      "step": 959
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6726136540919132,
+      "learning_rate": 1.8777197063090394e-05,
+      "loss": 1.0997,
+      "step": 960
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5439068068722722,
+      "learning_rate": 1.877420969741454e-05,
+      "loss": 1.1362,
+      "step": 961
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7160729860978975,
+      "learning_rate": 1.877121892528838e-05,
+      "loss": 1.0377,
+      "step": 962
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7265577160591808,
+      "learning_rate": 1.876822474787303e-05,
+      "loss": 1.1358,
+      "step": 963
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8003576575005957,
+      "learning_rate": 1.8765227166330933e-05,
+      "loss": 1.0789,
+      "step": 964
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6660501666053384,
+      "learning_rate": 1.8762226181825857e-05,
+      "loss": 1.1345,
+      "step": 965
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5473001275218682,
+      "learning_rate": 1.875922179552288e-05,
+      "loss": 1.1498,
+      "step": 966
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7268692616054572,
+      "learning_rate": 1.875621400858842e-05,
+      "loss": 1.0763,
+      "step": 967
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5308912720932011,
+      "learning_rate": 1.875320282219019e-05,
+      "loss": 1.0572,
+      "step": 968
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7608306195486156,
+      "learning_rate": 1.8750188237497247e-05,
+      "loss": 1.099,
+      "step": 969
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5534209773278882,
+      "learning_rate": 1.874717025567995e-05,
+      "loss": 1.1224,
+      "step": 970
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7778551330128594,
+      "learning_rate": 1.874414887790999e-05,
+      "loss": 1.0653,
+      "step": 971
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.635799824589882,
+      "learning_rate": 1.8741124105360363e-05,
+      "loss": 1.0944,
+      "step": 972
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6937741761697317,
+      "learning_rate": 1.873809593920539e-05,
+      "loss": 1.1372,
+      "step": 973
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6636378492629963,
+      "learning_rate": 1.8735064380620717e-05,
+      "loss": 1.0915,
+      "step": 974
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5594643483730036,
+      "learning_rate": 1.873202943078329e-05,
+      "loss": 1.1639,
+      "step": 975
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7283103562037442,
+      "learning_rate": 1.8728991090871387e-05,
+      "loss": 1.06,
+      "step": 976
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7182160367860503,
+      "learning_rate": 1.8725949362064596e-05,
+      "loss": 1.0857,
+      "step": 977
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7441333061378663,
+      "learning_rate": 1.8722904245543817e-05,
+      "loss": 1.0784,
+      "step": 978
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.644655154075292,
+      "learning_rate": 1.871985574249127e-05,
+      "loss": 1.1124,
+      "step": 979
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7075844541817132,
+      "learning_rate": 1.8716803854090495e-05,
+      "loss": 1.0064,
+      "step": 980
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7717008205712617,
+      "learning_rate": 1.8713748581526334e-05,
+      "loss": 1.0988,
+      "step": 981
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6913477093170859,
+      "learning_rate": 1.871068992598495e-05,
+      "loss": 1.1194,
+      "step": 982
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6076360739193601,
+      "learning_rate": 1.8707627888653816e-05,
+      "loss": 1.1182,
+      "step": 983
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6558502051649092,
+      "learning_rate": 1.8704562470721728e-05,
+      "loss": 1.0189,
+      "step": 984
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6453914148830651,
+      "learning_rate": 1.870149367337878e-05,
+      "loss": 1.1737,
+      "step": 985
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5807416170678041,
+      "learning_rate": 1.8698421497816386e-05,
+      "loss": 1.1906,
+      "step": 986
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7195113409731955,
+      "learning_rate": 1.869534594522727e-05,
+      "loss": 1.1503,
+      "step": 987
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7008166781362775,
+      "learning_rate": 1.8692267016805473e-05,
+      "loss": 1.0795,
+      "step": 988
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6233098848099466,
+      "learning_rate": 1.8689184713746333e-05,
+      "loss": 1.0762,
+      "step": 989
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.721106574506225,
+      "learning_rate": 1.868609903724651e-05,
+      "loss": 1.0623,
+      "step": 990
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7740944843907716,
+      "learning_rate": 1.8683009988503972e-05,
+      "loss": 1.1507,
+      "step": 991
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6625699776943638,
+      "learning_rate": 1.867991756871799e-05,
+      "loss": 1.0506,
+      "step": 992
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5399126720232642,
+      "learning_rate": 1.867682177908915e-05,
+      "loss": 1.1198,
+      "step": 993
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6396682297101186,
+      "learning_rate": 1.867372262081934e-05,
+      "loss": 1.0778,
+      "step": 994
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6790692634378462,
+      "learning_rate": 1.8670620095111766e-05,
+      "loss": 1.0854,
+      "step": 995
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6615925663597524,
+      "learning_rate": 1.8667514203170934e-05,
+      "loss": 1.0552,
+      "step": 996
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6398526125475367,
+      "learning_rate": 1.8664404946202658e-05,
+      "loss": 0.9744,
+      "step": 997
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.649894775202031,
+      "learning_rate": 1.8661292325414058e-05,
+      "loss": 1.1217,
+      "step": 998
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7438754826321575,
+      "learning_rate": 1.865817634201356e-05,
+      "loss": 1.1112,
+      "step": 999
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7164802523119754,
+      "learning_rate": 1.8655056997210893e-05,
+      "loss": 1.084,
+      "step": 1000
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7037939087523987,
+      "learning_rate": 1.8651934292217097e-05,
+      "loss": 1.06,
+      "step": 1001
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6769238026517229,
+      "learning_rate": 1.864880822824452e-05,
+      "loss": 1.0267,
+      "step": 1002
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.35138460953932676,
+      "learning_rate": 1.8645678806506795e-05,
+      "loss": 1.0418,
+      "step": 1003
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.694603579302971,
+      "learning_rate": 1.864254602821888e-05,
+      "loss": 1.0697,
+      "step": 1004
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7501605096499845,
+      "learning_rate": 1.8639409894597026e-05,
+      "loss": 1.0612,
+      "step": 1005
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6901459174028317,
+      "learning_rate": 1.8636270406858786e-05,
+      "loss": 1.1615,
+      "step": 1006
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6646102133296238,
+      "learning_rate": 1.8633127566223023e-05,
+      "loss": 1.1109,
+      "step": 1007
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7080974229061955,
+      "learning_rate": 1.862998137390989e-05,
+      "loss": 1.1433,
+      "step": 1008
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7528621073578877,
+      "learning_rate": 1.8626831831140845e-05,
+      "loss": 1.0327,
+      "step": 1009
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6209029716588604,
+      "learning_rate": 1.8623678939138652e-05,
+      "loss": 1.0385,
+      "step": 1010
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.819058877690145,
+      "learning_rate": 1.8620522699127374e-05,
+      "loss": 0.9572,
+      "step": 1011
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7252871559957876,
+      "learning_rate": 1.8617363112332376e-05,
+      "loss": 1.0912,
+      "step": 1012
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.738378371219206,
+      "learning_rate": 1.8614200179980307e-05,
+      "loss": 1.0858,
+      "step": 1013
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8027254465341964,
+      "learning_rate": 1.8611033903299136e-05,
+      "loss": 1.0292,
+      "step": 1014
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5893333281880536,
+      "learning_rate": 1.8607864283518116e-05,
+      "loss": 1.1329,
+      "step": 1015
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7453079988171911,
+      "learning_rate": 1.8604691321867804e-05,
+      "loss": 1.1057,
+      "step": 1016
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6746370492115057,
+      "learning_rate": 1.8601515019580053e-05,
+      "loss": 1.1304,
+      "step": 1017
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6798156988256393,
+      "learning_rate": 1.8598335377888012e-05,
+      "loss": 1.0621,
+      "step": 1018
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7009243423003154,
+      "learning_rate": 1.8595152398026128e-05,
+      "loss": 1.0702,
+      "step": 1019
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7755868527579136,
+      "learning_rate": 1.8591966081230142e-05,
+      "loss": 1.0678,
+      "step": 1020
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8097134756332219,
+      "learning_rate": 1.8588776428737095e-05,
+      "loss": 0.9743,
+      "step": 1021
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7345659319563166,
+      "learning_rate": 1.858558344178532e-05,
+      "loss": 1.0874,
+      "step": 1022
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7162533985199239,
+      "learning_rate": 1.8582387121614437e-05,
+      "loss": 1.0465,
+      "step": 1023
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5466035084609447,
+      "learning_rate": 1.857918746946538e-05,
+      "loss": 1.1197,
+      "step": 1024
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6959024548265437,
+      "learning_rate": 1.8575984486580353e-05,
+      "loss": 1.0719,
+      "step": 1025
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7288389887337298,
+      "learning_rate": 1.857277817420287e-05,
+      "loss": 1.1556,
+      "step": 1026
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6630409330477112,
+      "learning_rate": 1.8569568533577727e-05,
+      "loss": 1.0275,
+      "step": 1027
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6861103053343653,
+      "learning_rate": 1.8566355565951023e-05,
+      "loss": 1.0609,
+      "step": 1028
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6943941584744144,
+      "learning_rate": 1.8563139272570142e-05,
+      "loss": 1.0389,
+      "step": 1029
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6951366829498306,
+      "learning_rate": 1.8559919654683756e-05,
+      "loss": 1.0854,
+      "step": 1030
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6366697991847831,
+      "learning_rate": 1.8556696713541833e-05,
+      "loss": 1.1117,
+      "step": 1031
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7041481241099057,
+      "learning_rate": 1.855347045039563e-05,
+      "loss": 1.1229,
+      "step": 1032
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6891274948784658,
+      "learning_rate": 1.8550240866497697e-05,
+      "loss": 1.0771,
+      "step": 1033
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7409265984624515,
+      "learning_rate": 1.854700796310186e-05,
+      "loss": 0.988,
+      "step": 1034
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6797284828407082,
+      "learning_rate": 1.8543771741463254e-05,
+      "loss": 1.0667,
+      "step": 1035
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.636515303057423,
+      "learning_rate": 1.8540532202838286e-05,
+      "loss": 1.1188,
+      "step": 1036
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7579868499314423,
+      "learning_rate": 1.8537289348484658e-05,
+      "loss": 1.0612,
+      "step": 1037
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6970015527698808,
+      "learning_rate": 1.8534043179661357e-05,
+      "loss": 1.131,
+      "step": 1038
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7752044622345973,
+      "learning_rate": 1.8530793697628658e-05,
+      "loss": 1.0967,
+      "step": 1039
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7414380006861955,
+      "learning_rate": 1.8527540903648122e-05,
+      "loss": 1.0146,
+      "step": 1040
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6882582954676307,
+      "learning_rate": 1.8524284798982595e-05,
+      "loss": 1.0438,
+      "step": 1041
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5973451359149509,
+      "learning_rate": 1.852102538489621e-05,
+      "loss": 1.1094,
+      "step": 1042
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7683255724944688,
+      "learning_rate": 1.8517762662654383e-05,
+      "loss": 1.1087,
+      "step": 1043
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6352900444808471,
+      "learning_rate": 1.851449663352381e-05,
+      "loss": 1.1061,
+      "step": 1044
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5348464283647189,
+      "learning_rate": 1.851122729877249e-05,
+      "loss": 1.0184,
+      "step": 1045
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7034344058520106,
+      "learning_rate": 1.8507954659669677e-05,
+      "loss": 1.0396,
+      "step": 1046
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.49456349795903803,
+      "learning_rate": 1.850467871748593e-05,
+      "loss": 1.1075,
+      "step": 1047
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6918742217958862,
+      "learning_rate": 1.850139947349308e-05,
+      "loss": 1.0855,
+      "step": 1048
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6685376961031991,
+      "learning_rate": 1.8498116928964244e-05,
+      "loss": 1.0853,
+      "step": 1049
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6869142168843096,
+      "learning_rate": 1.849483108517381e-05,
+      "loss": 1.0829,
+      "step": 1050
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.772718701548039,
+      "learning_rate": 1.849154194339747e-05,
+      "loss": 1.029,
+      "step": 1051
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6993281991802945,
+      "learning_rate": 1.8488249504912173e-05,
+      "loss": 1.114,
+      "step": 1052
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7940106014237412,
+      "learning_rate": 1.8484953770996163e-05,
+      "loss": 1.0413,
+      "step": 1053
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6476292198197371,
+      "learning_rate": 1.848165474292895e-05,
+      "loss": 1.0516,
+      "step": 1054
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7185603740231584,
+      "learning_rate": 1.8478352421991334e-05,
+      "loss": 0.9989,
+      "step": 1055
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7364308042063992,
+      "learning_rate": 1.847504680946539e-05,
+      "loss": 1.1285,
+      "step": 1056
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8799775862390946,
+      "learning_rate": 1.847173790663447e-05,
+      "loss": 1.1013,
+      "step": 1057
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6757574124989966,
+      "learning_rate": 1.8468425714783206e-05,
+      "loss": 1.0314,
+      "step": 1058
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5600069723660274,
+      "learning_rate": 1.84651102351975e-05,
+      "loss": 1.126,
+      "step": 1059
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6950983188644955,
+      "learning_rate": 1.846179146916454e-05,
+      "loss": 1.0352,
+      "step": 1060
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.559236268558715,
+      "learning_rate": 1.8458469417972783e-05,
+      "loss": 1.1155,
+      "step": 1061
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.673675294384761,
+      "learning_rate": 1.8455144082911965e-05,
+      "loss": 1.02,
+      "step": 1062
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7154648053970941,
+      "learning_rate": 1.8451815465273097e-05,
+      "loss": 1.0362,
+      "step": 1063
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7535682993016685,
+      "learning_rate": 1.8448483566348456e-05,
+      "loss": 1.0515,
+      "step": 1064
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6965444438685872,
+      "learning_rate": 1.8445148387431605e-05,
+      "loss": 1.0993,
+      "step": 1065
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7151080116881212,
+      "learning_rate": 1.8441809929817382e-05,
+      "loss": 1.0186,
+      "step": 1066
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7325065704245489,
+      "learning_rate": 1.8438468194801876e-05,
+      "loss": 1.1154,
+      "step": 1067
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7704862280383817,
+      "learning_rate": 1.8435123183682475e-05,
+      "loss": 1.069,
+      "step": 1068
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5394824641358492,
+      "learning_rate": 1.8431774897757824e-05,
+      "loss": 1.1188,
+      "step": 1069
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6991685590986054,
+      "learning_rate": 1.8428423338327847e-05,
+      "loss": 1.1127,
+      "step": 1070
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7234225168699902,
+      "learning_rate": 1.8425068506693727e-05,
+      "loss": 1.114,
+      "step": 1071
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6711398720984446,
+      "learning_rate": 1.842171040415793e-05,
+      "loss": 1.0876,
+      "step": 1072
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6472183218398868,
+      "learning_rate": 1.8418349032024185e-05,
+      "loss": 1.059,
+      "step": 1073
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6372835376515071,
+      "learning_rate": 1.8414984391597492e-05,
+      "loss": 1.0582,
+      "step": 1074
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7095899319401521,
+      "learning_rate": 1.8411616484184126e-05,
+      "loss": 1.1502,
+      "step": 1075
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6594904823169613,
+      "learning_rate": 1.8408245311091618e-05,
+      "loss": 1.0082,
+      "step": 1076
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7388210911554983,
+      "learning_rate": 1.8404870873628774e-05,
+      "loss": 1.113,
+      "step": 1077
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5172545275317738,
+      "learning_rate": 1.8401493173105675e-05,
+      "loss": 1.0405,
+      "step": 1078
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7928075397167125,
+      "learning_rate": 1.8398112210833648e-05,
+      "loss": 1.1497,
+      "step": 1079
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6762976776647012,
+      "learning_rate": 1.8394727988125308e-05,
+      "loss": 1.0889,
+      "step": 1080
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6942125319938758,
+      "learning_rate": 1.8391340506294524e-05,
+      "loss": 1.0842,
+      "step": 1081
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7724809361788191,
+      "learning_rate": 1.8387949766656434e-05,
+      "loss": 1.124,
+      "step": 1082
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6254146245937797,
+      "learning_rate": 1.8384555770527438e-05,
+      "loss": 1.0692,
+      "step": 1083
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6250987059555672,
+      "learning_rate": 1.8381158519225204e-05,
+      "loss": 1.1146,
+      "step": 1084
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7108269696294536,
+      "learning_rate": 1.8377758014068662e-05,
+      "loss": 1.1047,
+      "step": 1085
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.749202503415426,
+      "learning_rate": 1.8374354256378e-05,
+      "loss": 1.0386,
+      "step": 1086
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7502373394754928,
+      "learning_rate": 1.837094724747468e-05,
+      "loss": 0.9974,
+      "step": 1087
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6004486836215769,
+      "learning_rate": 1.8367536988681422e-05,
+      "loss": 1.1137,
+      "step": 1088
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6762235620583023,
+      "learning_rate": 1.83641234813222e-05,
+      "loss": 1.0675,
+      "step": 1089
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.663849945684039,
+      "learning_rate": 1.8360706726722253e-05,
+      "loss": 1.1492,
+      "step": 1090
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5377803225828871,
+      "learning_rate": 1.835728672620809e-05,
+      "loss": 1.1219,
+      "step": 1091
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7270258840640702,
+      "learning_rate": 1.8353863481107473e-05,
+      "loss": 1.1138,
+      "step": 1092
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6563000428402276,
+      "learning_rate": 1.835043699274942e-05,
+      "loss": 1.021,
+      "step": 1093
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6502900513405404,
+      "learning_rate": 1.8347007262464206e-05,
+      "loss": 1.0919,
+      "step": 1094
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.661801837490177,
+      "learning_rate": 1.8343574291583385e-05,
+      "loss": 1.0524,
+      "step": 1095
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6396147251055392,
+      "learning_rate": 1.8340138081439743e-05,
+      "loss": 1.077,
+      "step": 1096
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.611920996901279,
+      "learning_rate": 1.833669863336734e-05,
+      "loss": 1.1141,
+      "step": 1097
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.633085402000444,
+      "learning_rate": 1.833325594870148e-05,
+      "loss": 1.0821,
+      "step": 1098
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6632549558882758,
+      "learning_rate": 1.8329810028778747e-05,
+      "loss": 1.1064,
+      "step": 1099
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6789159618391225,
+      "learning_rate": 1.8326360874936952e-05,
+      "loss": 1.0511,
+      "step": 1100
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6983232442610912,
+      "learning_rate": 1.8322908488515182e-05,
+      "loss": 1.0815,
+      "step": 1101
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6630123731222816,
+      "learning_rate": 1.8319452870853772e-05,
+      "loss": 1.0126,
+      "step": 1102
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7851308981281575,
+      "learning_rate": 1.8315994023294306e-05,
+      "loss": 1.0758,
+      "step": 1103
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6991713845040827,
+      "learning_rate": 1.8312531947179634e-05,
+      "loss": 1.0427,
+      "step": 1104
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6730332663534416,
+      "learning_rate": 1.8309066643853854e-05,
+      "loss": 1.1017,
+      "step": 1105
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7249433729288663,
+      "learning_rate": 1.8305598114662312e-05,
+      "loss": 1.0403,
+      "step": 1106
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7061010907484814,
+      "learning_rate": 1.830212636095161e-05,
+      "loss": 1.0906,
+      "step": 1107
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6644898245183641,
+      "learning_rate": 1.8298651384069605e-05,
+      "loss": 1.0327,
+      "step": 1108
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6951558219492933,
+      "learning_rate": 1.8295173185365405e-05,
+      "loss": 1.0297,
+      "step": 1109
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6439686113434747,
+      "learning_rate": 1.829169176618936e-05,
+      "loss": 1.0281,
+      "step": 1110
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6972989610013408,
+      "learning_rate": 1.828820712789308e-05,
+      "loss": 1.0508,
+      "step": 1111
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7445598903209643,
+      "learning_rate": 1.828471927182942e-05,
+      "loss": 1.1514,
+      "step": 1112
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6067092517850581,
+      "learning_rate": 1.828122819935249e-05,
+      "loss": 1.1541,
+      "step": 1113
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7376021347296517,
+      "learning_rate": 1.8277733911817642e-05,
+      "loss": 1.0658,
+      "step": 1114
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6338605246728828,
+      "learning_rate": 1.8274236410581478e-05,
+      "loss": 1.0024,
+      "step": 1115
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.694453669758523,
+      "learning_rate": 1.827073569700185e-05,
+      "loss": 1.0729,
+      "step": 1116
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5371198511496511,
+      "learning_rate": 1.8267231772437854e-05,
+      "loss": 1.1084,
+      "step": 1117
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7004708340282615,
+      "learning_rate": 1.8263724638249834e-05,
+      "loss": 1.0964,
+      "step": 1118
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7579185077624004,
+      "learning_rate": 1.8260214295799382e-05,
+      "loss": 1.0662,
+      "step": 1119
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7422202263557931,
+      "learning_rate": 1.825670074644933e-05,
+      "loss": 1.027,
+      "step": 1120
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7222911443383554,
+      "learning_rate": 1.8253183991563768e-05,
+      "loss": 1.0032,
+      "step": 1121
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7402094597572579,
+      "learning_rate": 1.824966403250801e-05,
+      "loss": 1.0767,
+      "step": 1122
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6774983170685367,
+      "learning_rate": 1.8246140870648633e-05,
+      "loss": 1.0247,
+      "step": 1123
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6943567828543226,
+      "learning_rate": 1.8242614507353446e-05,
+      "loss": 1.1555,
+      "step": 1124
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6415903756452486,
+      "learning_rate": 1.8239084943991507e-05,
+      "loss": 1.0575,
+      "step": 1125
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7081809771695802,
+      "learning_rate": 1.823555218193311e-05,
+      "loss": 1.0861,
+      "step": 1126
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7250573563192066,
+      "learning_rate": 1.8232016222549797e-05,
+      "loss": 1.0646,
+      "step": 1127
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.792464307484445,
+      "learning_rate": 1.8228477067214352e-05,
+      "loss": 1.1575,
+      "step": 1128
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.701577286145075,
+      "learning_rate": 1.8224934717300794e-05,
+      "loss": 1.0314,
+      "step": 1129
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5599580005234396,
+      "learning_rate": 1.8221389174184385e-05,
+      "loss": 1.1508,
+      "step": 1130
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6517035597575832,
+      "learning_rate": 1.8217840439241633e-05,
+      "loss": 1.0355,
+      "step": 1131
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7446595682203213,
+      "learning_rate": 1.8214288513850267e-05,
+      "loss": 1.0767,
+      "step": 1132
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6668115095430587,
+      "learning_rate": 1.8210733399389277e-05,
+      "loss": 1.0974,
+      "step": 1133
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6778298187092117,
+      "learning_rate": 1.820717509723888e-05,
+      "loss": 0.9113,
+      "step": 1134
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.716972844869181,
+      "learning_rate": 1.8203613608780525e-05,
+      "loss": 1.027,
+      "step": 1135
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7502363050580045,
+      "learning_rate": 1.8200048935396908e-05,
+      "loss": 1.0817,
+      "step": 1136
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7682744983967533,
+      "learning_rate": 1.819648107847196e-05,
+      "loss": 1.0425,
+      "step": 1137
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3713371299606184,
+      "learning_rate": 1.8192910039390844e-05,
+      "loss": 1.0654,
+      "step": 1138
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7503688042330208,
+      "learning_rate": 1.8189335819539963e-05,
+      "loss": 0.9669,
+      "step": 1139
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6939472527313225,
+      "learning_rate": 1.8185758420306947e-05,
+      "loss": 1.0712,
+      "step": 1140
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9602585423778535,
+      "learning_rate": 1.818217784308067e-05,
+      "loss": 1.1191,
+      "step": 1141
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6919574659087846,
+      "learning_rate": 1.817859408925123e-05,
+      "loss": 1.09,
+      "step": 1142
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7741012199320503,
+      "learning_rate": 1.817500716020997e-05,
+      "loss": 1.0369,
+      "step": 1143
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6895607220913009,
+      "learning_rate": 1.8171417057349457e-05,
+      "loss": 0.9547,
+      "step": 1144
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5314551684721485,
+      "learning_rate": 1.816782378206349e-05,
+      "loss": 1.0832,
+      "step": 1145
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7048197285776628,
+      "learning_rate": 1.8164227335747108e-05,
+      "loss": 1.0253,
+      "step": 1146
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6969550647285951,
+      "learning_rate": 1.8160627719796568e-05,
+      "loss": 1.0316,
+      "step": 1147
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6664899298365394,
+      "learning_rate": 1.815702493560937e-05,
+      "loss": 1.0352,
+      "step": 1148
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5904541989602146,
+      "learning_rate": 1.8153418984584238e-05,
+      "loss": 1.0481,
+      "step": 1149
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7111655349826199,
+      "learning_rate": 1.8149809868121125e-05,
+      "loss": 1.0873,
+      "step": 1150
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6902691171792595,
+      "learning_rate": 1.8146197587621217e-05,
+      "loss": 1.1198,
+      "step": 1151
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7297909567405895,
+      "learning_rate": 1.814258214448692e-05,
+      "loss": 1.0704,
+      "step": 1152
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6364715632613904,
+      "learning_rate": 1.8138963540121878e-05,
+      "loss": 1.0453,
+      "step": 1153
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6759226468679618,
+      "learning_rate": 1.813534177593096e-05,
+      "loss": 1.0379,
+      "step": 1154
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7084058399237022,
+      "learning_rate": 1.8131716853320254e-05,
+      "loss": 1.0413,
+      "step": 1155
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7322755954272553,
+      "learning_rate": 1.8128088773697086e-05,
+      "loss": 1.0873,
+      "step": 1156
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5481961390577654,
+      "learning_rate": 1.8124457538469996e-05,
+      "loss": 1.1425,
+      "step": 1157
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6740501261893633,
+      "learning_rate": 1.8120823149048753e-05,
+      "loss": 1.0984,
+      "step": 1158
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6650892286679214,
+      "learning_rate": 1.811718560684436e-05,
+      "loss": 1.0495,
+      "step": 1159
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7213350941418664,
+      "learning_rate": 1.8113544913269025e-05,
+      "loss": 1.1152,
+      "step": 1160
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6966827930571241,
+      "learning_rate": 1.8109901069736202e-05,
+      "loss": 1.1266,
+      "step": 1161
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.681133526003358,
+      "learning_rate": 1.8106254077660552e-05,
+      "loss": 0.9766,
+      "step": 1162
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.735571571286572,
+      "learning_rate": 1.810260393845796e-05,
+      "loss": 1.1194,
+      "step": 1163
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8425513906477242,
+      "learning_rate": 1.809895065354554e-05,
+      "loss": 1.0632,
+      "step": 1164
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6866789134674715,
+      "learning_rate": 1.8095294224341622e-05,
+      "loss": 1.1125,
+      "step": 1165
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7208468209557509,
+      "learning_rate": 1.8091634652265755e-05,
+      "loss": 1.1318,
+      "step": 1166
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6628110859890566,
+      "learning_rate": 1.8087971938738715e-05,
+      "loss": 1.08,
+      "step": 1167
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6692492744830174,
+      "learning_rate": 1.808430608518249e-05,
+      "loss": 1.0548,
+      "step": 1168
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7136878303111165,
+      "learning_rate": 1.808063709302029e-05,
+      "loss": 1.0005,
+      "step": 1169
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5614127253489227,
+      "learning_rate": 1.807696496367655e-05,
+      "loss": 1.1351,
+      "step": 1170
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7618686806126905,
+      "learning_rate": 1.8073289698576913e-05,
+      "loss": 1.0213,
+      "step": 1171
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.706658600363527,
+      "learning_rate": 1.8069611299148236e-05,
+      "loss": 1.1281,
+      "step": 1172
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7190379888378651,
+      "learning_rate": 1.8065929766818617e-05,
+      "loss": 0.9985,
+      "step": 1173
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6545870784385032,
+      "learning_rate": 1.806224510301734e-05,
+      "loss": 1.0844,
+      "step": 1174
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6933229456717545,
+      "learning_rate": 1.8058557309174926e-05,
+      "loss": 1.0858,
+      "step": 1175
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6411079752410839,
+      "learning_rate": 1.8054866386723096e-05,
+      "loss": 1.0748,
+      "step": 1176
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.674680858228267,
+      "learning_rate": 1.80511723370948e-05,
+      "loss": 1.0795,
+      "step": 1177
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7267652017355117,
+      "learning_rate": 1.804747516172419e-05,
+      "loss": 1.1044,
+      "step": 1178
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7102476571714501,
+      "learning_rate": 1.8043774862046644e-05,
+      "loss": 1.0491,
+      "step": 1179
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6765426536691589,
+      "learning_rate": 1.804007143949874e-05,
+      "loss": 1.0345,
+      "step": 1180
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7406035525593763,
+      "learning_rate": 1.8036364895518272e-05,
+      "loss": 1.1459,
+      "step": 1181
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7345543684597512,
+      "learning_rate": 1.8032655231544253e-05,
+      "loss": 0.9984,
+      "step": 1182
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.805671964777949,
+      "learning_rate": 1.8028942449016903e-05,
+      "loss": 1.1381,
+      "step": 1183
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7159487604637391,
+      "learning_rate": 1.8025226549377647e-05,
+      "loss": 1.0616,
+      "step": 1184
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.670071372246047,
+      "learning_rate": 1.8021507534069133e-05,
+      "loss": 1.109,
+      "step": 1185
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5263508651738563,
+      "learning_rate": 1.8017785404535198e-05,
+      "loss": 1.0324,
+      "step": 1186
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6814757291910369,
+      "learning_rate": 1.8014060162220916e-05,
+      "loss": 0.9929,
+      "step": 1187
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7115479516295637,
+      "learning_rate": 1.801033180857254e-05,
+      "loss": 1.0303,
+      "step": 1188
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6705594162752795,
+      "learning_rate": 1.8006600345037558e-05,
+      "loss": 1.0899,
+      "step": 1189
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7423913456526471,
+      "learning_rate": 1.8002865773064644e-05,
+      "loss": 0.9689,
+      "step": 1190
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7135029498610271,
+      "learning_rate": 1.799912809410369e-05,
+      "loss": 1.0412,
+      "step": 1191
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6933881659007224,
+      "learning_rate": 1.799538730960579e-05,
+      "loss": 1.0425,
+      "step": 1192
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7710826871537211,
+      "learning_rate": 1.799164342102325e-05,
+      "loss": 1.1187,
+      "step": 1193
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7515023986929857,
+      "learning_rate": 1.7987896429809573e-05,
+      "loss": 1.0276,
+      "step": 1194
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7864361448217719,
+      "learning_rate": 1.798414633741947e-05,
+      "loss": 1.152,
+      "step": 1195
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5368940305559959,
+      "learning_rate": 1.7980393145308857e-05,
+      "loss": 1.1138,
+      "step": 1196
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7766624660265368,
+      "learning_rate": 1.797663685493485e-05,
+      "loss": 1.0633,
+      "step": 1197
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6797801638764922,
+      "learning_rate": 1.7972877467755777e-05,
+      "loss": 1.0665,
+      "step": 1198
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6798776520046556,
+      "learning_rate": 1.7969114985231152e-05,
+      "loss": 1.1161,
+      "step": 1199
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6544190108279894,
+      "learning_rate": 1.796534940882171e-05,
+      "loss": 1.0682,
+      "step": 1200
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6661001824798709,
+      "learning_rate": 1.7961580739989365e-05,
+      "loss": 1.1089,
+      "step": 1201
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6775715314280373,
+      "learning_rate": 1.795780898019726e-05,
+      "loss": 1.1008,
+      "step": 1202
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5406228990953943,
+      "learning_rate": 1.795403413090971e-05,
+      "loss": 0.9956,
+      "step": 1203
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8278922407386298,
+      "learning_rate": 1.7950256193592243e-05,
+      "loss": 1.0823,
+      "step": 1204
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7039607713643954,
+      "learning_rate": 1.794647516971159e-05,
+      "loss": 1.0878,
+      "step": 1205
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7382360184324924,
+      "learning_rate": 1.7942691060735666e-05,
+      "loss": 1.0392,
+      "step": 1206
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7364675028882501,
+      "learning_rate": 1.79389038681336e-05,
+      "loss": 1.0495,
+      "step": 1207
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5102860139373786,
+      "learning_rate": 1.7935113593375707e-05,
+      "loss": 1.0729,
+      "step": 1208
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6637768605165902,
+      "learning_rate": 1.7931320237933503e-05,
+      "loss": 1.0549,
+      "step": 1209
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6966539477427421,
+      "learning_rate": 1.79275238032797e-05,
+      "loss": 1.0739,
+      "step": 1210
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7201808176014152,
+      "learning_rate": 1.7923724290888205e-05,
+      "loss": 1.0042,
+      "step": 1211
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6978466391220802,
+      "learning_rate": 1.791992170223412e-05,
+      "loss": 1.0342,
+      "step": 1212
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7023199443421946,
+      "learning_rate": 1.791611603879374e-05,
+      "loss": 1.0827,
+      "step": 1213
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.711046524264807,
+      "learning_rate": 1.791230730204455e-05,
+      "loss": 1.0984,
+      "step": 1214
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7111882598432034,
+      "learning_rate": 1.7908495493465236e-05,
+      "loss": 1.0724,
+      "step": 1215
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7397042293671215,
+      "learning_rate": 1.7904680614535675e-05,
+      "loss": 1.1318,
+      "step": 1216
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7431005826800966,
+      "learning_rate": 1.7900862666736935e-05,
+      "loss": 1.0408,
+      "step": 1217
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.833495448845143,
+      "learning_rate": 1.789704165155127e-05,
+      "loss": 1.1236,
+      "step": 1218
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.718380789131345,
+      "learning_rate": 1.7893217570462134e-05,
+      "loss": 1.1259,
+      "step": 1219
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6816737633438438,
+      "learning_rate": 1.7889390424954168e-05,
+      "loss": 1.0476,
+      "step": 1220
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6896155018108607,
+      "learning_rate": 1.78855602165132e-05,
+      "loss": 1.1095,
+      "step": 1221
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6240774181225588,
+      "learning_rate": 1.7881726946626244e-05,
+      "loss": 1.0764,
+      "step": 1222
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7186047848457411,
+      "learning_rate": 1.787789061678151e-05,
+      "loss": 1.0133,
+      "step": 1223
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.623072029378063,
+      "learning_rate": 1.78740512284684e-05,
+      "loss": 1.0593,
+      "step": 1224
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6209074805878352,
+      "learning_rate": 1.787020878317749e-05,
+      "loss": 1.1005,
+      "step": 1225
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.737694532199722,
+      "learning_rate": 1.7866363282400555e-05,
+      "loss": 1.074,
+      "step": 1226
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7191220216867085,
+      "learning_rate": 1.7862514727630543e-05,
+      "loss": 1.0889,
+      "step": 1227
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6729674512209035,
+      "learning_rate": 1.7858663120361597e-05,
+      "loss": 1.1769,
+      "step": 1228
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7342667709752023,
+      "learning_rate": 1.785480846208905e-05,
+      "loss": 0.9748,
+      "step": 1229
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6654990608395835,
+      "learning_rate": 1.7850950754309405e-05,
+      "loss": 1.0695,
+      "step": 1230
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6977094026535325,
+      "learning_rate": 1.7847089998520365e-05,
+      "loss": 1.1003,
+      "step": 1231
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7005953434741174,
+      "learning_rate": 1.7843226196220803e-05,
+      "loss": 1.0336,
+      "step": 1232
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5374822212515398,
+      "learning_rate": 1.783935934891078e-05,
+      "loss": 1.1531,
+      "step": 1233
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.66331270329318,
+      "learning_rate": 1.7835489458091544e-05,
+      "loss": 1.0686,
+      "step": 1234
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6562887133139717,
+      "learning_rate": 1.7831616525265515e-05,
+      "loss": 1.0444,
+      "step": 1235
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7780370352110244,
+      "learning_rate": 1.7827740551936296e-05,
+      "loss": 1.0276,
+      "step": 1236
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.65875854613075,
+      "learning_rate": 1.7823861539608686e-05,
+      "loss": 1.0739,
+      "step": 1237
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7040676545283964,
+      "learning_rate": 1.7819979489788638e-05,
+      "loss": 1.071,
+      "step": 1238
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7319938783157933,
+      "learning_rate": 1.7816094403983298e-05,
+      "loss": 1.0494,
+      "step": 1239
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6441364055996427,
+      "learning_rate": 1.7812206283701002e-05,
+      "loss": 1.0391,
+      "step": 1240
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.662792719761601,
+      "learning_rate": 1.7808315130451244e-05,
+      "loss": 1.0414,
+      "step": 1241
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7000020060759544,
+      "learning_rate": 1.78044209457447e-05,
+      "loss": 1.1286,
+      "step": 1242
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7524744062550164,
+      "learning_rate": 1.7800523731093232e-05,
+      "loss": 1.0638,
+      "step": 1243
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6984111476663708,
+      "learning_rate": 1.7796623488009875e-05,
+      "loss": 1.0765,
+      "step": 1244
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.718703192021556,
+      "learning_rate": 1.7792720218008826e-05,
+      "loss": 1.098,
+      "step": 1245
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7251980263634293,
+      "learning_rate": 1.7788813922605488e-05,
+      "loss": 1.1129,
+      "step": 1246
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7232094903326964,
+      "learning_rate": 1.7784904603316402e-05,
+      "loss": 1.0889,
+      "step": 1247
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6134863203650668,
+      "learning_rate": 1.7780992261659305e-05,
+      "loss": 1.0936,
+      "step": 1248
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6620715833300803,
+      "learning_rate": 1.777707689915311e-05,
+      "loss": 1.1262,
+      "step": 1249
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.710224272257135,
+      "learning_rate": 1.777315851731789e-05,
+      "loss": 0.9873,
+      "step": 1250
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7373042291198082,
+      "learning_rate": 1.7769237117674893e-05,
+      "loss": 1.0176,
+      "step": 1251
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5339895539451469,
+      "learning_rate": 1.7765312701746543e-05,
+      "loss": 1.1075,
+      "step": 1252
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6504799385402074,
+      "learning_rate": 1.7761385271056436e-05,
+      "loss": 1.0797,
+      "step": 1253
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7223904238724231,
+      "learning_rate": 1.7757454827129338e-05,
+      "loss": 1.0865,
+      "step": 1254
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7686915263416053,
+      "learning_rate": 1.7753521371491174e-05,
+      "loss": 1.0204,
+      "step": 1255
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.678931166869248,
+      "learning_rate": 1.7749584905669057e-05,
+      "loss": 1.0541,
+      "step": 1256
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6793417023128082,
+      "learning_rate": 1.774564543119125e-05,
+      "loss": 1.0373,
+      "step": 1257
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6933584393208558,
+      "learning_rate": 1.7741702949587196e-05,
+      "loss": 1.0565,
+      "step": 1258
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.717370794189636,
+      "learning_rate": 1.7737757462387507e-05,
+      "loss": 1.0791,
+      "step": 1259
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.717277830474809,
+      "learning_rate": 1.7733808971123946e-05,
+      "loss": 1.0394,
+      "step": 1260
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7445743339638216,
+      "learning_rate": 1.7729857477329463e-05,
+      "loss": 1.1425,
+      "step": 1261
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7360163750170674,
+      "learning_rate": 1.7725902982538162e-05,
+      "loss": 0.9904,
+      "step": 1262
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6784143605587061,
+      "learning_rate": 1.772194548828531e-05,
+      "loss": 1.025,
+      "step": 1263
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.706658824483645,
+      "learning_rate": 1.7717984996107346e-05,
+      "loss": 1.0955,
+      "step": 1264
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6757921307197589,
+      "learning_rate": 1.771402150754187e-05,
+      "loss": 1.0004,
+      "step": 1265
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5540092677131035,
+      "learning_rate": 1.7710055024127637e-05,
+      "loss": 1.1366,
+      "step": 1266
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6810931544650077,
+      "learning_rate": 1.7706085547404582e-05,
+      "loss": 1.1257,
+      "step": 1267
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6714174059265096,
+      "learning_rate": 1.770211307891379e-05,
+      "loss": 1.0717,
+      "step": 1268
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7023834782840449,
+      "learning_rate": 1.769813762019751e-05,
+      "loss": 1.0774,
+      "step": 1269
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6676031097596058,
+      "learning_rate": 1.769415917279915e-05,
+      "loss": 1.075,
+      "step": 1270
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7078354330613669,
+      "learning_rate": 1.7690177738263284e-05,
+      "loss": 1.1329,
+      "step": 1271
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.56061565013086,
+      "learning_rate": 1.7686193318135635e-05,
+      "loss": 1.049,
+      "step": 1272
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6304547993462489,
+      "learning_rate": 1.76822059139631e-05,
+      "loss": 0.9847,
+      "step": 1273
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4881448587074394,
+      "learning_rate": 1.7678215527293724e-05,
+      "loss": 1.0832,
+      "step": 1274
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7295377465820255,
+      "learning_rate": 1.767422215967671e-05,
+      "loss": 1.0355,
+      "step": 1275
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6808606900866735,
+      "learning_rate": 1.767022581266242e-05,
+      "loss": 1.0236,
+      "step": 1276
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7616105937473209,
+      "learning_rate": 1.766622648780238e-05,
+      "loss": 1.1351,
+      "step": 1277
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.737447191381819,
+      "learning_rate": 1.766222418664926e-05,
+      "loss": 1.1041,
+      "step": 1278
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6549661254365107,
+      "learning_rate": 1.765821891075689e-05,
+      "loss": 1.0875,
+      "step": 1279
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.662278809638558,
+      "learning_rate": 1.7654210661680263e-05,
+      "loss": 1.0181,
+      "step": 1280
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7355807047760461,
+      "learning_rate": 1.765019944097551e-05,
+      "loss": 1.0176,
+      "step": 1281
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7593969597894784,
+      "learning_rate": 1.7646185250199936e-05,
+      "loss": 1.1021,
+      "step": 1282
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7507640916006856,
+      "learning_rate": 1.7642168090911976e-05,
+      "loss": 1.0825,
+      "step": 1283
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.752289376754265,
+      "learning_rate": 1.763814796467124e-05,
+      "loss": 1.0219,
+      "step": 1284
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7557365095354803,
+      "learning_rate": 1.763412487303847e-05,
+      "loss": 1.0925,
+      "step": 1285
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6491969566843623,
+      "learning_rate": 1.7630098817575578e-05,
+      "loss": 1.0398,
+      "step": 1286
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7024950056980033,
+      "learning_rate": 1.762606979984561e-05,
+      "loss": 1.0585,
+      "step": 1287
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7099145909574953,
+      "learning_rate": 1.7622037821412775e-05,
+      "loss": 1.0306,
+      "step": 1288
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6788480913589992,
+      "learning_rate": 1.7618002883842426e-05,
+      "loss": 1.1023,
+      "step": 1289
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7271405632839125,
+      "learning_rate": 1.7613964988701057e-05,
+      "loss": 1.1053,
+      "step": 1290
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7457955588015727,
+      "learning_rate": 1.7609924137556326e-05,
+      "loss": 1.0519,
+      "step": 1291
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5402713583438852,
+      "learning_rate": 1.7605880331977022e-05,
+      "loss": 1.1722,
+      "step": 1292
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5489218025690923,
+      "learning_rate": 1.76018335735331e-05,
+      "loss": 1.01,
+      "step": 1293
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6813192086915629,
+      "learning_rate": 1.7597783863795644e-05,
+      "loss": 1.017,
+      "step": 1294
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.713073783651929,
+      "learning_rate": 1.7593731204336895e-05,
+      "loss": 1.0671,
+      "step": 1295
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7602321290150248,
+      "learning_rate": 1.7589675596730233e-05,
+      "loss": 1.1028,
+      "step": 1296
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7420530277951629,
+      "learning_rate": 1.758561704255018e-05,
+      "loss": 1.1398,
+      "step": 1297
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7043416620489783,
+      "learning_rate": 1.7581555543372413e-05,
+      "loss": 1.0734,
+      "step": 1298
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.742983449608016,
+      "learning_rate": 1.7577491100773744e-05,
+      "loss": 1.0612,
+      "step": 1299
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6150495502021809,
+      "learning_rate": 1.7573423716332128e-05,
+      "loss": 1.1057,
+      "step": 1300
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7201756900407781,
+      "learning_rate": 1.7569353391626665e-05,
+      "loss": 1.0845,
+      "step": 1301
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7191864033778661,
+      "learning_rate": 1.7565280128237595e-05,
+      "loss": 1.0872,
+      "step": 1302
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5556662852574737,
+      "learning_rate": 1.75612039277463e-05,
+      "loss": 1.0299,
+      "step": 1303
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6848646820847262,
+      "learning_rate": 1.75571247917353e-05,
+      "loss": 1.0695,
+      "step": 1304
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7483496865371585,
+      "learning_rate": 1.7553042721788255e-05,
+      "loss": 1.1047,
+      "step": 1305
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7316793245368011,
+      "learning_rate": 1.754895771948997e-05,
+      "loss": 1.1149,
+      "step": 1306
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6489673644059881,
+      "learning_rate": 1.754486978642637e-05,
+      "loss": 1.0793,
+      "step": 1307
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7885768158153408,
+      "learning_rate": 1.7540778924184553e-05,
+      "loss": 0.9899,
+      "step": 1308
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7096998931600835,
+      "learning_rate": 1.7536685134352717e-05,
+      "loss": 1.0717,
+      "step": 1309
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.680006539781854,
+      "learning_rate": 1.7532588418520215e-05,
+      "loss": 1.0139,
+      "step": 1310
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7445693262828735,
+      "learning_rate": 1.7528488778277535e-05,
+      "loss": 1.1109,
+      "step": 1311
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5275565401758519,
+      "learning_rate": 1.75243862152163e-05,
+      "loss": 1.1233,
+      "step": 1312
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7400775809813693,
+      "learning_rate": 1.752028073092926e-05,
+      "loss": 1.103,
+      "step": 1313
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6485519102738981,
+      "learning_rate": 1.7516172327010314e-05,
+      "loss": 1.0844,
+      "step": 1314
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6545583761724091,
+      "learning_rate": 1.751206100505448e-05,
+      "loss": 1.0432,
+      "step": 1315
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6676928106300022,
+      "learning_rate": 1.7507946766657914e-05,
+      "loss": 1.023,
+      "step": 1316
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8677122493312068,
+      "learning_rate": 1.7503829613417905e-05,
+      "loss": 1.0765,
+      "step": 1317
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7560869190051522,
+      "learning_rate": 1.749970954693288e-05,
+      "loss": 1.1185,
+      "step": 1318
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6819847667925498,
+      "learning_rate": 1.7495586568802384e-05,
+      "loss": 1.1375,
+      "step": 1319
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7011623082255036,
+      "learning_rate": 1.7491460680627105e-05,
+      "loss": 1.0485,
+      "step": 1320
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7498993267107522,
+      "learning_rate": 1.7487331884008845e-05,
+      "loss": 1.0818,
+      "step": 1321
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6945366202487733,
+      "learning_rate": 1.7483200180550554e-05,
+      "loss": 1.0798,
+      "step": 1322
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6360793000591424,
+      "learning_rate": 1.74790655718563e-05,
+      "loss": 1.0867,
+      "step": 1323
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7104778345410463,
+      "learning_rate": 1.747492805953128e-05,
+      "loss": 1.1231,
+      "step": 1324
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.658002944731968,
+      "learning_rate": 1.7470787645181818e-05,
+      "loss": 1.0403,
+      "step": 1325
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6993676377755347,
+      "learning_rate": 1.7466644330415362e-05,
+      "loss": 0.9782,
+      "step": 1326
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6772683371088809,
+      "learning_rate": 1.7462498116840496e-05,
+      "loss": 1.0122,
+      "step": 1327
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6204746526902389,
+      "learning_rate": 1.745834900606692e-05,
+      "loss": 1.0847,
+      "step": 1328
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.686204746352357,
+      "learning_rate": 1.7454196999705458e-05,
+      "loss": 1.0065,
+      "step": 1329
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6869074664652423,
+      "learning_rate": 1.7450042099368066e-05,
+      "loss": 1.0231,
+      "step": 1330
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6864949347083187,
+      "learning_rate": 1.7445884306667823e-05,
+      "loss": 1.0165,
+      "step": 1331
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6100628115331344,
+      "learning_rate": 1.7441723623218917e-05,
+      "loss": 1.0871,
+      "step": 1332
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4933521419231461,
+      "learning_rate": 1.7437560050636678e-05,
+      "loss": 0.999,
+      "step": 1333
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7028294984710558,
+      "learning_rate": 1.7433393590537543e-05,
+      "loss": 1.069,
+      "step": 1334
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6797474240900023,
+      "learning_rate": 1.7429224244539077e-05,
+      "loss": 1.0198,
+      "step": 1335
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5551532945441033,
+      "learning_rate": 1.7425052014259965e-05,
+      "loss": 1.02,
+      "step": 1336
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6438641599358217,
+      "learning_rate": 1.7420876901320006e-05,
+      "loss": 1.0919,
+      "step": 1337
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.687121970853301,
+      "learning_rate": 1.7416698907340128e-05,
+      "loss": 1.1047,
+      "step": 1338
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7837836804464559,
+      "learning_rate": 1.741251803394237e-05,
+      "loss": 0.973,
+      "step": 1339
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6876023707237854,
+      "learning_rate": 1.740833428274989e-05,
+      "loss": 1.0804,
+      "step": 1340
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7083363288916842,
+      "learning_rate": 1.7404147655386966e-05,
+      "loss": 1.053,
+      "step": 1341
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7017662282368002,
+      "learning_rate": 1.739995815347899e-05,
+      "loss": 1.0798,
+      "step": 1342
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6908151057520625,
+      "learning_rate": 1.739576577865247e-05,
+      "loss": 1.0187,
+      "step": 1343
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6124484582215451,
+      "learning_rate": 1.739157053253503e-05,
+      "loss": 1.0689,
+      "step": 1344
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5495012760019465,
+      "learning_rate": 1.738737241675541e-05,
+      "loss": 1.085,
+      "step": 1345
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6923350838393342,
+      "learning_rate": 1.7383171432943466e-05,
+      "loss": 1.0419,
+      "step": 1346
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6481688772106067,
+      "learning_rate": 1.737896758273016e-05,
+      "loss": 0.9987,
+      "step": 1347
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7080509358991176,
+      "learning_rate": 1.7374760867747574e-05,
+      "loss": 1.0714,
+      "step": 1348
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6814097363955468,
+      "learning_rate": 1.7370551289628895e-05,
+      "loss": 1.0072,
+      "step": 1349
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7363246112475865,
+      "learning_rate": 1.7366338850008432e-05,
+      "loss": 0.9354,
+      "step": 1350
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7142978865294047,
+      "learning_rate": 1.73621235505216e-05,
+      "loss": 1.0334,
+      "step": 1351
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.669385723390627,
+      "learning_rate": 1.7357905392804918e-05,
+      "loss": 1.0527,
+      "step": 1352
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7099150029046482,
+      "learning_rate": 1.735368437849602e-05,
+      "loss": 1.049,
+      "step": 1353
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7025110597698314,
+      "learning_rate": 1.7349460509233654e-05,
+      "loss": 1.0595,
+      "step": 1354
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7216482170143179,
+      "learning_rate": 1.734523378665767e-05,
+      "loss": 1.056,
+      "step": 1355
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.49953934521073207,
+      "learning_rate": 1.7341004212409026e-05,
+      "loss": 1.0394,
+      "step": 1356
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6658297412284832,
+      "learning_rate": 1.7336771788129785e-05,
+      "loss": 1.0961,
+      "step": 1357
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.710747757227432,
+      "learning_rate": 1.7332536515463126e-05,
+      "loss": 0.9746,
+      "step": 1358
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7039081511141305,
+      "learning_rate": 1.7328298396053324e-05,
+      "loss": 1.0082,
+      "step": 1359
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7678637260100771,
+      "learning_rate": 1.7324057431545768e-05,
+      "loss": 1.0606,
+      "step": 1360
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6949937343858954,
+      "learning_rate": 1.7319813623586935e-05,
+      "loss": 1.0708,
+      "step": 1361
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6969398328649269,
+      "learning_rate": 1.7315566973824433e-05,
+      "loss": 1.0732,
+      "step": 1362
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7139800582140702,
+      "learning_rate": 1.7311317483906946e-05,
+      "loss": 1.0554,
+      "step": 1363
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6897124192133575,
+      "learning_rate": 1.730706515548427e-05,
+      "loss": 1.0723,
+      "step": 1364
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7418811656081683,
+      "learning_rate": 1.730280999020732e-05,
+      "loss": 1.0936,
+      "step": 1365
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5360804883715108,
+      "learning_rate": 1.729855198972808e-05,
+      "loss": 1.1119,
+      "step": 1366
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7242589496412213,
+      "learning_rate": 1.729429115569967e-05,
+      "loss": 1.067,
+      "step": 1367
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.820811826191187,
+      "learning_rate": 1.729002748977628e-05,
+      "loss": 1.0631,
+      "step": 1368
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6538212338766325,
+      "learning_rate": 1.7285760993613215e-05,
+      "loss": 1.0471,
+      "step": 1369
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.781570688848608,
+      "learning_rate": 1.7281491668866874e-05,
+      "loss": 1.0944,
+      "step": 1370
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7256273006528703,
+      "learning_rate": 1.727721951719476e-05,
+      "loss": 1.0906,
+      "step": 1371
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6396666471702329,
+      "learning_rate": 1.7272944540255468e-05,
+      "loss": 1.1009,
+      "step": 1372
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7141100798825905,
+      "learning_rate": 1.726866673970869e-05,
+      "loss": 1.0386,
+      "step": 1373
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6189008551368615,
+      "learning_rate": 1.7264386117215216e-05,
+      "loss": 0.9572,
+      "step": 1374
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6979834535577661,
+      "learning_rate": 1.7260102674436933e-05,
+      "loss": 1.0669,
+      "step": 1375
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5881434163914435,
+      "learning_rate": 1.7255816413036818e-05,
+      "loss": 1.0854,
+      "step": 1376
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6686045740834066,
+      "learning_rate": 1.7251527334678946e-05,
+      "loss": 1.0405,
+      "step": 1377
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6854815372896126,
+      "learning_rate": 1.7247235441028486e-05,
+      "loss": 1.1332,
+      "step": 1378
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7404317684377344,
+      "learning_rate": 1.7242940733751696e-05,
+      "loss": 0.9788,
+      "step": 1379
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7815007210125604,
+      "learning_rate": 1.7238643214515934e-05,
+      "loss": 1.0411,
+      "step": 1380
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6872365563491111,
+      "learning_rate": 1.7234342884989642e-05,
+      "loss": 1.0634,
+      "step": 1381
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6985197316082884,
+      "learning_rate": 1.7230039746842352e-05,
+      "loss": 1.0854,
+      "step": 1382
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6576333678572911,
+      "learning_rate": 1.7225733801744698e-05,
+      "loss": 1.1251,
+      "step": 1383
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7869108165030174,
+      "learning_rate": 1.7221425051368394e-05,
+      "loss": 1.0228,
+      "step": 1384
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6722529325650081,
+      "learning_rate": 1.7217113497386245e-05,
+      "loss": 1.0934,
+      "step": 1385
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7143501852548514,
+      "learning_rate": 1.721279914147214e-05,
+      "loss": 1.1249,
+      "step": 1386
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.591788950666582,
+      "learning_rate": 1.7208481985301065e-05,
+      "loss": 1.0574,
+      "step": 1387
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6995069912507279,
+      "learning_rate": 1.7204162030549093e-05,
+      "loss": 1.0383,
+      "step": 1388
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.599060331737342,
+      "learning_rate": 1.7199839278893368e-05,
+      "loss": 1.0861,
+      "step": 1389
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6523464056509851,
+      "learning_rate": 1.719551373201214e-05,
+      "loss": 1.1093,
+      "step": 1390
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.801697868262666,
+      "learning_rate": 1.7191185391584736e-05,
+      "loss": 1.0301,
+      "step": 1391
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.631330564057925,
+      "learning_rate": 1.7186854259291558e-05,
+      "loss": 1.1002,
+      "step": 1392
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6948729825741066,
+      "learning_rate": 1.7182520336814105e-05,
+      "loss": 1.0126,
+      "step": 1393
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5778914523983442,
+      "learning_rate": 1.717818362583496e-05,
+      "loss": 1.0779,
+      "step": 1394
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6447125444671892,
+      "learning_rate": 1.7173844128037777e-05,
+      "loss": 1.0231,
+      "step": 1395
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5402168445418607,
+      "learning_rate": 1.71695018451073e-05,
+      "loss": 1.1416,
+      "step": 1396
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6904367499643368,
+      "learning_rate": 1.7165156778729355e-05,
+      "loss": 1.1447,
+      "step": 1397
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7072362323753708,
+      "learning_rate": 1.7160808930590845e-05,
+      "loss": 1.1261,
+      "step": 1398
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7187070170997057,
+      "learning_rate": 1.7156458302379753e-05,
+      "loss": 1.0225,
+      "step": 1399
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6929482709514626,
+      "learning_rate": 1.7152104895785147e-05,
+      "loss": 1.0649,
+      "step": 1400
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5678765061078533,
+      "learning_rate": 1.7147748712497162e-05,
+      "loss": 1.0149,
+      "step": 1401
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5253151587885788,
+      "learning_rate": 1.7143389754207026e-05,
+      "loss": 1.0251,
+      "step": 1402
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7323677721112851,
+      "learning_rate": 1.713902802260703e-05,
+      "loss": 1.0369,
+      "step": 1403
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7379671939585081,
+      "learning_rate": 1.7134663519390557e-05,
+      "loss": 1.0761,
+      "step": 1404
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7447059407075705,
+      "learning_rate": 1.7130296246252048e-05,
+      "loss": 1.085,
+      "step": 1405
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5112603756803951,
+      "learning_rate": 1.7125926204887034e-05,
+      "loss": 1.0569,
+      "step": 1406
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7017814803575699,
+      "learning_rate": 1.712155339699211e-05,
+      "loss": 1.0461,
+      "step": 1407
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6885332568588394,
+      "learning_rate": 1.7117177824264962e-05,
+      "loss": 1.0723,
+      "step": 1408
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7776507871429365,
+      "learning_rate": 1.7112799488404327e-05,
+      "loss": 0.9577,
+      "step": 1409
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5538042215114002,
+      "learning_rate": 1.7108418391110033e-05,
+      "loss": 1.1111,
+      "step": 1410
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6982822923324169,
+      "learning_rate": 1.7104034534082968e-05,
+      "loss": 1.1465,
+      "step": 1411
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6415034315876358,
+      "learning_rate": 1.7099647919025096e-05,
+      "loss": 1.0206,
+      "step": 1412
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7119378939189794,
+      "learning_rate": 1.7095258547639456e-05,
+      "loss": 1.0428,
+      "step": 1413
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7918817684515215,
+      "learning_rate": 1.709086642163015e-05,
+      "loss": 1.0448,
+      "step": 1414
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7174357146310543,
+      "learning_rate": 1.7086471542702355e-05,
+      "loss": 1.0639,
+      "step": 1415
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6784192503931491,
+      "learning_rate": 1.708207391256231e-05,
+      "loss": 1.0345,
+      "step": 1416
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7078855522520854,
+      "learning_rate": 1.707767353291733e-05,
+      "loss": 1.0404,
+      "step": 1417
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7552949187615764,
+      "learning_rate": 1.7073270405475796e-05,
+      "loss": 1.062,
+      "step": 1418
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.746528526558957,
+      "learning_rate": 1.7068864531947147e-05,
+      "loss": 1.0752,
+      "step": 1419
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6772298413124057,
+      "learning_rate": 1.70644559140419e-05,
+      "loss": 1.0697,
+      "step": 1420
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5420640897491213,
+      "learning_rate": 1.706004455347163e-05,
+      "loss": 1.0583,
+      "step": 1421
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5703189945515927,
+      "learning_rate": 1.705563045194898e-05,
+      "loss": 1.123,
+      "step": 1422
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6734197314743375,
+      "learning_rate": 1.7051213611187657e-05,
+      "loss": 1.0117,
+      "step": 1423
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6642371606829365,
+      "learning_rate": 1.704679403290243e-05,
+      "loss": 1.0526,
+      "step": 1424
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6699719059764634,
+      "learning_rate": 1.7042371718809132e-05,
+      "loss": 1.0002,
+      "step": 1425
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.716544465556886,
+      "learning_rate": 1.7037946670624652e-05,
+      "loss": 1.0363,
+      "step": 1426
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7015619760297042,
+      "learning_rate": 1.7033518890066956e-05,
+      "loss": 1.0115,
+      "step": 1427
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.717960242849031,
+      "learning_rate": 1.7029088378855055e-05,
+      "loss": 1.09,
+      "step": 1428
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.767842528217814,
+      "learning_rate": 1.7024655138709025e-05,
+      "loss": 1.042,
+      "step": 1429
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6777152885184278,
+      "learning_rate": 1.7020219171350004e-05,
+      "loss": 0.9916,
+      "step": 1430
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6288311668679559,
+      "learning_rate": 1.7015780478500187e-05,
+      "loss": 1.0228,
+      "step": 1431
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6895691327662478,
+      "learning_rate": 1.701133906188283e-05,
+      "loss": 1.1115,
+      "step": 1432
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.756811491847227,
+      "learning_rate": 1.700689492322224e-05,
+      "loss": 1.0529,
+      "step": 1433
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6666877878679786,
+      "learning_rate": 1.700244806424379e-05,
+      "loss": 1.1152,
+      "step": 1434
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7668648441023393,
+      "learning_rate": 1.6997998486673893e-05,
+      "loss": 1.1189,
+      "step": 1435
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6072265925656888,
+      "learning_rate": 1.699354619224004e-05,
+      "loss": 1.0554,
+      "step": 1436
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.750785520696565,
+      "learning_rate": 1.698909118267076e-05,
+      "loss": 1.1261,
+      "step": 1437
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7097026088522629,
+      "learning_rate": 1.6984633459695646e-05,
+      "loss": 1.025,
+      "step": 1438
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.643783110140985,
+      "learning_rate": 1.6980173025045328e-05,
+      "loss": 1.0524,
+      "step": 1439
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.710066180724209,
+      "learning_rate": 1.697570988045151e-05,
+      "loss": 1.0517,
+      "step": 1440
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6733674156334266,
+      "learning_rate": 1.6971244027646937e-05,
+      "loss": 1.0797,
+      "step": 1441
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8206017442008449,
+      "learning_rate": 1.69667754683654e-05,
+      "loss": 1.0002,
+      "step": 1442
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6948496888812223,
+      "learning_rate": 1.6962304204341758e-05,
+      "loss": 1.0447,
+      "step": 1443
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6513096991741979,
+      "learning_rate": 1.6957830237311904e-05,
+      "loss": 1.0971,
+      "step": 1444
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7080349603095151,
+      "learning_rate": 1.6953353569012784e-05,
+      "loss": 1.0902,
+      "step": 1445
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6868514013251078,
+      "learning_rate": 1.6948874201182402e-05,
+      "loss": 0.9476,
+      "step": 1446
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.714980648210039,
+      "learning_rate": 1.6944392135559798e-05,
+      "loss": 1.0369,
+      "step": 1447
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4979430384765025,
+      "learning_rate": 1.6939907373885062e-05,
+      "loss": 1.0986,
+      "step": 1448
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.701994989966161,
+      "learning_rate": 1.6935419917899335e-05,
+      "loss": 1.0317,
+      "step": 1449
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5817958698627023,
+      "learning_rate": 1.6930929769344807e-05,
+      "loss": 1.129,
+      "step": 1450
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6798493924149119,
+      "learning_rate": 1.69264369299647e-05,
+      "loss": 1.0818,
+      "step": 1451
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8311662898534563,
+      "learning_rate": 1.692194140150329e-05,
+      "loss": 1.1542,
+      "step": 1452
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6600188146948317,
+      "learning_rate": 1.69174431857059e-05,
+      "loss": 1.0444,
+      "step": 1453
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.33750553137524775,
+      "learning_rate": 1.6912942284318898e-05,
+      "loss": 1.0677,
+      "step": 1454
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5076703477112331,
+      "learning_rate": 1.6908438699089674e-05,
+      "loss": 1.0834,
+      "step": 1455
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7241577337521642,
+      "learning_rate": 1.690393243176668e-05,
+      "loss": 0.9983,
+      "step": 1456
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7250722726638326,
+      "learning_rate": 1.6899423484099413e-05,
+      "loss": 1.0379,
+      "step": 1457
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7252949621929339,
+      "learning_rate": 1.6894911857838394e-05,
+      "loss": 1.04,
+      "step": 1458
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6252526476400133,
+      "learning_rate": 1.689039755473519e-05,
+      "loss": 1.0728,
+      "step": 1459
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7085680330631782,
+      "learning_rate": 1.6885880576542417e-05,
+      "loss": 1.0274,
+      "step": 1460
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5218947475931638,
+      "learning_rate": 1.6881360925013712e-05,
+      "loss": 1.1083,
+      "step": 1461
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6092007300231266,
+      "learning_rate": 1.6876838601903765e-05,
+      "loss": 1.123,
+      "step": 1462
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6977052917016024,
+      "learning_rate": 1.6872313608968296e-05,
+      "loss": 0.9983,
+      "step": 1463
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6812171777471692,
+      "learning_rate": 1.6867785947964065e-05,
+      "loss": 1.0692,
+      "step": 1464
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7519490227404066,
+      "learning_rate": 1.6863255620648866e-05,
+      "loss": 1.0439,
+      "step": 1465
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7106509562601268,
+      "learning_rate": 1.685872262878152e-05,
+      "loss": 1.0499,
+      "step": 1466
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6594872814778435,
+      "learning_rate": 1.6854186974121903e-05,
+      "loss": 1.0487,
+      "step": 1467
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.520765465879072,
+      "learning_rate": 1.68496486584309e-05,
+      "loss": 1.053,
+      "step": 1468
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6963115248907196,
+      "learning_rate": 1.6845107683470453e-05,
+      "loss": 1.0392,
+      "step": 1469
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.624878755884747,
+      "learning_rate": 1.6840564051003517e-05,
+      "loss": 1.0121,
+      "step": 1470
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.678167621121788,
+      "learning_rate": 1.6836017762794087e-05,
+      "loss": 1.0514,
+      "step": 1471
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7840985556485944,
+      "learning_rate": 1.6831468820607192e-05,
+      "loss": 1.0756,
+      "step": 1472
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6798702429514253,
+      "learning_rate": 1.6826917226208886e-05,
+      "loss": 1.1229,
+      "step": 1473
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6849995336662225,
+      "learning_rate": 1.6822362981366257e-05,
+      "loss": 0.9803,
+      "step": 1474
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6786508280061286,
+      "learning_rate": 1.6817806087847417e-05,
+      "loss": 1.0548,
+      "step": 1475
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6521259186901508,
+      "learning_rate": 1.681324654742151e-05,
+      "loss": 1.0912,
+      "step": 1476
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5255916352393779,
+      "learning_rate": 1.6808684361858706e-05,
+      "loss": 1.0058,
+      "step": 1477
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6544566954243467,
+      "learning_rate": 1.6804119532930202e-05,
+      "loss": 1.0873,
+      "step": 1478
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.747383767040596,
+      "learning_rate": 1.6799552062408225e-05,
+      "loss": 1.0226,
+      "step": 1479
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6847482008012444,
+      "learning_rate": 1.6794981952066018e-05,
+      "loss": 1.052,
+      "step": 1480
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7263357890717622,
+      "learning_rate": 1.6790409203677863e-05,
+      "loss": 1.1084,
+      "step": 1481
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6976867994249097,
+      "learning_rate": 1.6785833819019052e-05,
+      "loss": 1.04,
+      "step": 1482
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6779270598947827,
+      "learning_rate": 1.678125579986591e-05,
+      "loss": 1.0709,
+      "step": 1483
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6520501274315831,
+      "learning_rate": 1.677667514799578e-05,
+      "loss": 1.149,
+      "step": 1484
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.47340053741361643,
+      "learning_rate": 1.6772091865187032e-05,
+      "loss": 1.0134,
+      "step": 1485
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7234423502416503,
+      "learning_rate": 1.676750595321905e-05,
+      "loss": 1.041,
+      "step": 1486
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7173512942981822,
+      "learning_rate": 1.6762917413872246e-05,
+      "loss": 1.0844,
+      "step": 1487
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6805898953640681,
+      "learning_rate": 1.675832624892805e-05,
+      "loss": 1.0995,
+      "step": 1488
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7188839119242726,
+      "learning_rate": 1.6753732460168907e-05,
