diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,18132 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9997583723964626,
+  "eval_steps": 500,
+  "global_step": 2586,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 47.279807180167374,
+      "learning_rate": 1.282051282051282e-07,
+      "loss": 12.1123,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 29.234797993521592,
+      "learning_rate": 2.564102564102564e-07,
+      "loss": 12.1377,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 48.013302305052015,
+      "learning_rate": 3.846153846153847e-07,
+      "loss": 12.0928,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 49.648225967434605,
+      "learning_rate": 5.128205128205128e-07,
+      "loss": 12.1123,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 41.50977099753674,
+      "learning_rate": 6.41025641025641e-07,
+      "loss": 12.0527,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 43.3095044824057,
+      "learning_rate": 7.692307692307694e-07,
+      "loss": 12.0547,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 31.569078860107826,
+      "learning_rate": 8.974358974358975e-07,
+      "loss": 12.1328,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 42.593654219739186,
+      "learning_rate": 1.0256410256410257e-06,
+      "loss": 12.0615,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 33.585740478581016,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 12.0371,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 39.77975641784306,
+      "learning_rate": 1.282051282051282e-06,
+      "loss": 12.001,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 23.23163778179969,
+      "learning_rate": 1.4102564102564104e-06,
+      "loss": 12.1406,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 35.197855381569674,
+      "learning_rate": 1.5384615384615387e-06,
+      "loss": 11.9727,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 35.28289473472646,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 11.8623,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 22.964433832401195,
+      "learning_rate": 1.794871794871795e-06,
+      "loss": 11.8867,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 32.38348517511878,
+      "learning_rate": 1.9230769230769234e-06,
+      "loss": 11.7998,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 21.734040594578076,
+      "learning_rate": 2.0512820512820513e-06,
+      "loss": 11.7617,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 25.19673160652617,
+      "learning_rate": 2.1794871794871797e-06,
+      "loss": 11.6992,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 29.296531551199227,
+      "learning_rate": 2.307692307692308e-06,
+      "loss": 11.5586,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 46.13003375968801,
+      "learning_rate": 2.435897435897436e-06,
+      "loss": 11.5527,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 41.849036854455164,
+      "learning_rate": 2.564102564102564e-06,
+      "loss": 11.3779,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 39.537236073998145,
+      "learning_rate": 2.6923076923076923e-06,
+      "loss": 11.2012,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 34.57172567317474,
+      "learning_rate": 2.8205128205128207e-06,
+      "loss": 11.0215,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 506.69228168760816,
+      "learning_rate": 2.948717948717949e-06,
+      "loss": 12.9199,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 111.64311280501529,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 10.7695,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 316.1715029609909,
+      "learning_rate": 3.205128205128206e-06,
+      "loss": 12.2754,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 73.89514507362225,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 10.127,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 182.78134072736862,
+      "learning_rate": 3.4615384615384617e-06,
+      "loss": 10.1006,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 254.93513728578208,
+      "learning_rate": 3.58974358974359e-06,
+      "loss": 10.3701,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 148.58471389468403,
+      "learning_rate": 3.7179487179487184e-06,
+      "loss": 8.9219,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 108.77056549970071,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 9.1855,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 103.87787595236031,
+      "learning_rate": 3.974358974358974e-06,
+      "loss": 8.1724,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 102.31703480340848,
+      "learning_rate": 4.102564102564103e-06,
+      "loss": 6.9956,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 172.46373641294485,
+      "learning_rate": 4.230769230769231e-06,
+      "loss": 6.3066,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 87.5216207739019,
+      "learning_rate": 4.358974358974359e-06,
+      "loss": 5.5588,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 80.50686987902449,
+      "learning_rate": 4.487179487179488e-06,
+      "loss": 4.576,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 57.645390934076,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 3.7121,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 37.675481020558486,
+      "learning_rate": 4.743589743589744e-06,
+      "loss": 3.5498,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 27.516266863848507,
+      "learning_rate": 4.871794871794872e-06,
+      "loss": 3.0763,
+      "step": 38
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 30.54432213720788,
+      "learning_rate": 5e-06,
+      "loss": 2.5987,
+      "step": 39
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 25.419156365403072,
+      "learning_rate": 5.128205128205128e-06,
+      "loss": 2.578,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 22.94976417729944,
+      "learning_rate": 5.256410256410257e-06,
+      "loss": 2.5072,
+      "step": 41
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 19.709663468695748,
+      "learning_rate": 5.384615384615385e-06,
+      "loss": 2.2178,
+      "step": 42
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.624104489751552,
+      "learning_rate": 5.512820512820514e-06,
+      "loss": 1.9278,