+      "loss": 1.0677,
+      "step": 1489
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6906478538611095,
+      "learning_rate": 1.674913604937828e-05,
+      "loss": 1.1597,
+      "step": 1490
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7490268830241964,
+      "learning_rate": 1.6744537018340662e-05,
+      "loss": 1.0163,
+      "step": 1491
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5912505125717852,
+      "learning_rate": 1.6739935368841555e-05,
+      "loss": 1.0782,
+      "step": 1492
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7351730017342761,
+      "learning_rate": 1.6735331102667475e-05,
+      "loss": 1.002,
+      "step": 1493
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.748344734668813,
+      "learning_rate": 1.6730724221605955e-05,
+      "loss": 1.0616,
+      "step": 1494
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.687865683146646,
+      "learning_rate": 1.6726114727445547e-05,
+      "loss": 1.0673,
+      "step": 1495
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.641309216424545,
+      "learning_rate": 1.6721502621975813e-05,
+      "loss": 1.0353,
+      "step": 1496
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6979408538671221,
+      "learning_rate": 1.6716887906987332e-05,
+      "loss": 1.0835,
+      "step": 1497
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6609664396074753,
+      "learning_rate": 1.6712270584271703e-05,
+      "loss": 1.069,
+      "step": 1498
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6627379345075971,
+      "learning_rate": 1.670765065562152e-05,
+      "loss": 0.9969,
+      "step": 1499
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5906132646035732,
+      "learning_rate": 1.67030281228304e-05,
+      "loss": 1.0578,
+      "step": 1500
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6994823699713598,
+      "learning_rate": 1.6698402987692968e-05,
+      "loss": 1.1257,
+      "step": 1501
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7400286036973409,
+      "learning_rate": 1.6693775252004866e-05,
+      "loss": 1.024,
+      "step": 1502
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6955905133790391,
+      "learning_rate": 1.668914491756274e-05,
+      "loss": 1.0578,
+      "step": 1503
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6435624277511836,
+      "learning_rate": 1.668451198616424e-05,
+      "loss": 1.0648,
+      "step": 1504
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8082776331153433,
+      "learning_rate": 1.6679876459608033e-05,
+      "loss": 1.0073,
+      "step": 1505
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7068995925834857,
+      "learning_rate": 1.667523833969379e-05,
+      "loss": 1.0161,
+      "step": 1506
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.909277859077618,
+      "learning_rate": 1.667059762822219e-05,
+      "loss": 1.0864,
+      "step": 1507
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6593833868533819,
+      "learning_rate": 1.666595432699491e-05,
+      "loss": 1.0426,
+      "step": 1508
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6895885749293313,
+      "learning_rate": 1.6661308437814652e-05,
+      "loss": 1.0366,
+      "step": 1509
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6967292331061324,
+      "learning_rate": 1.6656659962485097e-05,
+      "loss": 1.0961,
+      "step": 1510
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7118660407987364,
+      "learning_rate": 1.6652008902810952e-05,
+      "loss": 0.9937,
+      "step": 1511
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5805584187910761,
+      "learning_rate": 1.6647355260597915e-05,
+      "loss": 1.0757,
+      "step": 1512
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6576099508688966,
+      "learning_rate": 1.664269903765269e-05,
+      "loss": 1.0471,
+      "step": 1513
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6979633747732483,
+      "learning_rate": 1.6638040235782983e-05,
+      "loss": 1.041,
+      "step": 1514
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7738246564689404,
+      "learning_rate": 1.6633378856797505e-05,
+      "loss": 1.1357,
+      "step": 1515
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6728580022714323,
+      "learning_rate": 1.662871490250596e-05,
+      "loss": 1.075,
+      "step": 1516
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7008822203850875,
+      "learning_rate": 1.662404837471905e-05,
+      "loss": 0.9653,
+      "step": 1517
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6029319037981064,
+      "learning_rate": 1.66193792752485e-05,
+      "loss": 1.0681,
+      "step": 1518
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6436713699986608,
+      "learning_rate": 1.6614707605906995e-05,
+      "loss": 1.0536,
+      "step": 1519
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5608581455481694,
+      "learning_rate": 1.661003336850825e-05,
+      "loss": 1.0414,
+      "step": 1520
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6940439982259281,
+      "learning_rate": 1.660535656486696e-05,
+      "loss": 1.0013,
+      "step": 1521
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.678778014573007,
+      "learning_rate": 1.660067719679882e-05,
+      "loss": 1.062,
+      "step": 1522
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5439319119133093,
+      "learning_rate": 1.6595995266120528e-05,
+      "loss": 1.1495,
+      "step": 1523
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7047973027710526,
+      "learning_rate": 1.6591310774649766e-05,
+      "loss": 1.0347,
+      "step": 1524
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6563687950153423,
+      "learning_rate": 1.6586623724205216e-05,
+      "loss": 1.0763,
+      "step": 1525
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6596410839663412,
+      "learning_rate": 1.6581934116606554e-05,
+      "loss": 1.0294,
+      "step": 1526
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7058491710892341,
+      "learning_rate": 1.657724195367444e-05,
+      "loss": 1.1053,
+      "step": 1527
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7863621657172727,
+      "learning_rate": 1.657254723723054e-05,
+      "loss": 1.0331,
+      "step": 1528
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7349596548200876,
+      "learning_rate": 1.6567849969097505e-05,
+      "loss": 1.0225,
+      "step": 1529
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7943006170075954,
+      "learning_rate": 1.6563150151098973e-05,
+      "loss": 1.0698,
+      "step": 1530
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7059732501744225,
+      "learning_rate": 1.6558447785059577e-05,
+      "loss": 1.1033,
+      "step": 1531
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.719743482032119,
+      "learning_rate": 1.655374287280494e-05,
+      "loss": 1.0549,
+      "step": 1532
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7166960654208971,
+      "learning_rate": 1.6549035416161662e-05,
+      "loss": 1.0203,
+      "step": 1533
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6361934191964745,
+      "learning_rate": 1.654432541695735e-05,
+      "loss": 1.024,
+      "step": 1534
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6951855364280902,
+      "learning_rate": 1.653961287702058e-05,
+      "loss": 1.1127,
+      "step": 1535
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7524514849313375,
+      "learning_rate": 1.653489779818093e-05,
+      "loss": 1.0562,
+      "step": 1536
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7293882815847489,
+      "learning_rate": 1.6530180182268946e-05,
+      "loss": 1.0312,
+      "step": 1537
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7129767446205147,
+      "learning_rate": 1.652546003111618e-05,
+      "loss": 1.0735,
+      "step": 1538
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7443513965545893,
+      "learning_rate": 1.652073734655515e-05,
+      "loss": 1.098,
+      "step": 1539
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6432169870021687,
+      "learning_rate": 1.6516012130419366e-05,
+      "loss": 0.9803,
+      "step": 1540
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7124491056704746,
+      "learning_rate": 1.6511284384543317e-05,
+      "loss": 0.9998,
+      "step": 1541
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7571258693975107,
+      "learning_rate": 1.6506554110762483e-05,
+      "loss": 1.0437,
+      "step": 1542
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.654841997200215,
+      "learning_rate": 1.650182131091332e-05,
+      "loss": 1.0051,
+      "step": 1543
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7241601293504484,
+      "learning_rate": 1.6497085986833252e-05,
+      "loss": 1.0437,
+      "step": 1544
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7143242308415929,
+      "learning_rate": 1.6492348140360704e-05,
+      "loss": 1.0226,
+      "step": 1545
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7738754828800314,
+      "learning_rate": 1.6487607773335074e-05,
+      "loss": 1.0168,
+      "step": 1546
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5593718512937528,
+      "learning_rate": 1.648286488759673e-05,
+      "loss": 1.0294,
+      "step": 1547
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6897109810380279,
+      "learning_rate": 1.6478119484987026e-05,
+      "loss": 1.0629,
+      "step": 1548
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7388329149132477,
+      "learning_rate": 1.6473371567348287e-05,
+      "loss": 1.0637,
+      "step": 1549
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7083185392338965,
+      "learning_rate": 1.6468621136523823e-05,
+      "loss": 1.0384,
+      "step": 1550
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7209675509605282,
+      "learning_rate": 1.646386819435791e-05,
+      "loss": 1.0023,
+      "step": 1551
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5536459323169263,
+      "learning_rate": 1.6459112742695807e-05,
+      "loss": 1.0644,
+      "step": 1552
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6991775305025228,
+      "learning_rate": 1.6454354783383748e-05,
+      "loss": 1.0208,
+      "step": 1553
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.725658771663436,
+      "learning_rate": 1.644959431826893e-05,
+      "loss": 1.0446,
+      "step": 1554
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7586565133943507,
+      "learning_rate": 1.6444831349199528e-05,
+      "loss": 1.0315,
+      "step": 1555
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7247721163437518,
+      "learning_rate": 1.6440065878024697e-05,
+      "loss": 1.0151,
+      "step": 1556
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7786562966148459,
+      "learning_rate": 1.6435297906594553e-05,
+      "loss": 1.0334,
+      "step": 1557
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5771461672200883,
+      "learning_rate": 1.643052743676019e-05,
+      "loss": 1.0706,
+      "step": 1558
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5006192837733717,
+      "learning_rate": 1.6425754470373667e-05,
+      "loss": 1.0872,
+      "step": 1559
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5170389575921605,
+      "learning_rate": 1.642097900928801e-05,
+      "loss": 1.1198,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5283672586177772,
+      "learning_rate": 1.6416201055357225e-05,
+      "loss": 1.0428,
+      "step": 1561
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.697144753294787,
+      "learning_rate": 1.641142061043627e-05,
+      "loss": 1.0241,
+      "step": 1562
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5285414738789772,
+      "learning_rate": 1.640663767638108e-05,
+      "loss": 1.0646,
+      "step": 1563
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.545438357305353,
+      "learning_rate": 1.6401852255048564e-05,
+      "loss": 1.1253,
+      "step": 1564
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7240416703406276,
+      "learning_rate": 1.6397064348296578e-05,
+      "loss": 1.0273,
+      "step": 1565
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6677544886826058,
+      "learning_rate": 1.6392273957983955e-05,
+      "loss": 1.0814,
+      "step": 1566
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6857022295197894,
+      "learning_rate": 1.638748108597049e-05,
+      "loss": 1.127,
+      "step": 1567
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6637025397891626,
+      "learning_rate": 1.6382685734116934e-05,
+      "loss": 1.0378,
+      "step": 1568
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6564170842283675,
+      "learning_rate": 1.6377887904285018e-05,
+      "loss": 1.073,
+      "step": 1569
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6377782603784687,
+      "learning_rate": 1.637308759833742e-05,
+      "loss": 1.0331,
+      "step": 1570
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5212942422371103,
+      "learning_rate": 1.6368284818137787e-05,
+      "loss": 1.0878,
+      "step": 1571
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7853965377131532,
+      "learning_rate": 1.636347956555072e-05,
+      "loss": 1.1165,
+      "step": 1572
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6888594133725358,
+      "learning_rate": 1.635867184244178e-05,
+      "loss": 1.078,
+      "step": 1573
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.712242188410015,
+      "learning_rate": 1.63538616506775e-05,
+      "loss": 1.0853,
+      "step": 1574
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7297004290183252,
+      "learning_rate": 1.6349048992125358e-05,
+      "loss": 1.092,
+      "step": 1575
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6627856134391341,
+      "learning_rate": 1.634423386865379e-05,
+      "loss": 1.0528,
+      "step": 1576
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5479750764873884,
+      "learning_rate": 1.6339416282132196e-05,
+      "loss": 1.0861,
+      "step": 1577
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6807582870022684,
+      "learning_rate": 1.633459623443093e-05,
+      "loss": 1.0627,
+      "step": 1578
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6841642325896862,
+      "learning_rate": 1.6329773727421297e-05,
+      "loss": 1.0027,
+      "step": 1579
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7602049726719236,
+      "learning_rate": 1.6324948762975567e-05,
+      "loss": 1.0574,
+      "step": 1580
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6540402808013834,
+      "learning_rate": 1.632012134296695e-05,
+      "loss": 1.0234,
+      "step": 1581
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6784244417481702,
+      "learning_rate": 1.6315291469269617e-05,
+      "loss": 0.9955,
+      "step": 1582
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7096828003774142,
+      "learning_rate": 1.63104591437587e-05,
+      "loss": 1.0292,
+      "step": 1583
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7184964945272848,
+      "learning_rate": 1.6305624368310265e-05,
+      "loss": 1.0808,
+      "step": 1584
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7211375766639364,
+      "learning_rate": 1.630078714480134e-05,
+      "loss": 1.0724,
+      "step": 1585
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6780784669620709,
+      "learning_rate": 1.6295947475109904e-05,
+      "loss": 1.0702,
+      "step": 1586
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6620690505395228,
+      "learning_rate": 1.629110536111488e-05,
+      "loss": 1.0617,
+      "step": 1587
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6795119586987924,
+      "learning_rate": 1.628626080469615e-05,
+      "loss": 1.0662,
+      "step": 1588
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6993870849754348,
+      "learning_rate": 1.628141380773453e-05,
+      "loss": 1.0916,
+      "step": 1589
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6482025216818765,
+      "learning_rate": 1.6276564372111797e-05,
+      "loss": 1.0666,
+      "step": 1590
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7190670102736694,
+      "learning_rate": 1.6271712499710663e-05,
+      "loss": 0.9336,
+      "step": 1591
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7336664103192392,
+      "learning_rate": 1.62668581924148e-05,
+      "loss": 1.0196,
+      "step": 1592
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.716456805286363,
+      "learning_rate": 1.6262001452108807e-05,
+      "loss": 0.9967,
+      "step": 1593
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7159791215272355,
+      "learning_rate": 1.6257142280678247e-05,
+      "loss": 1.0484,
+      "step": 1594
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6695912271873113,
+      "learning_rate": 1.6252280680009613e-05,
+      "loss": 1.0389,
+      "step": 1595
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6533082976488761,
+      "learning_rate": 1.6247416651990343e-05,
+      "loss": 1.0696,
+      "step": 1596
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6823700387819879,
+      "learning_rate": 1.624255019850883e-05,
+      "loss": 1.0552,
+      "step": 1597
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6715577038128042,
+      "learning_rate": 1.6237681321454387e-05,
+      "loss": 0.967,
+      "step": 1598
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6932765695340884,
+      "learning_rate": 1.623281002271729e-05,
+      "loss": 1.0756,
+      "step": 1599
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6394929244971032,
+      "learning_rate": 1.6227936304188738e-05,
+      "loss": 1.0909,
+      "step": 1600
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7058496054730022,
+      "learning_rate": 1.622306016776088e-05,
+      "loss": 1.1107,
+      "step": 1601
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6898931270299707,
+      "learning_rate": 1.6218181615326795e-05,
+      "loss": 1.0052,
+      "step": 1602
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7203931923446839,
+      "learning_rate": 1.6213300648780515e-05,
+      "loss": 0.9763,
+      "step": 1603
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.691240261384822,
+      "learning_rate": 1.620841727001699e-05,
+      "loss": 1.0397,
+      "step": 1604
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6851004619937164,
+      "learning_rate": 1.6203531480932114e-05,
+      "loss": 1.0638,
+      "step": 1605
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6413213687110341,
+      "learning_rate": 1.619864328342273e-05,
+      "loss": 1.0066,
+      "step": 1606
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7531163554761955,
+      "learning_rate": 1.6193752679386593e-05,
+      "loss": 0.9613,
+      "step": 1607
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.703617661238433,
+      "learning_rate": 1.6188859670722414e-05,
+      "loss": 1.0338,
+      "step": 1608
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7363695948279827,
+      "learning_rate": 1.6183964259329817e-05,
+      "loss": 0.9649,
+      "step": 1609
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6593866410966408,
+      "learning_rate": 1.6179066447109376e-05,
+      "loss": 1.0421,
+      "step": 1610
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7095367045888973,
+      "learning_rate": 1.6174166235962588e-05,
+      "loss": 1.0706,
+      "step": 1611
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.48293830417448463,
+      "learning_rate": 1.6169263627791886e-05,
+      "loss": 1.0922,
+      "step": 1612
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6722160674751712,
+      "learning_rate": 1.616435862450063e-05,
+      "loss": 1.0634,
+      "step": 1613
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5211004225458633,
+      "learning_rate": 1.615945122799311e-05,
+      "loss": 1.0833,
+      "step": 1614
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.33454431231098136,
+      "learning_rate": 1.6154541440174547e-05,
+      "loss": 1.0492,
+      "step": 1615
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6551004470550432,
+      "learning_rate": 1.614962926295109e-05,
+      "loss": 0.9574,
+      "step": 1616
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6871029058996572,
+      "learning_rate": 1.6144714698229814e-05,
+      "loss": 1.0346,
+      "step": 1617
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6907051847243573,
+      "learning_rate": 1.6139797747918725e-05,
+      "loss": 1.0542,
+      "step": 1618
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6484100553207965,
+      "learning_rate": 1.613487841392675e-05,
+      "loss": 0.9858,
+      "step": 1619
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6470732117494323,
+      "learning_rate": 1.612995669816375e-05,
+      "loss": 0.9318,
+      "step": 1620
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6715451903466869,
+      "learning_rate": 1.6125032602540492e-05,
+      "loss": 1.0626,
+      "step": 1621
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6946399900556762,
+      "learning_rate": 1.6120106128968686e-05,
+      "loss": 1.045,
+      "step": 1622
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7187421864796824,
+      "learning_rate": 1.6115177279360965e-05,
+      "loss": 0.9679,
+      "step": 1623
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.703171506913439,
+      "learning_rate": 1.611024605563087e-05,
+      "loss": 1.0245,
+      "step": 1624
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7263339209083162,
+      "learning_rate": 1.610531245969287e-05,
+      "loss": 1.0757,
+      "step": 1625
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.660162363728499,
+      "learning_rate": 1.6100376493462368e-05,
+      "loss": 1.0403,
+      "step": 1626
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7787151368622436,
+      "learning_rate": 1.6095438158855668e-05,
+      "loss": 1.0269,
+      "step": 1627
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7245712705222395,
+      "learning_rate": 1.609049745779e-05,
+      "loss": 1.0279,
+      "step": 1628
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6995052884298023,
+      "learning_rate": 1.6085554392183517e-05,
+      "loss": 1.0123,
+      "step": 1629
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7051043374824195,
+      "learning_rate": 1.608060896395529e-05,
+      "loss": 1.0744,
+      "step": 1630
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6449401579542466,
+      "learning_rate": 1.60756611750253e-05,
+      "loss": 1.0574,
+      "step": 1631
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7553704638806292,
+      "learning_rate": 1.6070711027314446e-05,
+      "loss": 1.0719,
+      "step": 1632
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6453705793836575,
+      "learning_rate": 1.606575852274456e-05,
+      "loss": 1.0396,
+      "step": 1633
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7105161203735318,
+      "learning_rate": 1.6060803663238357e-05,
+      "loss": 1.0456,
+      "step": 1634
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7091412668854613,
+      "learning_rate": 1.6055846450719498e-05,
+      "loss": 1.0754,
+      "step": 1635
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6977156226440419,
+      "learning_rate": 1.6050886887112535e-05,
+      "loss": 0.9565,
+      "step": 1636
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6417173959987239,
+      "learning_rate": 1.6045924974342945e-05,
+      "loss": 1.0536,
+      "step": 1637
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6812332063362817,
+      "learning_rate": 1.604096071433711e-05,
+      "loss": 0.9916,
+      "step": 1638
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7092193382744374,
+      "learning_rate": 1.6035994109022333e-05,
+      "loss": 1.0489,
+      "step": 1639
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7638347931235374,
+      "learning_rate": 1.6031025160326814e-05,
+      "loss": 1.0042,
+      "step": 1640
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5346520967104808,
+      "learning_rate": 1.6026053870179678e-05,
+      "loss": 1.0744,
+      "step": 1641
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8310234357888315,
+      "learning_rate": 1.6021080240510943e-05,
+      "loss": 1.0519,
+      "step": 1642
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5115481626200947,
+      "learning_rate": 1.601610427325155e-05,
+      "loss": 1.0226,
+      "step": 1643
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6904619802374696,
+      "learning_rate": 1.6011125970333333e-05,
+      "loss": 1.0308,
+      "step": 1644
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5106128978443402,
+      "learning_rate": 1.600614533368905e-05,
+      "loss": 0.9571,
+      "step": 1645
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6659129035691933,
+      "learning_rate": 1.6001162365252348e-05,
+      "loss": 0.9973,
+      "step": 1646
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.715896517568214,
+      "learning_rate": 1.5996177066957787e-05,
+      "loss": 1.0604,
+      "step": 1647
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7454006997707282,
+      "learning_rate": 1.5991189440740838e-05,
+      "loss": 1.017,
+      "step": 1648
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6368966081146623,
+      "learning_rate": 1.5986199488537867e-05,
+      "loss": 1.0301,
+      "step": 1649
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6960521842754035,
+      "learning_rate": 1.598120721228614e-05,
+      "loss": 1.065,
+      "step": 1650
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6720162299362776,
+      "learning_rate": 1.5976212613923836e-05,
+      "loss": 1.0941,
+      "step": 1651
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7051624920139059,
+      "learning_rate": 1.5971215695390026e-05,
+      "loss": 0.9902,
+      "step": 1652
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7435260707969217,
+      "learning_rate": 1.5966216458624692e-05,
+      "loss": 1.0875,
+      "step": 1653
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7789732767152322,
+      "learning_rate": 1.5961214905568705e-05,
+      "loss": 0.9624,
+      "step": 1654
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6474549309105138,
+      "learning_rate": 1.595621103816384e-05,
+      "loss": 0.9714,
+      "step": 1655
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8117396609588702,
+      "learning_rate": 1.5951204858352772e-05,
+      "loss": 1.0338,
+      "step": 1656
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7342731344442043,
+      "learning_rate": 1.594619636807907e-05,
+      "loss": 1.1064,
+      "step": 1657
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.739432040793497,
+      "learning_rate": 1.5941185569287206e-05,
+      "loss": 1.0002,
+      "step": 1658
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.530475250577182,
+      "learning_rate": 1.5936172463922542e-05,
+      "loss": 1.0677,
+      "step": 1659
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6419787394485569,
+      "learning_rate": 1.593115705393134e-05,
+      "loss": 1.0234,
+      "step": 1660
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6677713619465611,
+      "learning_rate": 1.5926139341260755e-05,
+      "loss": 1.0523,
+      "step": 1661
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6656517858432675,
+      "learning_rate": 1.5921119327858835e-05,
+      "loss": 1.0355,
+      "step": 1662
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7158520864912428,
+      "learning_rate": 1.5916097015674518e-05,
+      "loss": 1.074,
+      "step": 1663
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6774419575193262,
+      "learning_rate": 1.5911072406657646e-05,
+      "loss": 1.033,
+      "step": 1664
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8067846873162285,
+      "learning_rate": 1.5906045502758943e-05,
+      "loss": 0.987,
+      "step": 1665
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.48830022174358817,
+      "learning_rate": 1.590101630593002e-05,
+      "loss": 1.0493,
+      "step": 1666
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6401477602174555,
+      "learning_rate": 1.5895984818123392e-05,
+      "loss": 1.0692,
+      "step": 1667
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6798592789317833,
+      "learning_rate": 1.5890951041292453e-05,
+      "loss": 0.9943,
+      "step": 1668
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7241010256635895,
+      "learning_rate": 1.588591497739149e-05,
+      "loss": 1.0543,
+      "step": 1669
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7268313527378797,
+      "learning_rate": 1.5880876628375668e-05,
+      "loss": 0.9133,
+      "step": 1670
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7363012883022247,
+      "learning_rate": 1.587583599620106e-05,
+      "loss": 1.0719,
+      "step": 1671
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7698703845301761,
+      "learning_rate": 1.5870793082824604e-05,
+      "loss": 1.0854,
+      "step": 1672
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7402369594366365,
+      "learning_rate": 1.5865747890204138e-05,
+      "loss": 1.0983,
+      "step": 1673
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6828099417151442,
+      "learning_rate": 1.5860700420298377e-05,
+      "loss": 1.0236,
+      "step": 1674
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6811858297180327,
+      "learning_rate": 1.5855650675066924e-05,
+      "loss": 1.0282,
+      "step": 1675
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6793429043640745,
+      "learning_rate": 1.5850598656470265e-05,
+      "loss": 1.0584,
+      "step": 1676
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.666440740399364,
+      "learning_rate": 1.584554436646976e-05,
+      "loss": 1.0182,
+      "step": 1677
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6736435084998303,
+      "learning_rate": 1.5840487807027665e-05,
+      "loss": 1.0894,
+      "step": 1678
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6002559487763981,
+      "learning_rate": 1.5835428980107113e-05,
+      "loss": 1.0449,
+      "step": 1679
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7600458447053444,
+      "learning_rate": 1.583036788767211e-05,
+      "loss": 1.0616,
+      "step": 1680
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7056017288819727,
+      "learning_rate": 1.5825304531687548e-05,
+      "loss": 1.053,
+      "step": 1681
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6788126989904656,
+      "learning_rate": 1.5820238914119195e-05,
+      "loss": 1.0326,
+      "step": 1682
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7880314475168942,
+      "learning_rate": 1.5815171036933697e-05,
+      "loss": 1.1007,
+      "step": 1683
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5188011658596241,
+      "learning_rate": 1.5810100902098582e-05,
+      "loss": 1.0467,
+      "step": 1684
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6872623644235665,
+      "learning_rate": 1.580502851158225e-05,
+      "loss": 0.9984,
+      "step": 1685
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.690864613624932,
+      "learning_rate": 1.5799953867353975e-05,
+      "loss": 1.0784,
+      "step": 1686
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5649677841474663,
+      "learning_rate": 1.579487697138391e-05,
+      "loss": 1.0474,
+      "step": 1687
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5953271192065278,
+      "learning_rate": 1.5789797825643086e-05,
+      "loss": 1.048,
+      "step": 1688
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.75615468044511,
+      "learning_rate": 1.5784716432103394e-05,
+      "loss": 0.9844,
+      "step": 1689
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7043653830637469,
+      "learning_rate": 1.5779632792737608e-05,
+      "loss": 1.0199,
+      "step": 1690
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6590240141021487,
+      "learning_rate": 1.5774546909519376e-05,
+      "loss": 1.0397,
+      "step": 1691
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7594345654844806,
+      "learning_rate": 1.5769458784423206e-05,
+      "loss": 1.1103,
+      "step": 1692
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8067350570066786,
+      "learning_rate": 1.5764368419424488e-05,
+      "loss": 0.9961,
+      "step": 1693
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7359582449562777,
+      "learning_rate": 1.575927581649948e-05,
+      "loss": 1.0953,
+      "step": 1694
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6996709375493255,
+      "learning_rate": 1.5754180977625303e-05,
+      "loss": 1.0486,
+      "step": 1695
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7014684569729547,
+      "learning_rate": 1.574908390477995e-05,
+      "loss": 1.0289,
+      "step": 1696
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6257137054066295,
+      "learning_rate": 1.5743984599942273e-05,
+      "loss": 0.9976,
+      "step": 1697
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.712839143316764,
+      "learning_rate": 1.5738883065092005e-05,
+      "loss": 1.0571,
+      "step": 1698
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7082020686156301,
+      "learning_rate": 1.5733779302209735e-05,
+      "loss": 0.9736,
+      "step": 1699
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7183320589177471,
+      "learning_rate": 1.572867331327692e-05,
+      "loss": 1.0263,
+      "step": 1700
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7144553862616476,
+      "learning_rate": 1.5723565100275884e-05,
+      "loss": 0.9494,
+      "step": 1701
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6703995452632951,
+      "learning_rate": 1.5718454665189806e-05,
+      "loss": 1.1112,
+      "step": 1702
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7702903206297822,
+      "learning_rate": 1.5713342010002733e-05,
+      "loss": 1.0609,
+      "step": 1703
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6740513490540173,
+      "learning_rate": 1.5708227136699578e-05,
+      "loss": 0.9521,
+      "step": 1704
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6228006100799971,
+      "learning_rate": 1.5703110047266105e-05,
+      "loss": 0.9829,
+      "step": 1705
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7255232918808024,
+      "learning_rate": 1.569799074368895e-05,
+      "loss": 1.1731,
+      "step": 1706
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7487137818951232,
+      "learning_rate": 1.5692869227955603e-05,
+      "loss": 1.0361,
+      "step": 1707
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.545392806046918,
+      "learning_rate": 1.5687745502054407e-05,
+      "loss": 1.0458,
+      "step": 1708
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7238037778182084,
+      "learning_rate": 1.5682619567974575e-05,
+      "loss": 0.9442,
+      "step": 1709
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7161419244771123,
+      "learning_rate": 1.567749142770617e-05,
+      "loss": 1.0082,
+      "step": 1710
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6418085051939154,
+      "learning_rate": 1.5672361083240106e-05,
+      "loss": 1.0355,
+      "step": 1711
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7271418038384121,
+      "learning_rate": 1.5667228536568167e-05,
+      "loss": 0.988,
+      "step": 1712
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7916207267299596,
+      "learning_rate": 1.566209378968298e-05,
+      "loss": 1.0176,
+      "step": 1713
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6708983332974949,
+      "learning_rate": 1.565695684457803e-05,
+      "loss": 1.0756,
+      "step": 1714
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7125982868649606,
+      "learning_rate": 1.5651817703247666e-05,
+      "loss": 1.0974,
+      "step": 1715
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6266353029549251,
+      "learning_rate": 1.5646676367687067e-05,
+      "loss": 1.1074,
+      "step": 1716
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6864832486915363,
+      "learning_rate": 1.564153283989228e-05,
+      "loss": 1.1072,
+      "step": 1717
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7070546915730359,
+      "learning_rate": 1.5636387121860207e-05,
+      "loss": 1.0814,
+      "step": 1718
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.716501098577317,
+      "learning_rate": 1.5631239215588578e-05,
+      "loss": 1.069,
+      "step": 1719
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7376031411830272,
+      "learning_rate": 1.5626089123076004e-05,
+      "loss": 0.9944,
+      "step": 1720
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7343334708851381,
+      "learning_rate": 1.5620936846321917e-05,
+      "loss": 1.0007,
+      "step": 1721
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.586157317019247,
+      "learning_rate": 1.561578238732661e-05,
+      "loss": 1.1089,
+      "step": 1722
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7148358593917418,
+      "learning_rate": 1.561062574809123e-05,
+      "loss": 1.0345,
+      "step": 1723
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.718866624448896,
+      "learning_rate": 1.5605466930617747e-05,
+      "loss": 1.0004,
+      "step": 1724
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6848970964123652,
+      "learning_rate": 1.5600305936909005e-05,
+      "loss": 1.0636,
+      "step": 1725
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6376178949544583,
+      "learning_rate": 1.559514276896867e-05,
+      "loss": 1.0847,
+      "step": 1726
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7559675973726596,
+      "learning_rate": 1.558997742880127e-05,
+      "loss": 1.033,
+      "step": 1727
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7346523184569304,
+      "learning_rate": 1.5584809918412158e-05,
+      "loss": 1.059,
+      "step": 1728
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.699926627923378,
+      "learning_rate": 1.557964023980755e-05,
+      "loss": 0.9559,
+      "step": 1729
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7424484441894904,
+      "learning_rate": 1.5574468394994486e-05,
+      "loss": 1.1169,
+      "step": 1730
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6988126704591366,
+      "learning_rate": 1.5569294385980856e-05,
+      "loss": 0.9833,
+      "step": 1731
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7337283730443588,
+      "learning_rate": 1.556411821477539e-05,
+      "loss": 1.0771,
+      "step": 1732
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6486469951497986,
+      "learning_rate": 1.5558939883387657e-05,
+      "loss": 1.1376,
+      "step": 1733
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7763266458051565,
+      "learning_rate": 1.5553759393828058e-05,
+      "loss": 1.0256,
+      "step": 1734
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.690214417006169,
+      "learning_rate": 1.554857674810784e-05,
+      "loss": 1.0031,
+      "step": 1735
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6989925059618346,
+      "learning_rate": 1.554339194823909e-05,
+      "loss": 1.0687,
+      "step": 1736
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7587635444636018,
+      "learning_rate": 1.553820499623472e-05,
+      "loss": 1.0663,
+      "step": 1737
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7321007474416522,
+      "learning_rate": 1.553301589410848e-05,
+      "loss": 1.1216,
+      "step": 1738
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6983174960330968,
+      "learning_rate": 1.5527824643874968e-05,
+      "loss": 0.986,
+      "step": 1739
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8078489177398049,
+      "learning_rate": 1.5522631247549598e-05,
+      "loss": 1.0635,
+      "step": 1740
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7705011340101637,
+      "learning_rate": 1.5517435707148628e-05,
+      "loss": 1.0725,
+      "step": 1741
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7127590262527242,
+      "learning_rate": 1.5512238024689144e-05,
+      "loss": 1.0341,
+      "step": 1742
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6702468492785513,
+      "learning_rate": 1.550703820218907e-05,
+      "loss": 1.0275,
+      "step": 1743
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7307805067834912,
+      "learning_rate": 1.550183624166715e-05,
+      "loss": 1.097,
+      "step": 1744
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7245614355320019,
+      "learning_rate": 1.549663214514297e-05,
+      "loss": 1.1442,
+      "step": 1745
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7871589569420303,
+      "learning_rate": 1.5491425914636934e-05,
+      "loss": 1.0906,
+      "step": 1746
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6875358908366739,
+      "learning_rate": 1.5486217552170283e-05,
+      "loss": 1.0099,
+      "step": 1747
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.715735522240056,
+      "learning_rate": 1.548100705976508e-05,
+      "loss": 1.076,
+      "step": 1748
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6514923591773827,
+      "learning_rate": 1.5475794439444226e-05,
+      "loss": 1.0839,
+      "step": 1749
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5770961900223569,
+      "learning_rate": 1.5470579693231432e-05,
+      "loss": 1.0562,
+      "step": 1750
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6737817111068336,
+      "learning_rate": 1.5465362823151245e-05,
+      "loss": 1.1042,
+      "step": 1751
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.725098680375221,
+      "learning_rate": 1.5460143831229026e-05,
+      "loss": 1.0298,
+      "step": 1752
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5496698222048226,
+      "learning_rate": 1.545492271949098e-05,
+      "loss": 1.0546,
+      "step": 1753
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7152667211851019,
+      "learning_rate": 1.544969948996411e-05,
+      "loss": 1.0235,
+      "step": 1754
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7125062050206552,
+      "learning_rate": 1.544447414467626e-05,
+      "loss": 1.066,
+      "step": 1755
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6727503850271641,
+      "learning_rate": 1.5439246685656093e-05,
+      "loss": 1.0364,
+      "step": 1756
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6955173636313178,
+      "learning_rate": 1.5434017114933082e-05,
+      "loss": 1.0233,
+      "step": 1757
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6847940118434955,
+      "learning_rate": 1.5428785434537527e-05,
+      "loss": 0.9996,
+      "step": 1758
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.724016405147181,
+      "learning_rate": 1.542355164650055e-05,
+      "loss": 1.0524,
+      "step": 1759
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7177292874251536,
+      "learning_rate": 1.541831575285408e-05,
+      "loss": 1.0695,
+      "step": 1760
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.716565078031637,
+      "learning_rate": 1.541307775563088e-05,
+      "loss": 1.0714,
+      "step": 1761
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5470457049554611,
+      "learning_rate": 1.540783765686452e-05,
+      "loss": 1.0644,
+      "step": 1762
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7930101938145686,
+      "learning_rate": 1.540259545858938e-05,
+      "loss": 1.052,
+      "step": 1763
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6623612186492775,
+      "learning_rate": 1.539735116284067e-05,
+      "loss": 1.0856,
+      "step": 1764
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6841672196017935,
+      "learning_rate": 1.53921047716544e-05,
+      "loss": 0.9412,
+      "step": 1765
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6994287136934914,
+      "learning_rate": 1.53868562870674e-05,
+      "loss": 1.0234,
+      "step": 1766
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7089842252245602,
+      "learning_rate": 1.5381605711117318e-05,
+      "loss": 1.0087,
+      "step": 1767
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7320427405491433,
+      "learning_rate": 1.5376353045842604e-05,
+      "loss": 1.1392,
+      "step": 1768
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6582547222920343,
+      "learning_rate": 1.5371098293282526e-05,
+      "loss": 1.1064,
+      "step": 1769
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.677573391750281,
+      "learning_rate": 1.5365841455477158e-05,
+      "loss": 1.0187,
+      "step": 1770
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7713924669867702,
+      "learning_rate": 1.5360582534467382e-05,
+      "loss": 1.0258,
+      "step": 1771
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6593048281847949,
+      "learning_rate": 1.5355321532294897e-05,
+      "loss": 1.0181,
+      "step": 1772
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.607616286411078,
+      "learning_rate": 1.5350058451002204e-05,
+      "loss": 1.0159,
+      "step": 1773
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6460274117091304,
+      "learning_rate": 1.5344793292632614e-05,
+      "loss": 1.0039,
+      "step": 1774
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6810150916334677,
+      "learning_rate": 1.533952605923024e-05,
+      "loss": 1.0591,
+      "step": 1775
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6309075873132964,
+      "learning_rate": 1.5334256752840007e-05,
+      "loss": 1.0251,
+      "step": 1776
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7808876379057069,
+      "learning_rate": 1.532898537550764e-05,
+      "loss": 1.0403,
+      "step": 1777
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7169727895660671,
+      "learning_rate": 1.532371192927966e-05,
+      "loss": 0.9992,
+      "step": 1778
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7510713361999458,
+      "learning_rate": 1.5318436416203412e-05,
+      "loss": 1.033,
+      "step": 1779
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6557875001518657,
+      "learning_rate": 1.531315883832703e-05,
+      "loss": 1.0024,
+      "step": 1780
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.720774248678274,
+      "learning_rate": 1.530787919769945e-05,
+      "loss": 1.049,
+      "step": 1781
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6980820236281271,
+      "learning_rate": 1.5302597496370408e-05,
+      "loss": 1.0632,
+      "step": 1782
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7168743547717811,
+      "learning_rate": 1.5297313736390447e-05,
+      "loss": 1.0544,
+      "step": 1783
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6432194742794173,
+      "learning_rate": 1.5292027919810898e-05,
+      "loss": 1.0741,
+      "step": 1784
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8143264986711487,
+      "learning_rate": 1.52867400486839e-05,
+      "loss": 1.0566,
+      "step": 1785
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7669172610257309,
+      "learning_rate": 1.528145012506239e-05,
+      "loss": 1.0007,
+      "step": 1786
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6501979119196134,
+      "learning_rate": 1.5276158151000096e-05,
+      "loss": 1.0574,
+      "step": 1787
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.738020041090958,
+      "learning_rate": 1.5270864128551542e-05,
+      "loss": 1.0211,
+      "step": 1788
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.696070004290682,
+      "learning_rate": 1.5265568059772053e-05,
+      "loss": 1.0055,
+      "step": 1789
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7178524809239045,
+      "learning_rate": 1.5260269946717746e-05,
+      "loss": 1.1423,
+      "step": 1790
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6979187471939786,
+      "learning_rate": 1.5254969791445526e-05,
+      "loss": 0.9507,
+      "step": 1791
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6897344438825576,
+      "learning_rate": 1.5249667596013102e-05,
+      "loss": 1.0753,
+      "step": 1792
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7657199773748681,
+      "learning_rate": 1.5244363362478967e-05,
+      "loss": 1.06,
+      "step": 1793
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5435614487558563,
+      "learning_rate": 1.5239057092902404e-05,
+      "loss": 1.0641,
+      "step": 1794
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.720212832234242,
+      "learning_rate": 1.523374878934349e-05,
+      "loss": 1.1118,
+      "step": 1795
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7732073474472889,
+      "learning_rate": 1.5228438453863095e-05,
+      "loss": 1.051,
+      "step": 1796
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6668376376188788,
+      "learning_rate": 1.522312608852287e-05,
+      "loss": 1.0527,
+      "step": 1797
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6756887996998949,
+      "learning_rate": 1.5217811695385263e-05,
+      "loss": 1.0758,
+      "step": 1798
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6911283612928538,
+      "learning_rate": 1.52124952765135e-05,
+      "loss": 1.0579,
+      "step": 1799
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6865636470732135,
+      "learning_rate": 1.5207176833971598e-05,
+      "loss": 1.0326,
+      "step": 1800
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6596334080603731,
+      "learning_rate": 1.520185636982436e-05,
+      "loss": 1.0392,
+      "step": 1801
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5092553085176,
+      "learning_rate": 1.5196533886137376e-05,
+      "loss": 1.0174,
+      "step": 1802
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4787085647068552,
+      "learning_rate": 1.5191209384977014e-05,
+      "loss": 1.0956,
+      "step": 1803
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5530944077215147,
+      "learning_rate": 1.5185882868410431e-05,
+      "loss": 1.1085,
+      "step": 1804
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6799767535669762,
+      "learning_rate": 1.5180554338505564e-05,
+      "loss": 0.9811,
+      "step": 1805
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6998981222574717,
+      "learning_rate": 1.517522379733113e-05,
+      "loss": 1.0501,
+      "step": 1806
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7018492138463424,
+      "learning_rate": 1.5169891246956629e-05,
+      "loss": 1.0283,
+      "step": 1807
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5512991688397215,
+      "learning_rate": 1.5164556689452346e-05,
+      "loss": 1.0446,
+      "step": 1808
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6736528843315573,
+      "learning_rate": 1.5159220126889329e-05,
+      "loss": 1.0512,
+      "step": 1809
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.719754680745962,
+      "learning_rate": 1.5153881561339426e-05,
+      "loss": 0.9935,
+      "step": 1810
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7339989756845233,
+      "learning_rate": 1.5148540994875242e-05,
+      "loss": 1.027,
+      "step": 1811
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7484967740515159,
+      "learning_rate": 1.5143198429570181e-05,
+      "loss": 1.0781,
+      "step": 1812
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6789244633930611,
+      "learning_rate": 1.5137853867498403e-05,
+      "loss": 1.0351,
+      "step": 1813
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7485076043528918,
+      "learning_rate": 1.5132507310734847e-05,
+      "loss": 1.0626,
+      "step": 1814
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7302896410106126,
+      "learning_rate": 1.5127158761355241e-05,
+      "loss": 1.0312,
+      "step": 1815
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7893064486781924,
+      "learning_rate": 1.512180822143607e-05,
+      "loss": 1.0667,
+      "step": 1816
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7047328393936543,
+      "learning_rate": 1.5116455693054594e-05,
+      "loss": 1.12,
+      "step": 1817
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7358767578802544,
+      "learning_rate": 1.5111101178288858e-05,
+      "loss": 0.9743,
+      "step": 1818
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6980337658422958,
+      "learning_rate": 1.510574467921766e-05,
+      "loss": 1.0038,
+      "step": 1819
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7911863602737563,
+      "learning_rate": 1.5100386197920585e-05,
+      "loss": 1.0293,
+      "step": 1820
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7228251652493624,
+      "learning_rate": 1.5095025736477977e-05,
+      "loss": 1.1112,
+      "step": 1821
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6976245901070317,
+      "learning_rate": 1.5089663296970952e-05,
+      "loss": 1.0325,
+      "step": 1822
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7344493969982706,
+      "learning_rate": 1.5084298881481388e-05,
+      "loss": 1.0101,
+      "step": 1823
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6356095221022264,
+      "learning_rate": 1.5078932492091942e-05,
+      "loss": 1.0165,
+      "step": 1824
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6981342774215605,
+      "learning_rate": 1.5073564130886032e-05,
+      "loss": 1.056,
+      "step": 1825
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7154203433402017,
+      "learning_rate": 1.506819379994784e-05,
+      "loss": 1.1021,
+      "step": 1826
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7628124667738517,
+      "learning_rate": 1.5062821501362308e-05,
+      "loss": 1.0308,
+      "step": 1827
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6729591989705712,
+      "learning_rate": 1.5057447237215152e-05,
+      "loss": 1.0688,
+      "step": 1828
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6674001356469933,
+      "learning_rate": 1.5052071009592846e-05,
+      "loss": 0.9651,
+      "step": 1829
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6698447271308224,
+      "learning_rate": 1.5046692820582625e-05,
+      "loss": 1.0038,
+      "step": 1830
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7055124566763724,
+      "learning_rate": 1.504131267227249e-05,
+      "loss": 1.0254,
+      "step": 1831
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6768216343095864,
+      "learning_rate": 1.5035930566751198e-05,
+      "loss": 1.0318,
+      "step": 1832
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7045794910041042,
+      "learning_rate": 1.5030546506108268e-05,
+      "loss": 1.0919,
+      "step": 1833
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7095469628856077,
+      "learning_rate": 1.5025160492433976e-05,
+      "loss": 1.0587,
+      "step": 1834
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7173523084729694,
+      "learning_rate": 1.501977252781936e-05,
+      "loss": 1.0908,
+      "step": 1835
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7121900227195952,
+      "learning_rate": 1.5014382614356213e-05,
+      "loss": 0.9841,
+      "step": 1836
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.706767957403425,
+      "learning_rate": 1.5008990754137088e-05,
+      "loss": 0.9954,
+      "step": 1837
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7251989775364378,
+      "learning_rate": 1.5003596949255284e-05,
+      "loss": 1.0726,
+      "step": 1838
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7362080386089945,
+      "learning_rate": 1.4998201201804867e-05,
+      "loss": 1.0806,
+      "step": 1839
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5161553894181269,
+      "learning_rate": 1.499280351388065e-05,
+      "loss": 1.0341,
+      "step": 1840
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6916909638452348,
+      "learning_rate": 1.49874038875782e-05,
+      "loss": 0.9904,
+      "step": 1841
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.35999881645641635,
+      "learning_rate": 1.498200232499384e-05,
+      "loss": 1.1356,
+      "step": 1842
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6526220393567851,
+      "learning_rate": 1.4976598828224643e-05,
+      "loss": 1.0536,
+      "step": 1843
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5310275363117933,
+      "learning_rate": 1.497119339936843e-05,
+      "loss": 1.1199,
+      "step": 1844
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7320873642396646,
+      "learning_rate": 1.4965786040523779e-05,
+      "loss": 1.0537,
+      "step": 1845
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6359104452942798,
+      "learning_rate": 1.496037675379001e-05,
+      "loss": 1.0906,
+      "step": 1846
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6733184490460234,
+      "learning_rate": 1.4954965541267192e-05,
+      "loss": 1.0299,
+      "step": 1847
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6749500891545874,
+      "learning_rate": 1.494955240505615e-05,
+      "loss": 1.0521,
+      "step": 1848
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7064425027625233,
+      "learning_rate": 1.494413734725844e-05,
+      "loss": 1.0865,
+      "step": 1849
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7906933362534012,
+      "learning_rate": 1.4938720369976385e-05,
+      "loss": 1.0671,
+      "step": 1850
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6446867605391627,
+      "learning_rate": 1.4933301475313036e-05,
+      "loss": 1.0662,
+      "step": 1851
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6564673541647501,
+      "learning_rate": 1.4927880665372197e-05,
+      "loss": 0.9843,
+      "step": 1852
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7509503467439869,
+      "learning_rate": 1.4922457942258411e-05,
+      "loss": 1.0283,
+      "step": 1853
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5254578424412988,
+      "learning_rate": 1.4917033308076967e-05,
+      "loss": 1.0457,
+      "step": 1854
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.688685067471726,
+      "learning_rate": 1.4911606764933892e-05,
+      "loss": 1.0434,
+      "step": 1855
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6626719955655539,
+      "learning_rate": 1.490617831493596e-05,
+      "loss": 0.9938,
+      "step": 1856
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7274686586559642,
+      "learning_rate": 1.4900747960190682e-05,
+      "loss": 1.0389,
+      "step": 1857
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6911077896020353,
+      "learning_rate": 1.489531570280631e-05,
+      "loss": 1.0588,
+      "step": 1858
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6520887344502216,
+      "learning_rate": 1.488988154489183e-05,
+      "loss": 1.0672,
+      "step": 1859
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6673961622122663,
+      "learning_rate": 1.4884445488556972e-05,
+      "loss": 0.9423,
+      "step": 1860
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7067867370478602,
+      "learning_rate": 1.4879007535912198e-05,
+      "loss": 1.0145,
+      "step": 1861
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6707717017563541,
+      "learning_rate": 1.4873567689068708e-05,
+      "loss": 0.991,
+      "step": 1862
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5102251636362444,
+      "learning_rate": 1.4868125950138442e-05,
+      "loss": 1.0586,
+      "step": 1863
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.740846394755799,
+      "learning_rate": 1.4862682321234064e-05,
+      "loss": 1.066,
+      "step": 1864
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6828012783202885,
+      "learning_rate": 1.4857236804468983e-05,
+      "loss": 1.0471,
+      "step": 1865
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6672653825465179,
+      "learning_rate": 1.4851789401957338e-05,
+      "loss": 1.1119,
+      "step": 1866
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5146545963199965,
+      "learning_rate": 1.4846340115813993e-05,
+      "loss": 1.0317,
+      "step": 1867
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5788124500254163,
+      "learning_rate": 1.484088894815455e-05,
+      "loss": 1.1475,
+      "step": 1868
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5188305371664926,
+      "learning_rate": 1.4835435901095341e-05,
+      "loss": 1.0339,
+      "step": 1869
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6643211476690544,
+      "learning_rate": 1.4829980976753426e-05,
+      "loss": 1.0367,
+      "step": 1870
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7441687081199582,
+      "learning_rate": 1.4824524177246597e-05,
+      "loss": 1.0245,
+      "step": 1871
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7017899812976058,
+      "learning_rate": 1.4819065504693365e-05,
+      "loss": 1.0763,
+      "step": 1872
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6635927527663326,
+      "learning_rate": 1.4813604961212984e-05,
+      "loss": 1.0005,
+      "step": 1873
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6753723879181233,
+      "learning_rate": 1.4808142548925417e-05,
+      "loss": 0.9714,
+      "step": 1874
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7103799762273827,
+      "learning_rate": 1.4802678269951365e-05,
+      "loss": 1.0414,
+      "step": 1875
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6698435809584508,
+      "learning_rate": 1.4797212126412243e-05,
+      "loss": 1.0491,
+      "step": 1876
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6941215217079543,
+      "learning_rate": 1.4791744120430202e-05,
+      "loss": 1.0287,
+      "step": 1877
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7419817658801122,
+      "learning_rate": 1.4786274254128112e-05,
+      "loss": 0.9996,
+      "step": 1878
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5666930359732919,
+      "learning_rate": 1.4780802529629559e-05,
+      "loss": 1.0737,
+      "step": 1879
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.789216741357579,
+      "learning_rate": 1.4775328949058856e-05,
+      "loss": 1.0493,
+      "step": 1880
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7080518194994964,
+      "learning_rate": 1.4769853514541037e-05,
+      "loss": 1.0349,
+      "step": 1881
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6992874709099824,
+      "learning_rate": 1.4764376228201848e-05,
+      "loss": 0.9888,
+      "step": 1882
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6186408187321577,
+      "learning_rate": 1.475889709216777e-05,
+      "loss": 1.0329,
+      "step": 1883
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6283165806276544,
+      "learning_rate": 1.4753416108565985e-05,
+      "loss": 1.0378,
+      "step": 1884
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7239441350755581,
+      "learning_rate": 1.47479332795244e-05,
+      "loss": 1.0465,
+      "step": 1885
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7967623936104518,
+      "learning_rate": 1.4742448607171644e-05,
+      "loss": 1.0585,
+      "step": 1886
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7829621195593062,
+      "learning_rate": 1.473696209363705e-05,
+      "loss": 1.0106,
+      "step": 1887
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6771888221103793,
+      "learning_rate": 1.4731473741050673e-05,
+      "loss": 1.0875,
+      "step": 1888
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.702249405015459,
+      "learning_rate": 1.4725983551543279e-05,
+      "loss": 1.0466,
+      "step": 1889
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6924301682657289,
+      "learning_rate": 1.472049152724635e-05,
+      "loss": 0.9764,
+      "step": 1890
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.668190627297071,
+      "learning_rate": 1.471499767029208e-05,
+      "loss": 1.0325,
+      "step": 1891
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6665058870425178,
+      "learning_rate": 1.470950198281337e-05,
+      "loss": 0.9941,
+      "step": 1892
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5745497176823923,
+      "learning_rate": 1.470400446694384e-05,
+      "loss": 1.0702,
+      "step": 1893
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6685424156666162,
+      "learning_rate": 1.4698505124817811e-05,
+      "loss": 1.058,
+      "step": 1894
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.64850935180756,
+      "learning_rate": 1.4693003958570318e-05,
+      "loss": 1.041,
+      "step": 1895
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6514205959469292,
+      "learning_rate": 1.4687500970337103e-05,
+      "loss": 1.045,
+      "step": 1896
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7059486108391244,
+      "learning_rate": 1.4681996162254618e-05,
+      "loss": 1.0066,
+      "step": 1897
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8007654135614347,
+      "learning_rate": 1.4676489536460015e-05,
+      "loss": 1.0742,
+      "step": 1898
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6795664036177733,
+      "learning_rate": 1.467098109509116e-05,
+      "loss": 1.0135,
+      "step": 1899
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6836314435896479,
+      "learning_rate": 1.4665470840286614e-05,
+      "loss": 1.0381,
+      "step": 1900
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.686123518420662,
+      "learning_rate": 1.4659958774185654e-05,
+      "loss": 0.9772,
+      "step": 1901
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6864339962724637,
+      "learning_rate": 1.4654444898928249e-05,
+      "loss": 1.0037,
+      "step": 1902
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6973958065266027,
+      "learning_rate": 1.4648929216655077e-05,
+      "loss": 1.0808,
+      "step": 1903
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6844080347981105,
+      "learning_rate": 1.4643411729507517e-05,
+      "loss": 1.0205,
+      "step": 1904
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6951520141410149,
+      "learning_rate": 1.4637892439627644e-05,
+      "loss": 0.9855,
+      "step": 1905
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7675416411800134,
+      "learning_rate": 1.4632371349158241e-05,
+      "loss": 1.0959,
+      "step": 1906
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.540787477565592,
+      "learning_rate": 1.4626848460242782e-05,
+      "loss": 1.097,
+      "step": 1907
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7247678203387045,
+      "learning_rate": 1.4621323775025444e-05,
+      "loss": 1.0195,
+      "step": 1908
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7421887984561576,
+      "learning_rate": 1.4615797295651099e-05,
+      "loss": 1.0369,
+      "step": 1909
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5194467782525433,
+      "learning_rate": 1.4610269024265317e-05,
+      "loss": 0.975,
+      "step": 1910
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7287403021951159,
+      "learning_rate": 1.4604738963014365e-05,
+      "loss": 1.0545,
+      "step": 1911
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6904541354452836,
+      "learning_rate": 1.4599207114045202e-05,
+      "loss": 1.058,
+      "step": 1912
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.779979309856644,
+      "learning_rate": 1.4593673479505482e-05,
+      "loss": 1.0195,
+      "step": 1913
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7400634893857747,
+      "learning_rate": 1.4588138061543551e-05,
+      "loss": 1.0252,
+      "step": 1914
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6595098021338824,
+      "learning_rate": 1.458260086230845e-05,
+      "loss": 1.028,
+      "step": 1915
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6788009995597014,
+      "learning_rate": 1.4577061883949912e-05,
+      "loss": 1.015,
+      "step": 1916
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7072388076232807,
+      "learning_rate": 1.4571521128618358e-05,
+      "loss": 1.0447,
+      "step": 1917
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3672261418473378,
+      "learning_rate": 1.4565978598464895e-05,
+      "loss": 1.135,
+      "step": 1918
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6694162443892546,
+      "learning_rate": 1.4560434295641338e-05,
+      "loss": 0.9843,
+      "step": 1919
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6909236142541647,
+      "learning_rate": 1.455488822230016e-05,
+      "loss": 1.0351,
+      "step": 1920
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5349200990222406,
+      "learning_rate": 1.4549340380594545e-05,
+      "loss": 1.027,
+      "step": 1921
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6723959016343993,
+      "learning_rate": 1.454379077267836e-05,
+      "loss": 0.961,
+      "step": 1922
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.737700925234033,
+      "learning_rate": 1.4538239400706147e-05,
+      "loss": 0.9077,
+      "step": 1923
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6881118093255555,
+      "learning_rate": 1.4532686266833143e-05,
+      "loss": 1.0065,
+      "step": 1924
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6345424611466187,
+      "learning_rate": 1.4527131373215265e-05,
+      "loss": 1.0339,
+      "step": 1925
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7019932664283679,
+      "learning_rate": 1.4521574722009115e-05,
+      "loss": 1.0158,
+      "step": 1926
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6346316373088203,
+      "learning_rate": 1.4516016315371974e-05,
+      "loss": 1.0616,
+      "step": 1927
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7557687813335799,
+      "learning_rate": 1.4510456155461807e-05,
+      "loss": 1.0804,
+      "step": 1928
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6563342328533653,
+      "learning_rate": 1.4504894244437264e-05,
+      "loss": 1.0309,
+      "step": 1929
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7574344120379441,
+      "learning_rate": 1.4499330584457667e-05,
+      "loss": 0.9662,
+      "step": 1930
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7013357944474894,
+      "learning_rate": 1.4493765177683017e-05,
+      "loss": 1.0405,
+      "step": 1931
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6685585765385241,
+      "learning_rate": 1.4488198026274007e-05,
+      "loss": 1.0352,
+      "step": 1932
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6936465468887111,
+      "learning_rate": 1.4482629132391985e-05,
+      "loss": 1.0098,
+      "step": 1933
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6686918253119841,
+      "learning_rate": 1.4477058498198993e-05,
+      "loss": 1.0801,
+      "step": 1934
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7846103843470108,
+      "learning_rate": 1.4471486125857743e-05,
+      "loss": 1.0912,
+      "step": 1935
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7511673871447248,
+      "learning_rate": 1.446591201753162e-05,
+      "loss": 1.0032,
+      "step": 1936
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6519499580312266,
+      "learning_rate": 1.4460336175384688e-05,
+      "loss": 1.0441,
+      "step": 1937
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6956613015198333,
+      "learning_rate": 1.4454758601581675e-05,
+      "loss": 0.9517,
+      "step": 1938
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.713808607795664,
+      "learning_rate": 1.4449179298287999e-05,
+      "loss": 1.0511,
+      "step": 1939
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8435906026932146,
+      "learning_rate": 1.4443598267669723e-05,
+      "loss": 1.0889,
+      "step": 1940
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6632263052339153,
+      "learning_rate": 1.4438015511893602e-05,
+      "loss": 1.071,
+      "step": 1941
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6225511743186044,
+      "learning_rate": 1.4432431033127056e-05,
+      "loss": 1.0652,
+      "step": 1942
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6826580067686125,
+      "learning_rate": 1.442684483353817e-05,
+      "loss": 1.0703,
+      "step": 1943
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6839166686648893,
+      "learning_rate": 1.4421256915295697e-05,
+      "loss": 0.9963,
+      "step": 1944
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7252262444166802,
+      "learning_rate": 1.4415667280569064e-05,
+      "loss": 0.9639,
+      "step": 1945
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5031213105339801,
+      "learning_rate": 1.4410075931528356e-05,
+      "loss": 1.001,
+      "step": 1946
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6430444894690293,
+      "learning_rate": 1.4404482870344322e-05,
+      "loss": 1.0295,
+      "step": 1947
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6573601688137906,
+      "learning_rate": 1.4398888099188396e-05,
+      "loss": 1.0708,
+      "step": 1948
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6971366506487653,
+      "learning_rate": 1.4393291620232646e-05,
+      "loss": 1.076,
+      "step": 1949
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6817301576905301,
+      "learning_rate": 1.4387693435649826e-05,
+      "loss": 1.0494,
+      "step": 1950
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5466948424236525,
+      "learning_rate": 1.4382093547613338e-05,
+      "loss": 1.092,
+      "step": 1951
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6766553573711729,
+      "learning_rate": 1.4376491958297263e-05,
+      "loss": 1.0231,
+      "step": 1952
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6729336800893753,
+      "learning_rate": 1.4370888669876317e-05,
+      "loss": 1.0507,
+      "step": 1953
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6954464667085627,
+      "learning_rate": 1.4365283684525895e-05,
+      "loss": 1.0073,
+      "step": 1954
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6689962451616699,
+      "learning_rate": 1.4359677004422045e-05,
+      "loss": 0.9817,
+      "step": 1955
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6831353835788363,
+      "learning_rate": 1.4354068631741476e-05,
+      "loss": 1.0391,
+      "step": 1956
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6120689342781522,
+      "learning_rate": 1.4348458568661548e-05,
+      "loss": 1.0745,
+      "step": 1957
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5575591039961425,
+      "learning_rate": 1.434284681736028e-05,
+      "loss": 1.0652,
+      "step": 1958
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7012480021616487,
+      "learning_rate": 1.4337233380016354e-05,
+      "loss": 1.0113,
+      "step": 1959
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6420063666914898,
+      "learning_rate": 1.433161825880909e-05,
+      "loss": 1.057,
+      "step": 1960
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8022149587449,
+      "learning_rate": 1.432600145591848e-05,
+      "loss": 1.0544,
+      "step": 1961
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7065082400788613,
+      "learning_rate": 1.4320382973525151e-05,
+      "loss": 1.0082,
+      "step": 1962
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6614436057867726,
+      "learning_rate": 1.43147628138104e-05,
+      "loss": 1.0836,
+      "step": 1963
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7260182308907308,
+      "learning_rate": 1.4309140978956161e-05,
+      "loss": 0.9725,
+      "step": 1964
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7163459625259518,
+      "learning_rate": 1.430351747114503e-05,
+      "loss": 1.0191,
+      "step": 1965
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7358447578719567,
+      "learning_rate": 1.429789229256024e-05,
+      "loss": 1.0115,
+      "step": 1966
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7453647903121436,
+      "learning_rate": 1.429226544538568e-05,
+      "loss": 1.0631,
+      "step": 1967
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7139815151893676,
+      "learning_rate": 1.4286636931805887e-05,
+      "loss": 1.0346,
+      "step": 1968
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6838624056757039,
+      "learning_rate": 1.4281006754006045e-05,
+      "loss": 0.9977,
+      "step": 1969
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8184384282888667,
+      "learning_rate": 1.427537491417198e-05,
+      "loss": 0.9855,
+      "step": 1970
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6917130969914567,
+      "learning_rate": 1.426974141449017e-05,
+      "loss": 1.035,
+      "step": 1971
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6838688338010143,
+      "learning_rate": 1.4264106257147732e-05,
+      "loss": 1.0278,
+      "step": 1972
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6481869609987368,
+      "learning_rate": 1.4258469444332423e-05,
+      "loss": 0.9728,
+      "step": 1973
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.741287111462222,
+      "learning_rate": 1.4252830978232658e-05,
+      "loss": 1.0168,
+      "step": 1974
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.706064845738209,
+      "learning_rate": 1.4247190861037474e-05,
+      "loss": 0.9979,
+      "step": 1975
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6546943507736225,
+      "learning_rate": 1.4241549094936567e-05,
+      "loss": 1.058,
+      "step": 1976
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7778722187565387,
+      "learning_rate": 1.4235905682120255e-05,
+      "loss": 1.0034,
+      "step": 1977
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.72176831429739,
+      "learning_rate": 1.4230260624779512e-05,
+      "loss": 1.0491,
+      "step": 1978
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7077676508850748,
+      "learning_rate": 1.4224613925105947e-05,
+      "loss": 1.1021,
+      "step": 1979
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7528045940689131,
+      "learning_rate": 1.4218965585291792e-05,
+      "loss": 1.067,
+      "step": 1980
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6131289342923766,
+      "learning_rate": 1.4213315607529939e-05,
+      "loss": 1.1897,
+      "step": 1981
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7226734659158217,
+      "learning_rate": 1.4207663994013896e-05,
+      "loss": 1.0248,
+      "step": 1982
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6506370298590259,
+      "learning_rate": 1.4202010746937815e-05,
+      "loss": 1.037,
+      "step": 1983
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.49488191909608886,
+      "learning_rate": 1.4196355868496485e-05,
+      "loss": 1.0758,
+      "step": 1984
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.684449127921717,
+      "learning_rate": 1.4190699360885323e-05,
+      "loss": 1.0453,
+      "step": 1985
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.657192544626447,
+      "learning_rate": 1.4185041226300376e-05,
+      "loss": 1.0469,
+      "step": 1986
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.769862774152308,
+      "learning_rate": 1.4179381466938332e-05,
+      "loss": 0.9774,
+      "step": 1987
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5496098478449233,
+      "learning_rate": 1.4173720084996501e-05,
+      "loss": 1.005,
+      "step": 1988
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7412973965088676,
+      "learning_rate": 1.4168057082672828e-05,
+      "loss": 1.0503,
+      "step": 1989
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6770578634838542,
+      "learning_rate": 1.4162392462165884e-05,
+      "loss": 1.0611,
+      "step": 1990
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6912114865458542,
+      "learning_rate": 1.4156726225674874e-05,
+      "loss": 1.048,
+      "step": 1991
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7324584078409523,
+      "learning_rate": 1.415105837539962e-05,
+      "loss": 1.0262,
+      "step": 1992
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6671589695599974,
+      "learning_rate": 1.414538891354058e-05,
+      "loss": 1.03,
+      "step": 1993
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7584299422568842,
+      "learning_rate": 1.4139717842298835e-05,
+      "loss": 1.0757,
+      "step": 1994
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7233629944822249,
+      "learning_rate": 1.4134045163876086e-05,
+      "loss": 1.0548,
+      "step": 1995
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7011663156188885,
+      "learning_rate": 1.4128370880474667e-05,
+      "loss": 0.994,
+      "step": 1996
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7367674851952177,
+      "learning_rate": 1.412269499429753e-05,
+      "loss": 1.0046,
+      "step": 1997
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5360934694254629,
+      "learning_rate": 1.4117017507548244e-05,
+      "loss": 0.9955,
+      "step": 1998
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5758089339400189,
+      "learning_rate": 1.4111338422431013e-05,
+      "loss": 1.0347,
+      "step": 1999
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6959459043761054,
+      "learning_rate": 1.4105657741150648e-05,
+      "loss": 0.989,
+      "step": 2000
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6875304598474028,
+      "learning_rate": 1.4099975465912584e-05,
+      "loss": 1.0225,
+      "step": 2001
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6369594148062739,
+      "learning_rate": 1.4094291598922877e-05,
+      "loss": 1.0009,
+      "step": 2002
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.34738183418457125,
+      "learning_rate": 1.40886061423882e-05,
+      "loss": 0.9366,
+      "step": 2003
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.34738183418457125,
+      "learning_rate": 1.40886061423882e-05,
+      "loss": 1.0785,
+      "step": 2004
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.34738183418457125,
+      "learning_rate": 1.40886061423882e-05,
+      "loss": 1.0804,
+      "step": 2005
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.546331643558699,
+      "learning_rate": 1.4082919098515846e-05,
+      "loss": 1.0321,
+      "step": 2006
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5965620771397517,
+      "learning_rate": 1.407723046951372e-05,
+      "loss": 1.0536,
+      "step": 2007
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7157494223395775,
+      "learning_rate": 1.4071540257590341e-05,
+      "loss": 1.0784,
+      "step": 2008
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6686458242128729,
+      "learning_rate": 1.4065848464954848e-05,
+      "loss": 0.983,
+      "step": 2009
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6977227171168019,
+      "learning_rate": 1.4060155093816988e-05,
+      "loss": 1.035,
+      "step": 2010
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6614846195886491,
+      "learning_rate": 1.4054460146387124e-05,
+      "loss": 1.0239,
+      "step": 2011
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6659518870617972,
+      "learning_rate": 1.4048763624876233e-05,
+      "loss": 0.9844,
+      "step": 2012
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.673961098647442,
+      "learning_rate": 1.4043065531495904e-05,
+      "loss": 0.9826,
+      "step": 2013
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7011864658712502,
+      "learning_rate": 1.4037365868458325e-05,
+      "loss": 1.0477,
+      "step": 2014
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7116298636433479,
+      "learning_rate": 1.4031664637976305e-05,
+      "loss": 1.0468,
+      "step": 2015
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5436080067225092,
+      "learning_rate": 1.402596184226326e-05,
+      "loss": 1.0789,
+      "step": 2016
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5983445977218846,
+      "learning_rate": 1.4020257483533208e-05,
+      "loss": 1.0696,
+      "step": 2017
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7035722205133182,
+      "learning_rate": 1.401455156400078e-05,
+      "loss": 1.017,
+      "step": 2018
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6939634638674967,
+      "learning_rate": 1.400884408588121e-05,
+      "loss": 1.0417,
+      "step": 2019
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.743920685781744,
+      "learning_rate": 1.400313505139034e-05,
+      "loss": 1.0621,
+      "step": 2020
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6289803947319206,
+      "learning_rate": 1.3997424462744607e-05,
+      "loss": 0.9618,
+      "step": 2021
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6985956492422198,
+      "learning_rate": 1.3991712322161065e-05,
+      "loss": 1.033,
+      "step": 2022
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.667464215108415,
+      "learning_rate": 1.3985998631857359e-05,
+      "loss": 1.0516,
+      "step": 2023
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6686683299765762,
+      "learning_rate": 1.398028339405174e-05,
+      "loss": 1.1021,
+      "step": 2024
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.70486218488879,
+      "learning_rate": 1.3974566610963068e-05,
+      "loss": 1.084,
+      "step": 2025
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5367637334223724,
+      "learning_rate": 1.3968848284810785e-05,
+      "loss": 1.0023,
+      "step": 2026
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6588862263266596,
+      "learning_rate": 1.3963128417814951e-05,
+      "loss": 0.9883,
+      "step": 2027
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7365355891479503,
+      "learning_rate": 1.3957407012196204e-05,
+      "loss": 0.995,
+      "step": 2028
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6739685256998721,
+      "learning_rate": 1.3951684070175802e-05,
+      "loss": 1.058,
+      "step": 2029
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7270341740998144,
+      "learning_rate": 1.3945959593975582e-05,
+      "loss": 1.0134,
+      "step": 2030
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7031095546282154,
+      "learning_rate": 1.3940233585817984e-05,
+      "loss": 1.0195,
+      "step": 2031
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5315343753083975,
+      "learning_rate": 1.3934506047926042e-05,
+      "loss": 0.956,
+      "step": 2032
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6755087592893919,
+      "learning_rate": 1.3928776982523384e-05,
+      "loss": 0.9673,
+      "step": 2033
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5645910566085668,
+      "learning_rate": 1.3923046391834229e-05,
+      "loss": 1.0718,
+      "step": 2034
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7264739202129737,
+      "learning_rate": 1.3917314278083391e-05,
+      "loss": 0.9888,
+      "step": 2035
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7826015548757311,
+      "learning_rate": 1.3911580643496272e-05,
+      "loss": 1.0031,
+      "step": 2036
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6792323464887216,
+      "learning_rate": 1.3905845490298867e-05,
+      "loss": 1.0101,
+      "step": 2037
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7672300908721228,
+      "learning_rate": 1.390010882071776e-05,
+      "loss": 1.0236,
+      "step": 2038
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7235205834831843,
+      "learning_rate": 1.3894370636980128e-05,
+      "loss": 1.094,
+      "step": 2039
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7128665889093824,
+      "learning_rate": 1.3888630941313728e-05,
+      "loss": 1.1004,
+      "step": 2040
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5768323390732744,
+      "learning_rate": 1.3882889735946901e-05,
+      "loss": 1.0418,
+      "step": 2041
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7165112650075555,
+      "learning_rate": 1.3877147023108592e-05,
+      "loss": 1.0323,
+      "step": 2042
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6136280461639929,
+      "learning_rate": 1.3871402805028314e-05,
+      "loss": 1.0344,
+      "step": 2043
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6826806291843056,
+      "learning_rate": 1.3865657083936167e-05,
+      "loss": 1.0516,
+      "step": 2044
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7131470990608739,
+      "learning_rate": 1.3859909862062844e-05,
+      "loss": 1.0685,
+      "step": 2045
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6252617740959331,
+      "learning_rate": 1.385416114163961e-05,
+      "loss": 1.0193,
+      "step": 2046
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6869745153071759,
+      "learning_rate": 1.3848410924898321e-05,
+      "loss": 1.0172,
+      "step": 2047
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7058576144788816,
+      "learning_rate": 1.3842659214071406e-05,
+      "loss": 1.0592,
+      "step": 2048
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7415015256334325,
+      "learning_rate": 1.3836906011391878e-05,
+      "loss": 0.9775,
+      "step": 2049
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6734648478120551,
+      "learning_rate": 1.3831151319093323e-05,
+      "loss": 1.0572,
+      "step": 2050
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6696911700989102,
+      "learning_rate": 1.382539513940992e-05,
+      "loss": 1.0263,
+      "step": 2051
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6617125004627151,
+      "learning_rate": 1.3819637474576411e-05,
+      "loss": 1.0359,
+      "step": 2052
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6652973348133597,
+      "learning_rate": 1.381387832682812e-05,
+      "loss": 1.0472,
+      "step": 2053
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5512558894941982,
+      "learning_rate": 1.380811769840095e-05,
+      "loss": 1.1028,
+      "step": 2054
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7089179278908745,
+      "learning_rate": 1.3802355591531366e-05,
+      "loss": 1.0465,
+      "step": 2055
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6917237681669411,
+      "learning_rate": 1.3796592008456427e-05,
+      "loss": 0.9907,
+      "step": 2056
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6930154272417123,
+      "learning_rate": 1.3790826951413747e-05,
+      "loss": 1.016,
+      "step": 2057
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7700226341696728,
+      "learning_rate": 1.3785060422641526e-05,
+      "loss": 1.0842,
+      "step": 2058
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6866554336085727,
+      "learning_rate": 1.3779292424378521e-05,
+      "loss": 1.0904,
+      "step": 2059
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6600957181929765,
+      "learning_rate": 1.3773522958864076e-05,
+      "loss": 1.0298,
+      "step": 2060
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7041708763273234,
+      "learning_rate": 1.3767752028338091e-05,
+      "loss": 1.0908,
+      "step": 2061
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7121236337687779,
+      "learning_rate": 1.376197963504104e-05,
+      "loss": 1.0857,
+      "step": 2062
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.527538585030005,
+      "learning_rate": 1.3756205781213965e-05,
+      "loss": 1.0161,
+      "step": 2063
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7191123423521739,
+      "learning_rate": 1.375043046909848e-05,
+      "loss": 1.0266,
+      "step": 2064
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6590626599998657,
+      "learning_rate": 1.3744653700936752e-05,
+      "loss": 1.0195,
+      "step": 2065
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6738257396961046,
+      "learning_rate": 1.3738875478971526e-05,
+      "loss": 0.953,
+      "step": 2066
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7166064976302348,
+      "learning_rate": 1.3733095805446107e-05,
+      "loss": 1.0895,
+      "step": 2067
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7168099493462058,
+      "learning_rate": 1.372731468260436e-05,
+      "loss": 1.0308,
+      "step": 2068
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6700507828284331,
+      "learning_rate": 1.372153211269072e-05,
+      "loss": 1.0706,
+      "step": 2069
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6934414981341668,
+      "learning_rate": 1.3715748097950176e-05,
+      "loss": 1.0074,
+      "step": 2070
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7061781937222171,
+      "learning_rate": 1.3709962640628284e-05,
+      "loss": 0.9956,
+      "step": 2071
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6765104841443461,
+      "learning_rate": 1.3704175742971158e-05,
+      "loss": 1.0492,
+      "step": 2072
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6741239324682299,
+      "learning_rate": 1.369838740722547e-05,
+      "loss": 1.025,
+      "step": 2073
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6542694740045549,
+      "learning_rate": 1.3692597635638452e-05,
+      "loss": 1.0859,
+      "step": 2074
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.678143683897546,
+      "learning_rate": 1.368680643045789e-05,
+      "loss": 1.0208,
+      "step": 2075
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.48389008214514256,
+      "learning_rate": 1.3681013793932132e-05,
+      "loss": 1.076,
+      "step": 2076
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7504618142414987,
+      "learning_rate": 1.3675219728310076e-05,
+      "loss": 1.0452,
+      "step": 2077
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6331610102692875,
+      "learning_rate": 1.3669424235841185e-05,
+      "loss": 1.0047,
+      "step": 2078
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6705883908725303,
+      "learning_rate": 1.3663627318775459e-05,
+      "loss": 1.0997,
+      "step": 2079
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6926103871853956,
+      "learning_rate": 1.3657828979363468e-05,
+      "loss": 1.0239,
+      "step": 2080
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5600085479138334,
+      "learning_rate": 1.3652029219856324e-05,
+      "loss": 1.0673,
+      "step": 2081
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.636005124808239,
+      "learning_rate": 1.3646228042505694e-05,
+      "loss": 1.0144,
+      "step": 2082
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6733141514338372,
+      "learning_rate": 1.3640425449563793e-05,
+      "loss": 1.0215,
+      "step": 2083
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.742684879473069,
+      "learning_rate": 1.3634621443283389e-05,
+      "loss": 1.0632,
+      "step": 2084
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7057336557016747,
+      "learning_rate": 1.36288160259178e-05,
+      "loss": 1.0145,
+      "step": 2085
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6649710641270106,
+      "learning_rate": 1.3623009199720882e-05,
+      "loss": 1.0359,
+      "step": 2086
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7296726633685303,
+      "learning_rate": 1.3617200966947053e-05,
+      "loss": 1.0831,
+      "step": 2087
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6593329467348119,
+      "learning_rate": 1.3611391329851262e-05,
+      "loss": 1.0667,
+      "step": 2088
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6463969024551744,
+      "learning_rate": 1.3605580290689013e-05,
+      "loss": 0.9757,
+      "step": 2089
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7312751284386098,
+      "learning_rate": 1.3599767851716353e-05,
+      "loss": 1.0916,
+      "step": 2090
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7333761531760936,
+      "learning_rate": 1.3593954015189867e-05,
+      "loss": 1.0559,
+      "step": 2091
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6925761184922775,
+      "learning_rate": 1.3588138783366692e-05,
+      "loss": 0.9807,
+      "step": 2092
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.595217434613477,
+      "learning_rate": 1.3582322158504495e-05,
+      "loss": 0.9985,
+      "step": 2093
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5198131322852667,
+      "learning_rate": 1.3576504142861496e-05,
+      "loss": 1.1138,
+      "step": 2094
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7595770340201341,
+      "learning_rate": 1.3570684738696444e-05,
+      "loss": 1.1379,
+      "step": 2095
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6850963591004411,
+      "learning_rate": 1.3564863948268631e-05,
+      "loss": 1.0131,
+      "step": 2096
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7442631005940047,
+      "learning_rate": 1.3559041773837898e-05,
+      "loss": 1.0252,
+      "step": 2097
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.742503945339705,
+      "learning_rate": 1.3553218217664603e-05,
+      "loss": 1.0662,
+      "step": 2098
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6754470909461471,
+      "learning_rate": 1.3547393282009656e-05,
+      "loss": 1.0385,
+      "step": 2099
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7031075853967027,
+      "learning_rate": 1.3541566969134496e-05,
+      "loss": 1.0795,
+      "step": 2100
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7867994220039176,
+      "learning_rate": 1.3535739281301102e-05,
+      "loss": 1.0665,
+      "step": 2101
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7362144395012067,
+      "learning_rate": 1.3529910220771975e-05,
+      "loss": 1.0177,
+      "step": 2102
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6650799884382674,
+      "learning_rate": 1.3524079789810163e-05,
+      "loss": 1.0095,
+      "step": 2103
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5737628578719627,
+      "learning_rate": 1.3518247990679241e-05,
+      "loss": 1.0452,
+      "step": 2104
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7221753610587538,
+      "learning_rate": 1.3512414825643312e-05,
+      "loss": 0.9364,
+      "step": 2105
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6402356815681461,
+      "learning_rate": 1.3506580296967011e-05,
+      "loss": 1.0178,
+      "step": 2106
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6580550154599448,
+      "learning_rate": 1.3500744406915505e-05,
+      "loss": 1.0399,
+      "step": 2107
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5728850170846911,
+      "learning_rate": 1.3494907157754485e-05,
+      "loss": 1.0756,
+      "step": 2108
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7447198401049837,
+      "learning_rate": 1.348906855175017e-05,
+      "loss": 0.9918,
+      "step": 2109
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6771306376177437,
+      "learning_rate": 1.3483228591169315e-05,
+      "loss": 1.0524,
+      "step": 2110
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6714309439527901,
+      "learning_rate": 1.347738727827919e-05,
+      "loss": 1.0175,
+      "step": 2111
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6691582180204217,
+      "learning_rate": 1.3471544615347591e-05,
+      "loss": 1.0754,
+      "step": 2112
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6763112899172841,
+      "learning_rate": 1.3465700604642847e-05,
+      "loss": 1.0283,
+      "step": 2113
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7005014617568085,
+      "learning_rate": 1.34598552484338e-05,
+      "loss": 1.0104,
+      "step": 2114
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7221492351722754,
+      "learning_rate": 1.3454008548989816e-05,
+      "loss": 1.0254,
+      "step": 2115
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6459109221299123,
+      "learning_rate": 1.3448160508580789e-05,
+      "loss": 1.0391,
+      "step": 2116
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6836191016833953,
+      "learning_rate": 1.3442311129477133e-05,
+      "loss": 1.0726,
+      "step": 2117
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7544097114756214,
+      "learning_rate": 1.343646041394977e-05,
+      "loss": 1.0417,
+      "step": 2118
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7304783005692707,
+      "learning_rate": 1.3430608364270156e-05,
+      "loss": 1.0583,
+      "step": 2119
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8039602870083635,
+      "learning_rate": 1.3424754982710256e-05,
+      "loss": 1.0547,
+      "step": 2120
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7191723525197163,
+      "learning_rate": 1.3418900271542552e-05,
+      "loss": 1.0147,
+      "step": 2121
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6650358084044017,
+      "learning_rate": 1.3413044233040045e-05,
+      "loss": 1.0234,
+      "step": 2122
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6412739531067441,
+      "learning_rate": 1.3407186869476253e-05,
+      "loss": 0.986,
+      "step": 2123
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6796148002395715,
+      "learning_rate": 1.3401328183125208e-05,
+      "loss": 1.0172,
+      "step": 2124
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6829161501274852,
+      "learning_rate": 1.339546817626145e-05,
+      "loss": 1.0405,
+      "step": 2125
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.716810314612088,
+      "learning_rate": 1.3389606851160037e-05,
+      "loss": 1.0387,
+      "step": 2126
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7344103129646382,
+      "learning_rate": 1.3383744210096537e-05,
+      "loss": 0.9728,
+      "step": 2127
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6961519160948926,
+      "learning_rate": 1.3377880255347026e-05,
+      "loss": 1.0205,
+      "step": 2128
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6550493370947411,
+      "learning_rate": 1.3372014989188098e-05,
+      "loss": 1.0846,
+      "step": 2129
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7248569436285724,
+      "learning_rate": 1.3366148413896851e-05,
+      "loss": 1.0205,
+      "step": 2130
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7247183707645745,
+      "learning_rate": 1.3360280531750886e-05,
+      "loss": 0.9526,
+      "step": 2131
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6814465455226241,
+      "learning_rate": 1.3354411345028324e-05,
+      "loss": 0.9997,
+      "step": 2132
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7645252508088517,
+      "learning_rate": 1.3348540856007782e-05,
+      "loss": 0.9963,
+      "step": 2133
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7371545056122459,
+      "learning_rate": 1.3342669066968385e-05,
+      "loss": 1.0565,
+      "step": 2134
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7396063263136147,
+      "learning_rate": 1.3336795980189763e-05,
+      "loss": 0.9996,
+      "step": 2135
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6553129950261437,
+      "learning_rate": 1.3330921597952056e-05,
+      "loss": 1.0167,
+      "step": 2136
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7706531903031821,
+      "learning_rate": 1.3325045922535896e-05,
+      "loss": 1.0567,
+      "step": 2137
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7166953415051726,
+      "learning_rate": 1.3319168956222423e-05,
+      "loss": 0.9814,
+      "step": 2138
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7003284170460674,
+      "learning_rate": 1.331329070129328e-05,
+      "loss": 1.0648,
+      "step": 2139
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.712910753097322,
+      "learning_rate": 1.3307411160030608e-05,
+      "loss": 0.9811,
+      "step": 2140
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6449791037029224,
+      "learning_rate": 1.3301530334717046e-05,
+      "loss": 1.0358,
+      "step": 2141
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7529659716058533,
+      "learning_rate": 1.3295648227635729e-05,
+      "loss": 1.0117,
+      "step": 2142
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.661377394794602,
+      "learning_rate": 1.32897648410703e-05,
+      "loss": 0.9846,
+      "step": 2143
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7181115916175068,
+      "learning_rate": 1.328388017730489e-05,
+      "loss": 1.0649,
+      "step": 2144
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.518957320255548,
+      "learning_rate": 1.327799423862413e-05,
+      "loss": 1.0322,
+      "step": 2145
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7511200441917767,
+      "learning_rate": 1.3272107027313142e-05,
+      "loss": 1.0368,
+      "step": 2146
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6901250713631065,
+      "learning_rate": 1.3266218545657541e-05,
+      "loss": 1.0681,
+      "step": 2147
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7153238378524605,
+      "learning_rate": 1.326032879594344e-05,
+      "loss": 0.9694,
+      "step": 2148
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7320418925605218,
+      "learning_rate": 1.3254437780457448e-05,
+      "loss": 1.046,
+      "step": 2149
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6833257044861556,
+      "learning_rate": 1.3248545501486654e-05,
+      "loss": 1.009,
+      "step": 2150
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.825938851798866,
+      "learning_rate": 1.3242651961318646e-05,
+      "loss": 1.0154,
+      "step": 2151
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.700170490307021,
+      "learning_rate": 1.32367571622415e-05,
+      "loss": 1.0038,
+      "step": 2152
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6881561751247944,
+      "learning_rate": 1.3230861106543777e-05,
+      "loss": 1.0307,
+      "step": 2153
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4889474427938491,
+      "learning_rate": 1.3224963796514532e-05,
+      "loss": 1.0524,
+      "step": 2154
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6741095046654769,
+      "learning_rate": 1.32190652344433e-05,
+      "loss": 0.9844,
+      "step": 2155
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7002581415675572,
+      "learning_rate": 1.3213165422620111e-05,
+      "loss": 1.0182,
+      "step": 2156
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.686929070529953,
+      "learning_rate": 1.3207264363335472e-05,
+      "loss": 1.0427,
+      "step": 2157
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7311455209992324,
+      "learning_rate": 1.3201362058880375e-05,
+      "loss": 1.0142,
+      "step": 2158
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6720376592938886,
+      "learning_rate": 1.3195458511546307e-05,
+      "loss": 1.0284,
+      "step": 2159
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6996043285363488,
+      "learning_rate": 1.3189553723625217e-05,
+      "loss": 1.1006,
+      "step": 2160
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.671855869573462,
+      "learning_rate": 1.318364769740955e-05,
+      "loss": 1.0259,
+      "step": 2161
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6873407408162945,
+      "learning_rate": 1.3177740435192235e-05,
+      "loss": 1.0328,
+      "step": 2162
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5587761987274876,
+      "learning_rate": 1.3171831939266668e-05,
+      "loss": 0.9792,
+      "step": 2163
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7210736156682831,
+      "learning_rate": 1.3165922211926734e-05,
+      "loss": 1.0219,
+      "step": 2164
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7295182085363027,
+      "learning_rate": 1.3160011255466791e-05,
+      "loss": 0.9868,
+      "step": 2165
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6373483912122138,
+      "learning_rate": 1.3154099072181677e-05,
+      "loss": 1.0586,
+      "step": 2166
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6517187004530443,
+      "learning_rate": 1.3148185664366704e-05,
+      "loss": 0.9841,
+      "step": 2167
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7180171238741587,
+      "learning_rate": 1.314227103431766e-05,
+      "loss": 1.025,
+      "step": 2168
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.619119496879883,
+      "learning_rate": 1.3136355184330809e-05,
+      "loss": 0.9896,
+      "step": 2169
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7251593411060351,
+      "learning_rate": 1.3130438116702888e-05,
+      "loss": 1.0911,
+      "step": 2170
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7431734754308529,
+      "learning_rate": 1.3124519833731106e-05,
+      "loss": 1.0805,
+      "step": 2171
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6495333310184858,
+      "learning_rate": 1.3118600337713146e-05,
+      "loss": 1.037,
+      "step": 2172
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6052174530274317,
+      "learning_rate": 1.3112679630947156e-05,
+      "loss": 1.0742,
+      "step": 2173
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7218192396377928,
+      "learning_rate": 1.310675771573176e-05,
+      "loss": 0.9873,
+      "step": 2174
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8046873138800702,
+      "learning_rate": 1.310083459436605e-05,
+      "loss": 1.0303,
+      "step": 2175
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7276956719540028,
+      "learning_rate": 1.3094910269149587e-05,
+      "loss": 0.9237,
+      "step": 2176
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7350098308343977,
+      "learning_rate": 1.3088984742382395e-05,
+      "loss": 0.9841,
+      "step": 2177
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5228853153740933,
+      "learning_rate": 1.3083058016364972e-05,
+      "loss": 1.0019,
+      "step": 2178
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.749666235239996,
+      "learning_rate": 1.3077130093398274e-05,
+      "loss": 1.0948,
+      "step": 2179
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6951084063671508,
+      "learning_rate": 1.3071200975783725e-05,
+      "loss": 1.0014,
+      "step": 2180
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7510214785280342,
+      "learning_rate": 1.3065270665823206e-05,
+      "loss": 1.0338,
+      "step": 2181
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6798885343520276,
+      "learning_rate": 1.3059339165819082e-05,
+      "loss": 1.0622,
+      "step": 2182
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7282529640909775,
+      "learning_rate": 1.3053406478074155e-05,
+      "loss": 1.0557,
+      "step": 2183
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7302399508941932,
+      "learning_rate": 1.3047472604891701e-05,
+      "loss": 1.0804,
+      "step": 2184
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7239647249362372,
+      "learning_rate": 1.3041537548575455e-05,
+      "loss": 1.0219,
+      "step": 2185
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6427880772373721,
+      "learning_rate": 1.303560131142961e-05,
+      "loss": 1.0458,
+      "step": 2186
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6598852806083614,
+      "learning_rate": 1.3029663895758814e-05,
+      "loss": 1.0532,
+      "step": 2187
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.788090703296003,
+      "learning_rate": 1.3023725303868183e-05,
+      "loss": 1.024,
+      "step": 2188
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6453573596479981,
+      "learning_rate": 1.3017785538063277e-05,
+      "loss": 1.0114,
+      "step": 2189
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7193691705520131,
+      "learning_rate": 1.3011844600650121e-05,
+      "loss": 1.0402,
+      "step": 2190
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.664731120257474,
+      "learning_rate": 1.300590249393519e-05,
+      "loss": 0.9656,
+      "step": 2191
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7730175637722574,
+      "learning_rate": 1.2999959220225416e-05,
+      "loss": 1.0327,
+      "step": 2192
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6132088315676862,
+      "learning_rate": 1.299401478182818e-05,
+      "loss": 1.0914,
+      "step": 2193
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.689550007832144,
+      "learning_rate": 1.2988069181051314e-05,
+      "loss": 1.0006,
+      "step": 2194
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5383396324110697,
+      "learning_rate": 1.2982122420203114e-05,
+      "loss": 1.0901,
+      "step": 2195
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7915393219110828,
+      "learning_rate": 1.2976174501592313e-05,
+      "loss": 1.0112,
+      "step": 2196
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7791394142895768,
+      "learning_rate": 1.2970225427528098e-05,
+      "loss": 1.05,
+      "step": 2197
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7366823131429241,
+      "learning_rate": 1.2964275200320104e-05,
+      "loss": 1.0592,
+      "step": 2198
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6940503077644125,
+      "learning_rate": 1.2958323822278413e-05,
+      "loss": 1.0742,
+      "step": 2199
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7317411565044932,
+      "learning_rate": 1.2952371295713558e-05,
+      "loss": 1.0288,
+      "step": 2200
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7131048464434367,
+      "learning_rate": 1.2946417622936512e-05,
+      "loss": 0.9119,
+      "step": 2201
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7585896589756861,
+      "learning_rate": 1.2940462806258696e-05,
+      "loss": 0.9524,
+      "step": 2202
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6858480281855426,
+      "learning_rate": 1.2934506847991976e-05,
+      "loss": 1.006,
+      "step": 2203
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6362698545397273,
+      "learning_rate": 1.2928549750448661e-05,
+      "loss": 0.9514,
+      "step": 2204
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7254938249834952,
+      "learning_rate": 1.2922591515941498e-05,
+      "loss": 1.074,
+      "step": 2205
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7007395423199075,
+      "learning_rate": 1.2916632146783683e-05,
+      "loss": 0.982,
+      "step": 2206
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6664987185680775,
+      "learning_rate": 1.2910671645288841e-05,
+      "loss": 1.0247,
+      "step": 2207
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6705638252367738,
+      "learning_rate": 1.2904710013771054e-05,
+      "loss": 1.0248,
+      "step": 2208
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5620311613427013,
+      "learning_rate": 1.2898747254544826e-05,
+      "loss": 1.0822,
+      "step": 2209
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6867801459100877,
+      "learning_rate": 1.2892783369925105e-05,
+      "loss": 1.0482,
+      "step": 2210
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.809945753224828,
+      "learning_rate": 1.2886818362227283e-05,
+      "loss": 1.0375,
+      "step": 2211
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7090928785283929,
+      "learning_rate": 1.2880852233767174e-05,
+      "loss": 1.0624,
+      "step": 2212
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.759189652667529,
+      "learning_rate": 1.2874884986861038e-05,
+      "loss": 1.0306,
+      "step": 2213
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6501456848309569,
+      "learning_rate": 1.2868916623825561e-05,
+      "loss": 1.0817,
+      "step": 2214
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7939855835853051,
+      "learning_rate": 1.2862947146977876e-05,
+      "loss": 1.0579,
+      "step": 2215
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6548913329023502,
+      "learning_rate": 1.2856976558635532e-05,
+      "loss": 1.0408,
+      "step": 2216
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6951527488046872,
+      "learning_rate": 1.2851004861116519e-05,
+      "loss": 1.0395,
+      "step": 2217
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7946137373669365,
+      "learning_rate": 1.2845032056739257e-05,
+      "loss": 0.982,
+      "step": 2218
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6479499113795134,
+      "learning_rate": 1.2839058147822595e-05,
+      "loss": 1.1269,
+      "step": 2219
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.649907229181452,
+      "learning_rate": 1.2833083136685803e-05,
+      "loss": 1.0913,
+      "step": 2220
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.644354184024163,
+      "learning_rate": 1.2827107025648595e-05,
+      "loss": 1.0953,
+      "step": 2221
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7172812669900046,
+      "learning_rate": 1.2821129817031099e-05,
+      "loss": 0.9884,
+      "step": 2222
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.690475482509461,
+      "learning_rate": 1.2815151513153874e-05,
+      "loss": 1.0461,
+      "step": 2223
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7085769915458817,
+      "learning_rate": 1.2809172116337903e-05,
+      "loss": 0.9706,
+      "step": 2224
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7196842466441463,
+      "learning_rate": 1.2803191628904594e-05,
+      "loss": 1.0406,
+      "step": 2225
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7235165930445143,
+      "learning_rate": 1.2797210053175779e-05,
+      "loss": 1.0449,
+      "step": 2226
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.716563699024576,
+      "learning_rate": 1.2791227391473706e-05,
+      "loss": 0.9908,
+      "step": 2227
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6222806624711028,
+      "learning_rate": 1.2785243646121059e-05,
+      "loss": 1.0887,
+      "step": 2228
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5100784462287166,
+      "learning_rate": 1.277925881944093e-05,
+      "loss": 1.0659,
+      "step": 2229
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6718827985822374,
+      "learning_rate": 1.2773272913756833e-05,
+      "loss": 1.0579,
+      "step": 2230
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.680501145081669,
+      "learning_rate": 1.2767285931392705e-05,
+      "loss": 1.0168,
+      "step": 2231
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6991913720766361,
+      "learning_rate": 1.27612978746729e-05,
+      "loss": 1.0215,
+      "step": 2232
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7358983902812243,
+      "learning_rate": 1.2755308745922182e-05,
+      "loss": 0.9827,
+      "step": 2233
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7329834408869526,
+      "learning_rate": 1.2749318547465742e-05,
+      "loss": 1.0805,
+      "step": 2234
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6852137104591314,
+      "learning_rate": 1.2743327281629181e-05,
+      "loss": 0.9386,
+      "step": 2235
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6126256270858033,
+      "learning_rate": 1.2737334950738512e-05,
+      "loss": 1.0905,
+      "step": 2236
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8952844835127487,
+      "learning_rate": 1.273134155712017e-05,
+      "loss": 1.0973,
+      "step": 2237
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6964884678154376,
+      "learning_rate": 1.272534710310099e-05,
+      "loss": 1.0374,
+      "step": 2238
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.735034475251445,
+      "learning_rate": 1.2719351591008228e-05,
+      "loss": 0.9814,
+      "step": 2239
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6796685648300633,
+      "learning_rate": 1.2713355023169547e-05,
+      "loss": 1.0995,
+      "step": 2240
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7197908967605137,
+      "learning_rate": 1.2707357401913022e-05,
+      "loss": 1.0604,
+      "step": 2241
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7685828023374914,
+      "learning_rate": 1.270135872956714e-05,
+      "loss": 0.9815,
+      "step": 2242
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6998286125549072,
+      "learning_rate": 1.2695359008460785e-05,
+      "loss": 1.0974,
+      "step": 2243
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5486849305789346,
+      "learning_rate": 1.2689358240923264e-05,
+      "loss": 1.0826,
+      "step": 2244
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.81688917451559,
+      "learning_rate": 1.2683356429284273e-05,
+      "loss": 1.0626,
+      "step": 2245
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6655590848553856,
+      "learning_rate": 1.2677353575873926e-05,
+      "loss": 0.9906,
+      "step": 2246
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6900379546853953,
+      "learning_rate": 1.2671349683022736e-05,
+      "loss": 1.0647,
+      "step": 2247
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7317732963017956,
+      "learning_rate": 1.2665344753061622e-05,
+      "loss": 1.0122,
+      "step": 2248
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6604254949779725,
+      "learning_rate": 1.2659338788321904e-05,
+      "loss": 0.9628,
+      "step": 2249
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6767418769814952,
+      "learning_rate": 1.2653331791135308e-05,
+      "loss": 1.0976,
+      "step": 2250
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7105340914452729,
+      "learning_rate": 1.2647323763833952e-05,
+      "loss": 0.9742,
+      "step": 2251
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6497509247565182,
+      "learning_rate": 1.264131470875036e-05,
+      "loss": 1.0251,
+      "step": 2252
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6199598376096419,
+      "learning_rate": 1.2635304628217452e-05,
+      "loss": 0.9738,
+      "step": 2253
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5406237943105332,
+      "learning_rate": 1.2629293524568555e-05,
+      "loss": 1.0458,
+      "step": 2254
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6894868137467697,
+      "learning_rate": 1.2623281400137383e-05,
+      "loss": 1.0391,
+      "step": 2255
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6990583155690552,
+      "learning_rate": 1.2617268257258051e-05,
+      "loss": 1.0591,
+      "step": 2256
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7075414096708934,
+      "learning_rate": 1.2611254098265063e-05,
+      "loss": 0.9564,
+      "step": 2257
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6972404263999001,
+      "learning_rate": 1.2605238925493326e-05,
+      "loss": 0.9477,
+      "step": 2258
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.498980864328532,
+      "learning_rate": 1.2599222741278136e-05,
+      "loss": 1.1345,
+      "step": 2259
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5637235260951151,
+      "learning_rate": 1.2593205547955185e-05,
+      "loss": 1.0635,
+      "step": 2260
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7358797305709297,
+      "learning_rate": 1.2587187347860554e-05,
+      "loss": 1.0622,
+      "step": 2261
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5323216745558956,
+      "learning_rate": 1.2581168143330716e-05,
+      "loss": 1.067,
+      "step": 2262
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6841675914994454,
+      "learning_rate": 1.2575147936702531e-05,
+      "loss": 1.03,
+      "step": 2263
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6494778604044265,
+      "learning_rate": 1.2569126730313255e-05,
+      "loss": 0.9551,
+      "step": 2264
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6446948988230105,
+      "learning_rate": 1.2563104526500523e-05,
+      "loss": 1.0179,
+      "step": 2265
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6553313856041643,
+      "learning_rate": 1.2557081327602361e-05,
+      "loss": 1.0939,
+      "step": 2266
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7112188484067391,
+      "learning_rate": 1.2551057135957187e-05,
+      "loss": 1.1027,
+      "step": 2267
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7079833538282807,
+      "learning_rate": 1.2545031953903796e-05,
+      "loss": 1.0431,
+      "step": 2268
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7043570007664874,
+      "learning_rate": 1.2539005783781374e-05,
+      "loss": 1.0123,
+      "step": 2269
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7153130412259205,
+      "learning_rate": 1.2532978627929486e-05,
+      "loss": 1.0483,
+      "step": 2270
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7033279555229512,
+      "learning_rate": 1.2526950488688083e-05,
+      "loss": 0.9275,
+      "step": 2271
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7024151476878582,
+      "learning_rate": 1.2520921368397492e-05,
+      "loss": 1.0926,
+      "step": 2272
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7066143428623564,
+      "learning_rate": 1.2514891269398429e-05,
+      "loss": 1.0469,
+      "step": 2273
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7878380845337916,
+      "learning_rate": 1.2508860194031986e-05,
+      "loss": 0.969,
+      "step": 2274
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6622925832322312,
+      "learning_rate": 1.2502828144639629e-05,
+      "loss": 1.0572,
+      "step": 2275
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.731379498576077,
+      "learning_rate": 1.2496795123563218e-05,
+      "loss": 1.0067,
+      "step": 2276
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7201058898751265,
+      "learning_rate": 1.249076113314497e-05,
+      "loss": 1.0504,
+      "step": 2277
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7256268112696773,
+      "learning_rate": 1.248472617572749e-05,
+      "loss": 1.003,
+      "step": 2278
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7258363556913576,
+      "learning_rate": 1.2478690253653756e-05,
+      "loss": 1.0158,
+      "step": 2279
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7095876887785567,
+      "learning_rate": 1.2472653369267122e-05,
+      "loss": 1.0398,
+      "step": 2280
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7709010751095551,
+      "learning_rate": 1.2466615524911316e-05,
+      "loss": 1.0082,
+      "step": 2281
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6737387751461883,
+      "learning_rate": 1.2460576722930432e-05,
+      "loss": 1.0489,
+      "step": 2282
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6612912232174752,
+      "learning_rate": 1.2454536965668949e-05,
+      "loss": 1.0081,
+      "step": 2283
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7043900368891725,
+      "learning_rate": 1.24484962554717e-05,
+      "loss": 1.0281,
+      "step": 2284
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5382695182217136,
+      "learning_rate": 1.24424545946839e-05,
+      "loss": 1.0596,
+      "step": 2285
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7474136910632865,
+      "learning_rate": 1.2436411985651131e-05,
+      "loss": 1.0029,
+      "step": 2286
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6784205639534885,
+      "learning_rate": 1.2430368430719342e-05,
+      "loss": 1.0042,
+      "step": 2287
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6637158633362711,
+      "learning_rate": 1.242432393223485e-05,
+      "loss": 1.033,
+      "step": 2288
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6765053799883055,
+      "learning_rate": 1.2418278492544328e-05,
+      "loss": 1.0075,
+      "step": 2289
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5772166213238326,
+      "learning_rate": 1.2412232113994841e-05,
+      "loss": 1.0235,
+      "step": 2290
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7426104807517094,
+      "learning_rate": 1.2406184798933786e-05,
+      "loss": 1.0394,
+      "step": 2291
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7109609349789278,
+      "learning_rate": 1.2400136549708945e-05,
+      "loss": 1.0504,
+      "step": 2292
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7064225777400686,
+      "learning_rate": 1.239408736866846e-05,
+      "loss": 0.9557,
+      "step": 2293
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7091504774054717,
+      "learning_rate": 1.2388037258160823e-05,
+      "loss": 1.0486,
+      "step": 2294
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5743960736521975,
+      "learning_rate": 1.23819862205349e-05,
+      "loss": 1.0956,
+      "step": 2295
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6810173476340398,
+      "learning_rate": 1.2375934258139917e-05,
+      "loss": 1.095,
+      "step": 2296
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7047210493508029,
+      "learning_rate": 1.2369881373325448e-05,
+      "loss": 1.0958,
+      "step": 2297
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6713079439814668,
+      "learning_rate": 1.236382756844143e-05,
+      "loss": 1.0071,
+      "step": 2298
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.718711668489876,
+      "learning_rate": 1.2357772845838159e-05,
+      "loss": 0.96,
+      "step": 2299
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6511885437597705,
+      "learning_rate": 1.2351717207866292e-05,
+      "loss": 1.0803,
+      "step": 2300
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7109798813752426,
+      "learning_rate": 1.2345660656876832e-05,
+      "loss": 1.061,
+      "step": 2301
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5731884056676405,
+      "learning_rate": 1.233960319522114e-05,
+      "loss": 1.1075,
+      "step": 2302
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.764205997476855,
+      "learning_rate": 1.2333544825250938e-05,
+      "loss": 1.012,
+      "step": 2303
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6524650546889943,
+      "learning_rate": 1.2327485549318285e-05,
+      "loss": 1.0562,
+      "step": 2304
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6873930207945744,
+      "learning_rate": 1.2321425369775601e-05,
+      "loss": 1.0505,
+      "step": 2305
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7147197881515772,
+      "learning_rate": 1.2315364288975665e-05,
+      "loss": 1.084,
+      "step": 2306
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.677964015492752,
+      "learning_rate": 1.2309302309271587e-05,
+      "loss": 1.0,
+      "step": 2307
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6171467889490304,
+      "learning_rate": 1.2303239433016842e-05,
+      "loss": 1.0583,
+      "step": 2308
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7635146665893098,
+      "learning_rate": 1.2297175662565248e-05,
+      "loss": 1.0546,
+      "step": 2309
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7437059951028514,
+      "learning_rate": 1.229111100027097e-05,
+      "loss": 1.0273,
+      "step": 2310
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7096674087166296,
+      "learning_rate": 1.228504544848851e-05,
+      "loss": 1.0915,
+      "step": 2311
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8009209519351916,
+      "learning_rate": 1.2278979009572736e-05,
+      "loss": 0.9495,
+      "step": 2312
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7303862433969092,
+      "learning_rate": 1.2272911685878841e-05,
+      "loss": 1.1032,
+      "step": 2313
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.739938864013745,
+      "learning_rate": 1.2266843479762372e-05,
+      "loss": 1.0017,
+      "step": 2314
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7077672496750805,
+      "learning_rate": 1.2260774393579209e-05,
+      "loss": 0.9984,
+      "step": 2315
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7321291207633092,
+      "learning_rate": 1.2254704429685593e-05,
+      "loss": 1.1031,
+      "step": 2316
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6674741358689177,
+      "learning_rate": 1.2248633590438084e-05,
+      "loss": 1.0251,
+      "step": 2317
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7113928048631364,
+      "learning_rate": 1.2242561878193589e-05,
+      "loss": 0.9675,
+      "step": 2318
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6063845649072886,
+      "learning_rate": 1.2236489295309362e-05,
+      "loss": 1.0016,
+      "step": 2319
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.733733610436993,
+      "learning_rate": 1.2230415844142984e-05,
+      "loss": 1.0089,
+      "step": 2320
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7304080321017937,
+      "learning_rate": 1.2224341527052378e-05,
+      "loss": 1.0319,
+      "step": 2321
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5632031115842902,
+      "learning_rate": 1.2218266346395811e-05,
+      "loss": 1.0704,
+      "step": 2322
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7170242001837285,
+      "learning_rate": 1.221219030453187e-05,
+      "loss": 1.0427,
+      "step": 2323
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.736052487453604,
+      "learning_rate": 1.220611340381948e-05,
+      "loss": 0.937,
+      "step": 2324
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5366153465281427,
+      "learning_rate": 1.2200035646617912e-05,
+      "loss": 1.023,
+      "step": 2325
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.698682408253766,
+      "learning_rate": 1.2193957035286757e-05,
+      "loss": 1.0353,
+      "step": 2326
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6640701503835135,
+      "learning_rate": 1.2187877572185937e-05,
+      "loss": 1.0094,
+      "step": 2327
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7018184765814728,
+      "learning_rate": 1.2181797259675713e-05,
+      "loss": 1.0397,
+      "step": 2328
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6926267338781826,
+      "learning_rate": 1.2175716100116677e-05,
+      "loss": 1.0459,
+      "step": 2329
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7271037646732713,
+      "learning_rate": 1.2169634095869736e-05,
+      "loss": 1.0343,
+      "step": 2330
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6616629482663929,
+      "learning_rate": 1.2163551249296132e-05,
+      "loss": 1.0461,
+      "step": 2331
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6691408366464388,
+      "learning_rate": 1.2157467562757443e-05,
+      "loss": 1.0415,
+      "step": 2332
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7710656666120231,
+      "learning_rate": 1.2151383038615563e-05,
+      "loss": 1.041,
+      "step": 2333
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6580411659013911,
+      "learning_rate": 1.214529767923271e-05,
+      "loss": 1.0106,
+      "step": 2334
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.808213789727779,
+      "learning_rate": 1.2139211486971436e-05,
+      "loss": 1.0397,
+      "step": 2335
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7180533489031021,
+      "learning_rate": 1.213312446419461e-05,
+      "loss": 1.0115,
+      "step": 2336
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5179349567854843,
+      "learning_rate": 1.2127036613265418e-05,
+      "loss": 1.0078,
+      "step": 2337
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7261382968929804,
+      "learning_rate": 1.2120947936547375e-05,
+      "loss": 1.0403,
+      "step": 2338
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8414917796575424,
+      "learning_rate": 1.2114858436404322e-05,
+      "loss": 1.0305,
+      "step": 2339
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5444264538781157,
+      "learning_rate": 1.2108768115200405e-05,
+      "loss": 1.0788,
+      "step": 2340
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6789012180796347,
+      "learning_rate": 1.2102676975300095e-05,
+      "loss": 1.0233,
+      "step": 2341
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.8125960151385193,
+      "learning_rate": 1.209658501906819e-05,
+      "loss": 1.0905,
+      "step": 2342
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6592522913111932,
+      "learning_rate": 1.2090492248869795e-05,
+      "loss": 0.9839,
+      "step": 2343
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7170599258757805,
+      "learning_rate": 1.2084398667070325e-05,
+      "loss": 0.9966,
+      "step": 2344
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7336621090877136,
+      "learning_rate": 1.2078304276035527e-05,
+      "loss": 1.0347,
+      "step": 2345
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7145778233762355,
+      "learning_rate": 1.2072209078131451e-05,
+      "loss": 1.0128,
+      "step": 2346
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7126587514205078,
+      "learning_rate": 1.2066113075724461e-05,
+      "loss": 1.0426,
+      "step": 2347
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6694913634949933,
+      "learning_rate": 1.206001627118124e-05,
+      "loss": 1.0626,
+      "step": 2348
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7554545974025539,
+      "learning_rate": 1.2053918666868776e-05,
+      "loss": 1.1035,
+      "step": 2349
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6226089562010468,
+      "learning_rate": 1.2047820265154362e-05,
+      "loss": 0.9496,
+      "step": 2350
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6598932069643314,
+      "learning_rate": 1.2041721068405614e-05,
+      "loss": 1.0429,
+      "step": 2351
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7705273127638652,
+      "learning_rate": 1.203562107899045e-05,
+      "loss": 1.0033,
+      "step": 2352
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6997018038923202,
+      "learning_rate": 1.2029520299277095e-05,
+      "loss": 0.9913,
+      "step": 2353
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6112664280353274,
+      "learning_rate": 1.2023418731634078e-05,
+      "loss": 0.9642,
+      "step": 2354
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.676779406673276,
+      "learning_rate": 1.2017316378430244e-05,
+      "loss": 1.0728,
+      "step": 2355
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5937583161923989,
+      "learning_rate": 1.2011213242034733e-05,
+      "loss": 1.0636,
+      "step": 2356
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7302524469910767,
+      "learning_rate": 1.2005109324816992e-05,
+      "loss": 1.0547,
+      "step": 2357
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7474906613924379,
+      "learning_rate": 1.1999004629146775e-05,
+      "loss": 1.0464,
+      "step": 2358
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5827194038957096,
+      "learning_rate": 1.1992899157394133e-05,
+      "loss": 1.1387,
+      "step": 2359
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7172731016852093,
+      "learning_rate": 1.1986792911929418e-05,
+      "loss": 0.9868,
+      "step": 2360
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5716400788314572,
+      "learning_rate": 1.198068589512329e-05,
+      "loss": 1.0359,
+      "step": 2361
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6919569482751787,
+      "learning_rate": 1.1974578109346702e-05,
+      "loss": 1.0425,
+      "step": 2362
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6973931535392067,
+      "learning_rate": 1.1968469556970905e-05,
+      "loss": 1.0695,
+      "step": 2363
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6844086250467035,
+      "learning_rate": 1.1962360240367445e-05,
+      "loss": 1.0287,
+      "step": 2364
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5193192175769301,
+      "learning_rate": 1.1956250161908179e-05,
+      "loss": 1.0439,
+      "step": 2365
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7196702271793837,
+      "learning_rate": 1.195013932396524e-05,
+      "loss": 1.0544,
+      "step": 2366
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6919240676063224,
+      "learning_rate": 1.1944027728911072e-05,
+      "loss": 0.9493,
+      "step": 2367
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6490276629941918,
+      "learning_rate": 1.1937915379118406e-05,
+      "loss": 1.0312,
+      "step": 2368
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6835754531912005,
+      "learning_rate": 1.1931802276960265e-05,
+      "loss": 1.0823,
+      "step": 2369
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7435597487462167,
+      "learning_rate": 1.1925688424809965e-05,
+      "loss": 1.0292,
+      "step": 2370
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.646415564667492,
+      "learning_rate": 1.1919573825041115e-05,
+      "loss": 1.0449,
+      "step": 2371
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6656086191274877,
+      "learning_rate": 1.1913458480027614e-05,
+      "loss": 1.0414,
+      "step": 2372
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6445661590990054,
+      "learning_rate": 1.1907342392143646e-05,
+      "loss": 0.9849,
+      "step": 2373
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7430821744792137,
+      "learning_rate": 1.1901225563763694e-05,
+      "loss": 1.0323,
+      "step": 2374
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7548793832273744,
+      "learning_rate": 1.1895107997262516e-05,
+      "loss": 0.9688,
+      "step": 2375
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7269921745814897,
+      "learning_rate": 1.1888989695015166e-05,
+      "loss": 1.0917,
+      "step": 2376
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8217405580816745,
+      "learning_rate": 1.1882870659396968e-05,
+      "loss": 0.9992,
+      "step": 2377
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7220497194633676,
+      "learning_rate": 1.1876750892783558e-05,
+      "loss": 0.9372,
+      "step": 2378
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6670625326434649,
+      "learning_rate": 1.1870630397550831e-05,
+      "loss": 1.0464,
+      "step": 2379
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6939141386487693,
+      "learning_rate": 1.1864509176074974e-05,
+      "loss": 1.0767,
+      "step": 2380
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7546420684960731,
+      "learning_rate": 1.185838723073246e-05,
+      "loss": 1.0037,
+      "step": 2381
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7324903305854644,
+      "learning_rate": 1.1852264563900038e-05,
+      "loss": 1.0726,
+      "step": 2382
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7548947787153292,
+      "learning_rate": 1.1846141177954733e-05,
+      "loss": 1.0344,
+      "step": 2383
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6526307062954657,
+      "learning_rate": 1.1840017075273861e-05,
+      "loss": 1.0152,
+      "step": 2384
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6507964577780937,
+      "learning_rate": 1.1833892258235008e-05,
+      "loss": 1.0018,
+      "step": 2385
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5651805698352278,
+      "learning_rate": 1.1827766729216035e-05,
+      "loss": 1.1089,
+      "step": 2386
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7473022075475,
+      "learning_rate": 1.1821640490595086e-05,
+      "loss": 1.0787,
+      "step": 2387
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5308259574065862,
+      "learning_rate": 1.181551354475058e-05,
+      "loss": 1.0836,
+      "step": 2388
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6483437651032647,
+      "learning_rate": 1.1809385894061206e-05,
+      "loss": 0.9825,
+      "step": 2389
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6754765354913135,
+      "learning_rate": 1.1803257540905926e-05,
+      "loss": 1.0527,
+      "step": 2390
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6976721289813248,
+      "learning_rate": 1.1797128487663982e-05,
+      "loss": 1.0728,
+      "step": 2391
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5961033118524163,
+      "learning_rate": 1.1790998736714882e-05,
+      "loss": 1.1854,
+      "step": 2392
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6825304620868191,
+      "learning_rate": 1.1784868290438404e-05,
+      "loss": 1.0678,
+      "step": 2393
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8058835613768393,
+      "learning_rate": 1.1778737151214606e-05,
+      "loss": 1.0573,
+      "step": 2394
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6851006310845926,
+      "learning_rate": 1.17726053214238e-05,
+      "loss": 0.9919,
+      "step": 2395
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7217972469433653,
+      "learning_rate": 1.1766472803446577e-05,
+      "loss": 1.0142,
+      "step": 2396
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6755891381322748,
+      "learning_rate": 1.1760339599663788e-05,
+      "loss": 1.0154,
+      "step": 2397
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.664665397018089,
+      "learning_rate": 1.1754205712456556e-05,
+      "loss": 1.0258,
+      "step": 2398
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6837791693002806,
+      "learning_rate": 1.1748071144206266e-05,
+      "loss": 0.9918,
+      "step": 2399
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7114954827574548,
+      "learning_rate": 1.1741935897294572e-05,
+      "loss": 1.0844,
+      "step": 2400
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6617275283291627,
+      "learning_rate": 1.1735799974103388e-05,
+      "loss": 1.0266,
+      "step": 2401
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7317970060620539,
+      "learning_rate": 1.1729663377014888e-05,
+      "loss": 1.0242,
+      "step": 2402
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6610604887710545,
+      "learning_rate": 1.172352610841151e-05,
+      "loss": 1.0681,
+      "step": 2403
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6609896867886342,
+      "learning_rate": 1.1717388170675954e-05,
+      "loss": 1.0023,
+      "step": 2404
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6520022401573565,
+      "learning_rate": 1.1711249566191179e-05,
+      "loss": 0.9845,
+      "step": 2405
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6728623317656051,
+      "learning_rate": 1.17051102973404e-05,
+      "loss": 1.0688,
+      "step": 2406
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7289883371258559,
+      "learning_rate": 1.1698970366507096e-05,
+      "loss": 0.9984,
+      "step": 2407
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6681044927081206,
+      "learning_rate": 1.1692829776074999e-05,
+      "loss": 0.9828,
+      "step": 2408
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7674361813508103,
+      "learning_rate": 1.1686688528428099e-05,
+      "loss": 0.9639,
+      "step": 2409
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7554350089355784,
+      "learning_rate": 1.1680546625950635e-05,
+      "loss": 1.0518,
+      "step": 2410
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7190544054581547,
+      "learning_rate": 1.167440407102711e-05,
+      "loss": 1.0855,
+      "step": 2411
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7361094348376571,
+      "learning_rate": 1.1668260866042271e-05,
+      "loss": 1.0368,
+      "step": 2412
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7115167706256681,
+      "learning_rate": 1.1662117013381126e-05,
+      "loss": 0.9593,
+      "step": 2413
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7253419534243358,
+      "learning_rate": 1.1655972515428928e-05,
+      "loss": 1.0685,
+      "step": 2414
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7525516655476682,
+      "learning_rate": 1.1649827374571182e-05,
+      "loss": 1.028,
+      "step": 2415
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7591630505253837,
+      "learning_rate": 1.1643681593193642e-05,
+      "loss": 0.9494,
+      "step": 2416
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5498351319090033,
+      "learning_rate": 1.1637535173682318e-05,
+      "loss": 1.0667,
+      "step": 2417
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6778273846114832,
+      "learning_rate": 1.1631388118423457e-05,
+      "loss": 1.0336,
+      "step": 2418
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6874876308615016,
+      "learning_rate": 1.1625240429803553e-05,
+      "loss": 0.9785,
+      "step": 2419
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7051799663897811,
+      "learning_rate": 1.1619092110209361e-05,
+      "loss": 1.0037,
+      "step": 2420
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7039883553870129,
+      "learning_rate": 1.1612943162027863e-05,
+      "loss": 1.0199,
+      "step": 2421
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7000998791481106,
+      "learning_rate": 1.1606793587646295e-05,
+      "loss": 1.0048,
+      "step": 2422
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.641000738609702,
+      "learning_rate": 1.160064338945213e-05,
+      "loss": 1.0261,
+      "step": 2423
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7557411539665355,
+      "learning_rate": 1.1594492569833093e-05,
+      "loss": 0.9774,
+      "step": 2424
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7038454487027319,
+      "learning_rate": 1.1588341131177137e-05,
+      "loss": 1.0174,
+      "step": 2425
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7056532784576763,
+      "learning_rate": 1.1582189075872467e-05,
+      "loss": 0.9884,
+      "step": 2426
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.668062616793114,
+      "learning_rate": 1.1576036406307523e-05,
+      "loss": 0.9995,
+      "step": 2427
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6466723418085567,
+      "learning_rate": 1.156988312487098e-05,
+      "loss": 1.0128,
+      "step": 2428
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6535868157692939,
+      "learning_rate": 1.1563729233951757e-05,
+      "loss": 0.9958,
+      "step": 2429
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.654009874731062,
+      "learning_rate": 1.1557574735939003e-05,
+      "loss": 0.9899,
+      "step": 2430
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6354248725389403,
+      "learning_rate": 1.1551419633222107e-05,
+      "loss": 0.953,
+      "step": 2431
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7149301883676448,
+      "learning_rate": 1.1545263928190692e-05,
+      "loss": 1.0491,
+      "step": 2432
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8165802756845808,
+      "learning_rate": 1.1539107623234618e-05,
+      "loss": 0.9796,
+      "step": 2433
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6919831189826264,
+      "learning_rate": 1.153295072074397e-05,
+      "loss": 1.0468,
+      "step": 2434
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7045723451388274,
+      "learning_rate": 1.1526793223109072e-05,
+      "loss": 1.037,
+      "step": 2435
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6938039751618345,
+      "learning_rate": 1.1520635132720475e-05,
+      "loss": 1.0352,
+      "step": 2436
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.692515095246435,
+      "learning_rate": 1.1514476451968961e-05,
+      "loss": 1.0413,
+      "step": 2437
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6439456930630795,
+      "learning_rate": 1.1508317183245545e-05,
+      "loss": 0.9855,
+      "step": 2438
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7489669372475095,
+      "learning_rate": 1.1502157328941466e-05,
+      "loss": 0.9943,
+      "step": 2439
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6602296723524124,
+      "learning_rate": 1.149599689144819e-05,
+      "loss": 1.0181,
+      "step": 2440
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6349372835870953,
+      "learning_rate": 1.1489835873157414e-05,
+      "loss": 1.003,
+      "step": 2441
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7034861590097552,
+      "learning_rate": 1.1483674276461053e-05,
+      "loss": 1.063,
+      "step": 2442
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6021269479790726,
+      "learning_rate": 1.1477512103751254e-05,
+      "loss": 1.071,
+      "step": 2443
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6873953907909276,
+      "learning_rate": 1.1471349357420384e-05,
+      "loss": 1.0228,
+      "step": 2444
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6581325456446048,
+      "learning_rate": 1.1465186039861033e-05,
+      "loss": 0.9531,
+      "step": 2445
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.691404737521745,
+      "learning_rate": 1.1459022153466016e-05,
+      "loss": 0.9501,
+      "step": 2446
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6651315312655791,
+      "learning_rate": 1.1452857700628362e-05,
+      "loss": 0.9214,
+      "step": 2447
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7241089549092644,
+      "learning_rate": 1.1446692683741326e-05,
+      "loss": 1.0699,
+      "step": 2448
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6652966394663107,
+      "learning_rate": 1.1440527105198377e-05,
+      "loss": 0.9708,
+      "step": 2449
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.660944750377448,
+      "learning_rate": 1.143436096739321e-05,
+      "loss": 1.0262,
+      "step": 2450
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6465991690221806,
+      "learning_rate": 1.1428194272719729e-05,
+      "loss": 1.0195,
+      "step": 2451
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7109864770357025,
+      "learning_rate": 1.1422027023572052e-05,
+      "loss": 1.0507,
+      "step": 2452
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8063250007190658,
+      "learning_rate": 1.1415859222344525e-05,
+      "loss": 1.0678,
+      "step": 2453
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6652663504751192,
+      "learning_rate": 1.14096908714317e-05,
+      "loss": 1.0734,
+      "step": 2454
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7238395295834184,
+      "learning_rate": 1.1403521973228342e-05,
+      "loss": 0.9802,
+      "step": 2455
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.706254435257777,
+      "learning_rate": 1.1397352530129428e-05,
+      "loss": 1.0353,
+      "step": 2456
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7602232749048063,
+      "learning_rate": 1.139118254453015e-05,
+      "loss": 1.015,
+      "step": 2457
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6206170619239374,
+      "learning_rate": 1.1385012018825907e-05,
+      "loss": 0.9722,
+      "step": 2458
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5032215508412121,
+      "learning_rate": 1.1378840955412313e-05,
+      "loss": 1.0245,
+      "step": 2459
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6743620637761332,
+      "learning_rate": 1.1372669356685185e-05,
+      "loss": 0.9809,
+      "step": 2460
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6459412751671945,
+      "learning_rate": 1.1366497225040549e-05,
+      "loss": 0.9896,
+      "step": 2461
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7734102920177123,
+      "learning_rate": 1.1360324562874643e-05,
+      "loss": 1.0264,
+      "step": 2462
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6752301908622097,
+      "learning_rate": 1.1354151372583901e-05,
+      "loss": 1.0555,
+      "step": 2463
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5659080695732166,
+      "learning_rate": 1.1347977656564974e-05,
+      "loss": 1.0714,
+      "step": 2464
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5415415710451493,
+      "learning_rate": 1.1341803417214705e-05,
+      "loss": 1.0863,
+      "step": 2465
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7317790008584674,
+      "learning_rate": 1.1335628656930153e-05,
+      "loss": 1.0306,
+      "step": 2466
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6965280828938263,
+      "learning_rate": 1.132945337810857e-05,
+      "loss": 0.9818,
+      "step": 2467
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7212532966166176,
+      "learning_rate": 1.132327758314741e-05,
+      "loss": 0.9829,
+      "step": 2468
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.670463428611195,
+      "learning_rate": 1.131710127444433e-05,
+      "loss": 1.1045,
+      "step": 2469
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.685068934879548,
+      "learning_rate": 1.1310924454397187e-05,
+      "loss": 1.0374,
+      "step": 2470
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7058314405942129,
+      "learning_rate": 1.1304747125404031e-05,
+      "loss": 0.9879,
+      "step": 2471
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6259322368514995,
+      "learning_rate": 1.129856928986312e-05,
+      "loss": 1.0265,
+      "step": 2472
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6544376922994798,
+      "learning_rate": 1.12923909501729e-05,
+      "loss": 1.0865,
+      "step": 2473
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5257791607557887,
+      "learning_rate": 1.1286212108732015e-05,
+      "loss": 1.14,
+      "step": 2474
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7580071594528006,
+      "learning_rate": 1.1280032767939302e-05,
+      "loss": 1.0597,
+      "step": 2475
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6696478098942166,
+      "learning_rate": 1.1273852930193798e-05,
+      "loss": 0.9993,
+      "step": 2476
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6504671250588543,
+      "learning_rate": 1.1267672597894725e-05,
+      "loss": 1.0015,
+      "step": 2477
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.810580324240718,
+      "learning_rate": 1.12614917734415e-05,
+      "loss": 1.0884,
+      "step": 2478
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6866386831206659,
+      "learning_rate": 1.1255310459233737e-05,
+      "loss": 1.0032,
+      "step": 2479
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7092110390764081,
+      "learning_rate": 1.1249128657671233e-05,
+      "loss": 0.9665,
+      "step": 2480
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7104702204291053,
+      "learning_rate": 1.1242946371153974e-05,
+      "loss": 0.9918,
+      "step": 2481
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6388025507067393,
+      "learning_rate": 1.1236763602082136e-05,
+      "loss": 1.0507,
+      "step": 2482
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6934185583854936,
+      "learning_rate": 1.1230580352856088e-05,
+      "loss": 1.0402,
+      "step": 2483
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.691890641551174,
+      "learning_rate": 1.1224396625876375e-05,
+      "loss": 1.0117,
+      "step": 2484
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6548023442777333,
+      "learning_rate": 1.1218212423543734e-05,
+      "loss": 1.0631,
+      "step": 2485
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7814789039789378,
+      "learning_rate": 1.1212027748259086e-05,
+      "loss": 1.0502,
+      "step": 2486
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.687821563773772,
+      "learning_rate": 1.1205842602423537e-05,
+      "loss": 1.0624,
+      "step": 2487
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6631274885904151,
+      "learning_rate": 1.1199656988438373e-05,
+      "loss": 1.0244,
+      "step": 2488
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6859647232783098,
+      "learning_rate": 1.1193470908705055e-05,
+      "loss": 1.0226,
+      "step": 2489
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6301526607949247,
+      "learning_rate": 1.1187284365625241e-05,
+      "loss": 1.0371,
+      "step": 2490
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5386099222859392,
+      "learning_rate": 1.1181097361600754e-05,
+      "loss": 1.1121,
+      "step": 2491
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6917960210408995,
+      "learning_rate": 1.1174909899033608e-05,
+      "loss": 1.0129,
+      "step": 2492
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7004728299513961,
+      "learning_rate": 1.1168721980325987e-05,
+      "loss": 0.9976,
+      "step": 2493
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6335857130451646,
+      "learning_rate": 1.1162533607880251e-05,
+      "loss": 0.955,
+      "step": 2494
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6996052478112758,
+      "learning_rate": 1.1156344784098942e-05,
+      "loss": 1.0613,
+      "step": 2495
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7214563644261359,
+      "learning_rate": 1.1150155511384772e-05,
+      "loss": 1.0504,
+      "step": 2496
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6916999785778614,
+      "learning_rate": 1.1143965792140631e-05,
+      "loss": 0.995,
+      "step": 2497
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6749592882688648,
+      "learning_rate": 1.1137775628769584e-05,
+      "loss": 1.0229,
+      "step": 2498
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6989647652822519,
+      "learning_rate": 1.1131585023674863e-05,
+      "loss": 1.0021,
+      "step": 2499
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7551709631035304,
+      "learning_rate": 1.1125393979259874e-05,
+      "loss": 0.9537,
+      "step": 2500
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7258190995296585,
+      "learning_rate": 1.1119202497928192e-05,
+      "loss": 1.0153,
+      "step": 2501
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6951303932426309,
+      "learning_rate": 1.1113010582083568e-05,
+      "loss": 0.9962,
+      "step": 2502
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5414772503021229,
+      "learning_rate": 1.1106818234129913e-05,
+      "loss": 1.0475,
+      "step": 2503
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6949288593400955,
+      "learning_rate": 1.1100625456471307e-05,
+      "loss": 1.0134,
+      "step": 2504
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6910019648740036,
+      "learning_rate": 1.1094432251512006e-05,
+      "loss": 1.0683,
+      "step": 2505
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6003776920997027,
+      "learning_rate": 1.1088238621656422e-05,
+      "loss": 1.0392,
+      "step": 2506
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.677663771047151,
+      "learning_rate": 1.1082044569309138e-05,
+      "loss": 0.9811,
+      "step": 2507
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6806531694600539,
+      "learning_rate": 1.1075850096874894e-05,
+      "loss": 1.0335,
+      "step": 2508
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.680391252145527,
+      "learning_rate": 1.1069655206758603e-05,
+      "loss": 1.024,
+      "step": 2509
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6748536102524181,
+      "learning_rate": 1.1063459901365325e-05,
+      "loss": 1.0586,
+      "step": 2510
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6878519050258323,
+      "learning_rate": 1.1057264183100303e-05,
+      "loss": 0.9805,
+      "step": 2511
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6935127643924386,
+      "learning_rate": 1.1051068054368921e-05,
+      "loss": 1.0785,
+      "step": 2512
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7630121465325813,
+      "learning_rate": 1.104487151757673e-05,
+      "loss": 1.0623,
+      "step": 2513
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6114478377074776,
+      "learning_rate": 1.1038674575129442e-05,
+      "loss": 0.9405,
+      "step": 2514
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7448666867984626,
+      "learning_rate": 1.1032477229432921e-05,
+      "loss": 0.9308,
+      "step": 2515
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.48860351743366925,
+      "learning_rate": 1.1026279482893187e-05,
+      "loss": 0.9965,
+      "step": 2516
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6289178260443701,
+      "learning_rate": 1.1020081337916425e-05,
+      "loss": 1.0504,
+      "step": 2517
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.742523273675205,
+      "learning_rate": 1.1013882796908963e-05,
+      "loss": 1.0239,
+      "step": 2518
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7021921670764146,
+      "learning_rate": 1.1007683862277292e-05,
+      "loss": 1.0151,
+      "step": 2519
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6444515886032911,
+      "learning_rate": 1.1001484536428052e-05,
+      "loss": 0.9545,
+      "step": 2520
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6666802336246664,
+      "learning_rate": 1.0995284821768029e-05,
+      "loss": 1.0371,
+      "step": 2521
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6782764019106223,
+      "learning_rate": 1.098908472070417e-05,
+      "loss": 1.0457,
+      "step": 2522
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6728036951530598,
+      "learning_rate": 1.0982884235643567e-05,
+      "loss": 1.015,
+      "step": 2523
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6627388972044479,
+      "learning_rate": 1.0976683368993464e-05,
+      "loss": 1.0547,
+      "step": 2524
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6490326965021136,
+      "learning_rate": 1.0970482123161249e-05,
+      "loss": 1.0582,
+      "step": 2525
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.673472442875784,
+      "learning_rate": 1.0964280500554459e-05,
+      "loss": 1.0048,
+      "step": 2526
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7393818109314274,
+      "learning_rate": 1.0958078503580776e-05,
+      "loss": 1.0845,
+      "step": 2527
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6769751450999085,
+      "learning_rate": 1.0951876134648032e-05,
+      "loss": 0.9953,
+      "step": 2528
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7244780149762527,
+      "learning_rate": 1.0945673396164198e-05,
+      "loss": 0.9982,
+      "step": 2529
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6884381334209104,
+      "learning_rate": 1.0939470290537389e-05,
+      "loss": 1.0181,
+      "step": 2530
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7435667945706768,
+      "learning_rate": 1.0933266820175868e-05,
+      "loss": 1.0374,
+      "step": 2531
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7598005435796099,
+      "learning_rate": 1.0927062987488035e-05,
+      "loss": 0.9216,
+      "step": 2532
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7494096717226252,
+      "learning_rate": 1.0920858794882429e-05,
+      "loss": 1.0207,
+      "step": 2533
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7054427164094734,
+      "learning_rate": 1.0914654244767736e-05,
+      "loss": 0.9852,
+      "step": 2534
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6493557439146374,
+      "learning_rate": 1.0908449339552769e-05,
+      "loss": 1.0035,
+      "step": 2535
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5260754269739927,
+      "learning_rate": 1.0902244081646489e-05,
+      "loss": 0.9963,
+      "step": 2536
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6823582823960768,
+      "learning_rate": 1.0896038473457993e-05,
+      "loss": 0.9706,
+      "step": 2537
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6283858930327149,
+      "learning_rate": 1.0889832517396511e-05,
+      "loss": 1.0582,
+      "step": 2538
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7120516280352904,
+      "learning_rate": 1.0883626215871408e-05,
+      "loss": 1.0256,
+      "step": 2539
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7071132514601605,
+      "learning_rate": 1.0877419571292183e-05,
+      "loss": 1.0289,
+      "step": 2540
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8195840245616401,
+      "learning_rate": 1.0871212586068469e-05,
+      "loss": 1.0059,
+      "step": 2541
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7406898125465105,
+      "learning_rate": 1.0865005262610033e-05,
+      "loss": 0.9324,
+      "step": 2542
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.810506636990608,
+      "learning_rate": 1.085879760332677e-05,
+      "loss": 1.0175,
+      "step": 2543
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6750473978556724,
+      "learning_rate": 1.085258961062871e-05,
+      "loss": 1.0701,
+      "step": 2544
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6988650974045061,
+      "learning_rate": 1.0846381286926007e-05,
+      "loss": 0.983,
+      "step": 2545
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7085688340829525,
+      "learning_rate": 1.0840172634628948e-05,
+      "loss": 1.0424,
+      "step": 2546
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.673966167098174,
+      "learning_rate": 1.0833963656147944e-05,
+      "loss": 1.0389,
+      "step": 2547