+      "step": 43
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 18.649553341368065,
+      "learning_rate": 5.641025641025641e-06,
+      "loss": 1.8549,
+      "step": 44
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 54.778542254207636,
+      "learning_rate": 5.769230769230769e-06,
+      "loss": 2.0851,
+      "step": 45
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 16.034705703498012,
+      "learning_rate": 5.897435897435898e-06,
+      "loss": 1.5248,
+      "step": 46
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.61319779654908,
+      "learning_rate": 6.025641025641026e-06,
+      "loss": 1.5049,
+      "step": 47
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.013952030578054,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 1.5219,
+      "step": 48
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 20.364691223375658,
+      "learning_rate": 6.282051282051282e-06,
+      "loss": 1.2063,
+      "step": 49
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.841850856770195,
+      "learning_rate": 6.410256410256412e-06,
+      "loss": 1.5725,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.029717943966945,
+      "learning_rate": 6.538461538461539e-06,
+      "loss": 1.1939,
+      "step": 51
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 34.08674532825822,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.0203,
+      "step": 52
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 16.254122121541265,
+      "learning_rate": 6.794871794871796e-06,
+      "loss": 0.9564,
+      "step": 53
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 17.88736857446396,
+      "learning_rate": 6.923076923076923e-06,
+      "loss": 1.443,
+      "step": 54
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.852289912395646,
+      "learning_rate": 7.051282051282053e-06,
+      "loss": 1.1138,
+      "step": 55
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 20.262325334929987,
+      "learning_rate": 7.17948717948718e-06,
+      "loss": 0.9219,
+      "step": 56
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 15.85832036827058,
+      "learning_rate": 7.307692307692308e-06,
+      "loss": 0.8468,
+      "step": 57
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.689120246753362,
+      "learning_rate": 7.435897435897437e-06,
+      "loss": 1.2285,
+      "step": 58
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.411046871864485,
+      "learning_rate": 7.564102564102564e-06,
+      "loss": 1.0675,
+      "step": 59
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.193089449601917,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 1.2942,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.848711317418662,
+      "learning_rate": 7.820512820512822e-06,
+      "loss": 0.7793,
+      "step": 61
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.254231397561643,
+      "learning_rate": 7.948717948717949e-06,
+      "loss": 1.0076,
+      "step": 62
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.224081466208796,
+      "learning_rate": 8.076923076923077e-06,
+      "loss": 0.7495,
+      "step": 63
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 19.198184160612673,
+      "learning_rate": 8.205128205128205e-06,
+      "loss": 0.7391,
+      "step": 64
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.603751018340077,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 1.1272,
+      "step": 65
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.648011495220087,
+      "learning_rate": 8.461538461538462e-06,
+      "loss": 0.9172,
+      "step": 66
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.597663609747956,
+      "learning_rate": 8.58974358974359e-06,
+      "loss": 1.0398,
+      "step": 67
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.80888600284641,
+      "learning_rate": 8.717948717948719e-06,
+      "loss": 0.8381,
+      "step": 68
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.265720281737142,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 1.0565,
+      "step": 69
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.327563244701436,
+      "learning_rate": 8.974358974358976e-06,
+      "loss": 0.7898,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.548464366276166,
+      "learning_rate": 9.102564102564104e-06,
+      "loss": 1.0255,
+      "step": 71
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.654369764317849,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 1.1506,
+      "step": 72
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.733536199369656,
+      "learning_rate": 9.358974358974359e-06,
+      "loss": 0.896,
+      "step": 73
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.074795791781662,
+      "learning_rate": 9.487179487179487e-06,
+      "loss": 1.217,
+      "step": 74
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.301488347706664,
+      "learning_rate": 9.615384615384616e-06,
+      "loss": 0.8091,
+      "step": 75
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.895001720574654,
+      "learning_rate": 9.743589743589744e-06,
+      "loss": 0.8315,
+      "step": 76
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.33028882065257,
+      "learning_rate": 9.871794871794872e-06,
+      "loss": 0.9963,
+      "step": 77
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.031516567458516,
+      "learning_rate": 1e-05,
+      "loss": 1.2161,
+      "step": 78
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.548517188835804,
+      "learning_rate": 9.999996077304179e-06,
+      "loss": 0.8166,
+      "step": 79
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.368075026226792,
+      "learning_rate": 9.999984309222866e-06,
+      "loss": 0.5367,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.017734661169415,
+      "learning_rate": 9.999964695774527e-06,
+      "loss": 0.7728,
+      "step": 81
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.113244343569303,
+      "learning_rate": 9.99993723698994e-06,
+      "loss": 0.9339,
+      "step": 82
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.808266144501516,
+      "learning_rate": 9.999901932912188e-06,
+      "loss": 0.591,
+      "step": 83
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.135445364844132,
+      "learning_rate": 9.999858783596665e-06,