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6506728907331333,
+      "learning_rate": 1.082775435389353e-05,
+      "loss": 1.017,
+      "step": 2548
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6269480913528078,
+      "learning_rate": 1.0821544730276379e-05,
+      "loss": 0.9133,
+      "step": 2549
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7281034846693851,
+      "learning_rate": 1.0815334787707277e-05,
+      "loss": 1.0132,
+      "step": 2550
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7558269196104037,
+      "learning_rate": 1.0809124528597138e-05,
+      "loss": 0.9599,
+      "step": 2551
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6966114352563489,
+      "learning_rate": 1.0802913955356998e-05,
+      "loss": 1.0073,
+      "step": 2552
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7050201683673749,
+      "learning_rate": 1.0796703070398016e-05,
+      "loss": 1.0374,
+      "step": 2553
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8016844847460913,
+      "learning_rate": 1.079049187613147e-05,
+      "loss": 0.9324,
+      "step": 2554
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7336360593814361,
+      "learning_rate": 1.0784280374968761e-05,
+      "loss": 1.0181,
+      "step": 2555
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5573948679217375,
+      "learning_rate": 1.0778068569321403e-05,
+      "loss": 0.9825,
+      "step": 2556
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.755251621379327,
+      "learning_rate": 1.077185646160104e-05,
+      "loss": 0.97,
+      "step": 2557
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7445607435824513,
+      "learning_rate": 1.0765644054219422e-05,
+      "loss": 1.1188,
+      "step": 2558
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6975966982557358,
+      "learning_rate": 1.0759431349588421e-05,
+      "loss": 1.02,
+      "step": 2559
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7054245528830946,
+      "learning_rate": 1.0753218350120023e-05,
+      "loss": 0.9985,
+      "step": 2560
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6703302796312501,
+      "learning_rate": 1.0747005058226325e-05,
+      "loss": 1.0509,
+      "step": 2561
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6550332021695965,
+      "learning_rate": 1.0740791476319543e-05,
+      "loss": 0.9335,
+      "step": 2562
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7314142203417278,
+      "learning_rate": 1.0734577606812007e-05,
+      "loss": 0.9769,
+      "step": 2563
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6794814931348546,
+      "learning_rate": 1.0728363452116149e-05,
+      "loss": 1.0339,
+      "step": 2564
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7369338272195185,
+      "learning_rate": 1.0722149014644523e-05,
+      "loss": 0.9871,
+      "step": 2565
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6793864929939818,
+      "learning_rate": 1.0715934296809782e-05,
+      "loss": 1.0078,
+      "step": 2566
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6670989557367216,
+      "learning_rate": 1.0709719301024698e-05,
+      "loss": 1.0131,
+      "step": 2567
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6005122464330594,
+      "learning_rate": 1.0703504029702148e-05,
+      "loss": 1.1057,
+      "step": 2568
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7591930334123423,
+      "learning_rate": 1.0697288485255107e-05,
+      "loss": 1.0059,
+      "step": 2569
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7127531446377676,
+      "learning_rate": 1.0691072670096669e-05,
+      "loss": 1.0258,
+      "step": 2570
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6629315578645026,
+      "learning_rate": 1.0684856586640026e-05,
+      "loss": 1.05,
+      "step": 2571
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6701812674455706,
+      "learning_rate": 1.0678640237298476e-05,
+      "loss": 1.0058,
+      "step": 2572
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5050888595782943,
+      "learning_rate": 1.0672423624485423e-05,
+      "loss": 1.0551,
+      "step": 2573
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7060666192703737,
+      "learning_rate": 1.0666206750614363e-05,
+      "loss": 0.9779,
+      "step": 2574
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6695121876405166,
+      "learning_rate": 1.0659989618098904e-05,
+      "loss": 0.9938,
+      "step": 2575
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7471359160021382,
+      "learning_rate": 1.065377222935275e-05,
+      "loss": 1.0573,
+      "step": 2576
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7362191092564309,
+      "learning_rate": 1.0647554586789708e-05,
+      "loss": 0.9837,
+      "step": 2577
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6784725153585726,
+      "learning_rate": 1.064133669282368e-05,
+      "loss": 1.0042,
+      "step": 2578
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.68733595601794,
+      "learning_rate": 1.0635118549868668e-05,
+      "loss": 0.9979,
+      "step": 2579
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7556638838921187,
+      "learning_rate": 1.0628900160338764e-05,
+      "loss": 0.9913,
+      "step": 2580
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6651537553006114,
+      "learning_rate": 1.0622681526648167e-05,
+      "loss": 1.0532,
+      "step": 2581
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5311799438286136,
+      "learning_rate": 1.0616462651211156e-05,
+      "loss": 1.0341,
+      "step": 2582
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7343549925084099,
+      "learning_rate": 1.0610243536442125e-05,
+      "loss": 1.0374,
+      "step": 2583
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6438628442321916,
+      "learning_rate": 1.0604024184755539e-05,
+      "loss": 1.0219,
+      "step": 2584
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6460004374029747,
+      "learning_rate": 1.0597804598565969e-05,
+      "loss": 0.9996,
+      "step": 2585
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6741043537915745,
+      "learning_rate": 1.0591584780288069e-05,
+      "loss": 1.0337,
+      "step": 2586
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.646931547547896,
+      "learning_rate": 1.0585364732336587e-05,
+      "loss": 0.9872,
+      "step": 2587
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7269272363338853,
+      "learning_rate": 1.0579144457126365e-05,
+      "loss": 0.9694,
+      "step": 2588
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6652723212581239,
+      "learning_rate": 1.057292395707232e-05,
+      "loss": 1.0048,
+      "step": 2589
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.700899575983468,
+      "learning_rate": 1.0566703234589471e-05,
+      "loss": 0.9794,
+      "step": 2590
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5738170512097963,
+      "learning_rate": 1.0560482292092912e-05,
+      "loss": 1.0544,
+      "step": 2591
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6969422848116191,
+      "learning_rate": 1.0554261131997833e-05,
+      "loss": 1.0118,
+      "step": 2592
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7718572050163138,
+      "learning_rate": 1.0548039756719497e-05,
+      "loss": 1.0405,
+      "step": 2593
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6477957205092532,
+      "learning_rate": 1.054181816867326e-05,
+      "loss": 0.9784,
+      "step": 2594
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6969866382767496,
+      "learning_rate": 1.053559637027455e-05,
+      "loss": 1.0178,
+      "step": 2595
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6879096175241006,
+      "learning_rate": 1.0529374363938888e-05,
+      "loss": 1.0494,
+      "step": 2596
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6881659682030795,
+      "learning_rate": 1.0523152152081875e-05,
+      "loss": 0.9747,
+      "step": 2597
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6762521101663181,
+      "learning_rate": 1.051692973711918e-05,
+      "loss": 1.0034,
+      "step": 2598
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5057526564180554,
+      "learning_rate": 1.0510707121466568e-05,
+      "loss": 1.0765,
+      "step": 2599
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6481958779414922,
+      "learning_rate": 1.0504484307539864e-05,
+      "loss": 1.0115,
+      "step": 2600
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7428806835362859,
+      "learning_rate": 1.0498261297754984e-05,
+      "loss": 1.0222,
+      "step": 2601
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7322372888090554,
+      "learning_rate": 1.0492038094527907e-05,
+      "loss": 0.9841,
+      "step": 2602
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6865486042082267,
+      "learning_rate": 1.0485814700274706e-05,
+      "loss": 1.0544,
+      "step": 2603
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7231873411038422,
+      "learning_rate": 1.047959111741151e-05,
+      "loss": 1.0027,
+      "step": 2604
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5328781767518035,
+      "learning_rate": 1.0473367348354529e-05,
+      "loss": 1.0707,
+      "step": 2605
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6861587011122535,
+      "learning_rate": 1.0467143395520044e-05,
+      "loss": 1.0262,
+      "step": 2606
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7258187615276758,
+      "learning_rate": 1.046091926132441e-05,
+      "loss": 1.0637,
+      "step": 2607
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6504894466314127,
+      "learning_rate": 1.0454694948184045e-05,
+      "loss": 1.0422,
+      "step": 2608
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4986754886169897,
+      "learning_rate": 1.044847045851545e-05,
+      "loss": 1.0159,
+      "step": 2609
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6784002343800077,
+      "learning_rate": 1.044224579473518e-05,
+      "loss": 1.0111,
+      "step": 2610
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.780674536406171,
+      "learning_rate": 1.0436020959259862e-05,
+      "loss": 1.0493,
+      "step": 2611
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6951776963922702,
+      "learning_rate": 1.0429795954506203e-05,
+      "loss": 1.0654,
+      "step": 2612
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7506641233507519,
+      "learning_rate": 1.0423570782890951e-05,
+      "loss": 1.0729,
+      "step": 2613
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7469760739459017,
+      "learning_rate": 1.0417345446830938e-05,
+      "loss": 0.9916,
+      "step": 2614
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5235721042142697,
+      "learning_rate": 1.0411119948743052e-05,
+      "loss": 1.022,
+      "step": 2615
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7236417182146672,
+      "learning_rate": 1.0404894291044247e-05,
+      "loss": 1.0087,
+      "step": 2616
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.805309496772351,
+      "learning_rate": 1.0398668476151538e-05,
+      "loss": 1.0353,
+      "step": 2617
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6994664880038085,
+      "learning_rate": 1.0392442506482e-05,
+      "loss": 0.973,
+      "step": 2618
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4958541528648602,
+      "learning_rate": 1.038621638445277e-05,
+      "loss": 0.9977,
+      "step": 2619
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7360178551127855,
+      "learning_rate": 1.037999011248104e-05,
+      "loss": 0.9623,
+      "step": 2620
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6534816340531777,
+      "learning_rate": 1.0373763692984062e-05,
+      "loss": 0.9971,
+      "step": 2621
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6828967631119581,
+      "learning_rate": 1.0367537128379154e-05,
+      "loss": 1.0703,
+      "step": 2622
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7041567135901811,
+      "learning_rate": 1.0361310421083677e-05,
+      "loss": 1.0303,
+      "step": 2623
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5664721336297104,
+      "learning_rate": 1.0355083573515052e-05,
+      "loss": 1.0733,
+      "step": 2624
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4970600968712528,
+      "learning_rate": 1.0348856588090764e-05,
+      "loss": 1.0342,
+      "step": 2625
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.699163096208254,
+      "learning_rate": 1.0342629467228331e-05,
+      "loss": 1.0138,
+      "step": 2626
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6345280844485682,
+      "learning_rate": 1.0336402213345345e-05,
+      "loss": 1.0148,
+      "step": 2627
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7037931316311775,
+      "learning_rate": 1.0330174828859434e-05,
+      "loss": 1.0479,
+      "step": 2628
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6574470886246415,
+      "learning_rate": 1.0323947316188288e-05,
+      "loss": 1.0787,
+      "step": 2629
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.815868665118103,
+      "learning_rate": 1.031771967774964e-05,
+      "loss": 0.9763,
+      "step": 2630
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7591080956707762,
+      "learning_rate": 1.0311491915961271e-05,
+      "loss": 1.0159,
+      "step": 2631
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6621606654696779,
+      "learning_rate": 1.030526403324102e-05,
+      "loss": 0.9899,
+      "step": 2632
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7771100538638763,
+      "learning_rate": 1.0299036032006759e-05,
+      "loss": 1.0187,
+      "step": 2633
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.75478578996768,
+      "learning_rate": 1.0292807914676412e-05,
+      "loss": 1.0389,
+      "step": 2634
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6348988141756742,
+      "learning_rate": 1.0286579683667952e-05,
+      "loss": 1.0046,
+      "step": 2635
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6808219382844073,
+      "learning_rate": 1.0280351341399392e-05,
+      "loss": 0.9497,
+      "step": 2636
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7704011616287824,
+      "learning_rate": 1.027412289028879e-05,
+      "loss": 1.0307,
+      "step": 2637
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7143220178944948,
+      "learning_rate": 1.0267894332754243e-05,
+      "loss": 1.0344,
+      "step": 2638
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7012567361373758,
+      "learning_rate": 1.0261665671213891e-05,
+      "loss": 0.9628,
+      "step": 2639
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7528648630404586,
+      "learning_rate": 1.0255436908085919e-05,
+      "loss": 0.9767,
+      "step": 2640
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7495814584779894,
+      "learning_rate": 1.024920804578854e-05,
+      "loss": 1.0176,
+      "step": 2641
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7996774491369094,
+      "learning_rate": 1.0242979086740019e-05,
+      "loss": 1.0084,
+      "step": 2642
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6979738849745181,
+      "learning_rate": 1.023675003335865e-05,
+      "loss": 1.0021,
+      "step": 2643
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6827660717421822,
+      "learning_rate": 1.0230520888062765e-05,
+      "loss": 0.999,
+      "step": 2644
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6988644917897641,
+      "learning_rate": 1.0224291653270739e-05,
+      "loss": 0.9992,
+      "step": 2645
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6467209566571834,
+      "learning_rate": 1.0218062331400969e-05,
+      "loss": 1.0596,
+      "step": 2646
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5679534719643543,
+      "learning_rate": 1.0211832924871889e-05,
+      "loss": 1.0326,
+      "step": 2647
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7183305589645208,
+      "learning_rate": 1.0205603436101978e-05,
+      "loss": 1.0627,
+      "step": 2648
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.730769620686021,
+      "learning_rate": 1.0199373867509734e-05,
+      "loss": 1.0192,
+      "step": 2649
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6702624980918108,
+      "learning_rate": 1.019314422151369e-05,
+      "loss": 1.0135,
+      "step": 2650
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7271811050783858,
+      "learning_rate": 1.0186914500532408e-05,
+      "loss": 0.9912,
+      "step": 2651
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6680399261492277,
+      "learning_rate": 1.0180684706984483e-05,
+      "loss": 1.0306,
+      "step": 2652
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5425545864095073,
+      "learning_rate": 1.0174454843288533e-05,
+      "loss": 1.074,
+      "step": 2653
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7659825244759423,
+      "learning_rate": 1.0168224911863205e-05,
+      "loss": 0.9746,
+      "step": 2654
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6562305550288603,
+      "learning_rate": 1.0161994915127173e-05,
+      "loss": 1.05,
+      "step": 2655
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6994419346137394,
+      "learning_rate": 1.015576485549914e-05,
+      "loss": 0.9705,
+      "step": 2656
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7740311068196506,
+      "learning_rate": 1.0149534735397823e-05,
+      "loss": 1.0505,
+      "step": 2657
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7505842993063178,
+      "learning_rate": 1.0143304557241979e-05,
+      "loss": 1.0789,
+      "step": 2658
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6519348724003564,
+      "learning_rate": 1.0137074323450372e-05,
+      "loss": 1.0113,
+      "step": 2659
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6571976242109744,
+      "learning_rate": 1.0130844036441787e-05,
+      "loss": 1.0079,
+      "step": 2660
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6501820611254552,
+      "learning_rate": 1.0124613698635043e-05,
+      "loss": 0.9918,
+      "step": 2661
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.682550672591527,
+      "learning_rate": 1.0118383312448973e-05,
+      "loss": 1.0212,
+      "step": 2662
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7079384668967243,
+      "learning_rate": 1.0112152880302426e-05,
+      "loss": 1.0576,
+      "step": 2663
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.572512301224634,
+      "learning_rate": 1.0105922404614265e-05,
+      "loss": 1.0015,
+      "step": 2664
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5658865579809078,
+      "learning_rate": 1.0099691887803385e-05,
+      "loss": 1.0928,
+      "step": 2665
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6982488086486153,
+      "learning_rate": 1.0093461332288678e-05,
+      "loss": 1.0509,
+      "step": 2666
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6426160555870879,
+      "learning_rate": 1.0087230740489065e-05,
+      "loss": 0.9517,
+      "step": 2667
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6915528596352477,
+      "learning_rate": 1.0081000114823473e-05,
+      "loss": 0.9604,
+      "step": 2668
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7499616442388781,
+      "learning_rate": 1.007476945771085e-05,
+      "loss": 0.9357,
+      "step": 2669
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5291718328999704,
+      "learning_rate": 1.006853877157015e-05,
+      "loss": 1.0953,
+      "step": 2670
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7301650818807365,
+      "learning_rate": 1.0062308058820337e-05,
+      "loss": 0.9684,
+      "step": 2671
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6402782838475677,
+      "learning_rate": 1.0056077321880393e-05,
+      "loss": 1.0057,
+      "step": 2672
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6355636197632697,
+      "learning_rate": 1.0049846563169297e-05,
+      "loss": 1.0039,
+      "step": 2673
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8340073444717687,
+      "learning_rate": 1.0043615785106051e-05,
+      "loss": 0.9553,
+      "step": 2674
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6174825774759735,
+      "learning_rate": 1.0037384990109658e-05,
+      "loss": 0.9868,
+      "step": 2675
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6952118855783822,
+      "learning_rate": 1.0031154180599123e-05,
+      "loss": 1.0621,
+      "step": 2676
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7290892192726779,
+      "learning_rate": 1.0024923358993458e-05,
+      "loss": 1.0338,
+      "step": 2677
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7767702425598536,
+      "learning_rate": 1.0018692527711695e-05,
+      "loss": 1.0032,
+      "step": 2678
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7091566328643946,
+      "learning_rate": 1.0012461689172846e-05,
+      "loss": 0.9561,
+      "step": 2679
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7184616670853056,
+      "learning_rate": 1.0006230845795937e-05,
+      "loss": 0.986,
+      "step": 2680
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7289537281869556,
+      "learning_rate": 1e-05,
+      "loss": 1.0072,
+      "step": 2681
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6412184984274094,
+      "learning_rate": 9.993769154204063e-06,
+      "loss": 0.9804,
+      "step": 2682
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7293502601998002,
+      "learning_rate": 9.987538310827159e-06,
+      "loss": 0.9651,
+      "step": 2683
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.709274080913764,
+      "learning_rate": 9.981307472288308e-06,
+      "loss": 1.0085,
+      "step": 2684
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6730903057094165,
+      "learning_rate": 9.975076641006542e-06,
+      "loss": 0.9615,
+      "step": 2685
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6299618111022383,
+      "learning_rate": 9.968845819400883e-06,
+      "loss": 1.0225,
+      "step": 2686
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6666231630792607,
+      "learning_rate": 9.962615009890346e-06,
+      "loss": 1.014,
+      "step": 2687
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7278679553083724,
+      "learning_rate": 9.956384214893949e-06,
+      "loss": 1.0038,
+      "step": 2688
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8288600266674654,
+      "learning_rate": 9.950153436830707e-06,
+      "loss": 1.0406,
+      "step": 2689
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6979859663825727,
+      "learning_rate": 9.94392267811961e-06,
+      "loss": 1.0623,
+      "step": 2690
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6697723285676964,
+      "learning_rate": 9.937691941179665e-06,
+      "loss": 1.006,
+      "step": 2691
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6591923603236733,
+      "learning_rate": 9.931461228429856e-06,
+      "loss": 0.9737,
+      "step": 2692
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7224361328571558,
+      "learning_rate": 9.925230542289151e-06,
+      "loss": 1.0383,
+      "step": 2693
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6924440024170467,
+      "learning_rate": 9.91899988517653e-06,
+      "loss": 0.9584,
+      "step": 2694
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5490480057743523,
+      "learning_rate": 9.912769259510938e-06,
+      "loss": 1.0575,
+      "step": 2695
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7591906666693742,
+      "learning_rate": 9.906538667711324e-06,
+      "loss": 1.0351,
+      "step": 2696
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6666989789306641,
+      "learning_rate": 9.90030811219662e-06,
+      "loss": 1.0967,
+      "step": 2697
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6871035305864773,
+      "learning_rate": 9.894077595385736e-06,
+      "loss": 0.9683,
+      "step": 2698
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6983298285686559,
+      "learning_rate": 9.887847119697577e-06,
+      "loss": 1.0187,
+      "step": 2699
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6892078394326652,
+      "learning_rate": 9.881616687551032e-06,
+      "loss": 1.0273,
+      "step": 2700
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7020095370702895,
+      "learning_rate": 9.875386301364958e-06,
+      "loss": 0.9233,
+      "step": 2701
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.758662357065725,
+      "learning_rate": 9.869155963558215e-06,
+      "loss": 0.9377,
+      "step": 2702
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7201647989039508,
+      "learning_rate": 9.862925676549635e-06,
+      "loss": 1.0179,
+      "step": 2703
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6125287303208936,
+      "learning_rate": 9.856695442758023e-06,
+      "loss": 1.0137,
+      "step": 2704
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5675738691986886,
+      "learning_rate": 9.850465264602175e-06,
+      "loss": 1.0588,
+      "step": 2705
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7010020289114334,
+      "learning_rate": 9.844235144500865e-06,
+      "loss": 1.0023,
+      "step": 2706
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6330982503800319,
+      "learning_rate": 9.83800508487283e-06,
+      "loss": 0.9965,
+      "step": 2707
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7835081554137434,
+      "learning_rate": 9.831775088136797e-06,
+      "loss": 1.1068,
+      "step": 2708
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.718562504423076,
+      "learning_rate": 9.82554515671147e-06,
+      "loss": 1.0104,
+      "step": 2709
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6902016522661416,
+      "learning_rate": 9.819315293015519e-06,
+      "loss": 1.0667,
+      "step": 2710
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6818094382113872,
+      "learning_rate": 9.813085499467594e-06,
+      "loss": 1.0062,
+      "step": 2711
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7300034675349791,
+      "learning_rate": 9.806855778486314e-06,
+      "loss": 1.0121,
+      "step": 2712
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.720486379771713,
+      "learning_rate": 9.800626132490268e-06,
+      "loss": 1.0815,
+      "step": 2713
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7179357266483631,
+      "learning_rate": 9.794396563898022e-06,
+      "loss": 1.025,
+      "step": 2714
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6829750523092889,
+      "learning_rate": 9.788167075128113e-06,
+      "loss": 1.0171,
+      "step": 2715
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7198042868861892,
+      "learning_rate": 9.781937668599035e-06,
+      "loss": 1.0129,
+      "step": 2716
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6835781518886104,
+      "learning_rate": 9.775708346729263e-06,
+      "loss": 0.9996,
+      "step": 2717
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6748200506439644,
+      "learning_rate": 9.769479111937238e-06,
+      "loss": 1.0568,
+      "step": 2718
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5276258268940244,
+      "learning_rate": 9.763249966641352e-06,
+      "loss": 1.0212,
+      "step": 2719
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7830410696417072,
+      "learning_rate": 9.757020913259986e-06,
+      "loss": 0.9719,
+      "step": 2720
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6399235080554683,
+      "learning_rate": 9.750791954211464e-06,
+      "loss": 1.0256,
+      "step": 2721
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7301952281686023,
+      "learning_rate": 9.744563091914085e-06,
+      "loss": 0.9994,
+      "step": 2722
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6505490384588629,
+      "learning_rate": 9.738334328786114e-06,
+      "loss": 0.9866,
+      "step": 2723
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7223187979586185,
+      "learning_rate": 9.732105667245759e-06,
+      "loss": 1.0248,
+      "step": 2724
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7310870207580493,
+      "learning_rate": 9.725877109711212e-06,
+      "loss": 1.0119,
+      "step": 2725
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6617432023994871,
+      "learning_rate": 9.719648658600611e-06,
+      "loss": 1.0676,
+      "step": 2726
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7332088993448104,
+      "learning_rate": 9.71342031633205e-06,
+      "loss": 1.0435,
+      "step": 2727
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6154219995751529,
+      "learning_rate": 9.70719208532359e-06,
+      "loss": 0.939,
+      "step": 2728
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7553548926799055,
+      "learning_rate": 9.700963967993246e-06,
+      "loss": 1.051,
+      "step": 2729
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6664540461791529,
+      "learning_rate": 9.694735966758982e-06,
+      "loss": 1.0498,
+      "step": 2730
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3296266409447374,
+      "learning_rate": 9.688508084038729e-06,
+      "loss": 1.0638,
+      "step": 2731
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6984830278237237,
+      "learning_rate": 9.682280322250365e-06,
+      "loss": 1.0297,
+      "step": 2732
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5907418930880511,
+      "learning_rate": 9.676052683811715e-06,
+      "loss": 1.1157,
+      "step": 2733
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6386297827010354,
+      "learning_rate": 9.669825171140568e-06,
+      "loss": 1.0195,
+      "step": 2734
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7333300283294666,
+      "learning_rate": 9.66359778665466e-06,
+      "loss": 1.0481,
+      "step": 2735
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6115923255274532,
+      "learning_rate": 9.657370532771672e-06,
+      "loss": 1.049,
+      "step": 2736
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6744784401429308,
+      "learning_rate": 9.651143411909241e-06,
+      "loss": 0.9681,
+      "step": 2737
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5887618438009865,
+      "learning_rate": 9.64491642648495e-06,
+      "loss": 1.0573,
+      "step": 2738
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.530126914678891,
+      "learning_rate": 9.638689578916326e-06,
+      "loss": 1.05,
+      "step": 2739
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6799066865750159,
+      "learning_rate": 9.632462871620847e-06,
+      "loss": 0.9626,
+      "step": 2740
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.725615584243556,
+      "learning_rate": 9.62623630701594e-06,
+      "loss": 1.0669,
+      "step": 2741
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6809728211796436,
+      "learning_rate": 9.620009887518963e-06,
+      "loss": 0.9738,
+      "step": 2742
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6714615260229706,
+      "learning_rate": 9.613783615547233e-06,
+      "loss": 0.9784,
+      "step": 2743
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6061308467331131,
+      "learning_rate": 9.607557493518006e-06,
+      "loss": 1.0159,
+      "step": 2744
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.745949885719098,
+      "learning_rate": 9.601331523848464e-06,
+      "loss": 1.0294,
+      "step": 2745
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6922378848090723,
+      "learning_rate": 9.595105708955758e-06,
+      "loss": 1.0087,
+      "step": 2746
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6751341659191323,
+      "learning_rate": 9.588880051256951e-06,
+      "loss": 1.0386,
+      "step": 2747
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7607550664837116,
+      "learning_rate": 9.582654553169064e-06,
+      "loss": 1.0392,
+      "step": 2748
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5909017368327416,
+      "learning_rate": 9.576429217109054e-06,
+      "loss": 1.096,
+      "step": 2749
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7640945229927498,
+      "learning_rate": 9.5702040454938e-06,
+      "loss": 0.9865,
+      "step": 2750
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7814634194881342,
+      "learning_rate": 9.563979040740138e-06,
+      "loss": 1.032,
+      "step": 2751
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6339931956025219,
+      "learning_rate": 9.557754205264826e-06,
+      "loss": 1.0008,
+      "step": 2752
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7926444943839768,
+      "learning_rate": 9.551529541484554e-06,
+      "loss": 1.0627,
+      "step": 2753
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7357293971337432,
+      "learning_rate": 9.545305051815957e-06,
+      "loss": 1.0232,
+      "step": 2754
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6735437919562587,
+      "learning_rate": 9.539080738675597e-06,
+      "loss": 1.0247,
+      "step": 2755
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.34784931730606977,
+      "learning_rate": 9.53285660447996e-06,
+      "loss": 1.093,
+      "step": 2756
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6599934561525368,
+      "learning_rate": 9.526632651645476e-06,
+      "loss": 0.9841,
+      "step": 2757
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5297846761221732,
+      "learning_rate": 9.520408882588497e-06,
+      "loss": 1.0789,
+      "step": 2758
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.752749843001113,
+      "learning_rate": 9.514185299725299e-06,
+      "loss": 0.9338,
+      "step": 2759
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6793717289650855,
+      "learning_rate": 9.507961905472093e-06,
+      "loss": 1.0374,
+      "step": 2760
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7395272371007782,
+      "learning_rate": 9.501738702245023e-06,
+      "loss": 1.0162,
+      "step": 2761
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7552658052478125,
+      "learning_rate": 9.495515692460138e-06,
+      "loss": 1.0129,
+      "step": 2762
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7473213603651846,
+      "learning_rate": 9.489292878533436e-06,
+      "loss": 0.99,
+      "step": 2763
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5038076774244258,
+      "learning_rate": 9.483070262880823e-06,
+      "loss": 1.0886,
+      "step": 2764
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6607588278990708,
+      "learning_rate": 9.476847847918126e-06,
+      "loss": 1.0082,
+      "step": 2765
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7127765745712328,
+      "learning_rate": 9.47062563606111e-06,
+      "loss": 1.0829,
+      "step": 2766
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.720829850811203,
+      "learning_rate": 9.464403629725454e-06,
+      "loss": 1.0438,
+      "step": 2767
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5273296097883243,
+      "learning_rate": 9.458181831326744e-06,
+      "loss": 1.1035,
+      "step": 2768
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6887744320363594,
+      "learning_rate": 9.451960243280506e-06,
+      "loss": 1.0403,
+      "step": 2769
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5179767412063826,
+      "learning_rate": 9.44573886800217e-06,
+      "loss": 1.0201,
+      "step": 2770
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6771806914840863,
+      "learning_rate": 9.43951770790709e-06,
+      "loss": 0.9987,
+      "step": 2771
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7529601129562475,
+      "learning_rate": 9.433296765410534e-06,
+      "loss": 1.0061,
+      "step": 2772
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7244019373798609,
+      "learning_rate": 9.427076042927683e-06,
+      "loss": 1.0257,
+      "step": 2773
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7420378798244263,
+      "learning_rate": 9.420855542873638e-06,
+      "loss": 1.0272,
+      "step": 2774
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6874421786400524,
+      "learning_rate": 9.414635267663416e-06,
+      "loss": 0.9489,
+      "step": 2775
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.710124650341572,
+      "learning_rate": 9.408415219711934e-06,
+      "loss": 1.0491,
+      "step": 2776
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6612210078085458,
+      "learning_rate": 9.402195401434036e-06,
+      "loss": 1.0635,
+      "step": 2777
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6865128888223424,
+      "learning_rate": 9.395975815244468e-06,
+      "loss": 1.0374,
+      "step": 2778
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7975457199358593,
+      "learning_rate": 9.389756463557878e-06,
+      "loss": 1.0223,
+      "step": 2779
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6525386917732303,
+      "learning_rate": 9.383537348788844e-06,
+      "loss": 0.9828,
+      "step": 2780
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7487147405122959,
+      "learning_rate": 9.377318473351838e-06,
+      "loss": 0.9883,
+      "step": 2781
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6972132718013511,
+      "learning_rate": 9.371099839661238e-06,
+      "loss": 0.9868,
+      "step": 2782
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6423232301474794,
+      "learning_rate": 9.364881450131335e-06,
+      "loss": 1.0161,
+      "step": 2783
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7125959304400583,
+      "learning_rate": 9.358663307176323e-06,
+      "loss": 0.9182,
+      "step": 2784
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7486468453201213,
+      "learning_rate": 9.352445413210294e-06,
+      "loss": 0.9982,
+      "step": 2785
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7900769324701314,
+      "learning_rate": 9.346227770647251e-06,
+      "loss": 0.9562,
+      "step": 2786
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7860517401893288,
+      "learning_rate": 9.3400103819011e-06,
+      "loss": 1.0336,
+      "step": 2787
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6168693314613356,
+      "learning_rate": 9.33379324938564e-06,
+      "loss": 1.0002,
+      "step": 2788
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.678347494102185,
+      "learning_rate": 9.327576375514582e-06,
+      "loss": 0.9805,
+      "step": 2789
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.765232591635616,
+      "learning_rate": 9.321359762701527e-06,
+      "loss": 1.0734,
+      "step": 2790
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6973430820464502,
+      "learning_rate": 9.315143413359975e-06,
+      "loss": 1.0023,
+      "step": 2791
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5835164729923606,
+      "learning_rate": 9.308927329903333e-06,
+      "loss": 1.0847,
+      "step": 2792
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.673273676806348,
+      "learning_rate": 9.302711514744897e-06,
+      "loss": 1.0078,
+      "step": 2793
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5470938492672813,
+      "learning_rate": 9.296495970297855e-06,
+      "loss": 1.0682,
+      "step": 2794
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7307981162507449,
+      "learning_rate": 9.290280698975307e-06,
+      "loss": 0.9316,
+      "step": 2795
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.668460395039042,
+      "learning_rate": 9.284065703190221e-06,
+      "loss": 1.059,
+      "step": 2796
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6739696479515326,
+      "learning_rate": 9.27785098535548e-06,
+      "loss": 0.9973,
+      "step": 2797
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5440410184572314,
+      "learning_rate": 9.271636547883856e-06,
+      "loss": 1.0507,
+      "step": 2798
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6658867831555576,
+      "learning_rate": 9.265422393187998e-06,
+      "loss": 1.0497,
+      "step": 2799
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6884828707029095,
+      "learning_rate": 9.259208523680457e-06,
+      "loss": 1.0477,
+      "step": 2800
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7239197505278342,
+      "learning_rate": 9.252994941773679e-06,
+      "loss": 1.0979,
+      "step": 2801
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6720651200647688,
+      "learning_rate": 9.24678164987998e-06,
+      "loss": 1.0292,
+      "step": 2802
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6940222703528672,
+      "learning_rate": 9.24056865041158e-06,
+      "loss": 1.002,
+      "step": 2803
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.516707363989533,
+      "learning_rate": 9.234355945780581e-06,
+      "loss": 1.1273,
+      "step": 2804
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6953616588173449,
+      "learning_rate": 9.228143538398963e-06,
+      "loss": 1.0451,
+      "step": 2805
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6415397802262867,
+      "learning_rate": 9.221931430678598e-06,
+      "loss": 1.018,
+      "step": 2806
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6855083869280946,
+      "learning_rate": 9.215719625031245e-06,
+      "loss": 0.9622,
+      "step": 2807
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6481806098434665,
+      "learning_rate": 9.209508123868534e-06,
+      "loss": 0.9312,
+      "step": 2808
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.758134302810316,
+      "learning_rate": 9.203296929601986e-06,
+      "loss": 0.921,
+      "step": 2809
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6504787630011741,
+      "learning_rate": 9.197086044643004e-06,
+      "loss": 1.0244,
+      "step": 2810
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7304022468489966,
+      "learning_rate": 9.190875471402865e-06,
+      "loss": 1.0081,
+      "step": 2811
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6783569427634901,
+      "learning_rate": 9.184665212292723e-06,
+      "loss": 1.0378,
+      "step": 2812
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.47849047981281434,
+      "learning_rate": 9.178455269723623e-06,
+      "loss": 1.0226,
+      "step": 2813
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.714868438111141,
+      "learning_rate": 9.172245646106471e-06,
+      "loss": 0.9489,
+      "step": 2814
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6744295036220724,
+      "learning_rate": 9.166036343852061e-06,
+      "loss": 0.9769,
+      "step": 2815
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6758276465659208,
+      "learning_rate": 9.159827365371055e-06,
+      "loss": 1.0128,
+      "step": 2816
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7945734144833224,
+      "learning_rate": 9.153618713073995e-06,
+      "loss": 1.0048,
+      "step": 2817
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8075879070779053,
+      "learning_rate": 9.14741038937129e-06,
+      "loss": 0.9722,
+      "step": 2818
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7200650271353413,
+      "learning_rate": 9.141202396673232e-06,
+      "loss": 1.0649,
+      "step": 2819
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6399502188520432,
+      "learning_rate": 9.13499473738997e-06,
+      "loss": 1.0303,
+      "step": 2820
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6684119043686292,
+      "learning_rate": 9.128787413931536e-06,
+      "loss": 0.9479,
+      "step": 2821
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6948384099945206,
+      "learning_rate": 9.122580428707822e-06,
+      "loss": 1.1047,
+      "step": 2822
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.48747241021416626,
+      "learning_rate": 9.116373784128597e-06,
+      "loss": 0.9761,
+      "step": 2823
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7767272499698675,
+      "learning_rate": 9.110167482603494e-06,
+      "loss": 1.059,
+      "step": 2824
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7343790735062616,
+      "learning_rate": 9.10396152654201e-06,
+      "loss": 1.0163,
+      "step": 2825
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6500662096023301,
+      "learning_rate": 9.097755918353513e-06,
+      "loss": 1.0217,
+      "step": 2826
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5360630662619251,
+      "learning_rate": 9.091550660447236e-06,
+      "loss": 1.0477,
+      "step": 2827
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8004318057350696,
+      "learning_rate": 9.08534575523227e-06,
+      "loss": 0.9805,
+      "step": 2828
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6691229736031918,
+      "learning_rate": 9.079141205117573e-06,
+      "loss": 1.0166,
+      "step": 2829
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6651828622392437,
+      "learning_rate": 9.072937012511968e-06,
+      "loss": 1.0143,
+      "step": 2830
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7083899590111805,
+      "learning_rate": 9.066733179824134e-06,
+      "loss": 1.0309,
+      "step": 2831
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7966347211124915,
+      "learning_rate": 9.060529709462613e-06,
+      "loss": 1.045,
+      "step": 2832
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7419510372334583,
+      "learning_rate": 9.054326603835807e-06,
+      "loss": 0.9991,
+      "step": 2833
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7183001472511829,
+      "learning_rate": 9.048123865351971e-06,
+      "loss": 1.043,
+      "step": 2834
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7170654446383782,
+      "learning_rate": 9.041921496419225e-06,
+      "loss": 1.0655,
+      "step": 2835
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4706030927415568,
+      "learning_rate": 9.035719499445545e-06,
+      "loss": 0.9994,
+      "step": 2836
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5867684751218317,
+      "learning_rate": 9.029517876838755e-06,
+      "loss": 1.0242,
+      "step": 2837
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6162338323728417,
+      "learning_rate": 9.023316631006536e-06,
+      "loss": 1.0045,
+      "step": 2838
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7698247352615882,
+      "learning_rate": 9.017115764356436e-06,
+      "loss": 1.047,
+      "step": 2839
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7129534292268087,
+      "learning_rate": 9.010915279295833e-06,
+      "loss": 1.0095,
+      "step": 2840
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6385000262540204,
+      "learning_rate": 9.004715178231975e-06,
+      "loss": 0.9973,
+      "step": 2841
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6121460439612054,
+      "learning_rate": 8.998515463571953e-06,
+      "loss": 0.9653,
+      "step": 2842
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7184890790319273,
+      "learning_rate": 8.992316137722711e-06,
+      "loss": 0.9494,
+      "step": 2843
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6725438631102264,
+      "learning_rate": 8.986117203091042e-06,
+      "loss": 1.0488,
+      "step": 2844
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6943796460584548,
+      "learning_rate": 8.97991866208358e-06,
+      "loss": 0.9636,
+      "step": 2845
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.616079274973886,
+      "learning_rate": 8.973720517106814e-06,
+      "loss": 0.8985,
+      "step": 2846
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7058299175712088,
+      "learning_rate": 8.967522770567086e-06,
+      "loss": 1.0563,
+      "step": 2847
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6714831563941568,
+      "learning_rate": 8.961325424870561e-06,
+      "loss": 1.0042,
+      "step": 2848
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6545226481194695,
+      "learning_rate": 8.955128482423271e-06,
+      "loss": 0.9755,
+      "step": 2849
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7380009996640953,
+      "learning_rate": 8.948931945631082e-06,
+      "loss": 1.1173,
+      "step": 2850
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6776591126728173,
+      "learning_rate": 8.9427358168997e-06,
+      "loss": 0.9932,
+      "step": 2851
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6691741823326085,
+      "learning_rate": 8.936540098634675e-06,
+      "loss": 0.9943,
+      "step": 2852
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6624790665954364,
+      "learning_rate": 8.930344793241404e-06,
+      "loss": 1.0604,
+      "step": 2853
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.647043868140034,
+      "learning_rate": 8.924149903125108e-06,
+      "loss": 1.0215,
+      "step": 2854
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7264868020707304,
+      "learning_rate": 8.917955430690865e-06,
+      "loss": 1.0126,
+      "step": 2855
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.73195395551241,
+      "learning_rate": 8.91176137834358e-06,
+      "loss": 1.0506,
+      "step": 2856
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6682483093749243,
+      "learning_rate": 8.905567748487997e-06,
+      "loss": 1.0139,
+      "step": 2857
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.683911366062445,
+      "learning_rate": 8.899374543528695e-06,
+      "loss": 1.0112,
+      "step": 2858
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7020244623134808,
+      "learning_rate": 8.893181765870094e-06,
+      "loss": 1.0716,
+      "step": 2859
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6549968096335095,
+      "learning_rate": 8.886989417916435e-06,
+      "loss": 1.0459,
+      "step": 2860
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6968096073635319,
+      "learning_rate": 8.88079750207181e-06,
+      "loss": 1.0413,
+      "step": 2861
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7494443307545299,
+      "learning_rate": 8.87460602074013e-06,
+      "loss": 1.0403,
+      "step": 2862
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6481973711717491,
+      "learning_rate": 8.86841497632514e-06,
+      "loss": 0.957,
+      "step": 2863
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5321674747020173,
+      "learning_rate": 8.862224371230418e-06,
+      "loss": 1.0318,
+      "step": 2864
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6042210885451216,
+      "learning_rate": 8.85603420785937e-06,
+      "loss": 1.0286,
+      "step": 2865
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6936907068576028,
+      "learning_rate": 8.84984448861523e-06,
+      "loss": 1.0223,
+      "step": 2866
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6546456455020174,
+      "learning_rate": 8.84365521590106e-06,
+      "loss": 1.0482,
+      "step": 2867
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7302163786105959,
+      "learning_rate": 8.837466392119752e-06,
+      "loss": 0.9,
+      "step": 2868
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6628405120101803,
+      "learning_rate": 8.831278019674017e-06,
+      "loss": 1.0722,
+      "step": 2869
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8302224389986428,
+      "learning_rate": 8.825090100966396e-06,
+      "loss": 0.9991,
+      "step": 2870
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.711866178672656,
+      "learning_rate": 8.818902638399247e-06,
+      "loss": 1.0276,
+      "step": 2871
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6645793175281257,
+      "learning_rate": 8.81271563437476e-06,
+      "loss": 0.9573,
+      "step": 2872
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6902907988648821,
+      "learning_rate": 8.806529091294948e-06,
+      "loss": 0.9575,
+      "step": 2873
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7265162388588772,
+      "learning_rate": 8.800343011561633e-06,
+      "loss": 0.9771,
+      "step": 2874
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8581595957257052,
+      "learning_rate": 8.794157397576464e-06,
+      "loss": 1.0962,
+      "step": 2875
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5360347871260089,
+      "learning_rate": 8.787972251740916e-06,
+      "loss": 1.0181,
+      "step": 2876
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7418345844125956,
+      "learning_rate": 8.781787576456269e-06,
+      "loss": 1.0546,
+      "step": 2877
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5757921656118279,
+      "learning_rate": 8.775603374123627e-06,
+      "loss": 1.0673,
+      "step": 2878
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6993978812556986,
+      "learning_rate": 8.769419647143917e-06,
+      "loss": 1.0775,
+      "step": 2879
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8052568556435675,
+      "learning_rate": 8.763236397917865e-06,
+      "loss": 1.0397,
+      "step": 2880
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6619100953990146,
+      "learning_rate": 8.757053628846028e-06,
+      "loss": 0.997,
+      "step": 2881
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.696409718077993,
+      "learning_rate": 8.75087134232877e-06,
+      "loss": 1.0591,
+      "step": 2882
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6589040254246431,
+      "learning_rate": 8.744689540766265e-06,
+      "loss": 0.9664,
+      "step": 2883
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6875234357191448,
+      "learning_rate": 8.738508226558499e-06,
+      "loss": 0.9713,
+      "step": 2884
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.500979182012775,
+      "learning_rate": 8.73232740210528e-06,
+      "loss": 1.0947,
+      "step": 2885
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6749859648069633,
+      "learning_rate": 8.726147069806206e-06,
+      "loss": 0.9521,
+      "step": 2886
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.759685382474957,
+      "learning_rate": 8.719967232060698e-06,
+      "loss": 1.0726,
+      "step": 2887
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6587788871961258,
+      "learning_rate": 8.713787891267988e-06,
+      "loss": 1.0924,
+      "step": 2888
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7835572338452272,
+      "learning_rate": 8.707609049827102e-06,
+      "loss": 0.9867,
+      "step": 2889
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6565703657045309,
+      "learning_rate": 8.70143071013688e-06,
+      "loss": 1.0754,
+      "step": 2890
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8015185069879246,
+      "learning_rate": 8.695252874595972e-06,
+      "loss": 1.0204,
+      "step": 2891
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.684838686084031,
+      "learning_rate": 8.689075545602816e-06,
+      "loss": 1.0119,
+      "step": 2892
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6826156437186252,
+      "learning_rate": 8.68289872555567e-06,
+      "loss": 1.0475,
+      "step": 2893
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6883164188408091,
+      "learning_rate": 8.676722416852594e-06,
+      "loss": 0.9814,
+      "step": 2894
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6738221507369813,
+      "learning_rate": 8.670546621891434e-06,
+      "loss": 0.9622,
+      "step": 2895
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6686092699392776,
+      "learning_rate": 8.66437134306985e-06,
+      "loss": 0.9922,
+      "step": 2896
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6463573236716781,
+      "learning_rate": 8.658196582785297e-06,
+      "loss": 1.0196,
+      "step": 2897
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.685951869267113,
+      "learning_rate": 8.652022343435027e-06,
+      "loss": 0.9762,
+      "step": 2898
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6793946174287565,
+      "learning_rate": 8.645848627416102e-06,
+      "loss": 0.9868,
+      "step": 2899
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7214201966807837,
+      "learning_rate": 8.63967543712536e-06,
+      "loss": 0.9941,
+      "step": 2900
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6726479190213511,
+      "learning_rate": 8.633502774959453e-06,
+      "loss": 1.0977,
+      "step": 2901
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7004395003558742,
+      "learning_rate": 8.627330643314818e-06,
+      "loss": 1.0026,
+      "step": 2902
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7633178721482706,
+      "learning_rate": 8.62115904458769e-06,
+      "loss": 0.9695,
+      "step": 2903
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7765000235562967,
+      "learning_rate": 8.614987981174093e-06,
+      "loss": 0.942,
+      "step": 2904
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7365927755425539,
+      "learning_rate": 8.608817455469854e-06,
+      "loss": 0.9891,
+      "step": 2905
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6960362118519852,
+      "learning_rate": 8.602647469870573e-06,
+      "loss": 0.9943,
+      "step": 2906
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7087635796399874,
+      "learning_rate": 8.596478026771658e-06,
+      "loss": 0.9966,
+      "step": 2907
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6303671300619411,
+      "learning_rate": 8.590309128568303e-06,
+      "loss": 1.0046,
+      "step": 2908
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5529469066740059,
+      "learning_rate": 8.584140777655476e-06,
+      "loss": 1.0584,
+      "step": 2909
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7414539497052505,
+      "learning_rate": 8.57797297642795e-06,
+      "loss": 1.0016,
+      "step": 2910
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.688046978084019,
+      "learning_rate": 8.571805727280278e-06,
+      "loss": 1.0448,
+      "step": 2911
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7320913658115389,
+      "learning_rate": 8.565639032606794e-06,
+      "loss": 1.017,
+      "step": 2912
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.665029926696437,
+      "learning_rate": 8.559472894801623e-06,
+      "loss": 1.0491,
+      "step": 2913
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5540502433063718,
+      "learning_rate": 8.553307316258678e-06,
+      "loss": 1.0266,
+      "step": 2914
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6646072401310875,
+      "learning_rate": 8.547142299371642e-06,
+      "loss": 1.0228,
+      "step": 2915
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7307679454268082,
+      "learning_rate": 8.540977846533986e-06,
+      "loss": 1.0682,
+      "step": 2916
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.63913736908407,
+      "learning_rate": 8.534813960138968e-06,
+      "loss": 0.9691,
+      "step": 2917
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7369753108344032,
+      "learning_rate": 8.528650642579618e-06,
+      "loss": 1.0448,
+      "step": 2918
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.693402481253445,
+      "learning_rate": 8.52248789624875e-06,
+      "loss": 1.0276,
+      "step": 2919
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6554803710314174,
+      "learning_rate": 8.516325723538949e-06,
+      "loss": 0.9884,
+      "step": 2920
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6618480519759248,
+      "learning_rate": 8.510164126842591e-06,
+      "loss": 1.0239,
+      "step": 2921
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8017928459050897,
+      "learning_rate": 8.504003108551814e-06,
+      "loss": 0.9377,
+      "step": 2922
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6780479408974802,
+      "learning_rate": 8.497842671058539e-06,
+      "loss": 1.0159,
+      "step": 2923
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6118599702777,
+      "learning_rate": 8.491682816754456e-06,
+      "loss": 0.9401,
+      "step": 2924
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6626948952585933,
+      "learning_rate": 8.485523548031044e-06,
+      "loss": 0.9904,
+      "step": 2925
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8566876507134639,
+      "learning_rate": 8.479364867279529e-06,
+      "loss": 0.9725,
+      "step": 2926
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7623451043476506,
+      "learning_rate": 8.47320677689093e-06,
+      "loss": 1.0518,
+      "step": 2927
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8166765114773078,
+      "learning_rate": 8.467049279256034e-06,
+      "loss": 0.9492,
+      "step": 2928
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6950290884436744,
+      "learning_rate": 8.460892376765387e-06,
+      "loss": 1.0296,
+      "step": 2929
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7050441094886446,
+      "learning_rate": 8.45473607180931e-06,
+      "loss": 0.987,
+      "step": 2930
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5472681816619468,
+      "learning_rate": 8.448580366777898e-06,
+      "loss": 1.0751,
+      "step": 2931
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6908531581146293,
+      "learning_rate": 8.442425264061e-06,
+      "loss": 1.0175,
+      "step": 2932
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7054642501282729,
+      "learning_rate": 8.436270766048245e-06,
+      "loss": 1.0275,
+      "step": 2933
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6966895702038708,
+      "learning_rate": 8.430116875129023e-06,
+      "loss": 1.0667,
+      "step": 2934
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7257921212400608,
+      "learning_rate": 8.42396359369248e-06,
+      "loss": 0.9853,
+      "step": 2935
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7229443229059331,
+      "learning_rate": 8.417810924127533e-06,
+      "loss": 0.9553,
+      "step": 2936
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5261406682650179,
+      "learning_rate": 8.411658868822866e-06,
+      "loss": 1.0322,
+      "step": 2937
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3596751373094446,
+      "learning_rate": 8.40550743016691e-06,
+      "loss": 1.1704,
+      "step": 2938
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7442603647080629,
+      "learning_rate": 8.39935661054787e-06,
+      "loss": 1.0467,
+      "step": 2939
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6682002039234305,
+      "learning_rate": 8.393206412353709e-06,
+      "loss": 0.9716,
+      "step": 2940
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7016662296520292,
+      "learning_rate": 8.38705683797214e-06,
+      "loss": 1.0197,
+      "step": 2941
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7515463432349359,
+      "learning_rate": 8.38090788979064e-06,
+      "loss": 1.0161,
+      "step": 2942
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5465235962959043,
+      "learning_rate": 8.374759570196448e-06,
+      "loss": 1.0281,
+      "step": 2943
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.72578356640662,
+      "learning_rate": 8.368611881576547e-06,
+      "loss": 1.0395,
+      "step": 2944
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.847108483947531,
+      "learning_rate": 8.362464826317687e-06,
+      "loss": 0.9732,
+      "step": 2945
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.72116205179911,
+      "learning_rate": 8.35631840680636e-06,
+      "loss": 1.049,
+      "step": 2946
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6419043262065883,
+      "learning_rate": 8.35017262542882e-06,
+      "loss": 1.0107,
+      "step": 2947
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6571958164067945,
+      "learning_rate": 8.344027484571075e-06,
+      "loss": 1.0295,
+      "step": 2948
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6808685903824765,
+      "learning_rate": 8.337882986618877e-06,
+      "loss": 1.0248,
+      "step": 2949
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6464708304795093,
+      "learning_rate": 8.331739133957729e-06,
+      "loss": 1.0485,
+      "step": 2950
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6440992085900532,
+      "learning_rate": 8.325595928972894e-06,
+      "loss": 1.0273,
+      "step": 2951
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6763774406363028,
+      "learning_rate": 8.319453374049367e-06,
+      "loss": 1.0234,
+      "step": 2952
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6945286845719398,
+      "learning_rate": 8.313311471571903e-06,
+      "loss": 1.0623,
+      "step": 2953
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6124105956106816,
+      "learning_rate": 8.307170223925003e-06,
+      "loss": 1.0474,
+      "step": 2954
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.675943811176531,
+      "learning_rate": 8.301029633492907e-06,
+      "loss": 1.0107,
+      "step": 2955
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6603525392748275,
+      "learning_rate": 8.294889702659602e-06,
+      "loss": 1.0168,
+      "step": 2956
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7673187475795384,
+      "learning_rate": 8.288750433808828e-06,
+      "loss": 1.1067,
+      "step": 2957
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7709749200704036,
+      "learning_rate": 8.282611829324049e-06,
+      "loss": 1.0484,
+      "step": 2958
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7062226618813797,
+      "learning_rate": 8.276473891588492e-06,
+      "loss": 0.9966,
+      "step": 2959
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.73629798614413,
+      "learning_rate": 8.270336622985116e-06,
+      "loss": 0.9478,
+      "step": 2960
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6970107217085887,
+      "learning_rate": 8.264200025896616e-06,
+      "loss": 1.079,
+      "step": 2961
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7337995406205559,
+      "learning_rate": 8.258064102705428e-06,
+      "loss": 1.0105,
+      "step": 2962
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6930687365465241,
+      "learning_rate": 8.251928855793736e-06,
+      "loss": 0.9658,
+      "step": 2963
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5664224249252707,
+      "learning_rate": 8.245794287543447e-06,
+      "loss": 1.0334,
+      "step": 2964
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.685834951084631,
+      "learning_rate": 8.239660400336213e-06,
+      "loss": 0.9521,
+      "step": 2965
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6664768929865765,
+      "learning_rate": 8.233527196553428e-06,
+      "loss": 1.0282,
+      "step": 2966
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7710859050042744,
+      "learning_rate": 8.227394678576204e-06,
+      "loss": 1.0222,
+      "step": 2967
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7493668236943558,
+      "learning_rate": 8.221262848785395e-06,
+      "loss": 0.9376,
+      "step": 2968
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6929080842889047,
+      "learning_rate": 8.215131709561597e-06,
+      "loss": 0.9971,
+      "step": 2969
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.654987611950482,
+      "learning_rate": 8.20900126328512e-06,
+      "loss": 1.0236,
+      "step": 2970
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5803571706428327,
+      "learning_rate": 8.202871512336023e-06,
+      "loss": 1.0615,
+      "step": 2971
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7195941310149719,
+      "learning_rate": 8.196742459094079e-06,
+      "loss": 0.9477,
+      "step": 2972
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6876323687593959,
+      "learning_rate": 8.190614105938796e-06,
+      "loss": 0.928,
+      "step": 2973
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7002782781774465,
+      "learning_rate": 8.184486455249424e-06,
+      "loss": 1.0463,
+      "step": 2974
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.593443531470797,
+      "learning_rate": 8.178359509404916e-06,
+      "loss": 1.0837,
+      "step": 2975
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.739478124067507,
+      "learning_rate": 8.172233270783966e-06,
+      "loss": 1.0851,
+      "step": 2976
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5169646580524231,
+      "learning_rate": 8.166107741764997e-06,
+      "loss": 1.074,
+      "step": 2977
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6615058888369425,
+      "learning_rate": 8.15998292472614e-06,
+      "loss": 1.1115,
+      "step": 2978
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.747389217484143,
+      "learning_rate": 8.153858822045267e-06,
+      "loss": 0.992,
+      "step": 2979
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.72397182714619,
+      "learning_rate": 8.147735436099967e-06,
+      "loss": 1.0938,
+      "step": 2980
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6961278650895178,
+      "learning_rate": 8.141612769267543e-06,
+      "loss": 1.0112,
+      "step": 2981
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7041602872993479,
+      "learning_rate": 8.135490823925027e-06,
+      "loss": 0.9784,
+      "step": 2982
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6625851445867049,
+      "learning_rate": 8.129369602449176e-06,
+      "loss": 1.0163,
+      "step": 2983
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6786879840530267,
+      "learning_rate": 8.123249107216446e-06,
+      "loss": 1.0234,
+      "step": 2984
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7155482070050896,
+      "learning_rate": 8.117129340603032e-06,
+      "loss": 1.0214,
+      "step": 2985
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7584886609930438,
+      "learning_rate": 8.111010304984841e-06,
+      "loss": 1.0906,
+      "step": 2986
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7617893572362664,
+      "learning_rate": 8.104892002737488e-06,
+      "loss": 0.9901,
+      "step": 2987
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7024136945612479,
+      "learning_rate": 8.098774436236308e-06,
+      "loss": 1.0469,
+      "step": 2988
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7347659769532766,
+      "learning_rate": 8.092657607856356e-06,
+      "loss": 0.9874,
+      "step": 2989
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6655258682372467,
+      "learning_rate": 8.086541519972388e-06,
+      "loss": 1.0026,
+      "step": 2990
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7089115469291215,
+      "learning_rate": 8.080426174958886e-06,
+      "loss": 0.9675,
+      "step": 2991
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.711037187689343,
+      "learning_rate": 8.074311575190039e-06,
+      "loss": 0.9565,
+      "step": 2992
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7722692436633872,
+      "learning_rate": 8.068197723039738e-06,
+      "loss": 1.0357,
+      "step": 2993
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6764917523749282,
+      "learning_rate": 8.062084620881598e-06,
+      "loss": 1.0809,
+      "step": 2994
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6972609429679513,
+      "learning_rate": 8.055972271088933e-06,
+      "loss": 0.9733,
+      "step": 2995
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6501447036224384,
+      "learning_rate": 8.049860676034762e-06,
+      "loss": 1.0356,
+      "step": 2996
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7886491186309991,
+      "learning_rate": 8.043749838091828e-06,
+      "loss": 1.007,
+      "step": 2997
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6824672271134491,
+      "learning_rate": 8.037639759632558e-06,
+      "loss": 1.0677,
+      "step": 2998
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6190678112157165,
+      "learning_rate": 8.031530443029099e-06,
+      "loss": 0.9924,
+      "step": 2999
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7288008674921438,
+      "learning_rate": 8.025421890653303e-06,
+      "loss": 1.0616,
+      "step": 3000
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6857469649456737,
+      "learning_rate": 8.019314104876712e-06,
+      "loss": 0.9887,
+      "step": 3001
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7206391178072069,
+      "learning_rate": 8.013207088070582e-06,
+      "loss": 1.0305,
+      "step": 3002
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7380715529053661,
+      "learning_rate": 8.007100842605872e-06,
+      "loss": 1.0755,
+      "step": 3003
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.488151978488106,
+      "learning_rate": 8.000995370853227e-06,
+      "loss": 1.0161,
+      "step": 3004
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.35698566042154145,
+      "learning_rate": 7.994890675183008e-06,
+      "loss": 0.9406,
+      "step": 3005
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.35698566042154145,
+      "learning_rate": 7.994890675183008e-06,
+      "loss": 1.0726,
+      "step": 3006
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.35698566042154145,
+      "learning_rate": 7.994890675183008e-06,
+      "loss": 0.9434,
+      "step": 3007
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7358315580369601,
+      "learning_rate": 7.98878675796527e-06,
+      "loss": 0.9844,
+      "step": 3008
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7548722917573736,
+      "learning_rate": 7.98268362156976e-06,
+      "loss": 0.9808,
+      "step": 3009
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.65923475584887,
+      "learning_rate": 7.976581268365924e-06,
+      "loss": 1.057,
+      "step": 3010
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5369632288140463,
+      "learning_rate": 7.97047970072291e-06,
+      "loss": 1.0291,
+      "step": 3011
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.646242373432174,
+      "learning_rate": 7.964378921009552e-06,
+      "loss": 1.023,
+      "step": 3012
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7029544098402843,
+      "learning_rate": 7.958278931594385e-06,
+      "loss": 1.0079,
+      "step": 3013
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7475188961349202,
+      "learning_rate": 7.952179734845642e-06,
+      "loss": 1.0715,
+      "step": 3014
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.550497039334602,
+      "learning_rate": 7.946081333131227e-06,
+      "loss": 1.0883,
+      "step": 3015
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6822755373686813,
+      "learning_rate": 7.93998372881876e-06,
+      "loss": 1.0449,
+      "step": 3016
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4994780999881114,
+      "learning_rate": 7.93388692427554e-06,
+      "loss": 1.0959,
+      "step": 3017
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7225615393243073,
+      "learning_rate": 7.92779092186855e-06,
+      "loss": 0.9637,
+      "step": 3018
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6952644218633557,
+      "learning_rate": 7.921695723964473e-06,
+      "loss": 0.955,
+      "step": 3019
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.651674482793734,
+      "learning_rate": 7.915601332929678e-06,
+      "loss": 1.0201,
+      "step": 3020
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.691237928519129,
+      "learning_rate": 7.90950775113021e-06,
+      "loss": 1.0293,
+      "step": 3021
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5172029080205045,
+      "learning_rate": 7.903414980931813e-06,
+      "loss": 1.0119,
+      "step": 3022
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6590248515130125,
+      "learning_rate": 7.897323024699907e-06,
+      "loss": 1.0242,
+      "step": 3023
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6598155887974556,
+      "learning_rate": 7.8912318847996e-06,
+      "loss": 0.9592,
+      "step": 3024
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7226908009509426,
+      "learning_rate": 7.885141563595685e-06,
+      "loss": 1.0559,
+      "step": 3025
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7333363471492739,
+      "learning_rate": 7.879052063452626e-06,
+      "loss": 0.9231,
+      "step": 3026
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6994674185923929,
+      "learning_rate": 7.872963386734584e-06,
+      "loss": 1.0291,
+      "step": 3027
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7502165691354159,
+      "learning_rate": 7.866875535805394e-06,
+      "loss": 1.0648,
+      "step": 3028
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.35646503173514904,
+      "learning_rate": 7.860788513028566e-06,
+      "loss": 1.1058,
+      "step": 3029
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6940312427878355,
+      "learning_rate": 7.85470232076729e-06,
+      "loss": 1.0151,
+      "step": 3030
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7102626464826105,
+      "learning_rate": 7.848616961384442e-06,
+      "loss": 1.0615,
+      "step": 3031
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6800138955627202,
+      "learning_rate": 7.842532437242559e-06,
+      "loss": 0.9727,
+      "step": 3032
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6868206668381635,
+      "learning_rate": 7.83644875070387e-06,
+      "loss": 0.9258,
+      "step": 3033
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6427921329816917,
+      "learning_rate": 7.83036590413027e-06,
+      "loss": 1.0294,
+      "step": 3034
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6986483989050711,
+      "learning_rate": 7.824283899883327e-06,
+      "loss": 0.9986,
+      "step": 3035
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5063048867564927,
+      "learning_rate": 7.818202740324287e-06,
+      "loss": 1.0311,
+      "step": 3036
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6520728892452963,
+      "learning_rate": 7.812122427814068e-06,
+      "loss": 1.0197,
+      "step": 3037
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7008573838919353,
+      "learning_rate": 7.806042964713248e-06,
+      "loss": 1.0676,
+      "step": 3038
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7971799212077398,
+      "learning_rate": 7.79996435338209e-06,
+      "loss": 1.0613,
+      "step": 3039
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6992122615768599,
+      "learning_rate": 7.793886596180521e-06,
+      "loss": 1.0359,
+      "step": 3040
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7186265055396525,
+      "learning_rate": 7.787809695468134e-06,
+      "loss": 1.0035,
+      "step": 3041
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7293748026665431,
+      "learning_rate": 7.78173365360419e-06,
+      "loss": 1.0051,
+      "step": 3042
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7586320414114917,
+      "learning_rate": 7.775658472947623e-06,
+      "loss": 0.9272,
+      "step": 3043
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7127921612626014,
+      "learning_rate": 7.769584155857019e-06,
+      "loss": 0.9693,
+      "step": 3044
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.553301111278523,
+      "learning_rate": 7.763510704690645e-06,
+      "loss": 1.0398,
+      "step": 3045
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6517240700231866,
+      "learning_rate": 7.757438121806414e-06,
+      "loss": 1.073,
+      "step": 3046
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6929928089515207,
+      "learning_rate": 7.75136640956192e-06,
+      "loss": 0.9893,
+      "step": 3047
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7082192971269086,
+      "learning_rate": 7.745295570314412e-06,
+      "loss": 1.0104,
+      "step": 3048
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5069933748882313,
+      "learning_rate": 7.739225606420793e-06,
+      "loss": 1.0042,
+      "step": 3049
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7292466806356952,
+      "learning_rate": 7.733156520237633e-06,
+      "loss": 0.9551,
+      "step": 3050
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7508560249376013,
+      "learning_rate": 7.727088314121165e-06,
+      "loss": 1.0258,
+      "step": 3051
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7266502458126012,
+      "learning_rate": 7.721020990427268e-06,
+      "loss": 1.0393,
+      "step": 3052
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7109965500151603,
+      "learning_rate": 7.714954551511489e-06,
+      "loss": 0.9919,
+      "step": 3053
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.580988377148931,
+      "learning_rate": 7.708888999729036e-06,
+      "loss": 1.0199,
+      "step": 3054
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6786852571563236,
+      "learning_rate": 7.702824337434756e-06,
+      "loss": 1.0288,
+      "step": 3055
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7221643883480929,
+      "learning_rate": 7.69676056698316e-06,
+      "loss": 0.9622,
+      "step": 3056
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6694303585001207,
+      "learning_rate": 7.690697690728417e-06,
+      "loss": 0.9522,
+      "step": 3057
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7238686788970029,
+      "learning_rate": 7.68463571102434e-06,
+      "loss": 0.9288,
+      "step": 3058
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6295364085480739,
+      "learning_rate": 7.678574630224399e-06,
+      "loss": 1.0095,
+      "step": 3059
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6991125979774875,
+      "learning_rate": 7.672514450681721e-06,
+      "loss": 0.9792,
+      "step": 3060
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7131893231735337,
+      "learning_rate": 7.666455174749066e-06,
+      "loss": 0.982,
+      "step": 3061
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7287422765247014,
+      "learning_rate": 7.66039680477886e-06,
+      "loss": 0.9626,
+      "step": 3062
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5393886045613562,
+      "learning_rate": 7.654339343123173e-06,
+      "loss": 1.0962,
+      "step": 3063
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7260412730328544,
+      "learning_rate": 7.648282792133711e-06,
+      "loss": 1.0573,
+      "step": 3064
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7095100123977217,
+      "learning_rate": 7.642227154161841e-06,
+      "loss": 0.955,
+      "step": 3065
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7076779437595146,
+      "learning_rate": 7.636172431558575e-06,
+      "loss": 1.0255,
+      "step": 3066
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7737062930380731,
+      "learning_rate": 7.630118626674557e-06,
+      "loss": 0.9765,
+      "step": 3067
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6655992691256399,
+      "learning_rate": 7.6240657418600846e-06,
+      "loss": 0.9891,
+      "step": 3068
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7393323015264008,
+      "learning_rate": 7.618013779465101e-06,
+      "loss": 1.0344,
+      "step": 3069
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6952589680102651,
+      "learning_rate": 7.611962741839178e-06,
+      "loss": 1.0263,
+      "step": 3070
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.687638855961795,
+      "learning_rate": 7.6059126313315466e-06,
+      "loss": 0.985,
+      "step": 3071
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.68480405244459,
+      "learning_rate": 7.599863450291056e-06,
+      "loss": 1.0547,
+      "step": 3072
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7056573199042806,
+      "learning_rate": 7.593815201066215e-06,
+      "loss": 1.0091,
+      "step": 3073
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.747703002480568,
+      "learning_rate": 7.587767886005164e-06,
+      "loss": 1.0118,
+      "step": 3074
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6489382831771812,
+      "learning_rate": 7.581721507455672e-06,
+      "loss": 1.0341,
+      "step": 3075
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6970426319488494,
+      "learning_rate": 7.575676067765154e-06,
+      "loss": 1.0218,
+      "step": 3076
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.547545318387781,
+      "learning_rate": 7.569631569280662e-06,
+      "loss": 1.0703,
+      "step": 3077
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6524610390941892,
+      "learning_rate": 7.563588014348871e-06,
+      "loss": 1.0113,
+      "step": 3078
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7398290050410108,
+      "learning_rate": 7.5575454053161e-06,
+      "loss": 1.0574,
+      "step": 3079
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6933444446287105,
+      "learning_rate": 7.551503744528304e-06,
+      "loss": 1.0057,
+      "step": 3080
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6915344572065044,
+      "learning_rate": 7.545463034331054e-06,
+      "loss": 1.0486,
+      "step": 3081
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6836906611053903,
+      "learning_rate": 7.539423277069568e-06,
+      "loss": 1.0327,
+      "step": 3082
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7906304650228294,
+      "learning_rate": 7.53338447508869e-06,
+      "loss": 1.0164,
+      "step": 3083
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6968059558385772,
+      "learning_rate": 7.52734663073288e-06,
+      "loss": 1.0266,
+      "step": 3084
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.678993626385086,
+      "learning_rate": 7.521309746346246e-06,
+      "loss": 1.0351,
+      "step": 3085
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5607548992839381,
+      "learning_rate": 7.515273824272516e-06,
+      "loss": 0.9811,
+      "step": 3086
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7052609160687989,
+      "learning_rate": 7.509238866855033e-06,
+      "loss": 0.9655,
+      "step": 3087
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6538269173620511,
+      "learning_rate": 7.503204876436785e-06,
+      "loss": 0.9993,
+      "step": 3088
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6853042876371188,
+      "learning_rate": 7.497171855360372e-06,
+      "loss": 1.0338,
+      "step": 3089
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7466524355274178,
+      "learning_rate": 7.491139805968018e-06,
+      "loss": 1.0173,
+      "step": 3090
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7624527230259859,
+      "learning_rate": 7.485108730601571e-06,
+      "loss": 1.0389,
+      "step": 3091
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6653568654334768,
+      "learning_rate": 7.4790786316025125e-06,
+      "loss": 1.0089,
+      "step": 3092
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7318589949164089,
+      "learning_rate": 7.473049511311921e-06,
+      "loss": 1.109,
+      "step": 3093
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6915218048964682,
+      "learning_rate": 7.467021372070515e-06,
+      "loss": 1.088,
+      "step": 3094
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7530916572539593,
+      "learning_rate": 7.46099421621863e-06,
+      "loss": 0.9822,
+      "step": 3095
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6704897814826802,
+      "learning_rate": 7.4549680460962044e-06,
+      "loss": 0.9659,
+      "step": 3096
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6876219950305585,
+      "learning_rate": 7.448942864042819e-06,
+      "loss": 1.0123,
+      "step": 3097
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7533942483410881,
+      "learning_rate": 7.4429186723976425e-06,
+      "loss": 0.9832,
+      "step": 3098
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6716257257250123,
+      "learning_rate": 7.43689547349948e-06,
+      "loss": 0.9125,
+      "step": 3099
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6807664855076975,
+      "learning_rate": 7.43087326968675e-06,
+      "loss": 0.9911,
+      "step": 3100
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6649940242749706,
+      "learning_rate": 7.42485206329747e-06,
+      "loss": 1.0267,
+      "step": 3101
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7634995937271253,
+      "learning_rate": 7.418831856669286e-06,
+      "loss": 1.0079,
+      "step": 3102
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7013400077592792,
+      "learning_rate": 7.41281265213945e-06,
+      "loss": 0.9886,
+      "step": 3103
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6639761971863174,
+      "learning_rate": 7.406794452044816e-06,
+      "loss": 1.019,
+      "step": 3104
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6988316512833077,
+      "learning_rate": 7.400777258721865e-06,
+      "loss": 0.9731,
+      "step": 3105
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5918372875896825,
+      "learning_rate": 7.394761074506679e-06,
+      "loss": 1.0476,
+      "step": 3106
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6998148634754812,
+      "learning_rate": 7.3887459017349405e-06,
+      "loss": 1.0186,
+      "step": 3107
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5350640162006853,
+      "learning_rate": 7.382731742741953e-06,
+      "loss": 1.0591,
+      "step": 3108
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.665423183236309,
+      "learning_rate": 7.376718599862621e-06,
+      "loss": 0.9094,
+      "step": 3109
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.751737093985409,
+      "learning_rate": 7.370706475431446e-06,
+      "loss": 0.9952,
+      "step": 3110
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6762867648031105,
+      "learning_rate": 7.364695371782547e-06,
+      "loss": 0.975,
+      "step": 3111
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6871576403484744,
+      "learning_rate": 7.358685291249644e-06,
+      "loss": 0.9768,
+      "step": 3112
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7386807604994574,
+      "learning_rate": 7.352676236166051e-06,
+      "loss": 0.9692,
+      "step": 3113
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6557679348453452,
+      "learning_rate": 7.346668208864695e-06,
+      "loss": 0.971,
+      "step": 3114
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7083217728261132,
+      "learning_rate": 7.3406612116781e-06,
+      "loss": 1.0074,
+      "step": 3115
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7052844268321736,
+      "learning_rate": 7.33465524693838e-06,
+      "loss": 1.0135,
+      "step": 3116
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6369004539479912,
+      "learning_rate": 7.328650316977265e-06,
+      "loss": 1.0628,
+      "step": 3117
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6890894497445553,
+      "learning_rate": 7.322646424126079e-06,
+      "loss": 1.0101,
+      "step": 3118
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6416494806899522,
+      "learning_rate": 7.316643570715729e-06,
+      "loss": 1.042,
+      "step": 3119
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5451636924503518,
+      "learning_rate": 7.310641759076742e-06,
+      "loss": 1.0911,
+      "step": 3120
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.663820911465433,
+      "learning_rate": 7.304640991539216e-06,
+      "loss": 0.9495,
+      "step": 3121
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7190093429152369,
+      "learning_rate": 7.2986412704328625e-06,
+      "loss": 1.0057,
+      "step": 3122
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7003568745549232,
+      "learning_rate": 7.292642598086982e-06,
+      "loss": 1.0476,
+      "step": 3123
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7943506568600248,
+      "learning_rate": 7.286644976830457e-06,
+      "loss": 1.0516,
+      "step": 3124
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.689702446697638,
+      "learning_rate": 7.280648408991775e-06,
+      "loss": 1.0198,
+      "step": 3125
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7308311196691709,
+      "learning_rate": 7.274652896899015e-06,
+      "loss": 1.0176,
+      "step": 3126
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7088888299001396,
+      "learning_rate": 7.268658442879834e-06,
+      "loss": 1.0167,
+      "step": 3127
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5903023872036481,
+      "learning_rate": 7.262665049261489e-06,
+      "loss": 1.0436,
+      "step": 3128
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7629993567628237,
+      "learning_rate": 7.256672718370824e-06,
+      "loss": 1.0202,
+      "step": 3129
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6643402585394107,
+      "learning_rate": 7.250681452534261e-06,
+      "loss": 1.0092,
+      "step": 3130
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7673000843294776,
+      "learning_rate": 7.2446912540778196e-06,
+      "loss": 0.959,
+      "step": 3131
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6657272591159614,
+      "learning_rate": 7.238702125327106e-06,
+      "loss": 1.0439,
+      "step": 3132
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7040439491587841,
+      "learning_rate": 7.232714068607296e-06,
+      "loss": 1.0046,
+      "step": 3133
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6497205737272971,
+      "learning_rate": 7.226727086243168e-06,
+      "loss": 0.9383,
+      "step": 3134
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.64713719638001,
+      "learning_rate": 7.220741180559074e-06,
+      "loss": 1.0328,
+      "step": 3135
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7881907786599643,
+      "learning_rate": 7.214756353878942e-06,
+      "loss": 0.989,
+      "step": 3136
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6409383659247135,
+      "learning_rate": 7.208772608526293e-06,
+      "loss": 0.969,
+      "step": 3137
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6829630997541241,
+      "learning_rate": 7.202789946824227e-06,
+      "loss": 0.9994,
+      "step": 3138
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6865892976766627,
+      "learning_rate": 7.1968083710954075e-06,
+      "loss": 0.9981,
+      "step": 3139
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7489111686198318,
+      "learning_rate": 7.1908278836621e-06,
+      "loss": 0.9719,
+      "step": 3140
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.691171969650933,
+      "learning_rate": 7.184848486846128e-06,
+      "loss": 0.9732,
+      "step": 3141
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7671294178036973,
+      "learning_rate": 7.178870182968904e-06,
+      "loss": 0.9736,
+      "step": 3142
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7563003038199797,
+      "learning_rate": 7.1728929743514065e-06,
+      "loss": 1.0226,
+      "step": 3143
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7418662348365136,
+      "learning_rate": 7.166916863314199e-06,
+      "loss": 0.9971,
+      "step": 3144
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7418725840863774,
+      "learning_rate": 7.1609418521774095e-06,
+      "loss": 0.9831,
+      "step": 3145
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5458390217797147,
+      "learning_rate": 7.154967943260748e-06,
+      "loss": 0.9961,
+      "step": 3146
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7130512904627322,
+      "learning_rate": 7.148995138883483e-06,
+      "loss": 0.9583,
+      "step": 3147
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6893379862549558,
+      "learning_rate": 7.143023441364471e-06,
+      "loss": 1.0603,
+      "step": 3148
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7273026948771776,
+      "learning_rate": 7.13705285302213e-06,
+      "loss": 0.9916,
+      "step": 3149
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7358851580302583,
+      "learning_rate": 7.131083376174441e-06,
+      "loss": 0.9521,
+      "step": 3150
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6772864854046634,
+      "learning_rate": 7.125115013138966e-06,
+      "loss": 0.9777,
+      "step": 3151
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7121712123148857,
+      "learning_rate": 7.119147766232832e-06,
+      "loss": 1.0764,
+      "step": 3152
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7105189771114611,
+      "learning_rate": 7.113181637772721e-06,
+      "loss": 0.9828,
+      "step": 3153
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6497115284005541,
+      "learning_rate": 7.107216630074895e-06,
+      "loss": 1.0482,
+      "step": 3154
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7536687892831504,
+      "learning_rate": 7.1012527454551795e-06,
+      "loss": 0.9868,
+      "step": 3155
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.74201011657749,
+      "learning_rate": 7.09528998622895e-06,
+      "loss": 0.9677,
+      "step": 3156
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7951013293786582,
+      "learning_rate": 7.089328354711159e-06,
+      "loss": 0.9844,
+      "step": 3157
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7365453661746053,
+      "learning_rate": 7.083367853216323e-06,
+      "loss": 1.0017,
+      "step": 3158
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5364072261444514,
+      "learning_rate": 7.077408484058505e-06,
+      "loss": 1.0864,
+      "step": 3159
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6740360172766017,
+      "learning_rate": 7.071450249551342e-06,
+      "loss": 0.9922,
+      "step": 3160
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6948294381070502,
+      "learning_rate": 7.065493152008026e-06,
+      "loss": 1.0361,
+      "step": 3161
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6862598721227758,
+      "learning_rate": 7.059537193741306e-06,
+      "loss": 0.9664,
+      "step": 3162
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6586788691709328,
+      "learning_rate": 7.053582377063489e-06,
+      "loss": 1.0408,
+      "step": 3163
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7407067579894558,
+      "learning_rate": 7.047628704286446e-06,
+      "loss": 0.9828,
+      "step": 3164
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7453175752126312,
+      "learning_rate": 7.041676177721588e-06,
+      "loss": 1.0079,
+      "step": 3165
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7151178891692445,
+      "learning_rate": 7.035724799679898e-06,
+      "loss": 0.9938,
+      "step": 3166
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.47988918307115225,
+      "learning_rate": 7.029774572471904e-06,
+      "loss": 1.011,
+      "step": 3167
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7326733208651374,
+      "learning_rate": 7.023825498407689e-06,
+      "loss": 1.0284,
+      "step": 3168
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7129534500289791,
+      "learning_rate": 7.0178775797968855e-06,
+      "loss": 0.9844,
+      "step": 3169
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6516822606075966,
+      "learning_rate": 7.011930818948688e-06,
+      "loss": 0.9423,
+      "step": 3170
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8374247088564003,
+      "learning_rate": 7.005985218171825e-06,
+      "loss": 0.969,
+      "step": 3171
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7349254145305562,
+      "learning_rate": 7.000040779774591e-06,
+      "loss": 0.9921,
+      "step": 3172
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5727983190367804,
+      "learning_rate": 6.994097506064812e-06,
+      "loss": 1.0182,
+      "step": 3173
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.765283666294845,
+      "learning_rate": 6.9881553993498805e-06,
+      "loss": 1.0294,
+      "step": 3174
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.738242090263811,
+      "learning_rate": 6.9822144619367275e-06,
+      "loss": 1.0327,
+      "step": 3175
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6786511445372556,
+      "learning_rate": 6.97627469613182e-06,
+      "loss": 0.9927,
+      "step": 3176
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7014363417426183,
+      "learning_rate": 6.970336104241186e-06,
+      "loss": 1.0684,
+      "step": 3177
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6270772365816405,
+      "learning_rate": 6.9643986885703955e-06,
+      "loss": 0.9839,
+      "step": 3178
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6969329046260675,
+      "learning_rate": 6.958462451424547e-06,
+      "loss": 0.9914,
+      "step": 3179
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7057868561394046,
+      "learning_rate": 6.952527395108302e-06,
+      "loss": 1.0137,
+      "step": 3180
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7102503629312942,
+      "learning_rate": 6.9465935219258504e-06,
+      "loss": 0.9884,
+      "step": 3181
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5254176922089159,
+      "learning_rate": 6.9406608341809215e-06,
+      "loss": 1.0129,
+      "step": 3182
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5981995996544919,
+      "learning_rate": 6.934729334176793e-06,
+      "loss": 1.0599,
+      "step": 3183
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.736873751711971,
+      "learning_rate": 6.928799024216282e-06,
+      "loss": 1.0213,
+      "step": 3184
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6915284366573967,
+      "learning_rate": 6.92286990660173e-06,
+      "loss": 1.041,
+      "step": 3185
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7074322025796578,
+      "learning_rate": 6.91694198363503e-06,
+      "loss": 0.9717,
+      "step": 3186
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6131899105892533,
+      "learning_rate": 6.911015257617606e-06,
+      "loss": 0.9841,
+      "step": 3187
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.72481145980892,
+      "learning_rate": 6.905089730850416e-06,
+      "loss": 1.0361,
+      "step": 3188
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5182966116405203,
+      "learning_rate": 6.8991654056339505e-06,
+      "loss": 1.1597,
+      "step": 3189
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7142822351959707,
+      "learning_rate": 6.893242284268244e-06,
+      "loss": 1.0409,
+      "step": 3190
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6765884838403508,
+      "learning_rate": 6.887320369052848e-06,
+      "loss": 1.0765,
+      "step": 3191
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.697573717294624,
+      "learning_rate": 6.8813996622868584e-06,
+      "loss": 0.9903,
+      "step": 3192
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6865841160198031,
+      "learning_rate": 6.8754801662688964e-06,
+      "loss": 0.996,
+      "step": 3193
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6935511297224438,
+      "learning_rate": 6.869561883297116e-06,
+      "loss": 0.9641,
+      "step": 3194
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7431816889209735,
+      "learning_rate": 6.863644815669197e-06,
+      "loss": 1.0262,
+      "step": 3195
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6727807957942876,
+      "learning_rate": 6.857728965682344e-06,
+      "loss": 0.9958,
+      "step": 3196
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7775842651346733,
+      "learning_rate": 6.851814335633298e-06,
+      "loss": 1.034,
+      "step": 3197
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5288500841421891,
+      "learning_rate": 6.8459009278183275e-06,
+      "loss": 0.9904,
+      "step": 3198
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6904605318637759,
+      "learning_rate": 6.839988744533211e-06,
+      "loss": 0.9305,
+      "step": 3199
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6556619102766692,
+      "learning_rate": 6.834077788073268e-06,
+      "loss": 1.0217,
+      "step": 3200
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5865289372196374,
+      "learning_rate": 6.828168060733336e-06,
+      "loss": 1.0122,
+      "step": 3201
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.677726352116388,
+      "learning_rate": 6.822259564807768e-06,
+      "loss": 0.9785,
+      "step": 3202
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6886471673026469,
+      "learning_rate": 6.81635230259045e-06,
+      "loss": 0.9901,
+      "step": 3203
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5292557377744805,
+      "learning_rate": 6.810446276374789e-06,
+      "loss": 0.9979,
+      "step": 3204
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7713362301518366,
+      "learning_rate": 6.8045414884536975e-06,
+      "loss": 1.002,
+      "step": 3205
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6958484384224979,
+      "learning_rate": 6.7986379411196255e-06,
+      "loss": 0.9637,
+      "step": 3206
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7192508265146454,
+      "learning_rate": 6.7927356366645315e-06,
+      "loss": 0.9938,
+      "step": 3207
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6539687156930817,
+      "learning_rate": 6.786834577379893e-06,
+      "loss": 0.9965,
+      "step": 3208
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6405356524019207,
+      "learning_rate": 6.780934765556702e-06,
+      "loss": 1.0074,
+      "step": 3209
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6876563729559056,
+      "learning_rate": 6.775036203485472e-06,
+      "loss": 1.0185,
+      "step": 3210
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7600163812031003,
+      "learning_rate": 6.769138893456225e-06,
+      "loss": 0.9798,
+      "step": 3211
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7278392484704517,
+      "learning_rate": 6.763242837758504e-06,
+      "loss": 0.9927,
+      "step": 3212
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.70984190970968,
+      "learning_rate": 6.757348038681357e-06,
+      "loss": 1.0411,
+      "step": 3213
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6902708695937897,
+      "learning_rate": 6.751454498513349e-06,
+      "loss": 0.9298,
+      "step": 3214
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6924039045542755,
+      "learning_rate": 6.745562219542554e-06,
+      "loss": 0.9662,
+      "step": 3215
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6849547304719548,
+      "learning_rate": 6.7396712040565625e-06,
+      "loss": 0.988,
+      "step": 3216
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6831451187528024,
+      "learning_rate": 6.733781454342463e-06,
+      "loss": 0.9948,
+      "step": 3217
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7337369675739607,
+      "learning_rate": 6.727892972686861e-06,
+      "loss": 0.9716,
+      "step": 3218
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6521182162789554,
+      "learning_rate": 6.722005761375873e-06,
+      "loss": 1.047,
+      "step": 3219
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.731788013058303,
+      "learning_rate": 6.716119822695111e-06,
+      "loss": 1.0256,
+      "step": 3220
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7024609206780646,
+      "learning_rate": 6.710235158929703e-06,
+      "loss": 1.0291,
+      "step": 3221
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6450679782418445,
+      "learning_rate": 6.704351772364274e-06,
+      "loss": 0.9886,
+      "step": 3222
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7530726985989601,
+      "learning_rate": 6.698469665282958e-06,
+      "loss": 0.9617,
+      "step": 3223
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6676852690942534,
+      "learning_rate": 6.692588839969397e-06,
+      "loss": 1.021,
+      "step": 3224
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7380523394007131,
+      "learning_rate": 6.6867092987067214e-06,
+      "loss": 0.9408,
+      "step": 3225
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7164448255382535,
+      "learning_rate": 6.680831043777579e-06,
+      "loss": 1.0651,
+      "step": 3226
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5392025067815471,
+      "learning_rate": 6.674954077464108e-06,
+      "loss": 1.1021,
+      "step": 3227
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6731884754805728,
+      "learning_rate": 6.6690784020479484e-06,
+      "loss": 1.0034,
+      "step": 3228
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7457105366438866,
+      "learning_rate": 6.6632040198102364e-06,
+      "loss": 1.0121,
+      "step": 3229
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8023043026086358,
+      "learning_rate": 6.657330933031619e-06,
+      "loss": 1.052,
+      "step": 3230
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.77027439531793,
+      "learning_rate": 6.651459143992221e-06,
+      "loss": 1.0542,
+      "step": 3231
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6872733649162897,
+      "learning_rate": 6.645588654971677e-06,
+      "loss": 0.9976,
+      "step": 3232
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5399765948398303,
+      "learning_rate": 6.639719468249115e-06,
+      "loss": 0.9801,
+      "step": 3233
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6456017949511895,
+      "learning_rate": 6.633851586103153e-06,
+      "loss": 1.0792,
+      "step": 3234
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.686306562631849,
+      "learning_rate": 6.627985010811903e-06,
+      "loss": 0.9984,
+      "step": 3235
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7165354843245196,
+      "learning_rate": 6.622119744652977e-06,
+      "loss": 0.9954,
+      "step": 3236
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.717628058553665,
+      "learning_rate": 6.616255789903467e-06,
+      "loss": 0.9973,
+      "step": 3237
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7151384814948639,
+      "learning_rate": 6.610393148839964e-06,
+      "loss": 1.0103,
+      "step": 3238
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6722042177513897,
+      "learning_rate": 6.6045318237385526e-06,
+      "loss": 0.9768,
+      "step": 3239
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.745015357041718,
+      "learning_rate": 6.598671816874794e-06,
+      "loss": 0.9509,
+      "step": 3240
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.69545233207071,
+      "learning_rate": 6.5928131305237465e-06,
+      "loss": 0.9616,
+      "step": 3241
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5163015342153089,
+      "learning_rate": 6.586955766959958e-06,
+      "loss": 1.0166,
+      "step": 3242
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6996083121033039,
+      "learning_rate": 6.581099728457451e-06,
+      "loss": 1.0652,
+      "step": 3243
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.689788790662944,
+      "learning_rate": 6.5752450172897466e-06,
+      "loss": 1.0264,
+      "step": 3244
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7734354616639213,
+      "learning_rate": 6.569391635729847e-06,
+      "loss": 1.0139,
+      "step": 3245
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6898303304649427,
+      "learning_rate": 6.563539586050233e-06,
+      "loss": 1.0209,
+      "step": 3246
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6707084277264541,
+      "learning_rate": 6.557688870522871e-06,
+      "loss": 0.9979,
+      "step": 3247
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6412535989624403,
+      "learning_rate": 6.551839491419213e-06,
+      "loss": 1.0101,
+      "step": 3248
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5701255537296795,
+      "learning_rate": 6.545991451010185e-06,
+      "loss": 1.0616,
+      "step": 3249
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5498081728673219,
+      "learning_rate": 6.5401447515662065e-06,
+      "loss": 1.0475,
+      "step": 3250
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5506484457075295,
+      "learning_rate": 6.5342993953571556e-06,
+      "loss": 1.0008,
+      "step": 3251
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6661050065001609,
+      "learning_rate": 6.52845538465241e-06,
+      "loss": 1.0434,
+      "step": 3252
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.625103101781463,
+      "learning_rate": 6.522612721720813e-06,
+      "loss": 1.0216,
+      "step": 3253
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.531201543155485,
+      "learning_rate": 6.5167714088306865e-06,
+      "loss": 1.0992,
+      "step": 3254
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.728453645820189,
+      "learning_rate": 6.51093144824983e-06,
+      "loss": 1.0636,
+      "step": 3255
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7092593877529008,
+      "learning_rate": 6.505092842245519e-06,
+      "loss": 1.0264,
+      "step": 3256
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7350650846812118,
+      "learning_rate": 6.499255593084498e-06,
+      "loss": 1.0356,
+      "step": 3257
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6975674360666722,
+      "learning_rate": 6.493419703032991e-06,
+      "loss": 1.0284,
+      "step": 3258
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6924880764922097,
+      "learning_rate": 6.487585174356691e-06,
+      "loss": 0.9806,
+      "step": 3259
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7404253372529904,
+      "learning_rate": 6.481752009320761e-06,
+      "loss": 1.0433,
+      "step": 3260
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6929024245668954,
+      "learning_rate": 6.4759202101898366e-06,
+      "loss": 0.988,
+      "step": 3261
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6259191634779272,
+      "learning_rate": 6.4700897792280285e-06,
+      "loss": 1.0071,
+      "step": 3262
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7064112958378802,
+      "learning_rate": 6.464260718698902e-06,
+      "loss": 1.0516,
+      "step": 3263
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6784152819814072,
+      "learning_rate": 6.458433030865503e-06,
+      "loss": 0.9772,
+      "step": 3264
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7624756358443702,
+      "learning_rate": 6.452606717990346e-06,
+      "loss": 1.0709,
+      "step": 3265
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7045188113309476,
+      "learning_rate": 6.4467817823354005e-06,
+      "loss": 0.9568,
+      "step": 3266
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5529736175626269,
+      "learning_rate": 6.440958226162104e-06,
+      "loss": 1.0104,
+      "step": 3267
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6802411759418387,
+      "learning_rate": 6.43513605173137e-06,
+      "loss": 1.0234,
+      "step": 3268
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5256023975927009,
+      "learning_rate": 6.4293152613035594e-06,
+      "loss": 1.0891,
+      "step": 3269
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6627315123601732,
+      "learning_rate": 6.4234958571385095e-06,
+      "loss": 1.0708,
+      "step": 3270
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7160695975660768,
+      "learning_rate": 6.4176778414955075e-06,
+      "loss": 1.041,
+      "step": 3271
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7689120021340042,
+      "learning_rate": 6.4118612166333124e-06,
+      "loss": 0.9456,
+      "step": 3272
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6498328269572798,
+      "learning_rate": 6.4060459848101354e-06,
+      "loss": 1.0383,
+      "step": 3273
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.676905036476696,
+      "learning_rate": 6.400232148283651e-06,
+      "loss": 1.0681,
+      "step": 3274
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6450774769535894,
+      "learning_rate": 6.3944197093109885e-06,
+      "loss": 1.0265,
+      "step": 3275
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7203808945144948,
+      "learning_rate": 6.388608670148741e-06,
+      "loss": 1.0125,
+      "step": 3276
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7395808736083611,
+      "learning_rate": 6.38279903305295e-06,
+      "loss": 1.0015,
+      "step": 3277
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8605811760221073,
+      "learning_rate": 6.376990800279119e-06,
+      "loss": 0.9614,
+      "step": 3278
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5396830024064441,
+      "learning_rate": 6.3711839740822035e-06,
+      "loss": 1.0927,
+      "step": 3279
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7189232304659116,
+      "learning_rate": 6.3653785567166125e-06,
+      "loss": 1.0234,
+      "step": 3280
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6656426977627078,
+      "learning_rate": 6.359574550436209e-06,
+      "loss": 0.9742,
+      "step": 3281
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5607812195900913,
+      "learning_rate": 6.3537719574943105e-06,
+      "loss": 1.0362,
+      "step": 3282
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5150870391920592,
+      "learning_rate": 6.347970780143678e-06,
+      "loss": 1.0553,
+      "step": 3283
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6707297975777325,
+      "learning_rate": 6.342171020636533e-06,
+      "loss": 1.0065,
+      "step": 3284
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6790872489742008,
+      "learning_rate": 6.336372681224543e-06,
+      "loss": 1.0447,
+      "step": 3285
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7794119899921153,
+      "learning_rate": 6.330575764158819e-06,
+      "loss": 1.0079,
+      "step": 3286
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5439492165628965,
+      "learning_rate": 6.324780271689923e-06,
+      "loss": 1.0083,
+      "step": 3287
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7434052746313046,
+      "learning_rate": 6.318986206067872e-06,
+      "loss": 0.9969,
+      "step": 3288
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.76320453731579,
+      "learning_rate": 6.313193569542113e-06,
+      "loss": 1.081,
+      "step": 3289
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6783557130247897,
+      "learning_rate": 6.30740236436155e-06,
+      "loss": 0.9785,
+      "step": 3290
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5345615488014696,
+      "learning_rate": 6.301612592774533e-06,
+      "loss": 0.9885,
+      "step": 3291
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7642848115116297,
+      "learning_rate": 6.295824257028844e-06,
+      "loss": 0.9891,
+      "step": 3292
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7565143383633908,
+      "learning_rate": 6.290037359371717e-06,
+      "loss": 0.972,
+      "step": 3293
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7469817392062956,
+      "learning_rate": 6.284251902049827e-06,
+      "loss": 1.0023,
+      "step": 3294
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6884385966131619,
+      "learning_rate": 6.278467887309283e-06,
+      "loss": 1.0318,
+      "step": 3295
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7711588024388251,
+      "learning_rate": 6.272685317395644e-06,
+      "loss": 1.0043,
+      "step": 3296
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7261615043994522,
+      "learning_rate": 6.266904194553896e-06,
+      "loss": 1.021,
+      "step": 3297
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6983943056964287,
+      "learning_rate": 6.261124521028477e-06,
+      "loss": 1.0342,
+      "step": 3298
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7279628923123839,
+      "learning_rate": 6.255346299063252e-06,
+      "loss": 1.019,
+      "step": 3299
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6831175157057413,
+      "learning_rate": 6.249569530901525e-06,
+      "loss": 1.0114,
+      "step": 3300
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7563962385634532,
+      "learning_rate": 6.243794218786034e-06,
+      "loss": 1.0084,
+      "step": 3301
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7683345616251922,
+      "learning_rate": 6.238020364958964e-06,
+      "loss": 0.9916,
+      "step": 3302
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6961120897463356,
+      "learning_rate": 6.232247971661912e-06,
+      "loss": 1.015,
+      "step": 3303
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7066506295742665,
+      "learning_rate": 6.2264770411359256e-06,
+      "loss": 0.918,
+      "step": 3304
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6899192457302725,
+      "learning_rate": 6.22070757562148e-06,
+      "loss": 1.0087,
+      "step": 3305
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7356737396143912,
+      "learning_rate": 6.214939577358479e-06,
+      "loss": 1.0997,
+      "step": 3306
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6828395603865239,
+      "learning_rate": 6.209173048586253e-06,
+      "loss": 1.035,
+      "step": 3307
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7167075433719418,
+      "learning_rate": 6.203407991543577e-06,
+      "loss": 1.0079,
+      "step": 3308
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6668449466000789,
+      "learning_rate": 6.197644408468635e-06,
+      "loss": 1.0137,
+      "step": 3309
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.708333430437442,
+      "learning_rate": 6.191882301599052e-06,
+      "loss": 1.0426,
+      "step": 3310
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7501968140464693,
+      "learning_rate": 6.186121673171882e-06,
+      "loss": 0.9415,
+      "step": 3311
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7026829011985805,
+      "learning_rate": 6.180362525423591e-06,
+      "loss": 1.0532,
+      "step": 3312
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6001042198808405,
+      "learning_rate": 6.174604860590081e-06,
+      "loss": 1.0255,
+      "step": 3313
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6808136347229677,
+      "learning_rate": 6.168848680906678e-06,
+      "loss": 1.0109,
+      "step": 3314
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6291636975418506,
+      "learning_rate": 6.163093988608127e-06,
+      "loss": 0.9822,
+      "step": 3315
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.646280285872295,
+      "learning_rate": 6.157340785928595e-06,
+      "loss": 1.0543,
+      "step": 3316
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6963363519998835,
+      "learning_rate": 6.151589075101681e-06,
+      "loss": 0.9478,
+      "step": 3317
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6473606302855501,
+      "learning_rate": 6.145838858360391e-06,
+      "loss": 0.9432,
+      "step": 3318
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6959767419612874,
+      "learning_rate": 6.140090137937158e-06,
+      "loss": 1.0087,
+      "step": 3319
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7676129455485303,
+      "learning_rate": 6.134342916063838e-06,
+      "loss": 1.0423,
+      "step": 3320
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7174153613851804,
+      "learning_rate": 6.128597194971691e-06,
+      "loss": 0.9718,
+      "step": 3321
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.665209283306251,
+      "learning_rate": 6.122852976891413e-06,
+      "loss": 0.9827,
+      "step": 3322
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5758641401949562,
+      "learning_rate": 6.117110264053101e-06,
+      "loss": 0.9847,
+      "step": 3323
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7188844712908248,
+      "learning_rate": 6.111369058686276e-06,
+      "loss": 1.0213,
+      "step": 3324
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7586553232490871,
+      "learning_rate": 6.105629363019875e-06,
+      "loss": 0.9896,
+      "step": 3325
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.554776498390738,
+      "learning_rate": 6.099891179282242e-06,
+      "loss": 1.0446,
+      "step": 3326
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7450328614421939,
+      "learning_rate": 6.094154509701133e-06,
+      "loss": 0.9987,
+      "step": 3327
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7129531493963954,
+      "learning_rate": 6.088419356503732e-06,
+      "loss": 1.0035,
+      "step": 3328
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6846966031463083,
+      "learning_rate": 6.082685721916612e-06,
+      "loss": 1.0583,
+      "step": 3329
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6938056570461172,
+      "learning_rate": 6.076953608165772e-06,
+      "loss": 1.0538,
+      "step": 3330
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4982044092635117,
+      "learning_rate": 6.07122301747662e-06,
+      "loss": 1.0527,
+      "step": 3331
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6581346983592224,
+      "learning_rate": 6.065493952073961e-06,
+      "loss": 0.9883,
+      "step": 3332
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6813464599953528,
+      "learning_rate": 6.0597664141820176e-06,
+      "loss": 1.0247,
+      "step": 3333
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7291656301721215,
+      "learning_rate": 6.054040406024422e-06,
+      "loss": 1.075,
+      "step": 3334
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7429758456276061,
+      "learning_rate": 6.0483159298242e-06,
+      "loss": 0.9807,
+      "step": 3335
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6435334891107368,
+      "learning_rate": 6.042592987803796e-06,
+      "loss": 0.9816,
+      "step": 3336
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6855778558346135,
+      "learning_rate": 6.036871582185054e-06,
+      "loss": 1.0443,
+      "step": 3337
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7763606700447241,
+      "learning_rate": 6.031151715189217e-06,
+      "loss": 0.9941,
+      "step": 3338
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.747169860143676,
+      "learning_rate": 6.025433389036935e-06,
+      "loss": 0.8857,
+      "step": 3339
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.694149583182159,
+      "learning_rate": 6.019716605948261e-06,
+      "loss": 1.0505,
+      "step": 3340
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5227538438935914,
+      "learning_rate": 6.014001368142643e-06,
+      "loss": 1.0842,
+      "step": 3341
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6956346128001405,
+      "learning_rate": 6.008287677838937e-06,
+      "loss": 0.9139,
+      "step": 3342
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7428760162873217,
+      "learning_rate": 6.002575537255395e-06,
+      "loss": 1.0433,
+      "step": 3343
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6401399342392492,
+      "learning_rate": 5.996864948609662e-06,
+      "loss": 1.0396,
+      "step": 3344
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6733095884355711,
+      "learning_rate": 5.9911559141187924e-06,
+      "loss": 0.9795,
+      "step": 3345
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7611233713738702,
+      "learning_rate": 5.9854484359992235e-06,
+      "loss": 0.9835,
+      "step": 3346
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7079643652072418,
+      "learning_rate": 5.979742516466793e-06,
+      "loss": 1.0518,
+      "step": 3347
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6917346572549162,
+      "learning_rate": 5.974038157736746e-06,
+      "loss": 0.9168,
+      "step": 3348
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7525107110421312,
+      "learning_rate": 5.968335362023697e-06,
+      "loss": 1.0359,
+      "step": 3349
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.662680468565591,
+      "learning_rate": 5.962634131541676e-06,
+      "loss": 0.9639,
+      "step": 3350
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7202672548417246,
+      "learning_rate": 5.956934468504101e-06,
+      "loss": 1.1107,
+      "step": 3351
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6616363071850219,
+      "learning_rate": 5.951236375123768e-06,
+      "loss": 1.0115,
+      "step": 3352
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6716469238030627,
+      "learning_rate": 5.945539853612876e-06,
+      "loss": 1.0038,
+      "step": 3353
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7630884782601061,
+      "learning_rate": 5.939844906183016e-06,
+      "loss": 1.0259,
+      "step": 3354
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6931063289797971,
+      "learning_rate": 5.934151535045156e-06,
+      "loss": 1.0186,
+      "step": 3355
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7270276065952443,
+      "learning_rate": 5.92845974240966e-06,
+      "loss": 0.9638,
+      "step": 3356
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.720950563985332,
+      "learning_rate": 5.922769530486283e-06,
+      "loss": 0.9625,
+      "step": 3357
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.570300476513978,
+      "learning_rate": 5.917080901484156e-06,
+      "loss": 1.0475,
+      "step": 3358
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7014387063679907,
+      "learning_rate": 5.9113938576118e-06,
+      "loss": 0.9774,
+      "step": 3359
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7390035505656519,
+      "learning_rate": 5.905708401077128e-06,
+      "loss": 1.0101,
+      "step": 3360
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6818664834162622,
+      "learning_rate": 5.900024534087421e-06,
+      "loss": 0.9975,
+      "step": 3361
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7359189490527985,
+      "learning_rate": 5.894342258849355e-06,
+      "loss": 0.9515,
+      "step": 3362
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7115820053660433,
+      "learning_rate": 5.88866157756899e-06,
+      "loss": 0.9821,
+      "step": 3363
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7023592023515657,
+      "learning_rate": 5.882982492451757e-06,
+      "loss": 1.1179,
+      "step": 3364
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5154053435372695,
+      "learning_rate": 5.877305005702471e-06,
+      "loss": 0.9335,
+      "step": 3365
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7390911632701431,
+      "learning_rate": 5.871629119525335e-06,
+      "loss": 0.8745,
+      "step": 3366
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5199952772882013,
+      "learning_rate": 5.865954836123915e-06,
+      "loss": 1.0648,
+      "step": 3367
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6898081629555887,
+      "learning_rate": 5.860282157701167e-06,
+      "loss": 1.1239,
+      "step": 3368
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6307266719250456,
+      "learning_rate": 5.854611086459423e-06,
+      "loss": 0.9859,
+      "step": 3369
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6733548742657821,
+      "learning_rate": 5.8489416246003814e-06,
+      "loss": 1.0158,
+      "step": 3370
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7241801710845588,
+      "learning_rate": 5.8432737743251315e-06,
+      "loss": 0.9793,
+      "step": 3371
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7375179479933756,
+      "learning_rate": 5.8376075378341194e-06,
+      "loss": 1.0314,
+      "step": 3372
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6946511913262264,
+      "learning_rate": 5.831942917327172e-06,
+      "loss": 1.0065,
+      "step": 3373
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6935317203457778,
+      "learning_rate": 5.826279915003503e-06,
+      "loss": 1.0384,
+      "step": 3374
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6942199931532365,
+      "learning_rate": 5.8206185330616725e-06,
+      "loss": 1.0365,
+      "step": 3375
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.802932532795961,
+      "learning_rate": 5.814958773699625e-06,
+      "loss": 1.0191,
+      "step": 3376
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6836865415452017,
+      "learning_rate": 5.809300639114683e-06,
+      "loss": 0.9961,
+      "step": 3377
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6750861427018733,
+      "learning_rate": 5.803644131503516e-06,
+      "loss": 1.0376,
+      "step": 3378
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7203364079599336,
+      "learning_rate": 5.797989253062186e-06,
+      "loss": 1.0037,
+      "step": 3379
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6918905963195212,
+      "learning_rate": 5.792336005986105e-06,
+      "loss": 1.0195,
+      "step": 3380
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6989002804274821,
+      "learning_rate": 5.786684392470064e-06,
+      "loss": 1.0061,
+      "step": 3381
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7384898654676817,
+      "learning_rate": 5.781034414708208e-06,
+      "loss": 1.0175,
+      "step": 3382
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5969802844657993,
+      "learning_rate": 5.775386074894058e-06,
+      "loss": 1.0323,
+      "step": 3383
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6399846922854087,
+      "learning_rate": 5.769739375220489e-06,
+      "loss": 1.0189,
+      "step": 3384
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7096267235395087,
+      "learning_rate": 5.7640943178797445e-06,
+      "loss": 1.0411,
+      "step": 3385
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6927930975936472,
+      "learning_rate": 5.7584509050634395e-06,
+      "loss": 1.0532,
+      "step": 3386
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.706487362172857,
+      "learning_rate": 5.752809138962525e-06,
+      "loss": 1.002,
+      "step": 3387
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8724811482712113,
+      "learning_rate": 5.747169021767342e-06,
+      "loss": 0.9782,
+      "step": 3388
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6526417253131938,
+      "learning_rate": 5.7415305556675805e-06,
+      "loss": 0.9844,
+      "step": 3389
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5782847942345573,
+      "learning_rate": 5.73589374285227e-06,
+      "loss": 1.0474,
+      "step": 3390
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.64831425129427,
+      "learning_rate": 5.730258585509832e-06,
+      "loss": 0.9266,
+      "step": 3391
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.72130077198718,
+      "learning_rate": 5.724625085828022e-06,
+      "loss": 0.9475,
+      "step": 3392
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6848865935488324,
+      "learning_rate": 5.718993245993958e-06,
+      "loss": 0.9787,
+      "step": 3393
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5943524707446184,
+      "learning_rate": 5.713363068194115e-06,
+      "loss": 1.0301,
+      "step": 3394
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7463377955692476,
+      "learning_rate": 5.7077345546143235e-06,
+      "loss": 1.0584,
+      "step": 3395
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7285993174650117,
+      "learning_rate": 5.702107707439766e-06,
+      "loss": 0.9658,
+      "step": 3396
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7668801758911366,
+      "learning_rate": 5.6964825288549745e-06,
+      "loss": 1.0569,
+      "step": 3397
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7368019096749046,
+      "learning_rate": 5.690859021043842e-06,
+      "loss": 1.0036,
+      "step": 3398
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6975506408811448,
+      "learning_rate": 5.685237186189601e-06,
+      "loss": 0.9901,
+      "step": 3399
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8281411010693752,
+      "learning_rate": 5.679617026474853e-06,
+      "loss": 1.09,
+      "step": 3400
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7595018088187021,
+      "learning_rate": 5.673998544081527e-06,
+      "loss": 1.0007,
+      "step": 3401
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.550169844670275,
+      "learning_rate": 5.6683817411909114e-06,
+      "loss": 1.0366,
+      "step": 3402
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7120358096638025,
+      "learning_rate": 5.662766619983653e-06,
+      "loss": 1.0035,
+      "step": 3403
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8138009018002459,
+      "learning_rate": 5.65715318263972e-06,
+      "loss": 0.983,
+      "step": 3404
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.687320233146181,
+      "learning_rate": 5.651541431338454e-06,
+      "loss": 1.0576,
+      "step": 3405
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5435507228109281,
+      "learning_rate": 5.645931368258527e-06,
+      "loss": 0.9949,
+      "step": 3406
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7355315521538679,
+      "learning_rate": 5.640322995577958e-06,
+      "loss": 0.9767,
+      "step": 3407
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7236675686311443,
+      "learning_rate": 5.634716315474109e-06,
+      "loss": 0.9958,
+      "step": 3408
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6384944938586377,
+      "learning_rate": 5.629111330123689e-06,
+      "loss": 0.9814,
+      "step": 3409
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5026707400338136,
+      "learning_rate": 5.623508041702743e-06,
+      "loss": 1.0583,
+      "step": 3410
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6603821847634231,
+      "learning_rate": 5.617906452386659e-06,
+      "loss": 1.0258,
+      "step": 3411
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.671100105996673,
+      "learning_rate": 5.612306564350179e-06,
+      "loss": 0.9949,
+      "step": 3412
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.649255446254863,
+      "learning_rate": 5.6067083797673535e-06,
+      "loss": 0.9998,
+      "step": 3413
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.519954617538613,
+      "learning_rate": 5.601111900811607e-06,
+      "loss": 1.0635,
+      "step": 3414
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7132438131580313,
+      "learning_rate": 5.595517129655681e-06,
+      "loss": 1.0065,
+      "step": 3415
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7229229217453443,
+      "learning_rate": 5.589924068471648e-06,
+      "loss": 1.0085,
+      "step": 3416
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6704025764822715,
+      "learning_rate": 5.58433271943094e-06,
+      "loss": 1.0321,
+      "step": 3417
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7073735735608659,
+      "learning_rate": 5.578743084704306e-06,
+      "loss": 1.0079,
+      "step": 3418
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7570148675974402,
+      "learning_rate": 5.573155166461833e-06,
+      "loss": 0.9595,
+      "step": 3419
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5444490415893991,
+      "learning_rate": 5.567568966872947e-06,
+      "loss": 1.0651,
+      "step": 3420
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6335251442909676,
+      "learning_rate": 5.5619844881064e-06,
+      "loss": 0.9705,
+      "step": 3421
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7175889738912888,
+      "learning_rate": 5.556401732330281e-06,
+      "loss": 0.94,
+      "step": 3422
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6941981773945436,
+      "learning_rate": 5.550820701712007e-06,
+      "loss": 0.9725,
+      "step": 3423
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7174774297357663,
+      "learning_rate": 5.545241398418326e-06,
+      "loss": 1.0,
+      "step": 3424
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.754912714552894,
+      "learning_rate": 5.539663824615312e-06,
+      "loss": 0.9337,
+      "step": 3425
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5859624269781861,
+      "learning_rate": 5.534087982468384e-06,
+      "loss": 0.9729,
+      "step": 3426
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5761316212940842,
+      "learning_rate": 5.5285138741422615e-06,
+      "loss": 1.0771,
+      "step": 3427
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5639330718796134,
+      "learning_rate": 5.522941501801008e-06,
+      "loss": 1.0897,
+      "step": 3428
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6944020705061729,
+      "learning_rate": 5.517370867608021e-06,
+      "loss": 0.993,
+      "step": 3429
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6305142944840246,
+      "learning_rate": 5.511801973725997e-06,
+      "loss": 0.9673,
+      "step": 3430
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8360803390453262,
+      "learning_rate": 5.506234822316983e-06,
+      "loss": 1.0147,
+      "step": 3431
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7895765470509913,
+      "learning_rate": 5.500669415542336e-06,
+      "loss": 1.0447,
+      "step": 3432
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6221384790685425,
+      "learning_rate": 5.495105755562738e-06,
+      "loss": 0.9372,
+      "step": 3433
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6203961181261893,
+      "learning_rate": 5.4895438445381945e-06,
+      "loss": 1.0626,
+      "step": 3434
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6786163852772198,
+      "learning_rate": 5.48398368462803e-06,
+      "loss": 1.0051,
+      "step": 3435
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6622098969803679,
+      "learning_rate": 5.4784252779908905e-06,
+      "loss": 1.0579,
+      "step": 3436
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7470737186503811,
+      "learning_rate": 5.4728686267847354e-06,
+      "loss": 1.0591,
+      "step": 3437
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.494879407321919,
+      "learning_rate": 5.467313733166863e-06,
+      "loss": 1.0322,
+      "step": 3438
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.755071003436004,
+      "learning_rate": 5.461760599293855e-06,
+      "loss": 0.9874,
+      "step": 3439
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6676292431200396,
+      "learning_rate": 5.456209227321643e-06,
+      "loss": 0.9835,
+      "step": 3440
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7172107627852051,
+      "learning_rate": 5.450659619405458e-06,
+      "loss": 0.9619,
+      "step": 3441
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6891406243482281,
+      "learning_rate": 5.445111777699842e-06,
+      "loss": 1.0036,
+      "step": 3442
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7656671019970316,
+      "learning_rate": 5.439565704358667e-06,
+      "loss": 0.9327,
+      "step": 3443
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7330061513022661,
+      "learning_rate": 5.434021401535105e-06,
+      "loss": 1.0278,
+      "step": 3444
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6810533477424304,
+      "learning_rate": 5.428478871381646e-06,
+      "loss": 0.9924,
+      "step": 3445
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7472070641240209,
+      "learning_rate": 5.422938116050092e-06,
+      "loss": 1.0048,
+      "step": 3446
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7380365505302401,
+      "learning_rate": 5.417399137691552e-06,
+      "loss": 1.0164,
+      "step": 3447
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6615045005565193,
+      "learning_rate": 5.411861938456453e-06,
+      "loss": 0.9417,
+      "step": 3448
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.697060085443091,
+      "learning_rate": 5.406326520494522e-06,
+      "loss": 0.9258,
+      "step": 3449
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6527428055443447,
+      "learning_rate": 5.400792885954802e-06,
+      "loss": 1.0641,
+      "step": 3450
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6704540737480477,
+      "learning_rate": 5.395261036985635e-06,
+      "loss": 0.9878,
+      "step": 3451
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6986621015962543,
+      "learning_rate": 5.389730975734686e-06,
+      "loss": 0.9938,
+      "step": 3452
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.67798716491922,
+      "learning_rate": 5.384202704348902e-06,
+      "loss": 1.0355,
+      "step": 3453
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6725342531513401,
+      "learning_rate": 5.378676224974557e-06,
+      "loss": 1.0375,
+      "step": 3454
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6993798248374231,
+      "learning_rate": 5.373151539757224e-06,
+      "loss": 0.9829,
+      "step": 3455
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7433964901191773,
+      "learning_rate": 5.367628650841761e-06,
+      "loss": 1.0046,
+      "step": 3456
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6682153453407754,
+      "learning_rate": 5.362107560372358e-06,
+      "loss": 1.0748,
+      "step": 3457
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.522068128160466,
+      "learning_rate": 5.356588270492487e-06,
+      "loss": 1.112,
+      "step": 3458
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5089282984319022,
+      "learning_rate": 5.351070783344926e-06,
+      "loss": 1.0129,
+      "step": 3459
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6765006037660617,
+      "learning_rate": 5.3455551010717545e-06,
+      "loss": 0.9987,
+      "step": 3460
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6824105551809044,
+      "learning_rate": 5.34004122581435e-06,
+      "loss": 1.0063,
+      "step": 3461
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.722599422664567,
+      "learning_rate": 5.334529159713389e-06,
+      "loss": 0.9774,
+      "step": 3462
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6478820240571177,
+      "learning_rate": 5.329018904908841e-06,
+      "loss": 0.9908,
+      "step": 3463
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6763257927676875,
+      "learning_rate": 5.323510463539989e-06,
+      "loss": 1.056,
+      "step": 3464
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6152237293590908,
+      "learning_rate": 5.318003837745382e-06,
+      "loss": 0.9602,
+      "step": 3465
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7479480740130711,
+      "learning_rate": 5.3124990296628974e-06,
+      "loss": 1.0046,
+      "step": 3466
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7194004838122752,
+      "learning_rate": 5.306996041429688e-06,
+      "loss": 1.0053,
+      "step": 3467
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6960771144351495,
+      "learning_rate": 5.301494875182192e-06,
+      "loss": 0.9905,
+      "step": 3468
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6811316799353592,
+      "learning_rate": 5.295995533056162e-06,
+      "loss": 1.0311,
+      "step": 3469
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7209684964089388,
+      "learning_rate": 5.290498017186631e-06,
+      "loss": 0.9199,
+      "step": 3470
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7234872871543844,
+      "learning_rate": 5.2850023297079235e-06,
+      "loss": 1.0537,
+      "step": 3471
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6965897556205899,
+      "learning_rate": 5.279508472753654e-06,
+      "loss": 0.9985,
+      "step": 3472
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6625532223140335,
+      "learning_rate": 5.274016448456725e-06,
+      "loss": 0.9948,
+      "step": 3473
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6554382281812451,
+      "learning_rate": 5.2685262589493314e-06,
+      "loss": 1.0088,
+      "step": 3474
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6893581094545993,
+      "learning_rate": 5.263037906362953e-06,
+      "loss": 0.9768,
+      "step": 3475
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7104551460947008,
+      "learning_rate": 5.257551392828359e-06,
+      "loss": 1.0102,
+      "step": 3476
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.582065098066592,
+      "learning_rate": 5.252066720475597e-06,
+      "loss": 1.0532,
+      "step": 3477
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6692836839965847,
+      "learning_rate": 5.246583891434018e-06,
+      "loss": 1.075,
+      "step": 3478
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6750401398979895,
+      "learning_rate": 5.241102907832232e-06,
+      "loss": 1.0896,
+      "step": 3479
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7350112026302745,
+      "learning_rate": 5.235623771798151e-06,
+      "loss": 1.011,
+      "step": 3480
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6781799162450315,
+      "learning_rate": 5.23014648545897e-06,
+      "loss": 1.0472,
+      "step": 3481
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6276571417396732,
+      "learning_rate": 5.224671050941146e-06,
+      "loss": 0.98,
+      "step": 3482
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7478433990902716,
+      "learning_rate": 5.2191974703704425e-06,
+      "loss": 0.9802,
+      "step": 3483
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.705845252031138,
+      "learning_rate": 5.213725745871889e-06,
+      "loss": 0.9911,
+      "step": 3484
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5276677870400116,
+      "learning_rate": 5.208255879569799e-06,
+      "loss": 0.9656,
+      "step": 3485
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7049070528278721,
+      "learning_rate": 5.20278787358776e-06,
+      "loss": 1.0569,
+      "step": 3486
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6677411826945,
+      "learning_rate": 5.197321730048641e-06,
+      "loss": 1.0472,
+      "step": 3487
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6767607902184276,
+      "learning_rate": 5.1918574510745865e-06,
+      "loss": 0.9403,
+      "step": 3488
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6819769654367783,
+      "learning_rate": 5.186395038787017e-06,
+      "loss": 0.9937,
+      "step": 3489
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7177520126548478,
+      "learning_rate": 5.180934495306638e-06,
+      "loss": 1.0334,
+      "step": 3490
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5523504113000391,
+      "learning_rate": 5.175475822753404e-06,
+      "loss": 1.0596,
+      "step": 3491
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7617061891121318,
+      "learning_rate": 5.170019023246574e-06,
+      "loss": 0.8973,
+      "step": 3492
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6839558293040676,
+      "learning_rate": 5.16456409890466e-06,
+      "loss": 1.0314,
+      "step": 3493
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6518333652795956,
+      "learning_rate": 5.159111051845451e-06,
+      "loss": 1.0501,
+      "step": 3494
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5759010713237691,
+      "learning_rate": 5.153659884186013e-06,
+      "loss": 1.0006,
+      "step": 3495
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7079466848106257,
+      "learning_rate": 5.148210598042665e-06,
+      "loss": 0.9794,
+      "step": 3496
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7652469470788226,
+      "learning_rate": 5.142763195531017e-06,
+      "loss": 1.0396,
+      "step": 3497
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8273947618196726,
+      "learning_rate": 5.137317678765939e-06,
+      "loss": 1.0196,
+      "step": 3498
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6992924207824058,
+      "learning_rate": 5.131874049861563e-06,
+      "loss": 1.0193,
+      "step": 3499
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7562626117449582,
+      "learning_rate": 5.126432310931295e-06,
+      "loss": 1.0175,
+      "step": 3500
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6925847238019572,
+      "learning_rate": 5.120992464087807e-06,
+      "loss": 1.0776,
+      "step": 3501
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5483901575543051,
+      "learning_rate": 5.115554511443033e-06,
+      "loss": 0.996,
+      "step": 3502
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7421547059820209,
+      "learning_rate": 5.1101184551081705e-06,
+      "loss": 0.9508,
+      "step": 3503
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.749077175307708,
+      "learning_rate": 5.104684297193694e-06,
+      "loss": 1.0131,
+      "step": 3504
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5826362717292622,
+      "learning_rate": 5.099252039809317e-06,
+      "loss": 1.0262,
+      "step": 3505
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.8283499901248375,
+      "learning_rate": 5.09382168506404e-06,
+      "loss": 1.0135,
+      "step": 3506
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6843061736488347,
+      "learning_rate": 5.088393235066114e-06,
+      "loss": 0.9919,
+      "step": 3507
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7480045423377633,
+      "learning_rate": 5.082966691923037e-06,
+      "loss": 1.0487,
+      "step": 3508
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.733536848129017,
+      "learning_rate": 5.077542057741592e-06,
+      "loss": 1.0065,
+      "step": 3509
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6976924795932766,
+      "learning_rate": 5.0721193346278066e-06,
+      "loss": 0.9629,
+      "step": 3510
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6562416731666926,
+      "learning_rate": 5.066698524686966e-06,
+      "loss": 0.9622,
+      "step": 3511
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7098280847152029,
+      "learning_rate": 5.061279630023618e-06,
+      "loss": 1.0049,
+      "step": 3512
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6995582511228816,
+      "learning_rate": 5.055862652741562e-06,
+      "loss": 1.0707,
+      "step": 3513
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7160068880029021,
+      "learning_rate": 5.050447594943856e-06,
+      "loss": 0.9886,
+      "step": 3514
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6623367761514074,
+      "learning_rate": 5.045034458732808e-06,
+      "loss": 0.9794,
+      "step": 3515
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7465983671169042,
+      "learning_rate": 5.0396232462099945e-06,
+      "loss": 0.9735,
+      "step": 3516
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.673685800017378,
+      "learning_rate": 5.034213959476222e-06,
+      "loss": 0.9575,
+      "step": 3517
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5245871270498862,
+      "learning_rate": 5.028806600631569e-06,
+      "loss": 1.013,
+      "step": 3518
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7699136053542786,
+      "learning_rate": 5.023401171775357e-06,
+      "loss": 0.9871,
+      "step": 3519
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6734140173257435,
+      "learning_rate": 5.017997675006161e-06,
+      "loss": 1.0818,
+      "step": 3520
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6744236705047184,
+      "learning_rate": 5.012596112421806e-06,
+      "loss": 1.001,
+      "step": 3521
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6563269031220615,
+      "learning_rate": 5.007196486119355e-06,
+      "loss": 0.9965,
+      "step": 3522
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8309960660483239,
+      "learning_rate": 5.001798798195136e-06,
+      "loss": 1.0301,
+      "step": 3523
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.697257539418082,
+      "learning_rate": 4.996403050744719e-06,
+      "loss": 1.1138,
+      "step": 3524
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7574938564061345,
+      "learning_rate": 4.991009245862917e-06,
+      "loss": 1.0412,
+      "step": 3525
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7318829493330837,
+      "learning_rate": 4.985617385643789e-06,
+      "loss": 0.9712,
+      "step": 3526
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5318052570796813,
+      "learning_rate": 4.980227472180643e-06,
+      "loss": 1.0751,
+      "step": 3527
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.726768341863207,
+      "learning_rate": 4.974839507566027e-06,
+      "loss": 1.0125,
+      "step": 3528
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7159292487061466,
+      "learning_rate": 4.969453493891733e-06,
+      "loss": 1.0859,
+      "step": 3529
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6857635340124313,
+      "learning_rate": 4.9640694332488075e-06,
+      "loss": 1.01,
+      "step": 3530
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7111152174490596,
+      "learning_rate": 4.958687327727511e-06,
+      "loss": 0.982,
+      "step": 3531
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7542097843625565,
+      "learning_rate": 4.953307179417376e-06,
+      "loss": 1.0043,
+      "step": 3532
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7707388167813115,
+      "learning_rate": 4.947928990407156e-06,
+      "loss": 1.083,
+      "step": 3533
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6739274104761501,
+      "learning_rate": 4.94255276278485e-06,
+      "loss": 0.9301,
+      "step": 3534
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7058350652792179,
+      "learning_rate": 4.937178498637696e-06,
+      "loss": 0.9758,
+      "step": 3535
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.690107160259968,
+      "learning_rate": 4.931806200052165e-06,
+      "loss": 1.022,
+      "step": 3536
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6909588548683578,
+      "learning_rate": 4.926435869113971e-06,
+      "loss": 1.0242,
+      "step": 3537
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7215290703147514,
+      "learning_rate": 4.92106750790806e-06,
+      "loss": 1.0231,
+      "step": 3538
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7401525804909878,
+      "learning_rate": 4.915701118518616e-06,
+      "loss": 0.9894,
+      "step": 3539
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.566926665600665,
+      "learning_rate": 4.910336703029055e-06,
+      "loss": 0.9996,
+      "step": 3540
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4922891016476344,
+      "learning_rate": 4.904974263522025e-06,
+      "loss": 0.9945,
+      "step": 3541
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7327859799413466,
+      "learning_rate": 4.899613802079419e-06,
+      "loss": 1.039,
+      "step": 3542
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6514824673745047,
+      "learning_rate": 4.8942553207823395e-06,
+      "loss": 1.0568,
+      "step": 3543
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6900345303543477,
+      "learning_rate": 4.888898821711144e-06,
+      "loss": 1.0182,
+      "step": 3544
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7586886801506428,
+      "learning_rate": 4.883544306945407e-06,
+      "loss": 0.9741,
+      "step": 3545
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7249598234449235,
+      "learning_rate": 4.878191778563934e-06,
+      "loss": 0.9872,
+      "step": 3546
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6604032452660689,
+      "learning_rate": 4.872841238644766e-06,
+      "loss": 1.0588,
+      "step": 3547
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7462138366517018,
+      "learning_rate": 4.867492689265154e-06,
+      "loss": 0.9896,
+      "step": 3548
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5658287871427904,
+      "learning_rate": 4.8621461325016015e-06,
+      "loss": 0.9979,
+      "step": 3549
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5006929857320251,
+      "learning_rate": 4.856801570429822e-06,
+      "loss": 1.0587,
+      "step": 3550
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7263055388578172,
+      "learning_rate": 4.851459005124759e-06,
+      "loss": 1.0678,
+      "step": 3551
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7002812802089103,
+      "learning_rate": 4.846118438660578e-06,
+      "loss": 1.082,
+      "step": 3552
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4838644407404475,
+      "learning_rate": 4.840779873110675e-06,
+      "loss": 1.0522,
+      "step": 3553
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.738258508118523,
+      "learning_rate": 4.83544331054766e-06,
+      "loss": 1.032,
+      "step": 3554
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5331695022142504,
+      "learning_rate": 4.83010875304337e-06,
+      "loss": 1.0379,
+      "step": 3555
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5502330350996123,
+      "learning_rate": 4.824776202668875e-06,
+      "loss": 0.9987,
+      "step": 3556
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6666320008087347,
+      "learning_rate": 4.819445661494437e-06,
+      "loss": 0.9944,
+      "step": 3557
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7406971165690485,
+      "learning_rate": 4.8141171315895694e-06,
+      "loss": 1.0514,
+      "step": 3558
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.65403590611579,
+      "learning_rate": 4.808790615022987e-06,
+      "loss": 1.004,
+      "step": 3559
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5723542883939616,
+      "learning_rate": 4.803466113862626e-06,
+      "loss": 0.9698,
+      "step": 3560
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7327314018766137,
+      "learning_rate": 4.798143630175642e-06,
+      "loss": 1.0821,
+      "step": 3561
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6616718883080016,
+      "learning_rate": 4.792823166028405e-06,
+      "loss": 0.9808,
+      "step": 3562
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7081815826704948,
+      "learning_rate": 4.787504723486505e-06,
+      "loss": 0.9274,
+      "step": 3563
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6993727624251341,
+      "learning_rate": 4.7821883046147414e-06,
+      "loss": 1.0562,
+      "step": 3564
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6022207832958608,
+      "learning_rate": 4.776873911477133e-06,
+      "loss": 1.0134,
+      "step": 3565
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7490416705804358,
+      "learning_rate": 4.771561546136908e-06,
+      "loss": 0.9524,
+      "step": 3566
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7168558338225287,
+      "learning_rate": 4.766251210656509e-06,
+      "loss": 1.0435,
+      "step": 3567
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7280657641459858,
+      "learning_rate": 4.760942907097601e-06,
+      "loss": 1.071,
+      "step": 3568
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7660550399870347,
+      "learning_rate": 4.755636637521035e-06,
+      "loss": 1.0041,
+      "step": 3569
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7704409628678288,
+      "learning_rate": 4.750332403986902e-06,
+      "loss": 0.9905,
+      "step": 3570
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5572596129734135,
+      "learning_rate": 4.7450302085544735e-06,
+      "loss": 1.0449,
+      "step": 3571
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6780825267677658,
+      "learning_rate": 4.739730053282255e-06,
+      "loss": 0.9669,
+      "step": 3572
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6579345954318435,
+      "learning_rate": 4.734431940227951e-06,
+      "loss": 0.9333,
+      "step": 3573
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6937706495356368,
+      "learning_rate": 4.7291358714484594e-06,
+      "loss": 1.0255,
+      "step": 3574
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7259258964369989,
+      "learning_rate": 4.723841848999907e-06,
+      "loss": 1.0005,
+      "step": 3575
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6671428721700202,
+      "learning_rate": 4.718549874937612e-06,
+      "loss": 0.9675,
+      "step": 3576
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7139568975471771,
+      "learning_rate": 4.713259951316103e-06,
+      "loss": 0.9918,
+      "step": 3577
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7553596900268956,
+      "learning_rate": 4.707972080189106e-06,
+      "loss": 0.975,
+      "step": 3578
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6951506545189178,
+      "learning_rate": 4.702686263609559e-06,
+      "loss": 1.0209,
+      "step": 3579
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6875003841963985,
+      "learning_rate": 4.697402503629596e-06,
+      "loss": 0.9567,
+      "step": 3580
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7694463237635863,
+      "learning_rate": 4.69212080230055e-06,
+      "loss": 1.0341,
+      "step": 3581
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7601021709590462,
+      "learning_rate": 4.686841161672974e-06,
+      "loss": 1.0272,
+      "step": 3582
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7669683664653477,
+      "learning_rate": 4.681563583796587e-06,
+      "loss": 0.9915,
+      "step": 3583
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6910139622624584,
+      "learning_rate": 4.67628807072034e-06,
+      "loss": 1.0137,
+      "step": 3584
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.682740582048147,
+      "learning_rate": 4.6710146244923645e-06,
+      "loss": 1.0566,
+      "step": 3585
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.74869507412423,
+      "learning_rate": 4.665743247159995e-06,
+      "loss": 1.0314,
+      "step": 3586
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6556752266936166,
+      "learning_rate": 4.660473940769761e-06,
+      "loss": 0.9608,
+      "step": 3587
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6434745532547088,
+      "learning_rate": 4.655206707367388e-06,
+      "loss": 1.0664,
+      "step": 3588
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7202386943796052,
+      "learning_rate": 4.649941548997797e-06,
+      "loss": 0.9456,
+      "step": 3589
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6292189179873506,
+      "learning_rate": 4.644678467705101e-06,
+      "loss": 0.9171,
+      "step": 3590
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7237865383932671,
+      "learning_rate": 4.639417465532622e-06,
+      "loss": 0.9586,
+      "step": 3591
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7091930072703635,
+      "learning_rate": 4.634158544522849e-06,
+      "loss": 0.9892,
+      "step": 3592
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7569887934397731,
+      "learning_rate": 4.628901706717476e-06,
+      "loss": 1.1639,
+      "step": 3593
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7374277397096882,
+      "learning_rate": 4.623646954157399e-06,
+      "loss": 0.9515,
+      "step": 3594
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6973829724701593,
+      "learning_rate": 4.618394288882681e-06,
+      "loss": 1.0133,
+      "step": 3595
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6774989458040039,
+      "learning_rate": 4.613143712932603e-06,
+      "loss": 0.9635,
+      "step": 3596
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5381828694680166,
+      "learning_rate": 4.607895228345603e-06,
+      "loss": 0.954,
+      "step": 3597
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6829819544643628,
+      "learning_rate": 4.602648837159333e-06,
+      "loss": 0.9375,
+      "step": 3598
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6562810178764917,
+      "learning_rate": 4.597404541410622e-06,
+      "loss": 0.9408,
+      "step": 3599
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5611654746223439,
+      "learning_rate": 4.592162343135483e-06,
+      "loss": 0.985,
+      "step": 3600
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.680266061438235,
+      "learning_rate": 4.586922244369122e-06,
+      "loss": 0.9974,
+      "step": 3601
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7668156941605602,
+      "learning_rate": 4.5816842471459224e-06,
+      "loss": 0.965,
+      "step": 3602
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7008176528358645,
+      "learning_rate": 4.576448353499457e-06,
+      "loss": 1.0091,
+      "step": 3603
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6769736568448869,
+      "learning_rate": 4.571214565462477e-06,
+      "loss": 0.9604,
+      "step": 3604
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7681970023723068,
+      "learning_rate": 4.565982885066923e-06,
+      "loss": 0.944,
+      "step": 3605
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5897750026203415,
+      "learning_rate": 4.560753314343912e-06,
+      "loss": 0.9761,
+      "step": 3606
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7177448211367081,
+      "learning_rate": 4.555525855323738e-06,
+      "loss": 1.0805,
+      "step": 3607
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7045897521691488,
+      "learning_rate": 4.5503005100358945e-06,
+      "loss": 1.0292,
+      "step": 3608
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.715979946306504,
+      "learning_rate": 4.545077280509022e-06,
+      "loss": 0.9472,
+      "step": 3609
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7366975533763862,
+      "learning_rate": 4.539856168770974e-06,
+      "loss": 1.0294,
+      "step": 3610
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7076648739533403,
+      "learning_rate": 4.534637176848758e-06,
+      "loss": 0.9962,
+      "step": 3611
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6073066881002027,
+      "learning_rate": 4.52942030676857e-06,
+      "loss": 1.0938,
+      "step": 3612
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7489841275220059,
+      "learning_rate": 4.524205560555774e-06,
+      "loss": 0.9402,
+      "step": 3613
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8300256276713525,
+      "learning_rate": 4.5189929402349175e-06,
+      "loss": 1.0031,
+      "step": 3614
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7038402133596386,
+      "learning_rate": 4.513782447829717e-06,
+      "loss": 1.0182,
+      "step": 3615
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.580764845527028,
+      "learning_rate": 4.508574085363065e-06,
+      "loss": 1.056,
+      "step": 3616
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6625642278052529,
+      "learning_rate": 4.503367854857035e-06,
+      "loss": 1.0198,
+      "step": 3617
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7358548053543208,
+      "learning_rate": 4.498163758332853e-06,
+      "loss": 1.0057,
+      "step": 3618
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7022022490683144,
+      "learning_rate": 4.492961797810932e-06,
+      "loss": 0.9758,
+      "step": 3619
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6747261577728421,
+      "learning_rate": 4.4877619753108605e-06,
+      "loss": 1.0476,
+      "step": 3620
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7269296714136069,
+      "learning_rate": 4.4825642928513746e-06,
+      "loss": 1.011,
+      "step": 3621
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6872066006467572,
+      "learning_rate": 4.477368752450409e-06,
+      "loss": 0.9973,
+      "step": 3622
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.714551515605922,
+      "learning_rate": 4.472175356125036e-06,
+      "loss": 1.0009,
+      "step": 3623
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6828742027636361,
+      "learning_rate": 4.466984105891521e-06,
+      "loss": 0.9996,
+      "step": 3624
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7427330652665075,
+      "learning_rate": 4.461795003765285e-06,
+      "loss": 1.0478,
+      "step": 3625
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7609589122240593,
+      "learning_rate": 4.456608051760914e-06,
+      "loss": 1.0086,
+      "step": 3626
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6812375219598787,
+      "learning_rate": 4.45142325189216e-06,
+      "loss": 0.9822,
+      "step": 3627
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7438780306693864,
+      "learning_rate": 4.446240606171945e-06,
+      "loss": 1.0308,
+      "step": 3628
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.738486223532834,
+      "learning_rate": 4.4410601166123475e-06,
+      "loss": 1.0386,
+      "step": 3629
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6954306336027196,
+      "learning_rate": 4.4358817852246124e-06,
+      "loss": 0.9776,
+      "step": 3630
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.659279884837092,
+      "learning_rate": 4.430705614019147e-06,
+      "loss": 0.9876,
+      "step": 3631
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6560524017966476,
+      "learning_rate": 4.425531605005519e-06,
+      "loss": 1.0685,
+      "step": 3632
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6808517152163238,
+      "learning_rate": 4.420359760192452e-06,
+      "loss": 1.0072,
+      "step": 3633
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6475652720297692,
+      "learning_rate": 4.4151900815878455e-06,
+      "loss": 0.9539,
+      "step": 3634
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6969777581620836,
+      "learning_rate": 4.410022571198734e-06,
+      "loss": 1.0091,
+      "step": 3635
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7134794746813319,
+      "learning_rate": 4.404857231031332e-06,
+      "loss": 1.0413,
+      "step": 3636
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.68261299846453,
+      "learning_rate": 4.399694063090999e-06,
+      "loss": 1.0447,
+      "step": 3637
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6233348234589379,
+      "learning_rate": 4.394533069382255e-06,
+      "loss": 0.9696,
+      "step": 3638
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7103265147626071,
+      "learning_rate": 4.3893742519087754e-06,
+      "loss": 1.0707,
+      "step": 3639
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5830797998799375,
+      "learning_rate": 4.3842176126733914e-06,
+      "loss": 1.0346,
+      "step": 3640
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7192226155503479,
+      "learning_rate": 4.379063153678087e-06,
+      "loss": 0.9651,
+      "step": 3641
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6786553294068518,
+      "learning_rate": 4.373910876923997e-06,
+      "loss": 1.0004,
+      "step": 3642
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7247166970466709,
+      "learning_rate": 4.368760784411423e-06,
+      "loss": 0.9652,
+      "step": 3643
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7135452277938712,
+      "learning_rate": 4.363612878139799e-06,
+      "loss": 1.0035,
+      "step": 3644
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7172847108748767,
+      "learning_rate": 4.3584671601077224e-06,
+      "loss": 1.0316,
+      "step": 3645
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7355847040448801,
+      "learning_rate": 4.353323632312938e-06,
+      "loss": 1.0363,
+      "step": 3646
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6730706101719125,
+      "learning_rate": 4.348182296752336e-06,
+      "loss": 1.0251,
+      "step": 3647
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8061580051678662,
+      "learning_rate": 4.343043155421971e-06,
+      "loss": 0.991,
+      "step": 3648
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8075693388160825,
+      "learning_rate": 4.3379062103170214e-06,
+      "loss": 0.9641,
+      "step": 3649
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.812902354761616,
+      "learning_rate": 4.332771463431837e-06,
+      "loss": 0.9359,
+      "step": 3650
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7183782348800838,
+      "learning_rate": 4.327638916759898e-06,
+      "loss": 1.0079,
+      "step": 3651
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6952870441399567,
+      "learning_rate": 4.322508572293836e-06,
+      "loss": 1.0133,
+      "step": 3652
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7443550412686207,
+      "learning_rate": 4.317380432025428e-06,
+      "loss": 1.0897,
+      "step": 3653
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7219388461414555,
+      "learning_rate": 4.312254497945595e-06,
+      "loss": 1.0562,
+      "step": 3654
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6828278806398872,
+      "learning_rate": 4.3071307720444015e-06,
+      "loss": 0.9688,
+      "step": 3655
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7347302747123163,
+      "learning_rate": 4.3020092563110485e-06,
+      "loss": 0.9841,
+      "step": 3656
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7988217242296614,
+      "learning_rate": 4.2968899527338984e-06,
+      "loss": 0.9979,
+      "step": 3657
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6806740189743767,
+      "learning_rate": 4.291772863300428e-06,
+      "loss": 0.9955,
+      "step": 3658
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8045561991045472,
+      "learning_rate": 4.2866579899972686e-06,
+      "loss": 0.9642,
+      "step": 3659
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7146042971440091,
+      "learning_rate": 4.281545334810201e-06,
+      "loss": 1.0514,
+      "step": 3660
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.8019841845279553,
+      "learning_rate": 4.276434899724119e-06,
+      "loss": 1.0039,
+      "step": 3661
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.763695829820284,
+      "learning_rate": 4.27132668672308e-06,
+      "loss": 0.9815,
+      "step": 3662
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6937615430336292,
+      "learning_rate": 4.266220697790266e-06,
+      "loss": 1.0217,
+      "step": 3663
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7131566469906305,
+      "learning_rate": 4.2611169349079985e-06,
+      "loss": 0.9862,
+      "step": 3664
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7117343234129718,
+      "learning_rate": 4.25601540005773e-06,
+      "loss": 1.0021,
+      "step": 3665
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6973933916139259,
+      "learning_rate": 4.250916095220056e-06,
+      "loss": 1.0077,
+      "step": 3666
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6968342929581836,
+      "learning_rate": 4.2458190223747e-06,
+      "loss": 0.9489,
+      "step": 3667
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7107830763343569,
+      "learning_rate": 4.240724183500518e-06,
+      "loss": 0.9865,
+      "step": 3668
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6657778254434593,
+      "learning_rate": 4.2356315805755135e-06,
+      "loss": 0.9822,
+      "step": 3669
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7500112238776381,
+      "learning_rate": 4.230541215576798e-06,
+      "loss": 1.0226,
+      "step": 3670
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7008254718580139,
+      "learning_rate": 4.225453090480631e-06,
+      "loss": 0.9974,
+      "step": 3671
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6919449333715475,
+      "learning_rate": 4.220367207262398e-06,
+      "loss": 1.0466,
+      "step": 3672
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.34289759307902146,
+      "learning_rate": 4.21528356789661e-06,
+      "loss": 1.0953,
+      "step": 3673
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7132531977420846,
+      "learning_rate": 4.210202174356922e-06,
+      "loss": 0.9935,
+      "step": 3674
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6296505739361508,
+      "learning_rate": 4.20512302861609e-06,
+      "loss": 1.0667,
+      "step": 3675
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8175776067441227,
+      "learning_rate": 4.2000461326460274e-06,
+      "loss": 1.0613,
+      "step": 3676
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7007218443570233,
+      "learning_rate": 4.194971488417753e-06,
+      "loss": 1.0152,
+      "step": 3677
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6813617855992131,
+      "learning_rate": 4.189899097901421e-06,
+      "loss": 1.0079,
+      "step": 3678
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6375211491846127,
+      "learning_rate": 4.184828963066305e-06,
+      "loss": 1.0123,
+      "step": 3679
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7682023230087113,
+      "learning_rate": 4.179761085880809e-06,
+      "loss": 0.9695,
+      "step": 3680
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6122158780955724,
+      "learning_rate": 4.174695468312456e-06,
+      "loss": 0.9748,
+      "step": 3681
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7216987075705249,
+      "learning_rate": 4.16963211232789e-06,
+      "loss": 0.9703,
+      "step": 3682
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7397817624476335,
+      "learning_rate": 4.16457101989289e-06,
+      "loss": 1.0004,
+      "step": 3683
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.746169035193071,
+      "learning_rate": 4.159512192972337e-06,
+      "loss": 0.9979,
+      "step": 3684
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5537321729761446,
+      "learning_rate": 4.15445563353024e-06,
+      "loss": 0.9918,
+      "step": 3685
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5564275393169168,
+      "learning_rate": 4.149401343529742e-06,
+      "loss": 1.0329,
+      "step": 3686
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.661205887291067,
+      "learning_rate": 4.144349324933077e-06,
+      "loss": 0.9728,
+      "step": 3687
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7274996561565894,
+      "learning_rate": 4.139299579701623e-06,
+      "loss": 0.9374,
+      "step": 3688
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5132119804427485,
+      "learning_rate": 4.134252109795863e-06,
+      "loss": 0.9484,
+      "step": 3689
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7143978156153326,
+      "learning_rate": 4.129206917175397e-06,
+      "loss": 0.986,
+      "step": 3690
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7380349875333174,
+      "learning_rate": 4.124164003798944e-06,
+      "loss": 0.9857,
+      "step": 3691
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7088240183989502,
+      "learning_rate": 4.119123371624335e-06,
+      "loss": 0.961,
+      "step": 3692
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7694522721762382,
+      "learning_rate": 4.114085022608517e-06,
+      "loss": 0.9664,
+      "step": 3693
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6301662524853905,
+      "learning_rate": 4.109048958707552e-06,
+      "loss": 1.0914,
+      "step": 3694
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6511016511025979,
+      "learning_rate": 4.104015181876613e-06,
+      "loss": 1.0547,
+      "step": 3695
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6885895891609856,
+      "learning_rate": 4.09898369406998e-06,
+      "loss": 1.0242,
+      "step": 3696
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.625497306201267,
+      "learning_rate": 4.0939544972410636e-06,
+      "loss": 1.0244,
+      "step": 3697
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7373498392979961,
+      "learning_rate": 4.0889275933423576e-06,
+      "loss": 1.0716,
+      "step": 3698
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6704027386111256,
+      "learning_rate": 4.0839029843254815e-06,
+      "loss": 1.0217,
+      "step": 3699
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7446706795504302,
+      "learning_rate": 4.078880672141171e-06,
+      "loss": 1.0208,
+      "step": 3700
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6567094311711646,
+      "learning_rate": 4.073860658739246e-06,
+      "loss": 1.0417,
+      "step": 3701
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7152144801028606,
+      "learning_rate": 4.068842946068661e-06,
+      "loss": 1.0313,
+      "step": 3702
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7042922897887437,
+      "learning_rate": 4.063827536077459e-06,
+      "loss": 1.0071,
+      "step": 3703
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7319262908314311,
+      "learning_rate": 4.058814430712796e-06,
+      "loss": 1.0461,
+      "step": 3704
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7565634849278138,
+      "learning_rate": 4.0538036319209325e-06,
+      "loss": 1.0908,
+      "step": 3705
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7113939287519501,
+      "learning_rate": 4.0487951416472324e-06,
+      "loss": 0.9203,
+      "step": 3706
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6971664806127201,
+      "learning_rate": 4.043788961836164e-06,
+      "loss": 1.0457,
+      "step": 3707
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7752950090892409,
+      "learning_rate": 4.038785094431295e-06,
+      "loss": 0.9935,
+      "step": 3708
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4978024195824826,
+      "learning_rate": 4.0337835413753116e-06,
+      "loss": 1.0016,
+      "step": 3709
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7138754440385064,
+      "learning_rate": 4.0287843046099765e-06,
+      "loss": 0.9831,
+      "step": 3710
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8157796991350923,
+      "learning_rate": 4.0237873860761645e-06,
+      "loss": 0.9183,
+      "step": 3711
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.713493010951712,
+      "learning_rate": 4.018792787713865e-06,
+      "loss": 1.0085,
+      "step": 3712
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6665712723567794,
+      "learning_rate": 4.013800511462135e-06,
+      "loss": 1.1067,
+      "step": 3713
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6966049573342664,
+      "learning_rate": 4.008810559259162e-06,
+      "loss": 0.9905,
+      "step": 3714
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7044370479508854,
+      "learning_rate": 4.003822933042213e-06,
+      "loss": 1.0079,
+      "step": 3715
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8029004499258044,
+      "learning_rate": 3.998837634747655e-06,
+      "loss": 0.9976,
+      "step": 3716
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7591990444113169,
+      "learning_rate": 3.993854666310955e-06,
+      "loss": 0.999,
+      "step": 3717
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6965118688568986,
+      "learning_rate": 3.98887402966667e-06,
+      "loss": 1.0005,
+      "step": 3718
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6620603558881324,
+      "learning_rate": 3.983895726748455e-06,
+      "loss": 0.9846,
+      "step": 3719
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6886917401306477,
+      "learning_rate": 3.97891975948906e-06,
+      "loss": 1.0131,
+      "step": 3720
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7328408143913073,
+      "learning_rate": 3.973946129820326e-06,
+      "loss": 1.0391,
+      "step": 3721
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7809614042474812,
+      "learning_rate": 3.968974839673186e-06,
+      "loss": 1.0186,
+      "step": 3722
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.67054432103886,
+      "learning_rate": 3.964005890977672e-06,
+      "loss": 1.0591,
+      "step": 3723
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.667746756373545,
+      "learning_rate": 3.9590392856628946e-06,
+      "loss": 0.9434,
+      "step": 3724
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7634003967228591,
+      "learning_rate": 3.954075025657058e-06,
+      "loss": 1.0764,
+      "step": 3725
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5650172875391584,
+      "learning_rate": 3.949113112887471e-06,
+      "loss": 1.0522,
+      "step": 3726
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7216725144832473,
+      "learning_rate": 3.944153549280506e-06,
+      "loss": 0.9929,
+      "step": 3727
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6337332442107844,
+      "learning_rate": 3.939196336761645e-06,
+      "loss": 0.9531,
+      "step": 3728
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6900083990501733,
+      "learning_rate": 3.934241477255445e-06,
+      "loss": 0.9722,
+      "step": 3729
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6824363173526004,
+      "learning_rate": 3.929288972685555e-06,
+      "loss": 0.996,
+      "step": 3730
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.678861680923812,
+      "learning_rate": 3.924338824974705e-06,
+      "loss": 0.9888,
+      "step": 3731
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6752951034051481,
+      "learning_rate": 3.919391036044715e-06,
+      "loss": 1.0359,
+      "step": 3732
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5456550775448756,
+      "learning_rate": 3.914445607816486e-06,
+      "loss": 1.0397,
+      "step": 3733
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7035969425208023,
+      "learning_rate": 3.909502542210001e-06,
+      "loss": 0.9943,
+      "step": 3734
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.567619260641649,
+      "learning_rate": 3.904561841144338e-06,
+      "loss": 1.0569,
+      "step": 3735
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8509443901283404,
+      "learning_rate": 3.899623506537635e-06,
+      "loss": 1.0177,
+      "step": 3736
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6970211942852913,
+      "learning_rate": 3.894687540307127e-06,
+      "loss": 0.993,
+      "step": 3737
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.8283846995092723,
+      "learning_rate": 3.8897539443691355e-06,
+      "loss": 1.0178,
+      "step": 3738
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7674713990198891,
+      "learning_rate": 3.884822720639036e-06,
+      "loss": 0.9791,
+      "step": 3739
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7876041555091876,
+      "learning_rate": 3.879893871031314e-06,
+      "loss": 0.9906,
+      "step": 3740
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7504913694669855,
+      "learning_rate": 3.874967397459511e-06,
+      "loss": 1.0104,
+      "step": 3741
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5942781026857323,
+      "learning_rate": 3.870043301836256e-06,
+      "loss": 1.0554,
+      "step": 3742
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3512941191791753,
+      "learning_rate": 3.86512158607325e-06,
+      "loss": 1.0978,
+      "step": 3743
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6831360181694436,
+      "learning_rate": 3.860202252081276e-06,
+      "loss": 1.0727,
+      "step": 3744
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.704816753700335,
+      "learning_rate": 3.855285301770188e-06,
+      "loss": 1.0107,
+      "step": 3745
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7168997133598803,
+      "learning_rate": 3.850370737048913e-06,
+      "loss": 0.9549,
+      "step": 3746
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6830781029365008,
+      "learning_rate": 3.8454585598254565e-06,
+      "loss": 0.909,
+      "step": 3747
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7409001275107512,
+      "learning_rate": 3.840548772006891e-06,
+      "loss": 1.0531,
+      "step": 3748
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7475521942217164,
+      "learning_rate": 3.835641375499375e-06,
+      "loss": 1.0168,
+      "step": 3749
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.683873860549624,
+      "learning_rate": 3.830736372208118e-06,
+      "loss": 1.0357,
+      "step": 3750
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7587807562059198,
+      "learning_rate": 3.8258337640374125e-06,
+      "loss": 1.0886,
+      "step": 3751
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7172950367386369,
+      "learning_rate": 3.820933552890629e-06,
+      "loss": 0.9831,
+      "step": 3752
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7110950730374792,
+      "learning_rate": 3.816035740670185e-06,
+      "loss": 1.0605,
+      "step": 3753
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7719334048195114,
+      "learning_rate": 3.811140329277591e-06,
+      "loss": 0.9588,
+      "step": 3754
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6604212911161128,
+      "learning_rate": 3.8062473206134088e-06,
+      "loss": 1.0143,
+      "step": 3755
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6759147796434437,
+      "learning_rate": 3.8013567165772735e-06,
+      "loss": 0.9405,
+      "step": 3756
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7551719056671864,
+      "learning_rate": 3.7964685190678874e-06,
+      "loss": 1.0333,
+      "step": 3757
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5489707972039082,
+      "learning_rate": 3.7915827299830154e-06,
+      "loss": 1.0767,
+      "step": 3758
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5413562341277568,
+      "learning_rate": 3.7866993512194895e-06,
+      "loss": 1.1315,
+      "step": 3759
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5373346925387027,
+      "learning_rate": 3.7818183846732024e-06,
+      "loss": 1.0403,
+      "step": 3760
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6758842736715105,
+      "learning_rate": 3.776939832239125e-06,
+      "loss": 0.9899,
+      "step": 3761
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7201800243358781,
+      "learning_rate": 3.7720636958112623e-06,
+      "loss": 1.0078,
+      "step": 3762
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6828411630881612,
+      "learning_rate": 3.7671899772827113e-06,
+      "loss": 0.9852,
+      "step": 3763
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6869993670505187,
+      "learning_rate": 3.7623186785456156e-06,
+      "loss": 1.0115,
+      "step": 3764
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6861551231078906,
+      "learning_rate": 3.757449801491172e-06,
+      "loss": 1.024,
+      "step": 3765
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7230818620733138,
+      "learning_rate": 3.7525833480096575e-06,
+      "loss": 0.9427,
+      "step": 3766
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7437114934371738,
+      "learning_rate": 3.7477193199903903e-06,
+      "loss": 1.0691,
+      "step": 3767
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6351744578076794,
+      "learning_rate": 3.7428577193217563e-06,
+      "loss": 0.9075,
+      "step": 3768
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.502972859115773,
+      "learning_rate": 3.737998547891195e-06,
+      "loss": 0.9341,
+      "step": 3769
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7058987811269037,
+      "learning_rate": 3.7331418075852053e-06,
+      "loss": 1.0132,
+      "step": 3770
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6574400157617302,
+      "learning_rate": 3.728287500289339e-06,
+      "loss": 1.0106,
+      "step": 3771
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.48109892721841674,
+      "learning_rate": 3.7234356278882076e-06,
+      "loss": 1.0792,
+      "step": 3772
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7786651665130059,
+      "learning_rate": 3.718586192265473e-06,
+      "loss": 0.9818,
+      "step": 3773
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6350493238833286,
+      "learning_rate": 3.7137391953038516e-06,
+      "loss": 1.0593,
+      "step": 3774
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6549991357941729,
+      "learning_rate": 3.7088946388851223e-06,
+      "loss": 0.9857,
+      "step": 3775
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6826207887464565,
+      "learning_rate": 3.7040525248901003e-06,
+      "loss": 0.9849,
+      "step": 3776
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7060455175479279,
+      "learning_rate": 3.6992128551986617e-06,
+      "loss": 1.0706,
+      "step": 3777
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6845702727936522,
+      "learning_rate": 3.6943756316897406e-06,
+      "loss": 1.0125,
+      "step": 3778
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7020697093251931,
+      "learning_rate": 3.6895408562413027e-06,
+      "loss": 0.9926,
+      "step": 3779
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6909029641726339,
+      "learning_rate": 3.684708530730382e-06,
+      "loss": 1.0051,
+      "step": 3780
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7373660878216534,
+      "learning_rate": 3.6798786570330526e-06,
+      "loss": 0.9631,
+      "step": 3781
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6357733770112702,
+      "learning_rate": 3.6750512370244363e-06,
+      "loss": 0.9782,
+      "step": 3782
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7154789549249564,
+      "learning_rate": 3.670226272578704e-06,
+      "loss": 1.0296,
+      "step": 3783
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6489895888344237,
+      "learning_rate": 3.6654037655690732e-06,
+      "loss": 0.9902,
+      "step": 3784
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5305400095956575,
+      "learning_rate": 3.660583717867807e-06,
+      "loss": 1.0561,
+      "step": 3785
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6789329230386612,
+      "learning_rate": 3.655766131346211e-06,
+      "loss": 1.0777,
+      "step": 3786
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6625971594490065,
+      "learning_rate": 3.650951007874648e-06,
+      "loss": 0.9663,
+      "step": 3787
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6816071985938356,
+      "learning_rate": 3.6461383493225012e-06,
+      "loss": 1.0147,
+      "step": 3788
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6974630153125908,
+      "learning_rate": 3.6413281575582194e-06,
+      "loss": 0.9883,
+      "step": 3789
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6637813758257979,
+      "learning_rate": 3.6365204344492867e-06,
+      "loss": 1.0433,
+      "step": 3790
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7107425557541132,
+      "learning_rate": 3.6317151818622154e-06,
+      "loss": 0.9725,
+      "step": 3791
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6493284141391567,
+      "learning_rate": 3.62691240166258e-06,
+      "loss": 0.9968,
+      "step": 3792
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7265916517362856,
+      "learning_rate": 3.6221120957149826e-06,
+      "loss": 1.0197,
+      "step": 3793
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6929288038977751,
+      "learning_rate": 3.617314265883066e-06,
+      "loss": 0.9287,
+      "step": 3794
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7191976083974657,
+      "learning_rate": 3.612518914029515e-06,
+      "loss": 0.9841,
+      "step": 3795
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.695084894961554,
+      "learning_rate": 3.6077260420160487e-06,
+      "loss": 1.0322,
+      "step": 3796
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5477568864202936,
+      "learning_rate": 3.602935651703424e-06,
+      "loss": 0.9846,
+      "step": 3797
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6885778534551694,
+      "learning_rate": 3.598147744951438e-06,
+      "loss": 0.9363,
+      "step": 3798
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6426553907276277,
+      "learning_rate": 3.5933623236189198e-06,
+      "loss": 1.0734,
+      "step": 3799
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5020125553514293,
+      "learning_rate": 3.58857938956373e-06,
+      "loss": 1.0233,
+      "step": 3800
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6893744298779351,
+      "learning_rate": 3.58379894464278e-06,
+      "loss": 1.0115,
+      "step": 3801
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6883820742786461,
+      "learning_rate": 3.57902099071199e-06,
+      "loss": 1.0489,
+      "step": 3802
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6995107261756545,
+      "learning_rate": 3.5742455296263346e-06,
+      "loss": 0.9506,
+      "step": 3803
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7194340395403451,
+      "learning_rate": 3.569472563239814e-06,
+      "loss": 1.0601,
+      "step": 3804
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7023857538477034,
+      "learning_rate": 3.5647020934054465e-06,
+      "loss": 0.9788,
+      "step": 3805
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7243807891926269,
+      "learning_rate": 3.559934121975304e-06,
+      "loss": 1.0889,
+      "step": 3806
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7423535180784094,
+      "learning_rate": 3.5551686508004735e-06,
+      "loss": 0.9208,
+      "step": 3807
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6961696904221171,
+      "learning_rate": 3.550405681731074e-06,
+      "loss": 0.994,
+      "step": 3808
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7347937911391319,
+      "learning_rate": 3.5456452166162547e-06,
+      "loss": 0.9871,
+      "step": 3809
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7493244125966732,
+      "learning_rate": 3.540887257304193e-06,
+      "loss": 0.9919,
+      "step": 3810
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7114907971061589,
+      "learning_rate": 3.5361318056420925e-06,
+      "loss": 0.9373,
+      "step": 3811
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.657578645901821,
+      "learning_rate": 3.531378863476178e-06,
+      "loss": 1.0319,
+      "step": 3812
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5386153792755648,
+      "learning_rate": 3.5266284326517165e-06,
+      "loss": 0.9982,
+      "step": 3813
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6527474062563816,
+      "learning_rate": 3.5218805150129755e-06,
+      "loss": 1.009,
+      "step": 3814
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5524510577290064,
+      "learning_rate": 3.5171351124032703e-06,
+      "loss": 1.0994,
+      "step": 3815
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.675849559890553,
+      "learning_rate": 3.51239222666493e-06,
+      "loss": 1.0046,
+      "step": 3816
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7153415654025453,
+      "learning_rate": 3.507651859639295e-06,
+      "loss": 0.9792,
+      "step": 3817
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7315710447387809,
+      "learning_rate": 3.5029140131667493e-06,
+      "loss": 1.037,
+      "step": 3818
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6931878100494723,
+      "learning_rate": 3.4981786890866853e-06,
+      "loss": 1.0319,
+      "step": 3819
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6690376775426783,
+      "learning_rate": 3.493445889237518e-06,
+      "loss": 0.9576,
+      "step": 3820
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5393129111728941,
+      "learning_rate": 3.4887156154566847e-06,
+      "loss": 1.0392,
+      "step": 3821
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6770370541850433,
+      "learning_rate": 3.4839878695806385e-06,
+      "loss": 0.9683,
+      "step": 3822
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7136150027948639,
+      "learning_rate": 3.4792626534448547e-06,
+      "loss": 0.932,
+      "step": 3823
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6527242165277886,
+      "learning_rate": 3.4745399688838243e-06,
+      "loss": 1.0194,
+      "step": 3824
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6839067741753972,
+      "learning_rate": 3.469819817731056e-06,
+      "loss": 0.9691,
+      "step": 3825
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6291962899625314,
+      "learning_rate": 3.4651022018190715e-06,
+      "loss": 1.0183,
+      "step": 3826
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6885222249392489,
+      "learning_rate": 3.460387122979423e-06,
+      "loss": 0.9589,
+      "step": 3827
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7267415050867368,
+      "learning_rate": 3.455674583042652e-06,
+      "loss": 1.0154,
+      "step": 3828
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6838882636853867,
+      "learning_rate": 3.4509645838383386e-06,
+      "loss": 1.0118,
+      "step": 3829
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6959827185229936,
+      "learning_rate": 3.446257127195066e-06,
+      "loss": 1.0121,
+      "step": 3830
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6595898000618172,
+      "learning_rate": 3.4415522149404233e-06,
+      "loss": 1.038,
+      "step": 3831
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7791798001346127,
+      "learning_rate": 3.436849848901028e-06,
+      "loss": 1.0117,
+      "step": 3832
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7107799611711336,
+      "learning_rate": 3.432150030902497e-06,
+      "loss": 1.0356,
+      "step": 3833
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7760520858875126,
+      "learning_rate": 3.427452762769462e-06,
+      "loss": 1.0103,
+      "step": 3834
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.657368638038207,
+      "learning_rate": 3.4227580463255628e-06,
+      "loss": 1.0129,
+      "step": 3835
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7398156533827939,
+      "learning_rate": 3.4180658833934523e-06,
+      "loss": 0.9993,
+      "step": 3836
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6727508789373342,
+      "learning_rate": 3.4133762757947873e-06,
+      "loss": 1.0027,
+      "step": 3837
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.698661619023552,
+      "learning_rate": 3.4086892253502344e-06,
+      "loss": 0.9549,
+      "step": 3838
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6552855087586159,
+      "learning_rate": 3.4040047338794756e-06,
+      "loss": 0.9952,
+      "step": 3839
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8053427695308925,
+      "learning_rate": 3.3993228032011784e-06,
+      "loss": 1.049,
+      "step": 3840
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6352011178602616,
+      "learning_rate": 3.3946434351330415e-06,
+      "loss": 0.9142,
+      "step": 3841
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.693491931186655,
+      "learning_rate": 3.3899666314917512e-06,
+      "loss": 0.9485,
+      "step": 3842
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6977657232309247,
+      "learning_rate": 3.385292394093006e-06,
+      "loss": 0.9633,
+      "step": 3843
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6981147792896297,
+      "learning_rate": 3.3806207247515068e-06,
+      "loss": 1.0375,
+      "step": 3844
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7678254545565455,
+      "learning_rate": 3.375951625280948e-06,
+      "loss": 1.0357,
+      "step": 3845
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7102402057605492,
+      "learning_rate": 3.3712850974940437e-06,
+      "loss": 1.0212,
+      "step": 3846
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6168032223034605,
+      "learning_rate": 3.3666211432024974e-06,
+      "loss": 1.0107,
+      "step": 3847
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7000442897223953,
+      "learning_rate": 3.361959764217018e-06,
+      "loss": 1.0674,
+      "step": 3848
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6426972704496761,
+      "learning_rate": 3.357300962347313e-06,
+      "loss": 0.9689,
+      "step": 3849
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7335730049229195,
+      "learning_rate": 3.3526447394020887e-06,
+      "loss": 0.9932,
+      "step": 3850
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6679150414406501,
+      "learning_rate": 3.3479910971890516e-06,
+      "loss": 1.0676,
+      "step": 3851
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6596109998735872,
+      "learning_rate": 3.343340037514903e-06,
+      "loss": 0.9358,
+      "step": 3852
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.49279129646538083,
+      "learning_rate": 3.3386915621853533e-06,
+      "loss": 0.9727,
+      "step": 3853
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6888018829973999,
+      "learning_rate": 3.3340456730050887e-06,
+      "loss": 0.946,
+      "step": 3854
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7057980663794398,
+      "learning_rate": 3.3294023717778122e-06,
+      "loss": 0.9846,
+      "step": 3855
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7793705108773431,
+      "learning_rate": 3.324761660306215e-06,
+      "loss": 0.8932,
+      "step": 3856
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.710890168551417,
+      "learning_rate": 3.3201235403919683e-06,
+      "loss": 0.9413,
+      "step": 3857
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5081820286523575,
+      "learning_rate": 3.3154880138357626e-06,
+      "loss": 0.9818,
+      "step": 3858
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5400579140132938,
+      "learning_rate": 3.3108550824372632e-06,
+      "loss": 1.0568,
+      "step": 3859
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6799413879571993,
+      "learning_rate": 3.306224747995136e-06,
+      "loss": 0.9309,
+      "step": 3860
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7083400299699079,
+      "learning_rate": 3.301597012307034e-06,
+      "loss": 0.9435,
+      "step": 3861
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6861848780286665,
+      "learning_rate": 3.2969718771696047e-06,
+      "loss": 1.0214,
+      "step": 3862
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6929749646218588,
+      "learning_rate": 3.292349344378486e-06,
+      "loss": 1.0368,
+      "step": 3863
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7417736118479517,
+      "learning_rate": 3.287729415728298e-06,
+      "loss": 1.0085,
+      "step": 3864
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8065818953730572,
+      "learning_rate": 3.283112093012669e-06,
+      "loss": 0.9368,
+      "step": 3865
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.547751558943972,
+      "learning_rate": 3.278497378024187e-06,
+      "loss": 1.0608,
+      "step": 3866
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7444645301048803,
+      "learning_rate": 3.2738852725544547e-06,
+      "loss": 1.019,
+      "step": 3867
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7068164712505297,
+      "learning_rate": 3.2692757783940467e-06,
+      "loss": 0.9867,
+      "step": 3868
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6682862368552421,
+      "learning_rate": 3.264668897332527e-06,
+      "loss": 0.9363,
+      "step": 3869
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6877008112659402,
+      "learning_rate": 3.2600646311584494e-06,
+      "loss": 0.9843,
+      "step": 3870
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7755397957026493,
+      "learning_rate": 3.2554629816593375e-06,
+      "loss": 1.0173,
+      "step": 3871
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.7011413451336832,
+      "learning_rate": 3.250863950621721e-06,
+      "loss": 1.0511,
+      "step": 3872
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7204590493308112,
+      "learning_rate": 3.2462675398310984e-06,
+      "loss": 0.9857,
+      "step": 3873
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6017776495521483,
+      "learning_rate": 3.241673751071954e-06,
+      "loss": 1.0666,
+      "step": 3874
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6783260680321055,
+      "learning_rate": 3.2370825861277567e-06,
+      "loss": 0.991,
+      "step": 3875
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6830552871180368,
+      "learning_rate": 3.2324940467809527e-06,
+      "loss": 1.0137,
+      "step": 3876
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.67973090113049,
+      "learning_rate": 3.2279081348129713e-06,
+      "loss": 0.983,
+      "step": 3877
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6945111805601392,
+      "learning_rate": 3.223324852004219e-06,
+      "loss": 0.968,
+      "step": 3878
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6847793363422666,
+      "learning_rate": 3.2187442001340942e-06,
+      "loss": 0.9628,
+      "step": 3879
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7774418083791287,
+      "learning_rate": 3.21416618098095e-06,
+      "loss": 1.0539,
+      "step": 3880
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6739790238816353,
+      "learning_rate": 3.2095907963221396e-06,
+      "loss": 0.9924,
+      "step": 3881
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.665203642192266,
+      "learning_rate": 3.2050180479339865e-06,
+      "loss": 1.0408,
+      "step": 3882
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.845957330832424,
+      "learning_rate": 3.2004479375917783e-06,
+      "loss": 1.0089,
+      "step": 3883
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7336214437267757,
+      "learning_rate": 3.1958804670698008e-06,
+      "loss": 1.0682,
+      "step": 3884
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7767381333408567,
+      "learning_rate": 3.191315638141297e-06,
+      "loss": 1.0488,
+      "step": 3885
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7213246387729697,
+      "learning_rate": 3.1867534525784937e-06,
+      "loss": 0.9084,
+      "step": 3886
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7772948984273944,
+      "learning_rate": 3.182193912152586e-06,
+      "loss": 0.9963,
+      "step": 3887
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6951415854949715,
+      "learning_rate": 3.177637018633746e-06,
+      "loss": 0.9495,
+      "step": 3888
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6906937232412179,
+      "learning_rate": 3.1730827737911163e-06,
+      "loss": 0.956,
+      "step": 3889
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6381087356294242,
+      "learning_rate": 3.1685311793928077e-06,
+      "loss": 0.9866,
+      "step": 3890
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7779171466133724,
+      "learning_rate": 3.163982237205917e-06,
+      "loss": 1.0225,
+      "step": 3891
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7304564822799382,
+      "learning_rate": 3.1594359489964853e-06,
+      "loss": 1.0031,
+      "step": 3892
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.31046447549019396,
+      "learning_rate": 3.15489231652955e-06,
+      "loss": 1.048,
+      "step": 3893
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6496898931137041,
+      "learning_rate": 3.150351341569101e-06,
+      "loss": 0.9822,
+      "step": 3894
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6976014667859574,
+      "learning_rate": 3.1458130258781006e-06,
+      "loss": 0.9517,
+      "step": 3895
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6815402201671195,
+      "learning_rate": 3.141277371218484e-06,
+      "loss": 1.031,
+      "step": 3896
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5499946054634832,
+      "learning_rate": 3.136744379351139e-06,
+      "loss": 1.134,
+      "step": 3897
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7270206074832515,
+      "learning_rate": 3.1322140520359366e-06,
+      "loss": 1.0281,
+      "step": 3898
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6787592252953908,
+      "learning_rate": 3.1276863910317057e-06,
+      "loss": 1.0291,
+      "step": 3899
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7227382433838516,
+      "learning_rate": 3.1231613980962373e-06,
+      "loss": 0.9229,
+      "step": 3900
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7645903701020741,
+      "learning_rate": 3.1186390749862904e-06,
+      "loss": 0.9471,
+      "step": 3901
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.666720994537687,
+      "learning_rate": 3.1141194234575878e-06,
+      "loss": 0.9618,
+      "step": 3902
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6947532028550232,
+      "learning_rate": 3.1096024452648123e-06,
+      "loss": 1.0313,
+      "step": 3903
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7123286872586693,
+      "learning_rate": 3.1050881421616076e-06,
+      "loss": 0.9571,
+      "step": 3904
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7503097150577439,
+      "learning_rate": 3.100576515900591e-06,
+      "loss": 1.0135,
+      "step": 3905
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.737465201427649,
+      "learning_rate": 3.0960675682333186e-06,
+      "loss": 0.8362,
+      "step": 3906
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7311260316999147,
+      "learning_rate": 3.0915613009103296e-06,
+      "loss": 0.9921,
+      "step": 3907
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6479457356736391,
+      "learning_rate": 3.0870577156811077e-06,
+      "loss": 1.0023,
+      "step": 3908
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6480340404866516,
+      "learning_rate": 3.0825568142940998e-06,
+      "loss": 1.0399,
+      "step": 3909
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6673245049268686,
+      "learning_rate": 3.0780585984967113e-06,
+      "loss": 0.9767,
+      "step": 3910
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6940670131515528,
+      "learning_rate": 3.073563070035305e-06,
+      "loss": 0.9893,
+      "step": 3911
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7402913231618361,
+      "learning_rate": 3.069070230655198e-06,
+      "loss": 1.0265,
+      "step": 3912
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7020704316235903,
+      "learning_rate": 3.0645800821006667e-06,
+      "loss": 0.9936,
+      "step": 3913
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7835345159685241,
+      "learning_rate": 3.060092626114941e-06,
+      "loss": 0.9624,
+      "step": 3914
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6885938142942603,
+      "learning_rate": 3.0556078644402066e-06,
+      "loss": 1.0226,
+      "step": 3915
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6693178267324995,
+      "learning_rate": 3.051125798817598e-06,
+      "loss": 1.0443,
+      "step": 3916
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7538005881118373,
+      "learning_rate": 3.0466464309872167e-06,
+      "loss": 1.015,
+      "step": 3917
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7573707089222954,
+      "learning_rate": 3.042169762688096e-06,
+      "loss": 1.0018,
+      "step": 3918
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6791428801025445,
+      "learning_rate": 3.0376957956582452e-06,
+      "loss": 0.9713,
+      "step": 3919
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7452584948177351,
+      "learning_rate": 3.0332245316346e-06,
+      "loss": 1.0204,
+      "step": 3920
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6830294496511087,
+      "learning_rate": 3.0287559723530667e-06,
+      "loss": 1.0014,
+      "step": 3921
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6925052386840641,
+      "learning_rate": 3.024290119548495e-06,
+      "loss": 0.9873,
+      "step": 3922
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6784948739746507,
+      "learning_rate": 3.019826974954674e-06,
+      "loss": 0.9952,
+      "step": 3923
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5463415210248711,
+      "learning_rate": 3.0153665403043586e-06,
+      "loss": 1.0341,
+      "step": 3924
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7259522351794332,
+      "learning_rate": 3.01090881732924e-06,
+      "loss": 0.9228,
+      "step": 3925
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6905873697200716,
+      "learning_rate": 3.0064538077599603e-06,
+      "loss": 1.0288,
+      "step": 3926
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6081979592835606,
+      "learning_rate": 3.002001513326107e-06,
+      "loss": 0.9617,
+      "step": 3927
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7656880498594775,
+      "learning_rate": 2.9975519357562155e-06,
+      "loss": 1.0391,
+      "step": 3928
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7275808938832182,
+      "learning_rate": 2.9931050767777626e-06,
+      "loss": 1.0323,
+      "step": 3929
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7427446866518261,
+      "learning_rate": 2.9886609381171703e-06,
+      "loss": 1.034,
+      "step": 3930
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.487551490604395,
+      "learning_rate": 2.984219521499816e-06,
+      "loss": 1.0762,
+      "step": 3931
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7114391722218861,
+      "learning_rate": 2.9797808286499976e-06,
+      "loss": 0.9758,
+      "step": 3932
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6892350618697908,
+      "learning_rate": 2.9753448612909775e-06,
+      "loss": 1.0517,
+      "step": 3933
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.701123813521507,
+      "learning_rate": 2.9709116211449484e-06,
+      "loss": 1.0264,
+      "step": 3934
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6795198965132045,
+      "learning_rate": 2.966481109933047e-06,
+      "loss": 0.9752,
+      "step": 3935
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5490363842122166,
+      "learning_rate": 2.9620533293753495e-06,
+      "loss": 1.071,
+      "step": 3936
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7682590366073734,
+      "learning_rate": 2.957628281190873e-06,
+      "loss": 1.0716,
+      "step": 3937
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6858943555136574,
+      "learning_rate": 2.9532059670975732e-06,
+      "loss": 1.0117,
+      "step": 3938
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.745792876034167,
+      "learning_rate": 2.948786388812346e-06,
+      "loss": 0.9563,
+      "step": 3939
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5374266854669676,
+      "learning_rate": 2.9443695480510225e-06,
+      "loss": 1.0295,
+      "step": 3940
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6830110522777012,
+      "learning_rate": 2.9399554465283742e-06,
+      "loss": 1.0779,
+      "step": 3941
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5639922385927677,
+      "learning_rate": 2.935544085958102e-06,
+      "loss": 1.0601,
+      "step": 3942
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6734398839711239,
+      "learning_rate": 2.931135468052858e-06,
+      "loss": 0.9096,
+      "step": 3943
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6731913880588585,
+      "learning_rate": 2.926729594524207e-06,
+      "loss": 1.0641,
+      "step": 3944
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7780494984074896,
+      "learning_rate": 2.9223264670826746e-06,
+      "loss": 1.0115,
+      "step": 3945
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7068265564141696,
+      "learning_rate": 2.9179260874376915e-06,
+      "loss": 1.061,
+      "step": 3946
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.736970602236801,
+      "learning_rate": 2.9135284572976486e-06,
+      "loss": 1.046,
+      "step": 3947
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7782117104937878,
+      "learning_rate": 2.9091335783698517e-06,
+      "loss": 1.0011,
+      "step": 3948
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6930385666516649,
+      "learning_rate": 2.9047414523605467e-06,
+      "loss": 0.9915,
+      "step": 3949
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7367667195182045,
+      "learning_rate": 2.9003520809749053e-06,
+      "loss": 0.9413,
+      "step": 3950
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8236866005814191,
+      "learning_rate": 2.8959654659170354e-06,
+      "loss": 1.0036,
+      "step": 3951
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7618209920752754,
+      "learning_rate": 2.8915816088899696e-06,
+      "loss": 1.0134,
+      "step": 3952
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7347752975861137,
+      "learning_rate": 2.8872005115956746e-06,
+      "loss": 0.9499,
+      "step": 3953
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7020408787425441,
+      "learning_rate": 2.8828221757350406e-06,
+      "loss": 1.0175,
+      "step": 3954
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5901837950960427,
+      "learning_rate": 2.8784466030078905e-06,
+      "loss": 1.0125,
+      "step": 3955
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.663301404737722,
+      "learning_rate": 2.874073795112967e-06,
+      "loss": 0.9694,
+      "step": 3956
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7214956071566875,
+      "learning_rate": 2.8697037537479565e-06,
+      "loss": 0.9926,
+      "step": 3957
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6995170009413028,
+      "learning_rate": 2.8653364806094454e-06,
+      "loss": 1.0266,
+      "step": 3958
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6755902475020336,
+      "learning_rate": 2.86097197739297e-06,
+      "loss": 1.0337,
+      "step": 3959
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.658801221196825,
+      "learning_rate": 2.856610245792976e-06,
+      "loss": 1.0322,
+      "step": 3960
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5510479687264632,
+      "learning_rate": 2.8522512875028396e-06,
+      "loss": 0.9447,
+      "step": 3961
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6868492954736612,
+      "learning_rate": 2.847895104214856e-06,
+      "loss": 1.0872,
+      "step": 3962
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7162175179775762,
+      "learning_rate": 2.843541697620249e-06,
+      "loss": 0.936,
+      "step": 3963
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6450334731599076,
+      "learning_rate": 2.8391910694091584e-06,
+      "loss": 0.9759,
+      "step": 3964
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.652736599360207,
+      "learning_rate": 2.8348432212706443e-06,
+      "loss": 1.0122,
+      "step": 3965
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7689390209946388,
+      "learning_rate": 2.8304981548927025e-06,
+      "loss": 0.9745,
+      "step": 3966
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6935560078417669,
+      "learning_rate": 2.826155871962227e-06,
+      "loss": 1.0256,
+      "step": 3967
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5630527510855587,
+      "learning_rate": 2.8218163741650415e-06,
+      "loss": 1.1154,
+      "step": 3968
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.74013489806575,
+      "learning_rate": 2.817479663185898e-06,
+      "loss": 1.0162,
+      "step": 3969
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6747677505171787,
+      "learning_rate": 2.813145740708445e-06,
+      "loss": 0.9551,
+      "step": 3970
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8110557059884406,
+      "learning_rate": 2.808814608415271e-06,
+      "loss": 1.0729,
+      "step": 3971
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6920232888174581,
+      "learning_rate": 2.8044862679878605e-06,
+      "loss": 0.9402,
+      "step": 3972
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7674922419077744,
+      "learning_rate": 2.800160721106633e-06,
+      "loss": 0.9833,
+      "step": 3973
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7388760052150057,
+      "learning_rate": 2.7958379694509108e-06,
+      "loss": 0.9802,
+      "step": 3974
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6942887558821561,
+      "learning_rate": 2.791518014698935e-06,
+      "loss": 0.9979,
+      "step": 3975
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6880326676698523,
+      "learning_rate": 2.787200858527862e-06,
+      "loss": 0.9885,
+      "step": 3976
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.71502969563593,
+      "learning_rate": 2.7828865026137584e-06,
+      "loss": 1.014,
+      "step": 3977
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7283962182332376,
+      "learning_rate": 2.7785749486316085e-06,
+      "loss": 1.0397,
+      "step": 3978
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8025884106931555,
+      "learning_rate": 2.774266198255303e-06,
+      "loss": 1.0003,
+      "step": 3979
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6827022650220963,
+      "learning_rate": 2.7699602531576496e-06,
+      "loss": 0.9278,
+      "step": 3980
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6807101627610335,
+      "learning_rate": 2.765657115010364e-06,
+      "loss": 0.9005,
+      "step": 3981
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7108664979522951,
+      "learning_rate": 2.7613567854840685e-06,
+      "loss": 1.0004,
+      "step": 3982
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7590954126844002,
+      "learning_rate": 2.7570592662483086e-06,
+      "loss": 1.0082,
+      "step": 3983
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6486485073636331,
+      "learning_rate": 2.752764558971517e-06,
+      "loss": 1.0294,
+      "step": 3984
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7474047271114612,
+      "learning_rate": 2.748472665321056e-06,
+      "loss": 0.9614,
+      "step": 3985
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8151172143264522,
+      "learning_rate": 2.744183586963185e-06,
+      "loss": 1.0718,
+      "step": 3986
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6878655479541296,
+      "learning_rate": 2.739897325563069e-06,
+      "loss": 1.0383,
+      "step": 3987
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6991254219242365,
+      "learning_rate": 2.7356138827847856e-06,
+      "loss": 1.0187,
+      "step": 3988
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7236300739245863,
+      "learning_rate": 2.731333260291311e-06,
+      "loss": 0.9731,
+      "step": 3989
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7151135749133489,
+      "learning_rate": 2.7270554597445343e-06,
+      "loss": 0.9583,
+      "step": 3990
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7210413763836482,
+      "learning_rate": 2.7227804828052384e-06,
+      "loss": 0.9885,
+      "step": 3991
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6920808166616742,
+      "learning_rate": 2.7185083311331283e-06,
+      "loss": 0.9675,
+      "step": 3992
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6941607259364625,
+      "learning_rate": 2.7142390063867896e-06,
+      "loss": 0.9575,
+      "step": 3993
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7185405156808451,
+      "learning_rate": 2.709972510223725e-06,
+      "loss": 0.9403,
+      "step": 3994
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7020742753645235,
+      "learning_rate": 2.7057088443003343e-06,
+      "loss": 1.0048,
+      "step": 3995
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6648830953080674,
+      "learning_rate": 2.7014480102719174e-06,
+      "loss": 0.9299,
+      "step": 3996
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7395514840570028,
+      "learning_rate": 2.697190009792685e-06,
+      "loss": 0.9943,
+      "step": 3997
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.700876088660091,
+      "learning_rate": 2.692934844515729e-06,
+      "loss": 1.0284,
+      "step": 3998
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8282129768805645,
+      "learning_rate": 2.6886825160930587e-06,
+      "loss": 1.0145,
+      "step": 3999
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6974683524233848,
+      "learning_rate": 2.6844330261755715e-06,
+      "loss": 0.9452,
+      "step": 4000
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7188566156892396,
+      "learning_rate": 2.6801863764130653e-06,
+      "loss": 1.0014,
+      "step": 4001
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7264886532665046,
+      "learning_rate": 2.675942568454236e-06,
+      "loss": 0.9818,
+      "step": 4002
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5311098603221205,
+      "learning_rate": 2.671701603946678e-06,
+      "loss": 1.0344,
+      "step": 4003
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6779947386138694,
+      "learning_rate": 2.667463484536876e-06,
+      "loss": 0.9916,
+      "step": 4004
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7054924065257672,
+      "learning_rate": 2.6632282118702147e-06,
+      "loss": 0.9977,
+      "step": 4005
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6866070451658728,
+      "learning_rate": 2.65899578759098e-06,
+      "loss": 1.0126,
+      "step": 4006
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.36889516927406996,
+      "learning_rate": 2.654766213342335e-06,
+      "loss": 0.9873,
+      "step": 4007
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.36889516927406996,
+      "learning_rate": 2.654766213342335e-06,
+      "loss": 0.9518,
+      "step": 4008
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.36889516927406996,
+      "learning_rate": 2.654766213342335e-06,
+      "loss": 0.967,
+      "step": 4009
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7030013987428231,
+      "learning_rate": 2.650539490766346e-06,
+      "loss": 1.0132,
+      "step": 4010
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6989387127416868,
+      "learning_rate": 2.646315621503983e-06,
+      "loss": 0.988,
+      "step": 4011
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5675155201336387,
+      "learning_rate": 2.642094607195085e-06,
+      "loss": 1.0839,
+      "step": 4012
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5577716723907341,
+      "learning_rate": 2.6378764494784027e-06,
+      "loss": 1.0162,
+      "step": 4013
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6903292930032268,
+      "learning_rate": 2.633661149991569e-06,
+      "loss": 1.0435,
+      "step": 4014
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7577193579664545,
+      "learning_rate": 2.6294487103711064e-06,
+      "loss": 1.006,
+      "step": 4015
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6249258554343641,
+      "learning_rate": 2.6252391322524297e-06,
+      "loss": 1.0414,
+      "step": 4016
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7753140573275196,
+      "learning_rate": 2.6210324172698432e-06,
+      "loss": 0.9979,
+      "step": 4017
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7289472352360493,
+      "learning_rate": 2.6168285670565374e-06,
+      "loss": 1.0484,
+      "step": 4018
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6934156495643583,
+      "learning_rate": 2.6126275832445892e-06,
+      "loss": 0.917,
+      "step": 4019
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7063187915616665,
+      "learning_rate": 2.6084294674649734e-06,
+      "loss": 1.0259,
+      "step": 4020
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7057509765051053,
+      "learning_rate": 2.6042342213475346e-06,
+      "loss": 0.9939,
+      "step": 4021
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.651254696296089,
+      "learning_rate": 2.6000418465210143e-06,
+      "loss": 1.0685,
+      "step": 4022
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6896009185926044,
+      "learning_rate": 2.595852344613038e-06,
+      "loss": 1.0071,
+      "step": 4023
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6194079507553828,
+      "learning_rate": 2.5916657172501103e-06,
+      "loss": 0.9766,
+      "step": 4024
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.8265264799597346,
+      "learning_rate": 2.587481966057633e-06,
+      "loss": 0.9964,
+      "step": 4025
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7237022455270777,
+      "learning_rate": 2.583301092659872e-06,
+      "loss": 0.9716,
+      "step": 4026
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7050444705971438,
+      "learning_rate": 2.5791230986799944e-06,
+      "loss": 0.976,
+      "step": 4027
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7117775798018094,
+      "learning_rate": 2.5749479857400383e-06,
+      "loss": 0.9872,
+      "step": 4028
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6461576503538017,
+      "learning_rate": 2.5707757554609247e-06,
+      "loss": 1.028,
+      "step": 4029
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5667212265341215,
+      "learning_rate": 2.56660640946246e-06,
+      "loss": 0.996,
+      "step": 4030
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7799510441008581,
+      "learning_rate": 2.5624399493633257e-06,
+      "loss": 0.9951,
+      "step": 4031
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5647466362731153,
+      "learning_rate": 2.558276376781086e-06,
+      "loss": 1.0363,
+      "step": 4032
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6894070798956028,
+      "learning_rate": 2.55411569333218e-06,
+      "loss": 1.0609,
+      "step": 4033
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7644703426767052,
+      "learning_rate": 2.5499579006319365e-06,
+      "loss": 1.0502,
+      "step": 4034
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6417121099596911,
+      "learning_rate": 2.5458030002945457e-06,
+      "loss": 0.92,
+      "step": 4035
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6614396508853796,
+      "learning_rate": 2.5416509939330836e-06,
+      "loss": 0.9343,
+      "step": 4036
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7286913431432976,
+      "learning_rate": 2.537501883159509e-06,
+      "loss": 0.9795,
+      "step": 4037
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7268224301577961,
+      "learning_rate": 2.5333556695846384e-06,
+      "loss": 0.9581,
+      "step": 4038
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7522299463775758,
+      "learning_rate": 2.5292123548181847e-06,
+      "loss": 1.1014,
+      "step": 4039
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5036392235733044,
+      "learning_rate": 2.525071940468722e-06,
+      "loss": 1.0457,
+      "step": 4040
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5738472201404123,
+      "learning_rate": 2.520934428143701e-06,
+      "loss": 1.0364,
+      "step": 4041
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7507861661749121,
+      "learning_rate": 2.5167998194494468e-06,
+      "loss": 0.9718,
+      "step": 4042
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7072784196145271,
+      "learning_rate": 2.5126681159911558e-06,
+      "loss": 0.963,
+      "step": 4043
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4969226978946792,
+      "learning_rate": 2.5085393193729e-06,
+      "loss": 1.0192,
+      "step": 4044
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7297314727068553,
+      "learning_rate": 2.5044134311976156e-06,
+      "loss": 1.0139,
+      "step": 4045
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6878572611648531,
+      "learning_rate": 2.5002904530671236e-06,
+      "loss": 1.0046,
+      "step": 4046
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5309523303773234,
+      "learning_rate": 2.4961703865820974e-06,
+      "loss": 1.0411,
+      "step": 4047
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7404213361004576,
+      "learning_rate": 2.492053233342091e-06,
+      "loss": 1.0558,
+      "step": 4048
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6991856687904835,
+      "learning_rate": 2.487938994945527e-06,
+      "loss": 0.9113,
+      "step": 4049
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7902191112092496,
+      "learning_rate": 2.4838276729896884e-06,
+      "loss": 0.958,
+      "step": 4050
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7946579195700284,
+      "learning_rate": 2.479719269070743e-06,
+      "loss": 1.0085,
+      "step": 4051
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7122545465584165,
+      "learning_rate": 2.4756137847837025e-06,
+      "loss": 1.0546,
+      "step": 4052
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7861291377353142,
+      "learning_rate": 2.4715112217224657e-06,
+      "loss": 1.002,
+      "step": 4053
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6856820885246632,
+      "learning_rate": 2.467411581479786e-06,
+      "loss": 1.0499,
+      "step": 4054
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6228928370558919,
+      "learning_rate": 2.463314865647286e-06,
+      "loss": 1.0737,
+      "step": 4055
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5883761537719913,
+      "learning_rate": 2.45922107581545e-06,
+      "loss": 1.0132,
+      "step": 4056
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7702104172809936,
+      "learning_rate": 2.4551302135736287e-06,
+      "loss": 0.9787,
+      "step": 4057
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7228369373533634,
+      "learning_rate": 2.4510422805100366e-06,
+      "loss": 0.9601,
+      "step": 4058
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5622726481567197,
+      "learning_rate": 2.446957278211746e-06,
+      "loss": 1.0168,
+      "step": 4059
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7404999243089211,
+      "learning_rate": 2.4428752082647044e-06,
+      "loss": 1.0071,
+      "step": 4060
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6696756929285862,
+      "learning_rate": 2.438796072253704e-06,
+      "loss": 1.0459,
+      "step": 4061
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7826946061875012,
+      "learning_rate": 2.4347198717624054e-06,
+      "loss": 0.9109,
+      "step": 4062
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7087933403718212,
+      "learning_rate": 2.4306466083733392e-06,
+      "loss": 0.9715,
+      "step": 4063
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5258197248481739,
+      "learning_rate": 2.426576283667873e-06,
+      "loss": 1.1025,
+      "step": 4064
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7622105694481667,
+      "learning_rate": 2.422508899226258e-06,
+      "loss": 0.9342,
+      "step": 4065
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6617835655942759,
+      "learning_rate": 2.418444456627589e-06,
+      "loss": 0.9891,
+      "step": 4066
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6390890785711514,
+      "learning_rate": 2.4143829574498224e-06,
+      "loss": 1.0023,
+      "step": 4067
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7264154942445409,
+      "learning_rate": 2.4103244032697717e-06,
+      "loss": 1.0869,
+      "step": 4068
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7304920286444583,
+      "learning_rate": 2.406268795663108e-06,
+      "loss": 0.9385,
+      "step": 4069
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7406623321373293,
+      "learning_rate": 2.4022161362043574e-06,
+      "loss": 1.0468,
+      "step": 4070
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7008712739730492,
+      "learning_rate": 2.3981664264669025e-06,
+      "loss": 0.9196,
+      "step": 4071
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7105160550678918,
+      "learning_rate": 2.3941196680229794e-06,
+      "loss": 1.0173,
+      "step": 4072
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5732062026962829,
+      "learning_rate": 2.3900758624436772e-06,
+      "loss": 1.0071,
+      "step": 4073
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6634761704032293,
+      "learning_rate": 2.3860350112989473e-06,
+      "loss": 1.0334,
+      "step": 4074
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7043020120492051,
+      "learning_rate": 2.3819971161575807e-06,
+      "loss": 0.9894,
+      "step": 4075
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6971315266775869,
+      "learning_rate": 2.3779621785872252e-06,
+      "loss": 1.0015,
+      "step": 4076
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7465694821580046,
+      "learning_rate": 2.3739302001543918e-06,
+      "loss": 0.9172,
+      "step": 4077
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6878667061661474,
+      "learning_rate": 2.3699011824244234e-06,
+      "loss": 0.95,
+      "step": 4078
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5524472734335212,
+      "learning_rate": 2.365875126961531e-06,
+      "loss": 0.9739,
+      "step": 4079
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6090175781049083,
+      "learning_rate": 2.3618520353287644e-06,
+      "loss": 1.0821,
+      "step": 4080
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.693137627521449,
+      "learning_rate": 2.3578319090880263e-06,
+      "loss": 1.0396,
+      "step": 4081
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6740093394230996,
+      "learning_rate": 2.3538147498000695e-06,
+      "loss": 1.0038,
+      "step": 4082
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7693188976300543,
+      "learning_rate": 2.349800559024492e-06,
+      "loss": 0.9665,
+      "step": 4083
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6679199629278575,
+      "learning_rate": 2.3457893383197415e-06,
+      "loss": 0.9999,
+      "step": 4084
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7123506258451144,
+      "learning_rate": 2.3417810892431104e-06,
+      "loss": 1.0029,
+      "step": 4085
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7141505540697668,
+      "learning_rate": 2.3377758133507455e-06,
+      "loss": 0.9647,
+      "step": 4086
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7520998826530798,
+      "learning_rate": 2.3337735121976247e-06,
+      "loss": 0.9747,
+      "step": 4087
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.517751087694311,
+      "learning_rate": 2.32977418733758e-06,
+      "loss": 1.043,
+      "step": 4088
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6618759353870345,
+      "learning_rate": 2.3257778403232954e-06,
+      "loss": 0.9814,
+      "step": 4089
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6905474313161903,
+      "learning_rate": 2.321784472706279e-06,
+      "loss": 1.0073,
+      "step": 4090
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6872507380645704,
+      "learning_rate": 2.317794086036901e-06,
+      "loss": 0.9604,
+      "step": 4091
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6583272099871696,
+      "learning_rate": 2.3138066818643647e-06,
+      "loss": 0.994,
+      "step": 4092
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7275332762389007,
+      "learning_rate": 2.3098222617367184e-06,
+      "loss": 0.9436,
+      "step": 4093
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7166399804216894,
+      "learning_rate": 2.30584082720085e-06,
+      "loss": 1.0306,
+      "step": 4094
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6273234533370079,
+      "learning_rate": 2.301862379802492e-06,
+      "loss": 0.9837,
+      "step": 4095
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7115054752930702,
+      "learning_rate": 2.297886921086211e-06,
+      "loss": 0.9482,
+      "step": 4096
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.737530146058759,
+      "learning_rate": 2.2939144525954194e-06,
+      "loss": 1.0042,
+      "step": 4097
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7332760646068746,
+      "learning_rate": 2.2899449758723657e-06,
+      "loss": 0.9381,
+      "step": 4098
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7654224981063974,
+      "learning_rate": 2.285978492458134e-06,
+      "loss": 1.0069,
+      "step": 4099
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.682211994273042,
+      "learning_rate": 2.282015003892659e-06,
+      "loss": 1.0013,
+      "step": 4100
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7900176506855398,
+      "learning_rate": 2.2780545117146947e-06,
+      "loss": 1.0746,
+      "step": 4101
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7115728101099756,
+      "learning_rate": 2.2740970174618405e-06,
+      "loss": 1.0035,
+      "step": 4102
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5998409941117221,
+      "learning_rate": 2.270142522670541e-06,
+      "loss": 1.0724,
+      "step": 4103
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6794471234751911,
+      "learning_rate": 2.2661910288760545e-06,
+      "loss": 0.9551,
+      "step": 4104
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.693048450216349,
+      "learning_rate": 2.262242537612497e-06,
+      "loss": 1.009,
+      "step": 4105
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6867844740675154,
+      "learning_rate": 2.258297050412804e-06,
+      "loss": 0.914,
+      "step": 4106
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7148814955019047,
+      "learning_rate": 2.254354568808752e-06,
+      "loss": 1.0029,
+      "step": 4107
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.721892952140486,
+      "learning_rate": 2.2504150943309455e-06,
+      "loss": 1.0183,
+      "step": 4108
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5699510175799746,
+      "learning_rate": 2.246478628508827e-06,
+      "loss": 1.0722,
+      "step": 4109
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6745148420359571,
+      "learning_rate": 2.242545172870665e-06,
+      "loss": 0.9874,
+      "step": 4110
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.715483525272215,
+      "learning_rate": 2.238614728943561e-06,
+      "loss": 0.9657,
+      "step": 4111
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6174712509561098,
+      "learning_rate": 2.2346872982534584e-06,
+      "loss": 0.9855,
+      "step": 4112
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6906662609015527,
+      "learning_rate": 2.2307628823251083e-06,
+      "loss": 0.9707,
+      "step": 4113
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5673035053890295,
+      "learning_rate": 2.2268414826821117e-06,
+      "loss": 1.0532,
+      "step": 4114
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7995697962315591,
+      "learning_rate": 2.222923100846893e-06,
+      "loss": 0.9631,
+      "step": 4115
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6476295674322303,
+      "learning_rate": 2.2190077383406938e-06,
+      "loss": 0.9578,
+      "step": 4116
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7454993416643558,
+      "learning_rate": 2.2150953966835996e-06,
+      "loss": 1.0018,
+      "step": 4117
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7972769063554922,
+      "learning_rate": 2.211186077394516e-06,
+      "loss": 1.002,
+      "step": 4118
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7214101951998333,
+      "learning_rate": 2.207279781991173e-06,
+      "loss": 1.0118,
+      "step": 4119
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5931430632552803,
+      "learning_rate": 2.2033765119901294e-06,
+      "loss": 1.0554,
+      "step": 4120
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6910075903081861,
+      "learning_rate": 2.1994762689067705e-06,
+      "loss": 1.03,
+      "step": 4121
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.672169975677519,
+      "learning_rate": 2.1955790542553036e-06,
+      "loss": 0.9876,
+      "step": 4122
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7044379271833833,
+      "learning_rate": 2.1916848695487615e-06,
+      "loss": 0.9875,
+      "step": 4123
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.545938168381319,
+      "learning_rate": 2.1877937162990015e-06,
+      "loss": 1.0047,
+      "step": 4124
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.703816120025338,
+      "learning_rate": 2.1839055960167e-06,
+      "loss": 0.9815,
+      "step": 4125
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6637757357879537,
+      "learning_rate": 2.180020510211367e-06,
+      "loss": 0.9144,
+      "step": 4126
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6587372512599382,
+      "learning_rate": 2.1761384603913203e-06,
+      "loss": 0.9852,
+      "step": 4127
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6653001684305244,
+      "learning_rate": 2.172259448063704e-06,
+      "loss": 0.9722,
+      "step": 4128
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.708710115930786,
+      "learning_rate": 2.1683834747344913e-06,
+      "loss": 1.0125,
+      "step": 4129
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7117374678867964,
+      "learning_rate": 2.1645105419084587e-06,
+      "loss": 0.8666,
+      "step": 4130
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6826583369941706,
+      "learning_rate": 2.160640651089221e-06,
+      "loss": 0.9771,
+      "step": 4131
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7070500977555362,
+      "learning_rate": 2.1567738037791998e-06,
+      "loss": 0.9637,
+      "step": 4132
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5694367608355257,
+      "learning_rate": 2.152910001479638e-06,
+      "loss": 1.0786,
+      "step": 4133
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6500254121252765,
+      "learning_rate": 2.1490492456905964e-06,
+      "loss": 1.0848,
+      "step": 4134
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5834083534754445,
+      "learning_rate": 2.1451915379109546e-06,
+      "loss": 1.0004,
+      "step": 4135
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.665671098822492,
+      "learning_rate": 2.141336879638406e-06,
+      "loss": 0.9967,
+      "step": 4136
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6941005967855159,
+      "learning_rate": 2.1374852723694595e-06,
+      "loss": 0.9946,
+      "step": 4137
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6089880269346717,
+      "learning_rate": 2.133636717599451e-06,
+      "loss": 0.9883,
+      "step": 4138
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7209122668818256,
+      "learning_rate": 2.1297912168225086e-06,
+      "loss": 1.0366,
+      "step": 4139
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7617509602715621,
+      "learning_rate": 2.1259487715316e-06,
+      "loss": 0.8806,
+      "step": 4140
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6923542465655459,
+      "learning_rate": 2.1221093832184903e-06,
+      "loss": 0.9578,
+      "step": 4141
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6774915024691484,
+      "learning_rate": 2.118273053373757e-06,
+      "loss": 1.0319,
+      "step": 4142
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7612711411303111,
+      "learning_rate": 2.1144397834868034e-06,
+      "loss": 0.9648,
+      "step": 4143
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7053871181631938,
+      "learning_rate": 2.1106095750458332e-06,
+      "loss": 1.0239,
+      "step": 4144
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.694178691719762,
+      "learning_rate": 2.106782429537866e-06,
+      "loss": 0.978,
+      "step": 4145
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7476687717222943,
+      "learning_rate": 2.1029583484487315e-06,
+      "loss": 0.9968,
+      "step": 4146
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7113803129903343,
+      "learning_rate": 2.0991373332630683e-06,
+      "loss": 1.0414,
+      "step": 4147
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7043307541994598,
+      "learning_rate": 2.0953193854643274e-06,
+      "loss": 0.9717,
+      "step": 4148
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6907264045191064,
+      "learning_rate": 2.0915045065347673e-06,
+      "loss": 1.0269,
+      "step": 4149
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6011561204980163,
+      "learning_rate": 2.0876926979554545e-06,
+      "loss": 1.0544,
+      "step": 4150
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7678700241858933,
+      "learning_rate": 2.0838839612062633e-06,
+      "loss": 0.999,
+      "step": 4151
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5028726442077214,
+      "learning_rate": 2.080078297765884e-06,
+      "loss": 0.9775,
+      "step": 4152
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7506407543969152,
+      "learning_rate": 2.0762757091117937e-06,
+      "loss": 1.0569,
+      "step": 4153
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7036962552751683,
+      "learning_rate": 2.0724761967202987e-06,
+      "loss": 0.9872,
+      "step": 4154
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6763659538120518,
+      "learning_rate": 2.0686797620664987e-06,
+      "loss": 0.9922,
+      "step": 4155
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7005726088557839,
+      "learning_rate": 2.0648864066242937e-06,
+      "loss": 0.9889,
+      "step": 4156
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6629583367708343,
+      "learning_rate": 2.0610961318664013e-06,
+      "loss": 1.0717,
+      "step": 4157
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.711342585822284,
+      "learning_rate": 2.0573089392643362e-06,
+      "loss": 1.002,
+      "step": 4158
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6659837926218636,
+      "learning_rate": 2.0535248302884147e-06,
+      "loss": 1.0428,
+      "step": 4159
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6405694818702985,
+      "learning_rate": 2.0497438064077603e-06,
+      "loss": 1.0113,
+      "step": 4160
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.734106119089583,
+      "learning_rate": 2.045965869090295e-06,
+      "loss": 0.9666,
+      "step": 4161
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.719668699407119,
+      "learning_rate": 2.0421910198027452e-06,
+      "loss": 0.986,
+      "step": 4162
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.702773952994995,
+      "learning_rate": 2.0384192600106335e-06,
+      "loss": 0.9921,
+      "step": 4163
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6684536861326277,
+      "learning_rate": 2.0346505911782956e-06,
+      "loss": 1.0219,
+      "step": 4164
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7114328627324363,
+      "learning_rate": 2.0308850147688484e-06,
+      "loss": 1.0359,
+      "step": 4165
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7766124679333646,
+      "learning_rate": 2.0271225322442255e-06,
+      "loss": 1.0063,
+      "step": 4166
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7070072850153122,
+      "learning_rate": 2.0233631450651525e-06,
+      "loss": 0.9716,
+      "step": 4167
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7706169958820955,
+      "learning_rate": 2.019606854691145e-06,
+      "loss": 1.0565,
+      "step": 4168
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7488147894920034,
+      "learning_rate": 2.0158536625805325e-06,
+      "loss": 1.0042,
+      "step": 4169
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6649294826612484,
+      "learning_rate": 2.01210357019043e-06,
+      "loss": 1.0419,
+      "step": 4170
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6804613936475561,
+      "learning_rate": 2.008356578976752e-06,
+      "loss": 1.0443,
+      "step": 4171
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7845736501648349,
+      "learning_rate": 2.004612690394212e-06,
+      "loss": 1.0177,
+      "step": 4172
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7024589688206426,
+      "learning_rate": 2.0008719058963144e-06,
+      "loss": 1.0529,
+      "step": 4173
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6726379459195879,
+      "learning_rate": 1.997134226935361e-06,
+      "loss": 1.0311,
+      "step": 4174
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7199527906870204,
+      "learning_rate": 1.9933996549624468e-06,
+      "loss": 1.0334,
+      "step": 4175
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5114664940141518,
+      "learning_rate": 1.9896681914274616e-06,
+      "loss": 1.0587,
+      "step": 4176
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.663067501330989,
+      "learning_rate": 1.9859398377790872e-06,
+      "loss": 0.9648,
+      "step": 4177
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6899265226872334,
+      "learning_rate": 1.982214595464804e-06,
+      "loss": 1.0068,
+      "step": 4178
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7025414286270798,
+      "learning_rate": 1.97849246593087e-06,
+      "loss": 1.0418,
+      "step": 4179
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6989726407443898,
+      "learning_rate": 1.9747734506223525e-06,
+      "loss": 1.0168,
+      "step": 4180
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6468310808535637,
+      "learning_rate": 1.9710575509831008e-06,
+      "loss": 0.9366,
+      "step": 4181
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6711282995392073,
+      "learning_rate": 1.967344768455747e-06,
+      "loss": 1.0347,
+      "step": 4182
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7144930172261669,
+      "learning_rate": 1.9636351044817292e-06,
+      "loss": 0.9233,
+      "step": 4183
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7737931569116756,
+      "learning_rate": 1.9599285605012643e-06,
+      "loss": 0.964,
+      "step": 4184
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6715793385128391,
+      "learning_rate": 1.9562251379533593e-06,
+      "loss": 1.0254,
+      "step": 4185
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7336113361103018,
+      "learning_rate": 1.952524838275811e-06,
+      "loss": 1.0127,
+      "step": 4186
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6917055482803859,
+      "learning_rate": 1.9488276629052026e-06,
+      "loss": 0.944,
+      "step": 4187
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5847283630187721,
+      "learning_rate": 1.945133613276907e-06,
+      "loss": 1.0413,
+      "step": 4188
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7207469452144776,
+      "learning_rate": 1.941442690825076e-06,
+      "loss": 0.9985,
+      "step": 4189
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7108370832275407,
+      "learning_rate": 1.937754896982663e-06,
+      "loss": 1.0472,
+      "step": 4190
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6328635062530131,
+      "learning_rate": 1.9340702331813842e-06,
+      "loss": 0.9556,
+      "step": 4191
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6614628345382523,
+      "learning_rate": 1.9303887008517618e-06,
+      "loss": 0.9556,
+      "step": 4192
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7532548516230946,
+      "learning_rate": 1.9267103014230935e-06,
+      "loss": 0.958,
+      "step": 4193
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6023081346991401,
+      "learning_rate": 1.923035036323452e-06,
+      "loss": 1.0063,
+      "step": 4194
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7046048558,
+      "learning_rate": 1.91936290697971e-06,
+      "loss": 1.0493,
+      "step": 4195
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5199022424351029,
+      "learning_rate": 1.9156939148175125e-06,
+      "loss": 0.981,
+      "step": 4196
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7349778575310931,
+      "learning_rate": 1.9120280612612873e-06,
+      "loss": 1.0056,
+      "step": 4197
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7066503347805392,
+      "learning_rate": 1.9083653477342467e-06,
+      "loss": 0.9551,
+      "step": 4198
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7148160528310613,
+      "learning_rate": 1.904705775658381e-06,
+      "loss": 0.9533,
+      "step": 4199
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7667182304530982,
+      "learning_rate": 1.9010493464544621e-06,
+      "loss": 1.0184,
+      "step": 4200
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.724539610361626,
+      "learning_rate": 1.8973960615420416e-06,
+      "loss": 1.0625,
+      "step": 4201
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5330266693682022,
+      "learning_rate": 1.8937459223394517e-06,
+      "loss": 1.0296,
+      "step": 4202
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5420524093045872,
+      "learning_rate": 1.8900989302637985e-06,
+      "loss": 1.0783,
+      "step": 4203
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6818041709459963,
+      "learning_rate": 1.8864550867309771e-06,
+      "loss": 0.9984,
+      "step": 4204
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.680457454584589,
+      "learning_rate": 1.8828143931556442e-06,
+      "loss": 1.0171,
+      "step": 4205
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7584788703917779,
+      "learning_rate": 1.8791768509512487e-06,
+      "loss": 0.9999,
+      "step": 4206
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6105335662925933,
+      "learning_rate": 1.875542461530011e-06,
+      "loss": 1.0496,
+      "step": 4207
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6601982552726617,
+      "learning_rate": 1.871911226302917e-06,
+      "loss": 1.0231,
+      "step": 4208
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7001668242863338,
+      "learning_rate": 1.868283146679747e-06,
+      "loss": 0.9614,
+      "step": 4209
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7019314748026566,
+      "learning_rate": 1.8646582240690414e-06,
+      "loss": 0.963,
+      "step": 4210
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6492084987848797,
+      "learning_rate": 1.8610364598781227e-06,
+      "loss": 0.9828,
+      "step": 4211
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.667684174363725,
+      "learning_rate": 1.8574178555130818e-06,
+      "loss": 1.0273,
+      "step": 4212
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.658857290616103,
+      "learning_rate": 1.8538024123787868e-06,
+      "loss": 1.0593,
+      "step": 4213
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6795618520264531,
+      "learning_rate": 1.8501901318788773e-06,
+      "loss": 0.9363,
+      "step": 4214
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7056871816839114,
+      "learning_rate": 1.8465810154157626e-06,
+      "loss": 1.025,
+      "step": 4215
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6865902994053135,
+      "learning_rate": 1.8429750643906331e-06,
+      "loss": 0.9322,
+      "step": 4216
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7183491486140359,
+      "learning_rate": 1.8393722802034331e-06,
+      "loss": 1.02,
+      "step": 4217
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7116151662409909,
+      "learning_rate": 1.835772664252895e-06,
+      "loss": 1.0584,
+      "step": 4218
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7521176262899772,
+      "learning_rate": 1.832176217936511e-06,
+      "loss": 0.9595,
+      "step": 4219
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.662645514843738,
+      "learning_rate": 1.8285829426505453e-06,
+      "loss": 1.0438,
+      "step": 4220
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7137210080528533,
+      "learning_rate": 1.8249928397900351e-06,
+      "loss": 0.856,
+      "step": 4221
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7059309693492468,
+      "learning_rate": 1.8214059107487726e-06,
+      "loss": 0.9448,
+      "step": 4222
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6852617678411793,
+      "learning_rate": 1.8178221569193343e-06,
+      "loss": 1.0266,
+      "step": 4223
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7583220371645347,
+      "learning_rate": 1.8142415796930568e-06,
+      "loss": 0.9635,
+      "step": 4224
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5955054164006133,
+      "learning_rate": 1.8106641804600411e-06,
+      "loss": 1.0458,
+      "step": 4225
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7969025163230418,
+      "learning_rate": 1.8070899606091586e-06,
+      "loss": 0.9067,
+      "step": 4226
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6828605703889805,
+      "learning_rate": 1.8035189215280423e-06,
+      "loss": 1.008,
+      "step": 4227
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7839694553741198,
+      "learning_rate": 1.799951064603095e-06,
+      "loss": 1.0562,
+      "step": 4228
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.852330349757284,
+      "learning_rate": 1.7963863912194768e-06,
+      "loss": 0.9935,
+      "step": 4229
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7066649704773694,
+      "learning_rate": 1.7928249027611255e-06,
+      "loss": 0.8951,
+      "step": 4230
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.694081364804734,
+      "learning_rate": 1.789266600610724e-06,
+      "loss": 0.9445,
+      "step": 4231
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7077639057027477,
+      "learning_rate": 1.7857114861497337e-06,
+      "loss": 1.0087,
+      "step": 4232
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6495131538450615,
+      "learning_rate": 1.782159560758373e-06,
+      "loss": 0.9533,
+      "step": 4233
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7209861279620374,
+      "learning_rate": 1.7786108258156154e-06,
+      "loss": 1.0038,
+      "step": 4234
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7245014610581344,
+      "learning_rate": 1.7750652826992077e-06,
+      "loss": 0.9028,
+      "step": 4235
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6193578216297779,
+      "learning_rate": 1.7715229327856498e-06,
+      "loss": 1.0349,
+      "step": 4236
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6566779371830048,
+      "learning_rate": 1.7679837774502052e-06,
+      "loss": 0.9883,
+      "step": 4237
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6651739270099257,
+      "learning_rate": 1.7644478180668945e-06,
+      "loss": 0.9724,
+      "step": 4238
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6824114344319011,
+      "learning_rate": 1.7609150560084986e-06,
+      "loss": 1.0716,
+      "step": 4239
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6842524297618214,
+      "learning_rate": 1.7573854926465582e-06,
+      "loss": 0.9788,
+      "step": 4240
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7013475970662498,
+      "learning_rate": 1.7538591293513685e-06,
+      "loss": 0.9738,
+      "step": 4241
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7103824317868388,
+      "learning_rate": 1.7503359674919929e-06,
+      "loss": 1.0098,
+      "step": 4242
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7545361543713333,
+      "learning_rate": 1.746816008436234e-06,
+      "loss": 1.0138,
+      "step": 4243
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7103878354023649,
+      "learning_rate": 1.7432992535506687e-06,
+      "loss": 1.0326,
+      "step": 4244
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6666598530359681,
+      "learning_rate": 1.7397857042006194e-06,
+      "loss": 1.0268,
+      "step": 4245
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8076213089441461,
+      "learning_rate": 1.736275361750167e-06,
+      "loss": 0.998,
+      "step": 4246
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7698459448638214,
+      "learning_rate": 1.7327682275621506e-06,
+      "loss": 0.9195,
+      "step": 4247
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7414973503032828,
+      "learning_rate": 1.7292643029981525e-06,
+      "loss": 0.9676,
+      "step": 4248
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.702032928200163,
+      "learning_rate": 1.7257635894185232e-06,
+      "loss": 1.027,
+      "step": 4249
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7938377582313234,
+      "learning_rate": 1.7222660881823594e-06,
+      "loss": 0.9717,
+      "step": 4250
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7697751167051596,
+      "learning_rate": 1.7187718006475117e-06,
+      "loss": 1.0034,
+      "step": 4251
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5446423795951058,
+      "learning_rate": 1.7152807281705809e-06,
+      "loss": 1.0159,
+      "step": 4252
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7192738742765996,
+      "learning_rate": 1.7117928721069233e-06,
+      "loss": 0.9875,
+      "step": 4253
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6936935912563628,
+      "learning_rate": 1.708308233810644e-06,
+      "loss": 1.0302,
+      "step": 4254
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7101738299876185,
+      "learning_rate": 1.704826814634597e-06,
+      "loss": 0.9643,
+      "step": 4255
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7022906509221933,
+      "learning_rate": 1.701348615930397e-06,
+      "loss": 0.951,
+      "step": 4256
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7404717100016978,
+      "learning_rate": 1.6978736390483896e-06,
+      "loss": 0.9833,
+      "step": 4257
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7495346652576309,
+      "learning_rate": 1.6944018853376898e-06,
+      "loss": 0.9669,
+      "step": 4258
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7034037651372977,
+      "learning_rate": 1.6909333561461471e-06,
+      "loss": 0.9302,
+      "step": 4259
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5681546349971209,
+      "learning_rate": 1.6874680528203657e-06,
+      "loss": 0.985,
+      "step": 4260
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6511876211088181,
+      "learning_rate": 1.6840059767056949e-06,
+      "loss": 0.8461,
+      "step": 4261
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6845234508626911,
+      "learning_rate": 1.6805471291462316e-06,
+      "loss": 0.9667,
+      "step": 4262
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6604471488948964,
+      "learning_rate": 1.6770915114848197e-06,
+      "loss": 0.9695,
+      "step": 4263
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.588469181731601,
+      "learning_rate": 1.67363912506305e-06,
+      "loss": 1.0029,
+      "step": 4264
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7258837644769525,
+      "learning_rate": 1.6701899712212565e-06,
+      "loss": 1.0872,
+      "step": 4265
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6727173141778895,
+      "learning_rate": 1.66674405129852e-06,
+      "loss": 0.9467,
+      "step": 4266
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6507306471588556,
+      "learning_rate": 1.6633013666326636e-06,
+      "loss": 0.9598,
+      "step": 4267
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.714503463318509,
+      "learning_rate": 1.6598619185602616e-06,
+      "loss": 0.9609,
+      "step": 4268
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5211766296326683,
+      "learning_rate": 1.656425708416617e-06,
+      "loss": 1.024,
+      "step": 4269
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7779103794501024,
+      "learning_rate": 1.6529927375357957e-06,
+      "loss": 0.9571,
+      "step": 4270
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7261716790199697,
+      "learning_rate": 1.6495630072505841e-06,
+      "loss": 0.9891,
+      "step": 4271
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8024714598402403,
+      "learning_rate": 1.6461365188925304e-06,
+      "loss": 1.0094,
+      "step": 4272
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6542196786401129,
+      "learning_rate": 1.642713273791914e-06,
+      "loss": 1.0,
+      "step": 4273
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6833484427225497,
+      "learning_rate": 1.6392932732777489e-06,
+      "loss": 1.0162,
+      "step": 4274
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7087034754906328,
+      "learning_rate": 1.6358765186778057e-06,
+      "loss": 0.9921,
+      "step": 4275
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6755922529822368,
+      "learning_rate": 1.6324630113185835e-06,
+      "loss": 0.9559,
+      "step": 4276
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6822203448304328,
+      "learning_rate": 1.629052752525323e-06,
+      "loss": 1.0273,
+      "step": 4277
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7482464917580245,
+      "learning_rate": 1.625645743622003e-06,
+      "loss": 1.0435,
+      "step": 4278
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6769005288819985,
+      "learning_rate": 1.6222419859313443e-06,
+      "loss": 0.9575,
+      "step": 4279
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5235469013106813,
+      "learning_rate": 1.6188414807747999e-06,
+      "loss": 1.065,
+      "step": 4280
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6938958192208065,
+      "learning_rate": 1.6154442294725636e-06,
+      "loss": 1.0477,
+      "step": 4281
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7098862743419923,
+      "learning_rate": 1.6120502333435695e-06,
+      "loss": 0.9917,
+      "step": 4282
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6656515479007469,
+      "learning_rate": 1.6086594937054767e-06,
+      "loss": 0.9403,
+      "step": 4283
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7784816543466212,
+      "learning_rate": 1.6052720118746923e-06,
+      "loss": 0.9982,
+      "step": 4284
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7395104133454578,
+      "learning_rate": 1.6018877891663521e-06,
+      "loss": 0.9804,
+      "step": 4285
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6595532120426914,
+      "learning_rate": 1.5985068268943283e-06,
+      "loss": 1.0717,
+      "step": 4286
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6823317874290151,
+      "learning_rate": 1.5951291263712255e-06,
+      "loss": 0.9823,
+      "step": 4287
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.691617773152197,
+      "learning_rate": 1.5917546889083834e-06,
+      "loss": 0.9311,
+      "step": 4288
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.734556486619831,
+      "learning_rate": 1.5883835158158767e-06,
+      "loss": 0.9086,
+      "step": 4289
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7193269742802025,
+      "learning_rate": 1.5850156084025091e-06,
+      "loss": 1.0363,
+      "step": 4290
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5549441122531691,
+      "learning_rate": 1.5816509679758185e-06,
+      "loss": 1.0983,
+      "step": 4291
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.719820455202011,
+      "learning_rate": 1.578289595842074e-06,
+      "loss": 0.9814,
+      "step": 4292
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7744872924998399,
+      "learning_rate": 1.5749314933062754e-06,
+      "loss": 0.9709,
+      "step": 4293
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7356735267200611,
+      "learning_rate": 1.5715766616721584e-06,
+      "loss": 0.9712,
+      "step": 4294
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5414969797064112,
+      "learning_rate": 1.5682251022421757e-06,
+      "loss": 1.0383,
+      "step": 4295
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.8062629026529549,
+      "learning_rate": 1.5648768163175277e-06,
+      "loss": 0.9547,
+      "step": 4296
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7085880258091346,
+      "learning_rate": 1.5615318051981243e-06,
+      "loss": 1.0184,
+      "step": 4297
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6182994811193931,
+      "learning_rate": 1.5581900701826226e-06,
+      "loss": 0.9313,
+      "step": 4298
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6809863683897431,
+      "learning_rate": 1.5548516125683976e-06,
+      "loss": 0.9965,
+      "step": 4299
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7399975659586786,
+      "learning_rate": 1.5515164336515465e-06,
+      "loss": 1.0753,
+      "step": 4300
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6651443190721565,
+      "learning_rate": 1.5481845347269077e-06,
+      "loss": 1.0068,
+      "step": 4301
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6917787551702872,
+      "learning_rate": 1.5448559170880373e-06,
+      "loss": 0.9817,
+      "step": 4302
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7085869844735849,
+      "learning_rate": 1.5415305820272198e-06,
+      "loss": 0.9941,
+      "step": 4303
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7027197120373849,
+      "learning_rate": 1.5382085308354633e-06,
+      "loss": 1.0259,
+      "step": 4304
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6847704733159696,
+      "learning_rate": 1.534889764802503e-06,
+      "loss": 1.0513,
+      "step": 4305
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7198378691857563,
+      "learning_rate": 1.5315742852167992e-06,
+      "loss": 1.0337,
+      "step": 4306
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7101975904456828,
+      "learning_rate": 1.5282620933655312e-06,
+      "loss": 1.0123,
+      "step": 4307
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5428406314400208,
+      "learning_rate": 1.5249531905346138e-06,
+      "loss": 1.0049,
+      "step": 4308
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6904723270452024,
+      "learning_rate": 1.521647578008667e-06,
+      "loss": 1.0099,
+      "step": 4309
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7541758104905149,
+      "learning_rate": 1.5183452570710522e-06,
+      "loss": 0.9751,
+      "step": 4310
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6971681295813466,
+      "learning_rate": 1.5150462290038392e-06,
+      "loss": 0.9733,
+      "step": 4311
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6757746633056257,
+      "learning_rate": 1.511750495087827e-06,
+      "loss": 0.9134,
+      "step": 4312
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5357076321866113,
+      "learning_rate": 1.5084580566025309e-06,
+      "loss": 1.0316,
+      "step": 4313
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6696233312321765,
+      "learning_rate": 1.5051689148261895e-06,
+      "loss": 0.9828,
+      "step": 4314
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7053672887103296,
+      "learning_rate": 1.5018830710357612e-06,
+      "loss": 1.0419,
+      "step": 4315
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6817662236628521,
+      "learning_rate": 1.4986005265069204e-06,
+      "loss": 1.0009,
+      "step": 4316
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7164269230447711,
+      "learning_rate": 1.4953212825140728e-06,
+      "loss": 0.9862,
+      "step": 4317
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7256496196300779,
+      "learning_rate": 1.4920453403303249e-06,
+      "loss": 0.9052,
+      "step": 4318
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6335620554357664,
+      "learning_rate": 1.4887727012275112e-06,
+      "loss": 0.9787,
+      "step": 4319
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5326226192972258,
+      "learning_rate": 1.4855033664761898e-06,
+      "loss": 1.0135,
+      "step": 4320
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6835202161029595,
+      "learning_rate": 1.48223733734562e-06,
+      "loss": 1.0164,
+      "step": 4321
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7203636991629484,
+      "learning_rate": 1.4789746151037942e-06,
+      "loss": 0.9544,
+      "step": 4322
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6638553230721231,
+      "learning_rate": 1.475715201017407e-06,
+      "loss": 0.9855,
+      "step": 4323
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6766439610007583,
+      "learning_rate": 1.4724590963518803e-06,
+      "loss": 1.0364,
+      "step": 4324
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7397650468753707,
+      "learning_rate": 1.4692063023713444e-06,
+      "loss": 1.0256,
+      "step": 4325
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.655475837394072,
+      "learning_rate": 1.4659568203386464e-06,
+      "loss": 1.0136,
+      "step": 4326
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7055498669762608,
+      "learning_rate": 1.4627106515153456e-06,
+      "loss": 1.0123,
+      "step": 4327
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6832570205411899,
+      "learning_rate": 1.4594677971617178e-06,
+      "loss": 1.062,
+      "step": 4328
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7532307055732128,
+      "learning_rate": 1.4562282585367493e-06,
+      "loss": 0.9828,
+      "step": 4329
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7378312668812032,
+      "learning_rate": 1.452992036898142e-06,
+      "loss": 1.0026,
+      "step": 4330
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5982393790466648,
+      "learning_rate": 1.4497591335023087e-06,
+      "loss": 1.0248,
+      "step": 4331
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6914099250696654,
+      "learning_rate": 1.446529549604373e-06,
+      "loss": 1.0631,
+      "step": 4332
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7429136898009422,
+      "learning_rate": 1.4433032864581687e-06,
+      "loss": 0.9683,
+      "step": 4333
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6791672862319058,
+      "learning_rate": 1.4400803453162482e-06,
+      "loss": 0.9962,
+      "step": 4334
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7416092059583261,
+      "learning_rate": 1.4368607274298596e-06,
+      "loss": 0.9709,
+      "step": 4335
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6640491131183098,
+      "learning_rate": 1.4336444340489775e-06,
+      "loss": 1.0574,
+      "step": 4336
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6787654169917252,
+      "learning_rate": 1.430431466422273e-06,
+      "loss": 0.9711,
+      "step": 4337
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7755650238678398,
+      "learning_rate": 1.4272218257971327e-06,
+      "loss": 1.0229,
+      "step": 4338
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7045846462270665,
+      "learning_rate": 1.4240155134196499e-06,
+      "loss": 0.9843,
+      "step": 4339
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.7034295417623551,
+      "learning_rate": 1.4208125305346232e-06,
+      "loss": 0.9286,
+      "step": 4340
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6910640893525605,
+      "learning_rate": 1.4176128783855636e-06,
+      "loss": 1.0073,
+      "step": 4341
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6870182723315864,
+      "learning_rate": 1.4144165582146819e-06,
+      "loss": 0.9943,
+      "step": 4342
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7576408858167888,
+      "learning_rate": 1.4112235712629063e-06,
+      "loss": 0.9858,
+      "step": 4343
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6569946354206478,
+      "learning_rate": 1.40803391876986e-06,
+      "loss": 1.0046,
+      "step": 4344
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7398343982306335,
+      "learning_rate": 1.4048476019738756e-06,
+      "loss": 1.0529,
+      "step": 4345
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7208771344749973,
+      "learning_rate": 1.4016646221119912e-06,
+      "loss": 1.0377,
+      "step": 4346
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7917999390457722,
+      "learning_rate": 1.3984849804199485e-06,
+      "loss": 1.0533,
+      "step": 4347
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6400514678207719,
+      "learning_rate": 1.395308678132199e-06,
+      "loss": 1.0339,
+      "step": 4348
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7178443812710148,
+      "learning_rate": 1.392135716481885e-06,
+      "loss": 0.9731,
+      "step": 4349
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6639624196569095,
+      "learning_rate": 1.3889660967008656e-06,
+      "loss": 0.9525,
+      "step": 4350
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.670966447878471,
+      "learning_rate": 1.3857998200196943e-06,
+      "loss": 1.0162,
+      "step": 4351
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6233133225831482,
+      "learning_rate": 1.3826368876676278e-06,
+      "loss": 1.0735,
+      "step": 4352
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7044283102071452,
+      "learning_rate": 1.379477300872626e-06,
+      "loss": 0.9424,
+      "step": 4353
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7053339662714533,
+      "learning_rate": 1.3763210608613497e-06,
+      "loss": 1.0506,
+      "step": 4354
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.49993361876176373,
+      "learning_rate": 1.3731681688591593e-06,
+      "loss": 1.038,
+      "step": 4355
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6772132705117069,
+      "learning_rate": 1.370018626090116e-06,
+      "loss": 0.9354,
+      "step": 4356
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7181856175801943,
+      "learning_rate": 1.3668724337769823e-06,
+      "loss": 1.0342,
+      "step": 4357
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7188405459339587,
+      "learning_rate": 1.3637295931412153e-06,
+      "loss": 1.0311,
+      "step": 4358
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8067739245111044,
+      "learning_rate": 1.3605901054029746e-06,
+      "loss": 1.0316,
+      "step": 4359
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7336377470512645,
+      "learning_rate": 1.3574539717811231e-06,
+      "loss": 0.9934,
+      "step": 4360
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7079730502384319,
+      "learning_rate": 1.3543211934932065e-06,
+      "loss": 0.9925,
+      "step": 4361
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5858862990807044,
+      "learning_rate": 1.3511917717554846e-06,
+      "loss": 1.0471,
+      "step": 4362
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5458041859593987,
+      "learning_rate": 1.348065707782904e-06,
+      "loss": 1.0557,
+      "step": 4363
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6833362982316744,
+      "learning_rate": 1.3449430027891096e-06,
+      "loss": 1.0047,
+      "step": 4364
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8271653785437397,
+      "learning_rate": 1.3418236579864452e-06,
+      "loss": 1.0121,
+      "step": 4365
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7047915842087129,
+      "learning_rate": 1.338707674585945e-06,
+      "loss": 0.9789,
+      "step": 4366
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.626576213972416,
+      "learning_rate": 1.3355950537973438e-06,
+      "loss": 0.9942,
+      "step": 4367
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5003572165245801,
+      "learning_rate": 1.332485796829065e-06,
+      "loss": 1.0273,
+      "step": 4368
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7641196221532645,
+      "learning_rate": 1.329379904888235e-06,
+      "loss": 1.0222,
+      "step": 4369
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6994119936005674,
+      "learning_rate": 1.3262773791806617e-06,
+      "loss": 0.9207,
+      "step": 4370
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7223974183830751,
+      "learning_rate": 1.3231782209108546e-06,
+      "loss": 0.9525,
+      "step": 4371
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6322122953009325,
+      "learning_rate": 1.3200824312820137e-06,
+      "loss": 1.0511,
+      "step": 4372
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7223901479841147,
+      "learning_rate": 1.3169900114960298e-06,
+      "loss": 1.0094,
+      "step": 4373
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8170749559225432,
+      "learning_rate": 1.3139009627534927e-06,
+      "loss": 1.0391,
+      "step": 4374
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7206889154189265,
+      "learning_rate": 1.3108152862536683e-06,
+      "loss": 1.0045,
+      "step": 4375
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5801782359616745,
+      "learning_rate": 1.3077329831945295e-06,
+      "loss": 1.0397,
+      "step": 4376
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6947434388203404,
+      "learning_rate": 1.3046540547727305e-06,
+      "loss": 0.9232,
+      "step": 4377
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7265393914205619,
+      "learning_rate": 1.3015785021836159e-06,
+      "loss": 1.0624,
+      "step": 4378
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6875638912604789,
+      "learning_rate": 1.2985063266212229e-06,
+      "loss": 0.9801,
+      "step": 4379
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.704467537920351,
+      "learning_rate": 1.295437529278275e-06,
+      "loss": 1.0048,
+      "step": 4380
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.8179312392222771,
+      "learning_rate": 1.2923721113461852e-06,
+      "loss": 1.009,
+      "step": 4381
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6881081790272294,
+      "learning_rate": 1.2893100740150522e-06,
+      "loss": 0.936,
+      "step": 4382
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5666577613644832,
+      "learning_rate": 1.2862514184736695e-06,
+      "loss": 1.0513,
+      "step": 4383
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.744459671056804,
+      "learning_rate": 1.2831961459095088e-06,
+      "loss": 1.0005,
+      "step": 4384
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6835703041497307,
+      "learning_rate": 1.2801442575087296e-06,
+      "loss": 1.0282,
+      "step": 4385
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.878691724401112,
+      "learning_rate": 1.2770957544561868e-06,
+      "loss": 1.03,
+      "step": 4386
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.56406359339663,
+      "learning_rate": 1.274050637935408e-06,
+      "loss": 1.0459,
+      "step": 4387
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7317489589356944,
+      "learning_rate": 1.2710089091286148e-06,
+      "loss": 0.9496,
+      "step": 4388
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5523955841391262,
+      "learning_rate": 1.2679705692167122e-06,
+      "loss": 1.0072,
+      "step": 4389
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5863316187209606,
+      "learning_rate": 1.2649356193792873e-06,
+      "loss": 1.0144,
+      "step": 4390
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7476510789048302,
+      "learning_rate": 1.261904060794612e-06,
+      "loss": 1.0005,
+      "step": 4391
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.742437593344893,
+      "learning_rate": 1.2588758946396417e-06,
+      "loss": 1.0215,
+      "step": 4392
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5849197988810795,
+      "learning_rate": 1.2558511220900138e-06,
+      "loss": 1.0445,
+      "step": 4393
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7221860107294237,
+      "learning_rate": 1.2528297443200489e-06,
+      "loss": 0.9762,
+      "step": 4394
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6697409873522996,
+      "learning_rate": 1.2498117625027562e-06,
+      "loss": 1.0287,
+      "step": 4395
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7156403891260525,
+      "learning_rate": 1.246797177809812e-06,
+      "loss": 1.0165,
+      "step": 4396
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6819125400674281,
+      "learning_rate": 1.2437859914115847e-06,
+      "loss": 0.9794,
+      "step": 4397
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6524122536618725,
+      "learning_rate": 1.2407782044771211e-06,
+      "loss": 1.0298,
+      "step": 4398
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6967216439613825,
+      "learning_rate": 1.237773818174146e-06,
+      "loss": 0.9797,
+      "step": 4399
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5503021929912681,
+      "learning_rate": 1.23477283366907e-06,
+      "loss": 1.0747,
+      "step": 4400
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6772134054181762,
+      "learning_rate": 1.2317752521269722e-06,
+      "loss": 0.9852,
+      "step": 4401
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7260602881535825,
+      "learning_rate": 1.2287810747116224e-06,
+      "loss": 0.977,
+      "step": 4402
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6710007274684218,
+      "learning_rate": 1.225790302585461e-06,
+      "loss": 1.0295,
+      "step": 4403
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6299615793976279,
+      "learning_rate": 1.2228029369096094e-06,
+      "loss": 0.9896,
+      "step": 4404
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.707293534181908,
+      "learning_rate": 1.2198189788438652e-06,
+      "loss": 0.9571,
+      "step": 4405
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7280062238340327,
+      "learning_rate": 1.216838429546704e-06,
+      "loss": 1.0148,
+      "step": 4406
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6572789871359338,
+      "learning_rate": 1.2138612901752777e-06,
+      "loss": 1.0209,
+      "step": 4407
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7200000428070026,
+      "learning_rate": 1.2108875618854122e-06,
+      "loss": 1.0559,
+      "step": 4408
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6351595475591574,
+      "learning_rate": 1.2079172458316168e-06,
+      "loss": 1.014,
+      "step": 4409
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6945897083096075,
+      "learning_rate": 1.204950343167065e-06,
+      "loss": 1.0309,
+      "step": 4410
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5781172568683186,
+      "learning_rate": 1.2019868550436099e-06,
+      "loss": 0.9995,
+      "step": 4411
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7438474910889724,
+      "learning_rate": 1.1990267826117874e-06,
+      "loss": 1.0192,
+      "step": 4412
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6387148964097528,
+      "learning_rate": 1.1960701270207885e-06,
+      "loss": 1.0084,
+      "step": 4413
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6556329491511151,
+      "learning_rate": 1.1931168894184974e-06,
+      "loss": 0.9581,
+      "step": 4414
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6349525384242678,
+      "learning_rate": 1.19016707095146e-06,
+      "loss": 0.9354,
+      "step": 4415
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7361597501330766,
+      "learning_rate": 1.187220672764897e-06,
+      "loss": 0.9763,
+      "step": 4416
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.657523680489175,
+      "learning_rate": 1.1842776960027014e-06,
+      "loss": 0.9214,
+      "step": 4417
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6615680807399498,
+      "learning_rate": 1.1813381418074388e-06,
+      "loss": 0.9434,
+      "step": 4418
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6881522261012931,
+      "learning_rate": 1.1784020113203453e-06,
+      "loss": 1.0423,
+      "step": 4419
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6139258257125039,
+      "learning_rate": 1.1754693056813272e-06,
+      "loss": 0.975,
+      "step": 4420
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6873491691982063,
+      "learning_rate": 1.172540026028962e-06,
+      "loss": 1.0463,
+      "step": 4421
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.8370019947741217,
+      "learning_rate": 1.169614173500494e-06,
+      "loss": 0.9294,
+      "step": 4422
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6374134404868147,
+      "learning_rate": 1.1666917492318486e-06,
+      "loss": 1.0063,
+      "step": 4423
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6466190405577855,
+      "learning_rate": 1.1637727543576027e-06,
+      "loss": 1.0292,
+      "step": 4424
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6545015733476341,
+      "learning_rate": 1.1608571900110122e-06,
+      "loss": 0.9244,
+      "step": 4425
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6737336527365093,
+      "learning_rate": 1.1579450573240058e-06,
+      "loss": 0.9849,
+      "step": 4426
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7036794573709129,
+      "learning_rate": 1.1550363574271638e-06,
+      "loss": 0.9473,
+      "step": 4427
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5224408396475477,
+      "learning_rate": 1.1521310914497518e-06,
+      "loss": 0.9971,
+      "step": 4428
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7121651757678407,
+      "learning_rate": 1.149229260519691e-06,
+      "loss": 0.9963,
+      "step": 4429
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5143584363569619,
+      "learning_rate": 1.1463308657635718e-06,
+      "loss": 1.0323,
+      "step": 4430
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7549302652564969,
+      "learning_rate": 1.1434359083066515e-06,
+      "loss": 0.86,
+      "step": 4431
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7401908748101782,
+      "learning_rate": 1.140544389272853e-06,
+      "loss": 0.9919,
+      "step": 4432
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6529087276714727,
+      "learning_rate": 1.1376563097847616e-06,
+      "loss": 1.0315,
+      "step": 4433
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7342882125453501,
+      "learning_rate": 1.1347716709636282e-06,
+      "loss": 0.9855,
+      "step": 4434
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.717036106896565,
+      "learning_rate": 1.1318904739293745e-06,
+      "loss": 0.9609,
+      "step": 4435
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5840190551788376,
+      "learning_rate": 1.129012719800575e-06,
+      "loss": 1.0734,
+      "step": 4436
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7295328045409671,
+      "learning_rate": 1.1261384096944728e-06,
+      "loss": 0.9409,
+      "step": 4437
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7140883852545754,
+      "learning_rate": 1.1232675447269803e-06,
+      "loss": 0.9796,
+      "step": 4438
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6698119283377796,
+      "learning_rate": 1.1204001260126574e-06,
+      "loss": 1.0347,
+      "step": 4439
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7325641490849334,
+      "learning_rate": 1.1175361546647413e-06,
+      "loss": 0.9493,
+      "step": 4440
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7454300881160001,
+      "learning_rate": 1.1146756317951224e-06,
+      "loss": 0.9737,
+      "step": 4441
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7361670134165963,
+      "learning_rate": 1.1118185585143536e-06,
+      "loss": 0.9167,
+      "step": 4442
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7052465201483643,
+      "learning_rate": 1.1089649359316501e-06,
+      "loss": 0.964,
+      "step": 4443
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7329556755583521,
+      "learning_rate": 1.1061147651548855e-06,
+      "loss": 0.978,
+      "step": 4444
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6967446770791484,
+      "learning_rate": 1.1032680472905932e-06,
+      "loss": 0.9995,
+      "step": 4445
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8037571438753034,
+      "learning_rate": 1.1004247834439697e-06,
+      "loss": 1.0681,
+      "step": 4446
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7138737494778168,
+      "learning_rate": 1.097584974718866e-06,
+      "loss": 1.063,
+      "step": 4447
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6587674958110458,
+      "learning_rate": 1.0947486222177928e-06,
+      "loss": 0.9828,
+      "step": 4448
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6816110356061432,
+      "learning_rate": 1.0919157270419257e-06,
+      "loss": 0.9631,
+      "step": 4449
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6979282853831922,
+      "learning_rate": 1.0890862902910849e-06,
+      "loss": 1.0024,
+      "step": 4450
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7105411372134726,
+      "learning_rate": 1.0862603130637562e-06,
+      "loss": 1.009,
+      "step": 4451
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7149912053664631,
+      "learning_rate": 1.0834377964570863e-06,
+      "loss": 1.0426,
+      "step": 4452
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6856576368027062,
+      "learning_rate": 1.0806187415668668e-06,
+      "loss": 0.9984,
+      "step": 4453
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6724296201744202,
+      "learning_rate": 1.0778031494875574e-06,
+      "loss": 1.024,
+      "step": 4454
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7020448479411264,
+      "learning_rate": 1.0749910213122649e-06,
+      "loss": 1.0236,
+      "step": 4455
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.781516360351214,
+      "learning_rate": 1.072182358132755e-06,
+      "loss": 1.014,
+      "step": 4456
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6964815633013745,
+      "learning_rate": 1.0693771610394477e-06,
+      "loss": 0.9534,
+      "step": 4457
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6984373307173073,
+      "learning_rate": 1.066575431121417e-06,
+      "loss": 0.9427,
+      "step": 4458
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5131463447487065,
+      "learning_rate": 1.06377716946639e-06,
+      "loss": 0.9564,
+      "step": 4459
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7082316393077398,
+      "learning_rate": 1.0609823771607487e-06,
+      "loss": 1.0034,
+      "step": 4460
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7655142592956533,
+      "learning_rate": 1.0581910552895302e-06,
+      "loss": 1.0248,
+      "step": 4461
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7154832274698829,
+      "learning_rate": 1.055403204936416e-06,
+      "loss": 1.0699,
+      "step": 4462
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7101836639461616,
+      "learning_rate": 1.0526188271837512e-06,
+      "loss": 0.9492,
+      "step": 4463
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6707799640482497,
+      "learning_rate": 1.0498379231125278e-06,
+      "loss": 0.9446,
+      "step": 4464
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6468666782531336,
+      "learning_rate": 1.047060493802381e-06,
+      "loss": 0.9592,
+      "step": 4465
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6587335410098857,
+      "learning_rate": 1.0442865403316117e-06,
+      "loss": 1.1318,
+      "step": 4466
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7651310032133888,
+      "learning_rate": 1.0415160637771604e-06,
+      "loss": 0.925,
+      "step": 4467
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.667534289163843,
+      "learning_rate": 1.0387490652146236e-06,
+      "loss": 0.9832,
+      "step": 4468
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6769187996783778,
+      "learning_rate": 1.0359855457182455e-06,
+      "loss": 1.0315,
+      "step": 4469
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7353873626924243,
+      "learning_rate": 1.0332255063609177e-06,
+      "loss": 0.9566,
+      "step": 4470
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6563077581985954,
+      "learning_rate": 1.0304689482141839e-06,
+      "loss": 1.0093,
+      "step": 4471
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7894319937673598,
+      "learning_rate": 1.027715872348234e-06,
+      "loss": 1.0353,
+      "step": 4472
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.692111539601145,
+      "learning_rate": 1.0249662798319072e-06,
+      "loss": 1.0244,
+      "step": 4473
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7526660659872044,
+      "learning_rate": 1.0222201717326885e-06,
+      "loss": 1.068,
+      "step": 4474
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7333164654210027,
+      "learning_rate": 1.0194775491167164e-06,
+      "loss": 0.9594,
+      "step": 4475
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6738078189079542,
+      "learning_rate": 1.0167384130487667e-06,
+      "loss": 1.0041,
+      "step": 4476
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7487928611189901,
+      "learning_rate": 1.0140027645922656e-06,
+      "loss": 1.0654,
+      "step": 4477
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6708189290846505,
+      "learning_rate": 1.0112706048092924e-06,
+      "loss": 1.0046,
+      "step": 4478
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6618883543190408,
+      "learning_rate": 1.0085419347605575e-06,
+      "loss": 0.9736,
+      "step": 4479
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7158197231071054,
+      "learning_rate": 1.00581675550543e-06,
+      "loss": 1.1042,
+      "step": 4480
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7101991881696998,
+      "learning_rate": 1.003095068101917e-06,
+      "loss": 0.9301,
+      "step": 4481
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6691846265765812,
+      "learning_rate": 1.0003768736066722e-06,
+      "loss": 0.9689,
+      "step": 4482
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.49370111112024667,
+      "learning_rate": 9.976621730749892e-07,
+      "loss": 1.1097,
+      "step": 4483
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8052977439191635,
+      "learning_rate": 9.949509675608115e-07,
+      "loss": 1.0074,
+      "step": 4484
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7382392727276668,
+      "learning_rate": 9.922432581167207e-07,
+      "loss": 0.9929,
+      "step": 4485
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7167760111282576,
+      "learning_rate": 9.895390457939414e-07,
+      "loss": 0.9761,
+      "step": 4486
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.696903352970271,
+      "learning_rate": 9.86838331642348e-07,
+      "loss": 1.0137,
+      "step": 4487
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.701774763486246,
+      "learning_rate": 9.84141116710442e-07,
+      "loss": 1.0608,
+      "step": 4488
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6318416857860532,
+      "learning_rate": 9.814474020453824e-07,
+      "loss": 1.0154,
+      "step": 4489
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7603397026013967,
+      "learning_rate": 9.787571886929604e-07,
+      "loss": 1.0029,
+      "step": 4490
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6684770552031326,
+      "learning_rate": 9.76070477697605e-07,
+      "loss": 1.0048,
+      "step": 4491
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7609510104717992,
+      "learning_rate": 9.733872701023938e-07,
+      "loss": 0.869,
+      "step": 4492
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7076872262503642,
+      "learning_rate": 9.707075669490407e-07,
+      "loss": 0.9471,
+      "step": 4493
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6712426546866042,
+      "learning_rate": 9.680313692778976e-07,
+      "loss": 0.993,
+      "step": 4494
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6937612431555864,
+      "learning_rate": 9.653586781279567e-07,
+      "loss": 1.0369,
+      "step": 4495
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6504139876704537,
+      "learning_rate": 9.626894945368492e-07,
+      "loss": 1.0121,
+      "step": 4496
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6612869650872917,
+      "learning_rate": 9.600238195408428e-07,
+      "loss": 1.0302,
+      "step": 4497
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7431077459749891,
+      "learning_rate": 9.573616541748464e-07,
+      "loss": 0.9624,
+      "step": 4498
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.720066632579313,
+      "learning_rate": 9.547029994724023e-07,
+      "loss": 1.0321,
+      "step": 4499
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.742119106889851,
+      "learning_rate": 9.520478564656898e-07,
+      "loss": 0.9918,
+      "step": 4500
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6761017381226446,
+      "learning_rate": 9.49396226185535e-07,
+      "loss": 1.0074,
+      "step": 4501
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7324640727877494,
+      "learning_rate": 9.467481096613829e-07,
+      "loss": 0.9359,
+      "step": 4502
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.68134212691625,
+      "learning_rate": 9.441035079213267e-07,
+      "loss": 0.9863,
+      "step": 4503
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6619667692978868,
+      "learning_rate": 9.414624219920953e-07,
+      "loss": 0.9765,
+      "step": 4504
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7055950145657969,
+      "learning_rate": 9.38824852899043e-07,
+      "loss": 1.0215,
+      "step": 4505
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6813926291737087,
+      "learning_rate": 9.361908016661703e-07,
+      "loss": 0.9672,
+      "step": 4506
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7452425773714629,
+      "learning_rate": 9.335602693161039e-07,
+      "loss": 1.0371,
+      "step": 4507
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6539948161404885,
+      "learning_rate": 9.309332568701079e-07,
+      "loss": 1.0128,
+      "step": 4508
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6365426861579337,
+      "learning_rate": 9.283097653480788e-07,
+      "loss": 0.9319,
+      "step": 4509
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6727865796423177,
+      "learning_rate": 9.256897957685463e-07,
+      "loss": 0.9309,
+      "step": 4510
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.702283021350003,
+      "learning_rate": 9.230733491486721e-07,
+      "loss": 1.0377,
+      "step": 4511
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7077100550771795,
+      "learning_rate": 9.204604265042505e-07,
+      "loss": 1.0013,
+      "step": 4512
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6743141816321336,
+      "learning_rate": 9.178510288497123e-07,
+      "loss": 0.9905,
+      "step": 4513
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6847485370075141,
+      "learning_rate": 9.15245157198108e-07,
+      "loss": 0.9956,
+      "step": 4514
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6700125871790724,
+      "learning_rate": 9.126428125611342e-07,
+      "loss": 1.007,
+      "step": 4515
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.554199165389211,
+      "learning_rate": 9.10043995949108e-07,
+      "loss": 1.0721,
+      "step": 4516
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7798428081998837,
+      "learning_rate": 9.074487083709759e-07,
+      "loss": 0.9415,
+      "step": 4517
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6903254601050316,
+      "learning_rate": 9.04856950834323e-07,
+      "loss": 0.9741,
+      "step": 4518
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6729427767123548,
+      "learning_rate": 9.022687243453554e-07,
+      "loss": 0.9465,
+      "step": 4519
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6741014283646936,
+      "learning_rate": 8.996840299089149e-07,
+      "loss": 0.9961,
+      "step": 4520
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6874492629805334,
+      "learning_rate": 8.971028685284655e-07,
+      "loss": 0.9583,
+      "step": 4521
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6863604302106502,
+      "learning_rate": 8.945252412061056e-07,
+      "loss": 0.9293,
+      "step": 4522
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.713674816215144,
+      "learning_rate": 8.91951148942557e-07,
+      "loss": 0.9944,
+      "step": 4523
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5501591221803986,
+      "learning_rate": 8.893805927371724e-07,
+      "loss": 1.0046,
+      "step": 4524
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7833958598153972,
+      "learning_rate": 8.868135735879291e-07,
+      "loss": 1.0071,
+      "step": 4525
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.746457786976404,
+      "learning_rate": 8.842500924914299e-07,
+      "loss": 0.997,
+      "step": 4526
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6650614670891734,
+      "learning_rate": 8.816901504429143e-07,
+      "loss": 1.0055,
+      "step": 4527
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7831232329534226,
+      "learning_rate": 8.791337484362305e-07,
+      "loss": 1.0771,
+      "step": 4528
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.757589566887932,
+      "learning_rate": 8.765808874638682e-07,
+      "loss": 1.0123,
+      "step": 4529
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7395484167407184,
+      "learning_rate": 8.740315685169364e-07,
+      "loss": 0.942,
+      "step": 4530
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4943195860614081,
+      "learning_rate": 8.714857925851617e-07,
+      "loss": 1.0778,
+      "step": 4531
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5715186816373864,
+      "learning_rate": 8.689435606569086e-07,
+      "loss": 1.0325,
+      "step": 4532
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7793201397155887,
+      "learning_rate": 8.664048737191566e-07,
+      "loss": 1.0077,
+      "step": 4533
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7323942441987953,
+      "learning_rate": 8.638697327575108e-07,
+      "loss": 1.0408,
+      "step": 4534
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7098535658687878,
+      "learning_rate": 8.613381387562015e-07,
+      "loss": 0.9767,
+      "step": 4535
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6670051197855311,
+      "learning_rate": 8.588100926980802e-07,
+      "loss": 0.9295,
+      "step": 4536
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7185775674129252,
+      "learning_rate": 8.56285595564621e-07,
+      "loss": 0.9865,
+      "step": 4537
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6067005059670424,
+      "learning_rate": 8.537646483359185e-07,
+      "loss": 1.0359,
+      "step": 4538
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5367060011034405,
+      "learning_rate": 8.512472519906978e-07,
+      "loss": 1.0272,
+      "step": 4539
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6676751363164487,
+      "learning_rate": 8.487334075062914e-07,
+      "loss": 1.0752,
+      "step": 4540
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5251947510388536,
+      "learning_rate": 8.462231158586654e-07,
+      "loss": 1.0458,
+      "step": 4541
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6497967077482942,
+      "learning_rate": 8.437163780224011e-07,
+      "loss": 0.9322,
+      "step": 4542
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7248927245345167,
+      "learning_rate": 8.412131949706958e-07,
+      "loss": 1.0175,
+      "step": 4543
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.741315999307165,
+      "learning_rate": 8.387135676753755e-07,
+      "loss": 0.9794,
+      "step": 4544
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6613794998414756,
+      "learning_rate": 8.362174971068804e-07,
+      "loss": 0.9686,
+      "step": 4545
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6787655659844015,
+      "learning_rate": 8.337249842342721e-07,
+      "loss": 0.9651,
+      "step": 4546
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6149255009490237,
+      "learning_rate": 8.312360300252287e-07,
+      "loss": 1.0453,
+      "step": 4547
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7945312736113969,
+      "learning_rate": 8.287506354460484e-07,
+      "loss": 0.9765,
+      "step": 4548
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6747322517983433,
+      "learning_rate": 8.26268801461646e-07,
+      "loss": 1.0631,
+      "step": 4549
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7066739879409205,
+      "learning_rate": 8.237905290355563e-07,
+      "loss": 1.0725,
+      "step": 4550
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7806401943484624,
+      "learning_rate": 8.213158191299297e-07,
+      "loss": 0.9633,
+      "step": 4551
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6789071434569027,
+      "learning_rate": 8.188446727055311e-07,
+      "loss": 0.9557,
+      "step": 4552
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6965167583367103,
+      "learning_rate": 8.163770907217506e-07,
+      "loss": 0.9673,
+      "step": 4553
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6703947903732256,
+      "learning_rate": 8.139130741365819e-07,
+      "loss": 0.9867,
+      "step": 4554
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5256678957002158,
+      "learning_rate": 8.114526239066456e-07,
+      "loss": 1.0926,
+      "step": 4555
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6628237827618902,
+      "learning_rate": 8.08995740987173e-07,
+      "loss": 0.9887,
+      "step": 4556
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.709018410328057,
+      "learning_rate": 8.065424263320054e-07,
+      "loss": 1.0256,
+      "step": 4557
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7607755164033556,
+      "learning_rate": 8.040926808936112e-07,
+      "loss": 1.0042,
+      "step": 4558
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6880227637734948,
+      "learning_rate": 8.016465056230616e-07,
+      "loss": 1.0038,
+      "step": 4559
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7644943534995521,
+      "learning_rate": 7.99203901470047e-07,
+      "loss": 1.0274,
+      "step": 4560
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6993346066947773,
+      "learning_rate": 7.967648693828712e-07,
+      "loss": 1.0126,
+      "step": 4561
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5307902924883608,
+      "learning_rate": 7.943294103084487e-07,
+      "loss": 0.9973,
+      "step": 4562
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7342246157333155,
+      "learning_rate": 7.9189752519231e-07,
+      "loss": 0.9813,
+      "step": 4563
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6625673717936391,
+      "learning_rate": 7.894692149785954e-07,
+      "loss": 1.0207,
+      "step": 4564
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7040515834914022,
+      "learning_rate": 7.870444806100619e-07,
+      "loss": 0.9799,
+      "step": 4565
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7317011633166202,
+      "learning_rate": 7.846233230280698e-07,
+      "loss": 0.9894,
+      "step": 4566
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6896549007978383,
+      "learning_rate": 7.822057431725994e-07,
+      "loss": 1.0352,
+      "step": 4567
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7319548257003174,
+      "learning_rate": 7.797917419822377e-07,
+      "loss": 0.9503,
+      "step": 4568
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7116397158293777,
+      "learning_rate": 7.773813203941827e-07,
+      "loss": 1.0487,
+      "step": 4569
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6857967608545276,
+      "learning_rate": 7.749744793442448e-07,
+      "loss": 1.0198,
+      "step": 4570
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6924812927343743,
+      "learning_rate": 7.725712197668378e-07,
+      "loss": 1.0089,
+      "step": 4571
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5439360740605597,
+      "learning_rate": 7.701715425949952e-07,
+      "loss": 1.0328,
+      "step": 4572
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6524215301268613,
+      "learning_rate": 7.677754487603517e-07,
+      "loss": 1.0253,
+      "step": 4573
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7096836439278381,
+      "learning_rate": 7.653829391931533e-07,
+      "loss": 1.039,
+      "step": 4574
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.707649838350082,
+      "learning_rate": 7.629940148222559e-07,
+      "loss": 0.9944,
+      "step": 4575
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7231726880752661,
+      "learning_rate": 7.606086765751209e-07,
+      "loss": 0.942,
+      "step": 4576
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6385680193082852,
+      "learning_rate": 7.582269253778185e-07,
+      "loss": 0.9703,
+      "step": 4577
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.698199064390837,
+      "learning_rate": 7.55848762155027e-07,
+      "loss": 0.9507,
+      "step": 4578
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5498079057580356,
+      "learning_rate": 7.534741878300333e-07,
+      "loss": 0.9904,
+      "step": 4579
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6904694291151372,
+      "learning_rate": 7.511032033247256e-07,
+      "loss": 0.9932,
+      "step": 4580
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7211011729037826,
+      "learning_rate": 7.487358095596031e-07,
+      "loss": 0.9354,
+      "step": 4581
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7015793057957702,
+      "learning_rate": 7.463720074537728e-07,
+      "loss": 0.9565,
+      "step": 4582
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7306476914311295,
+      "learning_rate": 7.440117979249362e-07,
+      "loss": 0.8967,
+      "step": 4583
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.727825188736029,
+      "learning_rate": 7.416551818894158e-07,
+      "loss": 0.9817,
+      "step": 4584
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7576451079589352,
+      "learning_rate": 7.393021602621264e-07,
+      "loss": 1.0582,
+      "step": 4585
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5264564802371658,
+      "learning_rate": 7.369527339565951e-07,
+      "loss": 1.0043,
+      "step": 4586
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7405967449027616,
+      "learning_rate": 7.346069038849469e-07,
+      "loss": 0.9645,
+      "step": 4587
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7540043451924867,
+      "learning_rate": 7.322646709579173e-07,
+      "loss": 0.9611,
+      "step": 4588
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5503643512637902,
+      "learning_rate": 7.299260360848382e-07,
+      "loss": 1.0954,
+      "step": 4589
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6652376734168736,
+      "learning_rate": 7.275910001736497e-07,
+      "loss": 0.9894,
+      "step": 4590
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6522846224988075,
+      "learning_rate": 7.252595641308957e-07,
+      "loss": 1.0102,
+      "step": 4591
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7093362388076019,
+      "learning_rate": 7.229317288617144e-07,
+      "loss": 0.9678,
+      "step": 4592
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7633863915462092,
+      "learning_rate": 7.20607495269856e-07,
+      "loss": 0.928,
+      "step": 4593
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6831946346523639,
+      "learning_rate": 7.182868642576679e-07,
+      "loss": 0.9796,
+      "step": 4594
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8076590095464804,
+      "learning_rate": 7.15969836726097e-07,
+      "loss": 1.0032,
+      "step": 4595
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6644610913021954,
+      "learning_rate": 7.13656413574696e-07,
+      "loss": 1.0255,
+      "step": 4596
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.7638663100980692,
+      "learning_rate": 7.113465957016097e-07,
+      "loss": 1.0324,
+      "step": 4597
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6929233251133426,
+      "learning_rate": 7.090403840035942e-07,
+      "loss": 1.0801,
+      "step": 4598
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6906842208952774,
+      "learning_rate": 7.067377793759999e-07,
+      "loss": 1.0323,
+      "step": 4599
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.729337199478985,
+      "learning_rate": 7.044387827127752e-07,
+      "loss": 1.0738,
+      "step": 4600
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7673733702017073,
+      "learning_rate": 7.021433949064704e-07,
+      "loss": 1.0572,
+      "step": 4601
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6804314298997381,
+      "learning_rate": 6.99851616848235e-07,
+      "loss": 0.9971,
+      "step": 4602
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7379797657526292,
+      "learning_rate": 6.975634494278149e-07,
+      "loss": 1.0513,
+      "step": 4603
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6549017915027311,
+      "learning_rate": 6.952788935335541e-07,
+      "loss": 0.9814,
+      "step": 4604
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7424308182302015,
+      "learning_rate": 6.92997950052402e-07,
+      "loss": 0.9655,
+      "step": 4605
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.694729104932026,
+      "learning_rate": 6.907206198698912e-07,
+      "loss": 1.0424,
+      "step": 4606
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6579340789537287,
+      "learning_rate": 6.884469038701646e-07,
+      "loss": 1.006,
+      "step": 4607
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7240539960320486,
+      "learning_rate": 6.861768029359595e-07,
+      "loss": 0.9936,
+      "step": 4608
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6866226613261507,
+      "learning_rate": 6.839103179485995e-07,
+      "loss": 0.9788,
+      "step": 4609
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.684612000376506,
+      "learning_rate": 6.816474497880177e-07,
+      "loss": 1.0328,
+      "step": 4610
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6753408236447749,
+      "learning_rate": 6.793881993327366e-07,
+      "loss": 0.9829,
+      "step": 4611
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7071948869901201,
+      "learning_rate": 6.77132567459875e-07,
+      "loss": 0.9781,
+      "step": 4612
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6741798197995329,
+      "learning_rate": 6.748805550451453e-07,
+      "loss": 1.0358,
+      "step": 4613
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5159119958800342,
+      "learning_rate": 6.726321629628585e-07,
+      "loss": 1.0715,
+      "step": 4614
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6946504992589453,
+      "learning_rate": 6.703873920859161e-07,
+      "loss": 1.032,
+      "step": 4615
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.759922389829466,
+      "learning_rate": 6.681462432858154e-07,
+      "loss": 0.9683,
+      "step": 4616
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6886659671609644,
+      "learning_rate": 6.659087174326506e-07,
+      "loss": 1.045,
+      "step": 4617
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7387629759745331,
+      "learning_rate": 6.636748153951e-07,
+      "loss": 0.9114,
+      "step": 4618
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7105164411515494,
+      "learning_rate": 6.614445380404478e-07,
+      "loss": 1.0558,
+      "step": 4619
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7928373940120127,
+      "learning_rate": 6.592178862345622e-07,
+      "loss": 0.9652,
+      "step": 4620
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6842891009443312,
+      "learning_rate": 6.569948608419041e-07,
+      "loss": 0.9587,
+      "step": 4621
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7231238284284551,
+      "learning_rate": 6.547754627255332e-07,
+      "loss": 0.9667,
+      "step": 4622
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7137619091082049,
+      "learning_rate": 6.52559692747089e-07,
+      "loss": 0.9562,
+      "step": 4623
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5477621143821743,
+      "learning_rate": 6.503475517668168e-07,
+      "loss": 1.0845,
+      "step": 4624
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7708923097234268,
+      "learning_rate": 6.481390406435417e-07,
+      "loss": 0.9885,
+      "step": 4625
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7666195109966111,
+      "learning_rate": 6.459341602346858e-07,
+      "loss": 1.0255,
+      "step": 4626
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5915113183954693,
+      "learning_rate": 6.437329113962576e-07,
+      "loss": 1.0578,
+      "step": 4627
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6556273230647027,
+      "learning_rate": 6.415352949828601e-07,
+      "loss": 1.0329,
+      "step": 4628
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6348573594490753,
+      "learning_rate": 6.393413118476821e-07,
+      "loss": 1.0073,
+      "step": 4629
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.726902002551384,
+      "learning_rate": 6.371509628425021e-07,
+      "loss": 0.9604,
+      "step": 4630
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6855608931896154,
+      "learning_rate": 6.349642488176943e-07,
+      "loss": 0.9883,
+      "step": 4631
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.740007883728244,
+      "learning_rate": 6.327811706222097e-07,
+      "loss": 0.9303,
+      "step": 4632
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.661489406438282,
+      "learning_rate": 6.306017291035981e-07,
+      "loss": 1.0167,
+      "step": 4633
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.8137590062680492,
+      "learning_rate": 6.284259251079939e-07,
+      "loss": 0.9888,
+      "step": 4634
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6734448531705914,
+      "learning_rate": 6.262537594801177e-07,
+      "loss": 1.0225,
+      "step": 4635
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.720176064589945,
+      "learning_rate": 6.240852330632796e-07,
+      "loss": 1.0138,
+      "step": 4636
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5466359431852882,
+      "learning_rate": 6.219203466993762e-07,
+      "loss": 1.0428,
+      "step": 4637
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7155739829444027,
+      "learning_rate": 6.197591012288918e-07,
+      "loss": 1.0531,
+      "step": 4638
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.685000146299168,
+      "learning_rate": 6.17601497490895e-07,
+      "loss": 0.9642,
+      "step": 4639
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7726908731859989,
+      "learning_rate": 6.154475363230417e-07,
+      "loss": 1.0841,
+      "step": 4640
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7030524323305678,
+      "learning_rate": 6.132972185615749e-07,
+      "loss": 1.0109,
+      "step": 4641
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7682449208108014,
+      "learning_rate": 6.111505450413202e-07,
+      "loss": 1.0107,
+      "step": 4642
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7244329055976658,
+      "learning_rate": 6.090075165956943e-07,
+      "loss": 0.9858,
+      "step": 4643
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7105655348391213,
+      "learning_rate": 6.068681340566896e-07,
+      "loss": 0.9847,
+      "step": 4644
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6843555419416323,
+      "learning_rate": 6.047323982548924e-07,
+      "loss": 1.0079,
+      "step": 4645
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.728983248742151,
+      "learning_rate": 6.026003100194633e-07,
+      "loss": 1.0425,
+      "step": 4646
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6828064072014266,
+      "learning_rate": 6.004718701781575e-07,
+      "loss": 1.032,
+      "step": 4647
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7245415163149723,
+      "learning_rate": 5.983470795573088e-07,
+      "loss": 1.0128,
+      "step": 4648
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6900008241414476,
+      "learning_rate": 5.962259389818292e-07,
+      "loss": 1.0507,
+      "step": 4649
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7037548100621187,
+      "learning_rate": 5.941084492752236e-07,
+      "loss": 1.0009,
+      "step": 4650
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6524214846749342,
+      "learning_rate": 5.91994611259572e-07,
+      "loss": 0.96,
+      "step": 4651
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7787875678795453,
+      "learning_rate": 5.898844257555392e-07,
+      "loss": 0.9781,
+      "step": 4652
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5190489609109183,
+      "learning_rate": 5.87777893582372e-07,
+      "loss": 1.0181,
+      "step": 4653
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6775745312072723,
+      "learning_rate": 5.856750155578983e-07,
+      "loss": 0.9819,
+      "step": 4654
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6389422950709348,
+      "learning_rate": 5.835757924985286e-07,
+      "loss": 0.9619,
+      "step": 4655
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6522648597621937,
+      "learning_rate": 5.81480225219252e-07,
+      "loss": 0.978,
+      "step": 4656
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6412436656536578,
+      "learning_rate": 5.793883145336443e-07,
+      "loss": 1.0673,
+      "step": 4657
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7480978445853821,
+      "learning_rate": 5.773000612538505e-07,
+      "loss": 1.0205,
+      "step": 4658
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7309198100506444,
+      "learning_rate": 5.752154661906085e-07,
+      "loss": 0.96,
+      "step": 4659
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7199825189244069,
+      "learning_rate": 5.731345301532265e-07,
+      "loss": 0.987,
+      "step": 4660
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8356904898330313,
+      "learning_rate": 5.710572539495962e-07,
+      "loss": 1.076,
+      "step": 4661
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7189179188903217,
+      "learning_rate": 5.68983638386188e-07,
+      "loss": 1.015,
+      "step": 4662
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8244842273437111,
+      "learning_rate": 5.669136842680512e-07,
+      "loss": 0.9612,
+      "step": 4663
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6777091527754101,
+      "learning_rate": 5.648473923988129e-07,
+      "loss": 0.9957,
+      "step": 4664
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7097245544745473,
+      "learning_rate": 5.627847635806771e-07,
+      "loss": 1.0234,
+      "step": 4665
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6652830342620484,
+      "learning_rate": 5.607257986144321e-07,
+      "loss": 0.9816,
+      "step": 4666
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5327999020652452,
+      "learning_rate": 5.58670498299434e-07,
+      "loss": 1.0955,
+      "step": 4667
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7018779822553545,
+      "learning_rate": 5.566188634336212e-07,
+      "loss": 1.0001,
+      "step": 4668
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7611654506603305,
+      "learning_rate": 5.545708948135142e-07,
+      "loss": 0.9057,
+      "step": 4669
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6735775871503369,
+      "learning_rate": 5.525265932341984e-07,
+      "loss": 1.0339,
+      "step": 4670
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.758486093062519,
+      "learning_rate": 5.504859594893475e-07,
+      "loss": 1.0054,
+      "step": 4671
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.795282600931398,
+      "learning_rate": 5.484489943712013e-07,
+      "loss": 1.0395,
+      "step": 4672
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.34940493397419103,
+      "learning_rate": 5.464156986705826e-07,
+      "loss": 1.079,
+      "step": 4673
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6476389515585496,
+      "learning_rate": 5.443860731768869e-07,
+      "loss": 0.9406,
+      "step": 4674
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6957683428822734,
+      "learning_rate": 5.423601186780836e-07,
+      "loss": 1.0366,
+      "step": 4675
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7580374495506091,
+      "learning_rate": 5.403378359607181e-07,
+      "loss": 0.937,
+      "step": 4676
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7172361759759813,
+      "learning_rate": 5.383192258099113e-07,
+      "loss": 1.0118,
+      "step": 4677
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7195136365949403,
+      "learning_rate": 5.36304289009355e-07,
+      "loss": 1.0123,
+      "step": 4678
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6192298892852286,
+      "learning_rate": 5.342930263413193e-07,
+      "loss": 1.0889,
+      "step": 4679
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7014074333786129,
+      "learning_rate": 5.322854385866439e-07,
+      "loss": 0.9716,
+      "step": 4680
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6693119170930254,
+      "learning_rate": 5.302815265247452e-07,
+      "loss": 1.0698,
+      "step": 4681
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5605267676618634,
+      "learning_rate": 5.282812909336077e-07,
+      "loss": 1.0247,
+      "step": 4682
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7229522844562835,
+      "learning_rate": 5.262847325897968e-07,
+      "loss": 1.0159,
+      "step": 4683
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5626505618129131,
+      "learning_rate": 5.242918522684392e-07,
+      "loss": 0.9988,
+      "step": 4684
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7038586692880677,
+      "learning_rate": 5.22302650743245e-07,
+      "loss": 0.979,
+      "step": 4685
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6717070293988415,
+      "learning_rate": 5.203171287864872e-07,
+      "loss": 0.9174,
+      "step": 4686
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6645351515881298,
+      "learning_rate": 5.183352871690162e-07,
+      "loss": 1.0132,
+      "step": 4687
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6752659293884078,
+      "learning_rate": 5.163571266602485e-07,
+      "loss": 0.998,
+      "step": 4688
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5286303318262598,
+      "learning_rate": 5.143826480281778e-07,
+      "loss": 0.9523,
+      "step": 4689
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6960964365765726,
+      "learning_rate": 5.124118520393606e-07,
+      "loss": 0.9551,
+      "step": 4690
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.70151821470644,
+      "learning_rate": 5.104447394589295e-07,
+      "loss": 0.9101,
+      "step": 4691
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7167856916876412,
+      "learning_rate": 5.084813110505871e-07,
+      "loss": 1.0087,
+      "step": 4692
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6803673164833558,
+      "learning_rate": 5.065215675766023e-07,
+      "loss": 1.0308,
+      "step": 4693
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.8348231425841811,
+      "learning_rate": 5.045655097978131e-07,
+      "loss": 1.0192,
+      "step": 4694
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.648668165822955,
+      "learning_rate": 5.026131384736321e-07,
+      "loss": 1.0059,
+      "step": 4695
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.542947810545555,
+      "learning_rate": 5.006644543620342e-07,
+      "loss": 1.0046,
+      "step": 4696
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7117627459682737,
+      "learning_rate": 4.987194582195687e-07,
+      "loss": 0.9924,
+      "step": 4697
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6292564874149326,
+      "learning_rate": 4.967781508013459e-07,
+      "loss": 1.0287,
+      "step": 4698
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6421623555759799,
+      "learning_rate": 4.948405328610506e-07,
+      "loss": 1.0276,
+      "step": 4699
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6562059886900476,
+      "learning_rate": 4.929066051509346e-07,
+      "loss": 1.0392,
+      "step": 4700
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7199255170650279,
+      "learning_rate": 4.909763684218116e-07,
+      "loss": 1.0289,
+      "step": 4701
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5244156562798313,
+      "learning_rate": 4.890498234230689e-07,
+      "loss": 1.0515,
+      "step": 4702
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6625977332496934,
+      "learning_rate": 4.871269709026561e-07,
+      "loss": 1.0697,
+      "step": 4703
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6449521563264882,
+      "learning_rate": 4.852078116070902e-07,
+      "loss": 1.0314,
+      "step": 4704
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6995004517473662,
+      "learning_rate": 4.832923462814565e-07,
+      "loss": 0.9409,
+      "step": 4705
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7064364318264713,
+      "learning_rate": 4.813805756694035e-07,
+      "loss": 1.0521,
+      "step": 4706
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.689036438487676,
+      "learning_rate": 4.794725005131462e-07,
+      "loss": 1.0156,
+      "step": 4707
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6505148062841808,
+      "learning_rate": 4.775681215534656e-07,
+      "loss": 0.9271,
+      "step": 4708
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6505023800180594,
+      "learning_rate": 4.7566743952970894e-07,
+      "loss": 1.0182,
+      "step": 4709
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6907573672141541,
+      "learning_rate": 4.7377045517978173e-07,
+      "loss": 1.0005,
+      "step": 4710
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7541696970852236,
+      "learning_rate": 4.7187716924016355e-07,
+      "loss": 0.9629,
+      "step": 4711
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7026452897290779,
+      "learning_rate": 4.6998758244588995e-07,
+      "loss": 0.9712,
+      "step": 4712
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7784540705528595,
+      "learning_rate": 4.6810169553056616e-07,
+      "loss": 1.0239,
+      "step": 4713
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.8102116636980395,
+      "learning_rate": 4.662195092263566e-07,
+      "loss": 1.0247,
+      "step": 4714
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5769116743211663,
+      "learning_rate": 4.643410242639912e-07,
+      "loss": 1.0602,
+      "step": 4715
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7742218141085726,
+      "learning_rate": 4.6246624137276206e-07,
+      "loss": 1.0198,
+      "step": 4716
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7029632286983292,
+      "learning_rate": 4.605951612805237e-07,
+      "loss": 0.9595,
+      "step": 4717
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6650661360545952,
+      "learning_rate": 4.587277847136984e-07,
+      "loss": 1.0112,
+      "step": 4718
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.527505159584153,
+      "learning_rate": 4.568641123972606e-07,
+      "loss": 0.95,
+      "step": 4719
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6741171821741049,
+      "learning_rate": 4.550041450547549e-07,
+      "loss": 1.0132,
+      "step": 4720
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6653521984190635,
+      "learning_rate": 4.5314788340828365e-07,
+      "loss": 1.0237,
+      "step": 4721
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6875856553186117,
+      "learning_rate": 4.512953281785104e-07,
+      "loss": 0.9817,
+      "step": 4722
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7765205213175914,
+      "learning_rate": 4.494464800846654e-07,
+      "loss": 0.9937,
+      "step": 4723
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7045772968105548,
+      "learning_rate": 4.476013398445289e-07,
+      "loss": 1.0362,
+      "step": 4724
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7250626970100557,
+      "learning_rate": 4.4575990817445234e-07,
+      "loss": 0.9456,
+      "step": 4725
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6715666563332392,
+      "learning_rate": 4.4392218578934164e-07,
+      "loss": 1.0058,
+      "step": 4726
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7149047256045353,
+      "learning_rate": 4.4208817340266385e-07,
+      "loss": 1.0259,
+      "step": 4727
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7132018160926684,
+      "learning_rate": 4.4025787172644495e-07,
+      "loss": 1.0094,
+      "step": 4728
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.650567668982881,
+      "learning_rate": 4.384312814712721e-07,
+      "loss": 0.9832,
+      "step": 4729
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6591647680740225,
+      "learning_rate": 4.366084033462914e-07,
+      "loss": 0.9728,
+      "step": 4730
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7082076069773613,
+      "learning_rate": 4.3478923805920335e-07,
+      "loss": 0.932,
+      "step": 4731
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6276676635732951,
+      "learning_rate": 4.329737863162753e-07,
+      "loss": 0.9526,
+      "step": 4732
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.715586331657728,
+      "learning_rate": 4.311620488223256e-07,
+      "loss": 1.0344,
+      "step": 4733
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7296300879104922,
+      "learning_rate": 4.2935402628073166e-07,
+      "loss": 1.0511,
+      "step": 4734
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6838957057391671,
+      "learning_rate": 4.27549719393433e-07,
+      "loss": 0.9728,
+      "step": 4735
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7287445104277475,
+      "learning_rate": 4.2574912886092166e-07,
+      "loss": 1.0492,
+      "step": 4736
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7437820428512151,
+      "learning_rate": 4.239522553822495e-07,
+      "loss": 0.984,
+      "step": 4737
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7044507487472745,
+      "learning_rate": 4.221590996550251e-07,
+      "loss": 1.0073,
+      "step": 4738
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.696379815018403,
+      "learning_rate": 4.203696623754139e-07,
+      "loss": 0.9976,
+      "step": 4739
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7376178851002855,
+      "learning_rate": 4.1858394423813563e-07,
+      "loss": 1.0063,
+      "step": 4740
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7546447149610264,
+      "learning_rate": 4.1680194593646696e-07,
+      "loss": 1.041,
+      "step": 4741
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5636285252094628,
+      "learning_rate": 4.1502366816224327e-07,
+      "loss": 1.0757,
+      "step": 4742
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6724852511482485,
+      "learning_rate": 4.1324911160585014e-07,
+      "loss": 1.0042,
+      "step": 4743
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6806321031108146,
+      "learning_rate": 4.1147827695623643e-07,
+      "loss": 1.0258,
+      "step": 4744
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7095847168187829,
+      "learning_rate": 4.097111649008967e-07,
+      "loss": 0.9654,
+      "step": 4745
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7324916074160469,
+      "learning_rate": 4.0794777612588543e-07,
+      "loss": 0.9947,
+      "step": 4746
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.655639168376011,
+      "learning_rate": 4.061881113158117e-07,
+      "loss": 0.9855,
+      "step": 4747
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6552467016248087,
+      "learning_rate": 4.044321711538368e-07,
+      "loss": 0.9886,
+      "step": 4748
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7223762123898397,
+      "learning_rate": 4.02679956321681e-07,
+      "loss": 1.072,
+      "step": 4749
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6808582653205603,
+      "learning_rate": 4.00931467499609e-07,
+      "loss": 1.04,
+      "step": 4750
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6533419195825673,
+      "learning_rate": 3.9918670536644776e-07,
+      "loss": 1.0059,
+      "step": 4751
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7076124216317186,
+      "learning_rate": 3.974456705995733e-07,
+      "loss": 0.9847,
+      "step": 4752
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6957739538691565,
+      "learning_rate": 3.9570836387491487e-07,
+      "loss": 1.0364,
+      "step": 4753
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7069904900212457,
+      "learning_rate": 3.9397478586695513e-07,
+      "loss": 1.0003,
+      "step": 4754
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7738700147924269,
+      "learning_rate": 3.9224493724872915e-07,
+      "loss": 0.9794,
+      "step": 4755
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6940063921509282,
+      "learning_rate": 3.90518818691823e-07,
+      "loss": 1.0204,
+      "step": 4756
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7180315247515171,
+      "learning_rate": 3.8879643086637384e-07,
+      "loss": 0.9975,
+      "step": 4757
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7337812855798762,
+      "learning_rate": 3.8707777444107697e-07,
+      "loss": 1.0058,
+      "step": 4758
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6978585196075165,
+      "learning_rate": 3.8536285008316854e-07,
+      "loss": 0.9994,
+      "step": 4759
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7171761469799023,
+      "learning_rate": 3.8365165845844266e-07,
+      "loss": 0.9707,
+      "step": 4760
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7526504926744862,
+      "learning_rate": 3.819442002312457e-07,
+      "loss": 1.0062,
+      "step": 4761
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6109459557828425,
+      "learning_rate": 3.8024047606446736e-07,
+      "loss": 0.9755,
+      "step": 4762
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7712581541817676,
+      "learning_rate": 3.785404866195552e-07,
+      "loss": 1.0169,
+      "step": 4763
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5331842626723312,
+      "learning_rate": 3.768442325565036e-07,
+      "loss": 1.0345,
+      "step": 4764
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.684703116799368,
+      "learning_rate": 3.751517145338546e-07,
+      "loss": 0.9945,
+      "step": 4765
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5136138230326253,
+      "learning_rate": 3.7346293320870363e-07,
+      "loss": 1.1183,
+      "step": 4766
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6744533424657874,
+      "learning_rate": 3.717778892366941e-07,
+      "loss": 1.0027,
+      "step": 4767
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7791701752226684,
+      "learning_rate": 3.700965832720171e-07,
+      "loss": 1.0276,
+      "step": 4768
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7325372070412808,
+      "learning_rate": 3.684190159674117e-07,
+      "loss": 1.0045,
+      "step": 4769
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7272974992740792,
+      "learning_rate": 3.6674518797417236e-07,
+      "loss": 0.9394,
+      "step": 4770
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7900288380725201,
+      "learning_rate": 3.6507509994213155e-07,
+      "loss": 0.9881,
+      "step": 4771
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6617758740266667,
+      "learning_rate": 3.6340875251967946e-07,
+      "loss": 1.0457,
+      "step": 4772
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7229415792462166,
+      "learning_rate": 3.617461463537464e-07,
+      "loss": 0.9081,
+      "step": 4773
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6867027844344548,
+      "learning_rate": 3.6008728208981157e-07,
+      "loss": 0.9558,
+      "step": 4774
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7067186734762161,
+      "learning_rate": 3.5843216037190873e-07,
+      "loss": 1.0395,
+      "step": 4775
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6375297668156472,
+      "learning_rate": 3.5678078184260834e-07,
+      "loss": 1.0359,
+      "step": 4776
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.609852860046067,
+      "learning_rate": 3.5513314714303524e-07,
+      "loss": 0.9695,
+      "step": 4777
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.577595468768199,
+      "learning_rate": 3.5348925691285675e-07,
+      "loss": 1.0954,
+      "step": 4778
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6196111767165503,
+      "learning_rate": 3.518491117902878e-07,
+      "loss": 0.98,
+      "step": 4779
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7578453915764272,
+      "learning_rate": 3.502127124120891e-07,
+      "loss": 0.9699,
+      "step": 4780
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6415429370367673,
+      "learning_rate": 3.48580059413568e-07,
+      "loss": 1.0377,
+      "step": 4781
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7578148704334245,
+      "learning_rate": 3.4695115342857524e-07,
+      "loss": 0.9747,
+      "step": 4782
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6990802345929796,
+      "learning_rate": 3.4532599508950826e-07,
+      "loss": 0.9658,
+      "step": 4783
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7278491028799754,
+      "learning_rate": 3.437045850273113e-07,
+      "loss": 0.9614,
+      "step": 4784
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5168837877148755,
+      "learning_rate": 3.420869238714708e-07,
+      "loss": 1.0068,
+      "step": 4785
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6867660445674689,
+      "learning_rate": 3.404730122500155e-07,
+      "loss": 1.0175,
+      "step": 4786
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6991608369012896,
+      "learning_rate": 3.3886285078952753e-07,
+      "loss": 1.0064,
+      "step": 4787
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6341954031902405,
+      "learning_rate": 3.3725644011512125e-07,
+      "loss": 0.9601,
+      "step": 4788
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7911033395395465,
+      "learning_rate": 3.356537808504634e-07,
+      "loss": 0.9656,
+      "step": 4789
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5461131053563906,
+      "learning_rate": 3.3405487361776177e-07,
+      "loss": 1.0268,
+      "step": 4790
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6807525023184327,
+      "learning_rate": 3.3245971903776654e-07,
+      "loss": 1.0872,
+      "step": 4791
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7331956393125677,
+      "learning_rate": 3.308683177297711e-07,
+      "loss": 0.9865,
+      "step": 4792
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6702529670435735,
+      "learning_rate": 3.292806703116125e-07,
+      "loss": 1.0412,
+      "step": 4793
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7067308809600085,
+      "learning_rate": 3.2769677739966975e-07,
+      "loss": 1.0306,
+      "step": 4794
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7214326705886995,
+      "learning_rate": 3.2611663960886665e-07,
+      "loss": 0.9615,
+      "step": 4795
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7533820703598917,
+      "learning_rate": 3.245402575526646e-07,
+      "loss": 0.9959,
+      "step": 4796
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6805684758690942,
+      "learning_rate": 3.2296763184306965e-07,
+      "loss": 0.9462,
+      "step": 4797
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7494318931749402,
+      "learning_rate": 3.2139876309063233e-07,
+      "loss": 0.9833,
+      "step": 4798
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6986948366005488,
+      "learning_rate": 3.198336519044376e-07,
+      "loss": 0.9266,
+      "step": 4799
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7042397855287869,
+      "learning_rate": 3.182722988921161e-07,
+      "loss": 1.0433,
+      "step": 4800
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6956424118047426,
+      "learning_rate": 3.167147046598418e-07,
+      "loss": 1.0246,
+      "step": 4801
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5505361642467467,
+      "learning_rate": 3.151608698123232e-07,
+      "loss": 1.0081,
+      "step": 4802
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7474970680294826,
+      "learning_rate": 3.1361079495281443e-07,
+      "loss": 0.9353,
+      "step": 4803
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6777289237203439,
+      "learning_rate": 3.1206448068310635e-07,
+      "loss": 0.9084,
+      "step": 4804
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7658798022108447,
+      "learning_rate": 3.1052192760353316e-07,
+      "loss": 0.9983,
+      "step": 4805
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6993050907807605,
+      "learning_rate": 3.0898313631296586e-07,
+      "loss": 1.0153,
+      "step": 4806
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.8090631419006316,
+      "learning_rate": 3.0744810740881646e-07,
+      "loss": 1.0549,
+      "step": 4807
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6847717048878439,
+      "learning_rate": 3.0591684148703617e-07,
+      "loss": 0.9617,
+      "step": 4808
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7716312134786415,
+      "learning_rate": 3.043893391421149e-07,
+      "loss": 0.9939,
+      "step": 4809
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7292386596898297,
+      "learning_rate": 3.0286560096708275e-07,
+      "loss": 0.9821,
+      "step": 4810
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6825338207727746,
+      "learning_rate": 3.013456275535054e-07,
+      "loss": 1.0522,
+      "step": 4811
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6980959190071999,
+      "learning_rate": 2.998294194914897e-07,
+      "loss": 1.006,
+      "step": 4812
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6978107275391059,
+      "learning_rate": 2.983169773696815e-07,
+      "loss": 1.001,
+      "step": 4813
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7161504477429815,
+      "learning_rate": 2.968083017752599e-07,
+      "loss": 0.9591,
+      "step": 4814
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6178757289605086,
+      "learning_rate": 2.953033932939464e-07,
+      "loss": 0.9506,
+      "step": 4815
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6904531699066077,
+      "learning_rate": 2.938022525099982e-07,
+      "loss": 0.9891,
+      "step": 4816
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6997883332951284,
+      "learning_rate": 2.9230488000621003e-07,
+      "loss": 0.9274,
+      "step": 4817
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7398466498164404,
+      "learning_rate": 2.908112763639137e-07,
+      "loss": 0.9807,
+      "step": 4818
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7702154247083209,
+      "learning_rate": 2.8932144216297643e-07,
+      "loss": 1.0031,
+      "step": 4819
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5605376500536755,
+      "learning_rate": 2.878353779818044e-07,
+      "loss": 1.0266,
+      "step": 4820
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5732767371212605,
+      "learning_rate": 2.863530843973372e-07,
+      "loss": 1.0374,
+      "step": 4821
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6955437740216819,
+      "learning_rate": 2.848745619850546e-07,
+      "loss": 1.0135,
+      "step": 4822
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5743968160743064,
+      "learning_rate": 2.833998113189662e-07,
+      "loss": 1.0334,
+      "step": 4823
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6756753148917045,
+      "learning_rate": 2.8192883297162634e-07,
+      "loss": 0.9634,
+      "step": 4824
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7119976342761614,
+      "learning_rate": 2.804616275141148e-07,
+      "loss": 0.941,
+      "step": 4825
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6929533770934069,
+      "learning_rate": 2.7899819551605256e-07,
+      "loss": 0.9797,
+      "step": 4826
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.660637672867622,
+      "learning_rate": 2.7753853754559634e-07,
+      "loss": 1.0599,
+      "step": 4827
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7275478092148285,
+      "learning_rate": 2.760826541694328e-07,
+      "loss": 0.9771,
+      "step": 4828
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6773747120749685,
+      "learning_rate": 2.746305459527876e-07,
+      "loss": 1.0411,
+      "step": 4829
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.685663960848418,
+      "learning_rate": 2.7318221345941865e-07,
+      "loss": 1.0552,
+      "step": 4830
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8223024903875321,
+      "learning_rate": 2.717376572516184e-07,
+      "loss": 0.929,
+      "step": 4831
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6483674304458409,
+      "learning_rate": 2.7029687789021377e-07,
+      "loss": 1.0533,
+      "step": 4832
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.49006255349585415,
+      "learning_rate": 2.688598759345651e-07,
+      "loss": 1.0347,
+      "step": 4833
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.747777317333441,
+      "learning_rate": 2.67426651942565e-07,
+      "loss": 1.0483,
+      "step": 4834
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7448351092494271,
+      "learning_rate": 2.659972064706406e-07,
+      "loss": 1.0168,
+      "step": 4835
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7007707948865946,
+      "learning_rate": 2.645715400737536e-07,
+      "loss": 1.0419,
+      "step": 4836
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7384039258119355,
+      "learning_rate": 2.631496533053934e-07,
+      "loss": 0.9664,
+      "step": 4837
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7068152269687238,
+      "learning_rate": 2.6173154671758847e-07,
+      "loss": 1.058,
+      "step": 4838
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.686919098743283,
+      "learning_rate": 2.603172208608962e-07,
+      "loss": 1.0463,
+      "step": 4839
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7105114024548153,
+      "learning_rate": 2.589066762844039e-07,
+      "loss": 1.0499,
+      "step": 4840
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7281793191135044,
+      "learning_rate": 2.57499913535737e-07,
+      "loss": 1.0112,
+      "step": 4841
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6468023789297755,
+      "learning_rate": 2.5609693316104745e-07,
+      "loss": 0.9701,
+      "step": 4842
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7471813255437972,
+      "learning_rate": 2.5469773570502063e-07,
+      "loss": 1.0248,
+      "step": 4843
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6737949898781308,
+      "learning_rate": 2.5330232171087433e-07,
+      "loss": 0.9957,
+      "step": 4844
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6414116891433825,
+      "learning_rate": 2.51910691720354e-07,
+      "loss": 0.991,
+      "step": 4845
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7770809478077053,
+      "learning_rate": 2.5052284627374077e-07,
+      "loss": 1.0179,
+      "step": 4846
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6755575141300205,
+      "learning_rate": 2.491387859098426e-07,
+      "loss": 1.0199,
+      "step": 4847
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6654741807315624,
+      "learning_rate": 2.477585111659997e-07,
+      "loss": 0.9704,
+      "step": 4848
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6787762345243606,
+      "learning_rate": 2.463820225780811e-07,
+      "loss": 0.9802,
+      "step": 4849
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7047598465843211,
+      "learning_rate": 2.4500932068049046e-07,
+      "loss": 1.0037,
+      "step": 4850
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6421632026658975,
+      "learning_rate": 2.4364040600615477e-07,
+      "loss": 0.9647,
+      "step": 4851
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6808506234516893,
+      "learning_rate": 2.422752790865346e-07,
+      "loss": 0.9675,
+      "step": 4852
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7548228374010852,
+      "learning_rate": 2.409139404516203e-07,
+      "loss": 0.9868,
+      "step": 4853
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.8299214042958676,
+      "learning_rate": 2.3955639062992696e-07,
+      "loss": 1.0914,
+      "step": 4854
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6441969773665156,
+      "learning_rate": 2.3820263014850741e-07,
+      "loss": 1.0064,
+      "step": 4855
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7121575875214258,
+      "learning_rate": 2.3685265953293345e-07,
+      "loss": 1.0027,
+      "step": 4856
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6754064997571217,
+      "learning_rate": 2.3550647930731362e-07,
+      "loss": 1.0312,
+      "step": 4857
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7144212841242327,
+      "learning_rate": 2.3416408999427876e-07,
+      "loss": 1.0327,
+      "step": 4858
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6774320962672542,
+      "learning_rate": 2.3282549211499307e-07,
+      "loss": 1.0335,
+      "step": 4859
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6393381245382949,
+      "learning_rate": 2.3149068618914417e-07,
+      "loss": 0.9711,
+      "step": 4860
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7606931203416509,
+      "learning_rate": 2.3015967273494867e-07,
+      "loss": 1.0508,
+      "step": 4861
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6606564037739452,
+      "learning_rate": 2.2883245226915652e-07,
+      "loss": 0.926,
+      "step": 4862
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6816373478579645,
+      "learning_rate": 2.2750902530703667e-07,
+      "loss": 0.9641,
+      "step": 4863
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7386111790436136,
+      "learning_rate": 2.2618939236238924e-07,
+      "loss": 0.983,
+      "step": 4864
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6945526471884161,
+      "learning_rate": 2.2487355394754328e-07,
+      "loss": 1.105,
+      "step": 4865
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7010751306243311,
+      "learning_rate": 2.2356151057334908e-07,
+      "loss": 1.0047,
+      "step": 4866
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6756770436276406,
+      "learning_rate": 2.2225326274919135e-07,
+      "loss": 1.0374,
+      "step": 4867
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7283871012566441,
+      "learning_rate": 2.209488109829727e-07,
+      "loss": 1.0246,
+      "step": 4868
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6739743707906984,
+      "learning_rate": 2.196481557811303e-07,
+      "loss": 0.9692,
+      "step": 4869
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6642341453649547,
+      "learning_rate": 2.1835129764861907e-07,
+      "loss": 1.0449,
+      "step": 4870
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6569505438956268,
+      "learning_rate": 2.1705823708892737e-07,
+      "loss": 0.9926,
+      "step": 4871
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8265896308638735,
+      "learning_rate": 2.1576897460406477e-07,
+      "loss": 0.9839,
+      "step": 4872
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7765847745893565,
+      "learning_rate": 2.144835106945664e-07,
+      "loss": 0.9692,
+      "step": 4873
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7040982991120179,
+      "learning_rate": 2.1320184585949532e-07,
+      "loss": 0.9736,
+      "step": 4874
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6929703703712026,
+      "learning_rate": 2.119239805964357e-07,
+      "loss": 0.9595,
+      "step": 4875
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7242088791512152,
+      "learning_rate": 2.106499154015018e-07,
+      "loss": 1.0536,
+      "step": 4876
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.694830580475166,
+      "learning_rate": 2.0937965076932576e-07,
+      "loss": 1.0292,
+      "step": 4877
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7487290563237385,
+      "learning_rate": 2.0811318719307194e-07,
+      "loss": 1.0109,
+      "step": 4878
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6458770133557371,
+      "learning_rate": 2.0685052516442373e-07,
+      "loss": 1.0086,
+      "step": 4879
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7701745336037907,
+      "learning_rate": 2.0559166517358787e-07,
+      "loss": 0.9861,
+      "step": 4880
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.688825515054837,
+      "learning_rate": 2.0433660770930009e-07,
+      "loss": 0.983,
+      "step": 4881
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7519841197959013,
+      "learning_rate": 2.0308535325881616e-07,
+      "loss": 0.8839,
+      "step": 4882
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.66842345573898,
+      "learning_rate": 2.0183790230791532e-07,
+      "loss": 0.9819,
+      "step": 4883
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.49346517604322143,
+      "learning_rate": 2.0059425534090128e-07,
+      "loss": 1.0301,
+      "step": 4884
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7463587709978408,
+      "learning_rate": 1.9935441284059998e-07,
+      "loss": 1.0432,
+      "step": 4885
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6838294229058018,
+      "learning_rate": 1.981183752883631e-07,
+      "loss": 1.051,
+      "step": 4886
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7596522041240548,
+      "learning_rate": 1.9688614316406006e-07,
+      "loss": 1.0035,
+      "step": 4887
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7393063684401018,
+      "learning_rate": 1.9565771694608937e-07,
+      "loss": 0.9271,
+      "step": 4888
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7021922029640952,
+      "learning_rate": 1.9443309711136393e-07,
+      "loss": 0.903,
+      "step": 4889
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6770513760543084,
+      "learning_rate": 1.9321228413532788e-07,
+      "loss": 1.0429,
+      "step": 4890
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6574437823308478,
+      "learning_rate": 1.9199527849194098e-07,
+      "loss": 0.9352,
+      "step": 4891
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6734175061975172,
+      "learning_rate": 1.907820806536842e-07,
+      "loss": 0.9579,
+      "step": 4892
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7504524601828498,
+      "learning_rate": 1.895726910915663e-07,
+      "loss": 0.8978,
+      "step": 4893
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6666773064273177,
+      "learning_rate": 1.883671102751128e-07,
+      "loss": 1.0337,
+      "step": 4894
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.670837201010672,
+      "learning_rate": 1.8716533867237153e-07,
+      "loss": 0.9224,
+      "step": 4895
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.743733692610289,
+      "learning_rate": 1.859673767499115e-07,
+      "loss": 0.9532,
+      "step": 4896
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7127940732777346,
+      "learning_rate": 1.847732249728218e-07,
+      "loss": 0.993,
+      "step": 4897
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6833045982662561,
+      "learning_rate": 1.83582883804716e-07,
+      "loss": 0.9753,
+      "step": 4898
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7624306154985523,
+      "learning_rate": 1.8239635370772223e-07,
+      "loss": 0.9706,
+      "step": 4899
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5069538575819497,
+      "learning_rate": 1.8121363514249534e-07,
+      "loss": 0.9896,
+      "step": 4900
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6969672250487415,
+      "learning_rate": 1.8003472856820469e-07,
+      "loss": 0.9659,
+      "step": 4901
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.682493353492685,
+      "learning_rate": 1.7885963444254528e-07,
+      "loss": 1.0093,
+      "step": 4902
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7175461207438696,
+      "learning_rate": 1.7768835322172552e-07,
+      "loss": 1.0225,
+      "step": 4903
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7131502145033622,
+      "learning_rate": 1.7652088536048052e-07,
+      "loss": 0.9291,
+      "step": 4904
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6988891544958861,
+      "learning_rate": 1.7535723131206106e-07,
+      "loss": 0.9927,
+      "step": 4905
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6747763748390545,
+      "learning_rate": 1.7419739152823468e-07,
+      "loss": 1.038,
+      "step": 4906
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7347105918396646,
+      "learning_rate": 1.7304136645929448e-07,
+      "loss": 0.9814,
+      "step": 4907
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7043151162480183,
+      "learning_rate": 1.7188915655404814e-07,
+      "loss": 1.0594,
+      "step": 4908
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5349932091587055,
+      "learning_rate": 1.707407622598223e-07,
+      "loss": 1.0331,
+      "step": 4909
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7013559364250999,
+      "learning_rate": 1.695961840224636e-07,
+      "loss": 1.0277,
+      "step": 4910
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4792181475434164,
+      "learning_rate": 1.6845542228633772e-07,
+      "loss": 1.0009,
+      "step": 4911
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7244385832462993,
+      "learning_rate": 1.6731847749432705e-07,
+      "loss": 0.9879,
+      "step": 4912
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6737902859208379,
+      "learning_rate": 1.6618535008783075e-07,
+      "loss": 1.0176,
+      "step": 4913
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.679111021957968,
+      "learning_rate": 1.6505604050677249e-07,
+      "loss": 0.9789,
+      "step": 4914
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7295970742443898,
+      "learning_rate": 1.6393054918958373e-07,
+      "loss": 0.9297,
+      "step": 4915
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6656754503394647,
+      "learning_rate": 1.6280887657322276e-07,
+      "loss": 0.9505,
+      "step": 4916
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7458523221176184,
+      "learning_rate": 1.616910230931612e-07,
+      "loss": 0.9692,
+      "step": 4917
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6833803654235421,
+      "learning_rate": 1.6057698918338526e-07,
+      "loss": 0.9287,
+      "step": 4918
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5487643536207344,
+      "learning_rate": 1.5946677527640563e-07,
+      "loss": 1.0453,
+      "step": 4919
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5680684503820416,
+      "learning_rate": 1.5836038180324198e-07,
+      "loss": 1.0184,
+      "step": 4920
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7637648821011519,
+      "learning_rate": 1.5725780919343624e-07,
+      "loss": 0.9744,
+      "step": 4921
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7316430566454282,
+      "learning_rate": 1.561590578750438e-07,
+      "loss": 1.0058,
+      "step": 4922
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7203797883648492,
+      "learning_rate": 1.55064128274639e-07,
+      "loss": 0.9651,
+      "step": 4923
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6675275702550553,
+      "learning_rate": 1.5397302081731069e-07,
+      "loss": 0.9406,
+      "step": 4924
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7129852777258013,
+      "learning_rate": 1.5288573592666445e-07,
+      "loss": 0.9823,
+      "step": 4925
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5468354070400592,
+      "learning_rate": 1.518022740248215e-07,
+      "loss": 0.9839,
+      "step": 4926
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4986138557720332,
+      "learning_rate": 1.5072263553241872e-07,
+      "loss": 1.0436,
+      "step": 4927
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6752732479483867,
+      "learning_rate": 1.4964682086861082e-07,
+      "loss": 1.043,
+      "step": 4928
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7486056446525091,
+      "learning_rate": 1.4857483045106258e-07,
+      "loss": 0.9561,
+      "step": 4929
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6935281451545463,
+      "learning_rate": 1.475066646959611e-07,
+      "loss": 0.9651,
+      "step": 4930
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7160513415590516,
+      "learning_rate": 1.4644232401800352e-07,
+      "loss": 1.0401,
+      "step": 4931
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8869014781540732,
+      "learning_rate": 1.4538180883040264e-07,
+      "loss": 1.0117,
+      "step": 4932
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7284426354409481,
+      "learning_rate": 1.4432511954488915e-07,
+      "loss": 1.0092,
+      "step": 4933
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.677363236323139,
+      "learning_rate": 1.4327225657170485e-07,
+      "loss": 1.0468,
+      "step": 4934
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6680757461066171,
+      "learning_rate": 1.4222322031960723e-07,
+      "loss": 0.9686,
+      "step": 4935
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7100916401709634,
+      "learning_rate": 1.411780111958694e-07,
+      "loss": 0.9859,
+      "step": 4936
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6687356329066683,
+      "learning_rate": 1.4013662960627562e-07,
+      "loss": 0.9989,
+      "step": 4937
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5457159962460258,
+      "learning_rate": 1.3909907595512806e-07,
+      "loss": 1.0612,
+      "step": 4938
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6887601254430199,
+      "learning_rate": 1.3806535064524006e-07,
+      "loss": 1.0323,
+      "step": 4939
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7884286821041013,
+      "learning_rate": 1.3703545407793951e-07,
+      "loss": 0.9474,
+      "step": 4940
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7095360626617134,
+      "learning_rate": 1.360093866530665e-07,
+      "loss": 0.986,
+      "step": 4941
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7498978317359263,
+      "learning_rate": 1.34987148768978e-07,
+      "loss": 0.9632,
+      "step": 4942
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.560598267860736,
+      "learning_rate": 1.3396874082253986e-07,
+      "loss": 0.9956,
+      "step": 4943
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.731475704880381,
+      "learning_rate": 1.3295416320913357e-07,
+      "loss": 1.0048,
+      "step": 4944
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6714460067865604,
+      "learning_rate": 1.3194341632265518e-07,
+      "loss": 0.9651,
+      "step": 4945
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6782191724870387,
+      "learning_rate": 1.3093650055550855e-07,
+      "loss": 1.0397,
+      "step": 4946
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6947410275181053,
+      "learning_rate": 1.2993341629861432e-07,
+      "loss": 1.0656,
+      "step": 4947
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7522481086019642,
+      "learning_rate": 1.2893416394140323e-07,
+      "loss": 1.0548,
+      "step": 4948
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7272888706836813,
+      "learning_rate": 1.279387438718216e-07,
+      "loss": 0.9931,
+      "step": 4949
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7296847785534154,
+      "learning_rate": 1.269471564763247e-07,
+      "loss": 1.066,
+      "step": 4950
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5623473936059561,
+      "learning_rate": 1.2595940213988024e-07,
+      "loss": 0.9706,
+      "step": 4951
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6511426170318156,
+      "learning_rate": 1.2497548124597026e-07,
+      "loss": 0.9068,
+      "step": 4952
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7061682348176272,
+      "learning_rate": 1.2399539417658368e-07,
+      "loss": 0.9922,
+      "step": 4953
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7312432465651436,
+      "learning_rate": 1.2301914131222726e-07,
+      "loss": 0.8941,
+      "step": 4954
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8039321060293796,
+      "learning_rate": 1.2204672303191335e-07,
+      "loss": 0.9966,
+      "step": 4955
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6711191054573821,
+      "learning_rate": 1.2107813971317106e-07,
+      "loss": 0.9985,
+      "step": 4956
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6796067640794842,
+      "learning_rate": 1.201133917320363e-07,
+      "loss": 1.009,
+      "step": 4957
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.699995530411749,
+      "learning_rate": 1.1915247946305498e-07,
+      "loss": 0.9646,
+      "step": 4958
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6861234440682134,
+      "learning_rate": 1.1819540327929092e-07,
+      "loss": 0.9787,
+      "step": 4959
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5180300919062789,
+      "learning_rate": 1.1724216355231022e-07,
+      "loss": 1.0237,
+      "step": 4960
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7502492751072997,
+      "learning_rate": 1.1629276065219575e-07,
+      "loss": 0.9667,
+      "step": 4961
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6380308457616732,
+      "learning_rate": 1.1534719494753821e-07,
+      "loss": 0.9599,
+      "step": 4962
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6878165318171452,
+      "learning_rate": 1.144054668054373e-07,
+      "loss": 1.0019,
+      "step": 4963
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6673356008487864,
+      "learning_rate": 1.1346757659150498e-07,
+      "loss": 0.8856,
+      "step": 4964
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.714289727408765,
+      "learning_rate": 1.1253352466986334e-07,
+      "loss": 1.072,
+      "step": 4965
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5818864875849686,
+      "learning_rate": 1.116033114031434e-07,
+      "loss": 1.0309,
+      "step": 4966
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7356121714842194,
+      "learning_rate": 1.1067693715248406e-07,
+      "loss": 1.0131,
+      "step": 4967
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7290966945016711,
+      "learning_rate": 1.0975440227753764e-07,
+      "loss": 0.987,
+      "step": 4968
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7517974547394243,
+      "learning_rate": 1.0883570713646318e-07,
+      "loss": 0.9677,
+      "step": 4969
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7024456379116321,
+      "learning_rate": 1.0792085208593095e-07,
+      "loss": 0.905,
+      "step": 4970
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.8237951199227637,
+      "learning_rate": 1.0700983748111792e-07,
+      "loss": 1.0311,
+      "step": 4971
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7337199542521688,
+      "learning_rate": 1.061026636757101e-07,
+      "loss": 1.0364,
+      "step": 4972
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7190336025669252,
+      "learning_rate": 1.0519933102190682e-07,
+      "loss": 1.0789,
+      "step": 4973
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5883263518680493,
+      "learning_rate": 1.0429983987041092e-07,
+      "loss": 1.0527,
+      "step": 4974
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6081756646699255,
+      "learning_rate": 1.0340419057043527e-07,
+      "loss": 0.9963,
+      "step": 4975
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6451741816844714,
+      "learning_rate": 1.0251238346970393e-07,
+      "loss": 1.0,
+      "step": 4976
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7148773958421688,
+      "learning_rate": 1.0162441891444441e-07,
+      "loss": 1.0072,
+      "step": 4977
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6905567531389444,
+      "learning_rate": 1.007402972493976e-07,
+      "loss": 0.9407,
+      "step": 4978
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.34988724332444154,
+      "learning_rate": 9.986001881780783e-08,
+      "loss": 1.0027,
+      "step": 4979
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7133597236032166,
+      "learning_rate": 9.898358396143171e-08,
+      "loss": 1.0196,
+      "step": 4980
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6815444641491755,
+      "learning_rate": 9.811099302052928e-08,
+      "loss": 1.0367,
+      "step": 4981
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7006126753540293,
+      "learning_rate": 9.72422463338718e-08,
+      "loss": 0.9316,
+      "step": 4982
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7257283839479504,
+      "learning_rate": 9.637734423873612e-08,
+      "loss": 0.9893,
+      "step": 4983
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7886108885120439,
+      "learning_rate": 9.55162870709081e-08,
+      "loss": 0.9229,
+      "step": 4984
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7419211642350928,
+      "learning_rate": 9.465907516467698e-08,
+      "loss": 1.0317,
+      "step": 4985
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6790378013829521,
+      "learning_rate": 9.380570885284546e-08,
+      "loss": 0.9789,
+      "step": 4986
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5749404348254519,
+      "learning_rate": 9.295618846671739e-08,
+      "loss": 1.0627,
+      "step": 4987
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7070864529006307,
+      "learning_rate": 9.211051433610674e-08,
+      "loss": 0.9915,
+      "step": 4988
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.607300366125942,
+      "learning_rate": 9.126868678933198e-08,
+      "loss": 1.0001,
+      "step": 4989
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5841800222541268,
+      "learning_rate": 9.04307061532217e-08,
+      "loss": 0.9996,
+      "step": 4990
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6858080109467463,
+      "learning_rate": 8.959657275310674e-08,
+      "loss": 0.9969,
+      "step": 4991
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7438083384629024,
+      "learning_rate": 8.876628691282918e-08,
+      "loss": 1.0472,
+      "step": 4992
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6114113516721676,
+      "learning_rate": 8.793984895473117e-08,
+      "loss": 0.9399,
+      "step": 4993
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7268900154031076,
+      "learning_rate": 8.711725919966718e-08,
+      "loss": 0.978,
+      "step": 4994
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6747825592096192,
+      "learning_rate": 8.629851796699284e-08,
+      "loss": 1.0331,
+      "step": 4995
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.729138674075179,
+      "learning_rate": 8.54836255745728e-08,
+      "loss": 0.9573,
+      "step": 4996
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6941709934093022,
+      "learning_rate": 8.467258233877728e-08,
+      "loss": 0.9476,
+      "step": 4997
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5491707326610836,
+      "learning_rate": 8.386538857447779e-08,
+      "loss": 0.9664,
+      "step": 4998
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5259174611883856,
+      "learning_rate": 8.306204459505807e-08,
+      "loss": 1.0865,
+      "step": 4999
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7115350346354932,
+      "learning_rate": 8.226255071240308e-08,
+      "loss": 0.9812,
+      "step": 5000
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7244785440653115,
+      "learning_rate": 8.146690723690342e-08,
+      "loss": 0.9971,
+      "step": 5001
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6939280401157272,
+      "learning_rate": 8.067511447745535e-08,
+      "loss": 0.9957,
+      "step": 5002
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6653861389504097,
+      "learning_rate": 7.988717274146074e-08,
+      "loss": 0.9987,
+      "step": 5003
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7667800774217337,
+      "learning_rate": 7.910308233482488e-08,
+      "loss": 1.0045,
+      "step": 5004
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6818729925585109,
+      "learning_rate": 7.832284356195764e-08,
+      "loss": 1.0317,
+      "step": 5005
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7888189653453217,
+      "learning_rate": 7.754645672577776e-08,
+      "loss": 1.0346,
+      "step": 5006
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6833706586249868,
+      "learning_rate": 7.677392212770196e-08,
+      "loss": 0.9756,
+      "step": 5007
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6667113487475931,
+      "learning_rate": 7.600524006765808e-08,
+      "loss": 1.0083,
+      "step": 5008
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3272683976331049,
+      "learning_rate": 7.524041084407185e-08,
+      "loss": 1.0681,
+      "step": 5009
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3272683976331049,
+      "learning_rate": 7.524041084407185e-08,
+      "loss": 1.1034,
+      "step": 5010
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3272683976331049,
+      "learning_rate": 7.524041084407185e-08,
+      "loss": 0.9768,
+      "step": 5011
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7960889913051786,
+      "learning_rate": 7.447943475387797e-08,
+      "loss": 0.8782,
+      "step": 5012
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6936147294849554,
+      "learning_rate": 7.372231209251346e-08,
+      "loss": 1.0233,
+      "step": 5013
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7422166974556376,
+      "learning_rate": 7.296904315391873e-08,
+      "loss": 0.9388,
+      "step": 5014
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5450844279375318,
+      "learning_rate": 7.221962823053874e-08,
+      "loss": 1.0913,
+      "step": 5015
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5031574168271759,
+      "learning_rate": 7.147406761332298e-08,
+      "loss": 1.0527,
+      "step": 5016
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6769886485610069,
+      "learning_rate": 7.073236159172325e-08,
+      "loss": 0.9775,
+      "step": 5017
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.75259388442019,
+      "learning_rate": 6.999451045369587e-08,
+      "loss": 1.0298,
+      "step": 5018
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7295229281975644,
+      "learning_rate": 6.926051448569948e-08,
+      "loss": 1.0377,
+      "step": 5019
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5799990413353053,
+      "learning_rate": 6.853037397269724e-08,
+      "loss": 0.9446,
+      "step": 5020
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4887452072832034,
+      "learning_rate": 6.78040891981524e-08,
+      "loss": 0.9944,
+      "step": 5021
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7636317627326282,
+      "learning_rate": 6.70816604440383e-08,
+      "loss": 1.0018,
+      "step": 5022
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6902217948007174,
+      "learning_rate": 6.63630879908217e-08,
+      "loss": 0.9662,
+      "step": 5023
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6868901565256662,
+      "learning_rate": 6.564837211748054e-08,
+      "loss": 0.9606,
+      "step": 5024
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7312548869544808,
+      "learning_rate": 6.493751310149177e-08,
+      "loss": 0.9849,
+      "step": 5025
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7244127794006132,
+      "learning_rate": 6.42305112188335e-08,
+      "loss": 1.0556,
+      "step": 5026
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6888402924950102,
+      "learning_rate": 6.352736674398951e-08,
+      "loss": 0.9941,
+      "step": 5027
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6985887315335906,
+      "learning_rate": 6.282807994994477e-08,
+      "loss": 0.9968,
+      "step": 5028
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7779007267621976,
+      "learning_rate": 6.213265110818656e-08,
+      "loss": 1.0559,
+      "step": 5029
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6326354647359925,
+      "learning_rate": 6.144108048870335e-08,
+      "loss": 0.9811,
+      "step": 5030
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7321084777198723,
+      "learning_rate": 6.075336835998813e-08,
+      "loss": 0.9777,
+      "step": 5031
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6787389030776171,
+      "learning_rate": 6.00695149890329e-08,
+      "loss": 0.9566,
+      "step": 5032
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6840615710676595,
+      "learning_rate": 5.938952064133419e-08,
+      "loss": 0.9729,
+      "step": 5033
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6767233947023481,
+      "learning_rate": 5.871338558088857e-08,
+      "loss": 0.9456,
+      "step": 5034
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6536038046564747,
+      "learning_rate": 5.8041110070194976e-08,
+      "loss": 0.974,
+      "step": 5035
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7314315434207944,
+      "learning_rate": 5.7372694370254614e-08,
+      "loss": 1.0174,
+      "step": 5036
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6877788133671443,
+      "learning_rate": 5.67081387405688e-08,
+      "loss": 1.0535,
+      "step": 5037
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7079523624030477,
+      "learning_rate": 5.6047443439141146e-08,
+      "loss": 0.9886,
+      "step": 5038
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6796385379722891,
+      "learning_rate": 5.539060872247537e-08,
+      "loss": 1.0391,
+      "step": 5039
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7130119532526179,
+      "learning_rate": 5.47376348455797e-08,
+      "loss": 0.9665,
+      "step": 5040
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6922717909472074,
+      "learning_rate": 5.408852206195914e-08,
+      "loss": 1.0099,
+      "step": 5041
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.603685795129296,
+      "learning_rate": 5.344327062362098e-08,
+      "loss": 1.0618,
+      "step": 5042
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7611242351030946,
+      "learning_rate": 5.2801880781075954e-08,
+      "loss": 0.97,
+      "step": 5043
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.695853243111935,
+      "learning_rate": 5.216435278333376e-08,
+      "loss": 0.9385,
+      "step": 5044
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6537728639966415,
+      "learning_rate": 5.153068687790197e-08,
+      "loss": 0.9852,
+      "step": 5045
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7151797375056249,
+      "learning_rate": 5.0900883310794903e-08,
+      "loss": 1.0076,
+      "step": 5046
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.695582447955905,
+      "learning_rate": 5.0274942326521414e-08,
+      "loss": 0.9358,
+      "step": 5047
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7722708251709799,
+      "learning_rate": 4.9652864168096e-08,
+      "loss": 1.0397,
+      "step": 5048
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.632362422889261,
+      "learning_rate": 4.9034649077027706e-08,
+      "loss": 1.0046,
+      "step": 5049
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6837349970673708,
+      "learning_rate": 4.84202972933312e-08,
+      "loss": 1.05,
+      "step": 5050
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6817521429337349,
+      "learning_rate": 4.7809809055517906e-08,
+      "loss": 1.06,
+      "step": 5051
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6839575908404362,
+      "learning_rate": 4.720318460060047e-08,
+      "loss": 1.0035,
+      "step": 5052
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7364704013794777,
+      "learning_rate": 4.6600424164091606e-08,
+      "loss": 1.0423,
+      "step": 5053
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.685152020841155,
+      "learning_rate": 4.6001527980004125e-08,
+      "loss": 0.971,
+      "step": 5054
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7078865289173325,
+      "learning_rate": 4.54064962808487e-08,
+      "loss": 1.0535,
+      "step": 5055
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7198934337144755,
+      "learning_rate": 4.4815329297639434e-08,
+      "loss": 0.9853,
+      "step": 5056
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6848867673481766,
+      "learning_rate": 4.422802725988606e-08,
+      "loss": 1.0165,
+      "step": 5057
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7255852715472542,
+      "learning_rate": 4.364459039559843e-08,
+      "loss": 1.0159,
+      "step": 5058
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.681236562820168,
+      "learning_rate": 4.3065018931289784e-08,
+      "loss": 0.9846,
+      "step": 5059
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7006189818435471,
+      "learning_rate": 4.248931309196791e-08,
+      "loss": 1.027,
+      "step": 5060
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5918754917825644,
+      "learning_rate": 4.1917473101140696e-08,
+      "loss": 1.0481,
+      "step": 5061
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6402736293714691,
+      "learning_rate": 4.134949918081832e-08,
+      "loss": 1.0264,
+      "step": 5062
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7297350359795163,
+      "learning_rate": 4.0785391551506626e-08,
+      "loss": 0.9327,
+      "step": 5063
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.697229658239369,
+      "learning_rate": 4.022515043221154e-08,
+      "loss": 1.0217,
+      "step": 5064
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7555142963418877,
+      "learning_rate": 3.966877604043795e-08,
+      "loss": 1.0008,
+      "step": 5065
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6667900887163185,
+      "learning_rate": 3.9116268592189755e-08,
+      "loss": 1.021,
+      "step": 5066
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7632939623349013,
+      "learning_rate": 3.8567628301969806e-08,
+      "loss": 0.8868,
+      "step": 5067
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7562877560526701,
+      "learning_rate": 3.802285538277772e-08,
+      "loss": 1.0451,
+      "step": 5068
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6901714236371055,
+      "learning_rate": 3.748195004611543e-08,
+      "loss": 1.1124,
+      "step": 5069
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6757956485405063,
+      "learning_rate": 3.69449125019794e-08,
+      "loss": 0.987,
+      "step": 5070
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7903296564829994,
+      "learning_rate": 3.6411742958866184e-08,
+      "loss": 0.9919,
+      "step": 5071
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6950866665895584,
+      "learning_rate": 3.588244162377019e-08,
+      "loss": 1.0416,
+      "step": 5072
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6682381197720808,
+      "learning_rate": 3.5357008702185945e-08,
+      "loss": 0.9957,
+      "step": 5073
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.745398320894013,
+      "learning_rate": 3.483544439810249e-08,
+      "loss": 0.9597,
+      "step": 5074
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7613272704151002,
+      "learning_rate": 3.4317748914011187e-08,
+      "loss": 1.0201,
+      "step": 5075
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6974552728855099,
+      "learning_rate": 3.3803922450897917e-08,
+      "loss": 1.0005,
+      "step": 5076
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6859816743980273,
+      "learning_rate": 3.329396520824757e-08,
+      "loss": 0.9603,
+      "step": 5077
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7157609165862128,
+      "learning_rate": 3.2787877384045095e-08,
+      "loss": 1.0206,
+      "step": 5078
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6798187286477865,
+      "learning_rate": 3.228565917476889e-08,
+      "loss": 1.095,
+      "step": 5079
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6705213110782559,
+      "learning_rate": 3.178731077539743e-08,
+      "loss": 0.9858,
+      "step": 5080
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6682001025415432,
+      "learning_rate": 3.129283237940928e-08,
+      "loss": 1.0286,
+      "step": 5081
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7253579066896588,
+      "learning_rate": 3.080222417877421e-08,
+      "loss": 1.0345,
+      "step": 5082
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6923290694920561,
+      "learning_rate": 3.031548636396764e-08,
+      "loss": 0.8899,
+      "step": 5083
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7137897294381143,
+      "learning_rate": 2.983261912395397e-08,
+      "loss": 0.9199,
+      "step": 5084
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7354403952035664,
+      "learning_rate": 2.9353622646199898e-08,
+      "loss": 0.9684,
+      "step": 5085
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5421906133359983,
+      "learning_rate": 2.8878497116671124e-08,
+      "loss": 1.1039,
+      "step": 5086
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7200196484789824,
+      "learning_rate": 2.8407242719823424e-08,
+      "loss": 0.9991,
+      "step": 5087
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6570378930380982,
+      "learning_rate": 2.7939859638617118e-08,
+      "loss": 0.9897,
+      "step": 5088
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7040825077678265,
+      "learning_rate": 2.7476348054504832e-08,
+      "loss": 1.0296,
+      "step": 5089
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6450393386102231,
+      "learning_rate": 2.7016708147439285e-08,
+      "loss": 1.0099,
+      "step": 5090
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6541343413578153,
+      "learning_rate": 2.6560940095866626e-08,
+      "loss": 0.9838,
+      "step": 5091
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7108977763432106,
+      "learning_rate": 2.6109044076733092e-08,
+      "loss": 1.0532,
+      "step": 5092
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7090808154160321,
+      "learning_rate": 2.5661020265479452e-08,
+      "loss": 1.0334,
+      "step": 5093
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7179524593309223,
+      "learning_rate": 2.5216868836043242e-08,
+      "loss": 1.0071,
+      "step": 5094
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7242020058503096,
+      "learning_rate": 2.4776589960862074e-08,
+      "loss": 0.9833,
+      "step": 5095
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7483196995692523,
+      "learning_rate": 2.434018381086589e-08,
+      "loss": 0.99,
+      "step": 5096
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7513700135270045,
+      "learning_rate": 2.3907650555481387e-08,
+      "loss": 0.9359,
+      "step": 5097
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7120749116966201,
+      "learning_rate": 2.3478990362634235e-08,
+      "loss": 1.0717,
+      "step": 5098
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6942357642725627,
+      "learning_rate": 2.3054203398743537e-08,
+      "loss": 0.9881,
+      "step": 5099
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6516295242829304,
+      "learning_rate": 2.263328982872959e-08,
+      "loss": 1.0015,
+      "step": 5100
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7171876828740739,
+      "learning_rate": 2.221624981600168e-08,
+      "loss": 0.9851,
+      "step": 5101
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6647470423518678,
+      "learning_rate": 2.1803083522471402e-08,
+      "loss": 1.0024,
+      "step": 5102
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.636787420958267,
+      "learning_rate": 2.1393791108542672e-08,
+      "loss": 0.9744,
+      "step": 5103
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7161462880641183,
+      "learning_rate": 2.098837273311838e-08,
+      "loss": 0.9838,
+      "step": 5104
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6925314761867294,
+      "learning_rate": 2.058682855359595e-08,
+      "loss": 1.0178,
+      "step": 5105
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7372792098633482,
+      "learning_rate": 2.0189158725867353e-08,
+      "loss": 1.028,
+      "step": 5106
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6922041149598059,
+      "learning_rate": 1.979536340432131e-08,
+      "loss": 0.9907,
+      "step": 5107
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6831313581971656,
+      "learning_rate": 1.9405442741844415e-08,
+      "loss": 0.9539,
+      "step": 5108
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6611245157246511,
+      "learning_rate": 1.9019396889816688e-08,
+      "loss": 0.9785,
+      "step": 5109
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6652653693310717,
+      "learning_rate": 1.8637225998114904e-08,
+      "loss": 0.9688,
+      "step": 5110
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.72867028453072,
+      "learning_rate": 1.825893021510927e-08,
+      "loss": 0.9598,
+      "step": 5111
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7178169676772269,
+      "learning_rate": 1.7884509687668972e-08,
+      "loss": 0.9038,
+      "step": 5112
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5603287541301961,
+      "learning_rate": 1.7513964561156617e-08,
+      "loss": 1.047,
+      "step": 5113
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5403026702439107,
+      "learning_rate": 1.714729497942935e-08,
+      "loss": 1.0329,
+      "step": 5114
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5747575709597237,
+      "learning_rate": 1.6784501084843307e-08,
+      "loss": 1.0967,
+      "step": 5115
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7294595330677918,
+      "learning_rate": 1.6425583018244706e-08,
+      "loss": 0.9549,
+      "step": 5116
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7511700266072752,
+      "learning_rate": 1.607054091897986e-08,
+      "loss": 0.9976,
+      "step": 5117
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6889945388038763,
+      "learning_rate": 1.57193749248874e-08,
+      "loss": 1.0187,
+      "step": 5118
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7287249516199531,
+      "learning_rate": 1.537208517230271e-08,
+      "loss": 1.0236,
+      "step": 5119
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7282338861559292,
+      "learning_rate": 1.5028671796055715e-08,
+      "loss": 0.9658,
+      "step": 5120
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6630277131102853,
+      "learning_rate": 1.4689134929470884e-08,
+      "loss": 1.0538,
+      "step": 5121
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7240418495744013,
+      "learning_rate": 1.435347470436832e-08,
+      "loss": 0.97,
+      "step": 5122
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6518755045519495,
+      "learning_rate": 1.4021691251062675e-08,
+      "loss": 0.9401,
+      "step": 5123
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7384718511046704,
+      "learning_rate": 1.3693784698363133e-08,
+      "loss": 1.0393,
+      "step": 5124
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6855907985227943,
+      "learning_rate": 1.3369755173575639e-08,
+      "loss": 0.9562,
+      "step": 5125
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6804115002026423,
+      "learning_rate": 1.3049602802498451e-08,
+      "loss": 0.9879,
+      "step": 5126
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6968723108409004,
+      "learning_rate": 1.273332770942659e-08,
+      "loss": 1.0416,
+      "step": 5127
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6572432860766092,
+      "learning_rate": 1.2420930017148503e-08,
+      "loss": 0.9411,
+      "step": 5128
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6860933557168902,
+      "learning_rate": 1.2112409846947171e-08,
+      "loss": 0.9383,
+      "step": 5129
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7260107448664092,
+      "learning_rate": 1.1807767318602337e-08,
+      "loss": 0.9874,
+      "step": 5130
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6686793357642703,
+      "learning_rate": 1.150700255038606e-08,
+      "loss": 1.0537,
+      "step": 5131
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6974314590218886,
+      "learning_rate": 1.1210115659063825e-08,
+      "loss": 1.0021,
+      "step": 5132
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7291542797349717,
+      "learning_rate": 1.0917106759900097e-08,
+      "loss": 1.015,
+      "step": 5133
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7346500524607595,
+      "learning_rate": 1.0627975966649439e-08,
+      "loss": 0.9999,
+      "step": 5134
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6611481618590891,
+      "learning_rate": 1.034272339156206e-08,
+      "loss": 1.002,
+      "step": 5135
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6917497005405219,
+      "learning_rate": 1.0061349145383814e-08,
+      "loss": 0.9598,
+      "step": 5136
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7650870829109853,
+      "learning_rate": 9.783853337353987e-09,
+      "loss": 1.012,
+      "step": 5137
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6565608055401074,
+      "learning_rate": 9.510236075205292e-09,
+      "loss": 0.993,
+      "step": 5138
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6849734018708395,
+      "learning_rate": 9.240497465164978e-09,
+      "loss": 0.9898,
+      "step": 5139
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6916128422280186,
+      "learning_rate": 8.974637611955939e-09,
+      "loss": 1.0112,
+      "step": 5140
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6870988797368203,
+      "learning_rate": 8.712656618793391e-09,
+      "loss": 0.9808,
+      "step": 5141
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6905249603756547,
+      "learning_rate": 8.454554587388198e-09,
+      "loss": 1.043,
+      "step": 5142
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5182127704275881,
+      "learning_rate": 8.200331617943535e-09,
+      "loss": 1.0021,
+      "step": 5143
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7066009665273244,
+      "learning_rate": 7.949987809158232e-09,
+      "loss": 1.0005,
+      "step": 5144
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.736818379326403,
+      "learning_rate": 7.703523258223433e-09,
+      "loss": 1.0701,
+      "step": 5145
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6344270096827967,
+      "learning_rate": 7.460938060825929e-09,
+      "loss": 1.0834,
+      "step": 5146
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7192320976562581,
+      "learning_rate": 7.222232311145938e-09,
+      "loss": 1.0688,
+      "step": 5147
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6542969434400744,
+      "learning_rate": 6.987406101855998e-09,
+      "loss": 0.9951,
+      "step": 5148
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6661922753648826,
+      "learning_rate": 6.756459524125403e-09,
+      "loss": 1.0532,
+      "step": 5149
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7153709921111918,
+      "learning_rate": 6.5293926676135434e-09,
+      "loss": 0.9467,
+      "step": 5150
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6641789869154785,
+      "learning_rate": 6.306205620477679e-09,
+      "loss": 0.9388,
+      "step": 5151
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5705821089585555,
+      "learning_rate": 6.086898469365166e-09,
+      "loss": 1.0508,
+      "step": 5152
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6344634042995427,
+      "learning_rate": 5.871471299419007e-09,
+      "loss": 0.997,
+      "step": 5153
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6549856042179418,
+      "learning_rate": 5.6599241942767445e-09,
+      "loss": 0.9518,
+      "step": 5154
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.684997541937484,
+      "learning_rate": 5.452257236066017e-09,
+      "loss": 0.9743,
+      "step": 5155
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6678235203740501,
+      "learning_rate": 5.248470505412328e-09,
+      "loss": 0.9569,
+      "step": 5156
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7522931526936539,
+      "learning_rate": 5.0485640814312844e-09,
+      "loss": 0.9306,
+      "step": 5157
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6877578291905502,
+      "learning_rate": 4.8525380417330234e-09,
+      "loss": 0.9684,
+      "step": 5158
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5513137516302983,
+      "learning_rate": 4.660392462424446e-09,
+      "loss": 0.996,
+      "step": 5159
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7140804225679507,
+      "learning_rate": 4.472127418099215e-09,
+      "loss": 1.0259,
+      "step": 5160
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5524340244974044,
+      "learning_rate": 4.287742981851084e-09,
+      "loss": 1.1008,
+      "step": 5161
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5076547463790719,
+      "learning_rate": 4.1072392252639034e-09,
+      "loss": 1.0527,
+      "step": 5162
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6926144519676927,
+      "learning_rate": 3.930616218414951e-09,
+      "loss": 0.9849,
+      "step": 5163
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7806161628712427,
+      "learning_rate": 3.757874029874931e-09,
+      "loss": 1.107,
+      "step": 5164
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5348564648523596,
+      "learning_rate": 3.5890127267090844e-09,
+      "loss": 1.0753,
+      "step": 5165
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7882232661928668,
+      "learning_rate": 3.424032374476083e-09,
+      "loss": 1.0262,
+      "step": 5166
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.690510954610818,
+      "learning_rate": 3.2629330372246915e-09,
+      "loss": 0.9285,
+      "step": 5167
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6881557750868693,
+      "learning_rate": 3.105714777501545e-09,
+      "loss": 0.9064,
+      "step": 5168
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6729273261556528,
+      "learning_rate": 2.9523776563422644e-09,
+      "loss": 1.0201,
+      "step": 5169
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7250973094635889,
+      "learning_rate": 2.802921733278119e-09,
+      "loss": 0.9686,
+      "step": 5170
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6646767392960927,
+      "learning_rate": 2.657347066333804e-09,
+      "loss": 0.9147,
+      "step": 5171
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7663316745115161,
+      "learning_rate": 2.5156537120263335e-09,
+      "loss": 0.9339,
+      "step": 5172
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6881102682386533,
+      "learning_rate": 2.3778417253650376e-09,
+      "loss": 0.9878,
+      "step": 5173
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7121401499495653,
+      "learning_rate": 2.2439111598537844e-09,
+      "loss": 1.0036,
+      "step": 5174
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7284156619929044,
+      "learning_rate": 2.113862067488759e-09,
+      "loss": 1.0427,
+      "step": 5175
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7161973707820296,
+      "learning_rate": 1.987694498760684e-09,
+      "loss": 0.9915,
+      "step": 5176
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6929534699237861,
+      "learning_rate": 1.865408502650379e-09,
+      "loss": 1.0068,
+      "step": 5177
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6757996038521413,
+      "learning_rate": 1.747004126635421e-09,
+      "loss": 1.0042,
+      "step": 5178
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6585110038568899,
+      "learning_rate": 1.6324814166823744e-09,
+      "loss": 0.9883,
+      "step": 5179
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7561638301866249,
+      "learning_rate": 1.5218404172545609e-09,
+      "loss": 1.0404,
+      "step": 5180
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7233494423929635,
+      "learning_rate": 1.415081171305399e-09,
+      "loss": 1.0598,
+      "step": 5181
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6542000443819873,
+      "learning_rate": 1.3122037202828452e-09,
+      "loss": 1.0067,
+      "step": 5182
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6236247604906923,
+      "learning_rate": 1.2132081041282829e-09,
+      "loss": 1.0048,
+      "step": 5183
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5793562779388024,
+      "learning_rate": 1.1180943612754124e-09,
+      "loss": 1.0079,
+      "step": 5184
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7197895113602266,
+      "learning_rate": 1.026862528649142e-09,
+      "loss": 1.0334,
+      "step": 5185
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6846545698929309,
+      "learning_rate": 9.39512641668916e-10,
+      "loss": 1.0045,
+      "step": 5186
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7423122838371287,
+      "learning_rate": 8.560447342487177e-10,
+      "loss": 1.0577,
+      "step": 5187
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7041080263253731,
+      "learning_rate": 7.764588387915161e-10,
+      "loss": 0.965,
+      "step": 5188
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6444555785841867,
+      "learning_rate": 7.007549861970387e-10,
+      "loss": 1.0129,
+      "step": 5189
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7265446301784352,
+      "learning_rate": 6.289332058551089e-10,
+      "loss": 0.9568,
+      "step": 5190
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6557999056144201,
+      "learning_rate": 5.609935256500887e-10,
+      "loss": 0.984,
+      "step": 5191
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6313881697171624,
+      "learning_rate": 4.969359719586563e-10,
+      "loss": 0.9525,
+      "step": 5192
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6840602146291601,
+      "learning_rate": 4.3676056964869764e-10,
+      "loss": 0.9781,
+      "step": 5193
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6762265470846909,
+      "learning_rate": 3.804673420837457e-10,
+      "loss": 0.9731,
+      "step": 5194
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.713653052530832,
+      "learning_rate": 3.2805631111743064e-10,
+      "loss": 0.9266,
+      "step": 5195
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6811173029718917,
+      "learning_rate": 2.795274971001405e-10,
+      "loss": 1.0062,
+      "step": 5196
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6817899720946584,
+      "learning_rate": 2.3488091886902933e-10,
+      "loss": 1.0544,
+      "step": 5197
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6457665459848412,
+      "learning_rate": 1.941165937602296e-10,
+      "loss": 1.0338,
+      "step": 5198
+    },
+    {
+      "epoch": 1.0,
+      "step": 5198,
+      "total_flos": 3907018767597568.0,
+      "train_loss": 1.0487468054589055,
+      "train_runtime": 24096.2279,
+      "train_samples_per_second": 27.61,
+      "train_steps_per_second": 0.216
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 5198,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 50000,
+  "total_flos": 3907018767597568.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}