+      "loss": 0.9249,
+      "step": 84
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.530424283253378,
+      "learning_rate": 9.999807789111075e-06,
+      "loss": 0.643,
+      "step": 85
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.558115278249298,
+      "learning_rate": 9.999748949535436e-06,
+      "loss": 1.0037,
+      "step": 86
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.46631822932675,
+      "learning_rate": 9.99968226496207e-06,
+      "loss": 0.9391,
+      "step": 87
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.533803457871283,
+      "learning_rate": 9.999607735495609e-06,
+      "loss": 0.4386,
+      "step": 88
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.575415222368854,
+      "learning_rate": 9.999525361252996e-06,
+      "loss": 0.4872,
+      "step": 89
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 12.862480239438767,
+      "learning_rate": 9.999435142363484e-06,
+      "loss": 0.997,
+      "step": 90
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.265338092891238,
+      "learning_rate": 9.999337078968633e-06,
+      "loss": 0.4861,
+      "step": 91
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.813132934737774,
+      "learning_rate": 9.999231171222312e-06,
+      "loss": 0.4695,
+      "step": 92
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.494494367234264,
+      "learning_rate": 9.999117419290698e-06,
+      "loss": 0.9507,
+      "step": 93
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.29731472008422,
+      "learning_rate": 9.998995823352276e-06,
+      "loss": 1.0835,
+      "step": 94
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.973003505943797,
+      "learning_rate": 9.998866383597842e-06,
+      "loss": 0.6378,
+      "step": 95
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.378602523102169,
+      "learning_rate": 9.998729100230497e-06,
+      "loss": 1.014,
+      "step": 96
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.339949662257846,
+      "learning_rate": 9.998583973465647e-06,
+      "loss": 0.3508,
+      "step": 97
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.54455719121349,
+      "learning_rate": 9.998431003531008e-06,
+      "loss": 0.5341,
+      "step": 98
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.967610734692199,
+      "learning_rate": 9.998270190666602e-06,
+      "loss": 0.5081,
+      "step": 99
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.376613736854886,
+      "learning_rate": 9.998101535124758e-06,
+      "loss": 0.3209,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.601722192201839,
+      "learning_rate": 9.99792503717011e-06,
+      "loss": 0.4846,
+      "step": 101
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.22359026928023,
+      "learning_rate": 9.997740697079595e-06,
+      "loss": 0.7585,
+      "step": 102
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.040968238862975,
+      "learning_rate": 9.997548515142457e-06,
+      "loss": 0.5073,
+      "step": 103
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.084715042296548,
+      "learning_rate": 9.997348491660247e-06,
+      "loss": 0.6983,
+      "step": 104
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.4086997542404776,
+      "learning_rate": 9.997140626946815e-06,
+      "loss": 0.3524,
+      "step": 105
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 11.077089617894178,
+      "learning_rate": 9.99692492132832e-06,
+      "loss": 0.6197,
+      "step": 106
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.61237049405628,
+      "learning_rate": 9.996701375143216e-06,
+      "loss": 0.4866,
+      "step": 107
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.230360068207208,
+      "learning_rate": 9.99646998874227e-06,
+      "loss": 0.5656,
+      "step": 108
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.425448570927497,
+      "learning_rate": 9.996230762488544e-06,
+      "loss": 0.8637,
+      "step": 109
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.744697312303776,
+      "learning_rate": 9.9959836967574e-06,
+      "loss": 0.5646,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.3022800651898585,
+      "learning_rate": 9.995728791936505e-06,
+      "loss": 0.3901,
+      "step": 111
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.15302821871786,
+      "learning_rate": 9.995466048425825e-06,
+      "loss": 0.65,
+      "step": 112
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.607942057954628,
+      "learning_rate": 9.995195466637629e-06,
+      "loss": 0.4055,
+      "step": 113
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.6398697815205265,
+      "learning_rate": 9.994917046996472e-06,
+      "loss": 0.9344,
+      "step": 114
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.122555689230712,
+      "learning_rate": 9.994630789939223e-06,
+      "loss": 0.8358,
+      "step": 115
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.094408808052112,
+      "learning_rate": 9.994336695915041e-06,
+      "loss": 0.8405,
+      "step": 116
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.05662306983693,
+      "learning_rate": 9.99403476538538e-06,
+      "loss": 0.8575,
+      "step": 117
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.143200299838425,
+      "learning_rate": 9.993724998823995e-06,
+      "loss": 0.8526,
+      "step": 118
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.508983052481922,
+      "learning_rate": 9.993407396716935e-06,
+      "loss": 0.7197,
+      "step": 119
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.253313336966283,
+      "learning_rate": 9.993081959562539e-06,
+      "loss": 0.6145,
+      "step": 120
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.400637991921391,
+      "learning_rate": 9.992748687871445e-06,
+      "loss": 0.686,
+      "step": 121
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.266223281104883,
+      "learning_rate": 9.992407582166582e-06,
+      "loss": 0.2203,
+      "step": 122
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.173864805127318,
+      "learning_rate": 9.992058642983173e-06,
+      "loss": 0.5649,
+      "step": 123
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.134578213259644,
+      "learning_rate": 9.991701870868731e-06,
+      "loss": 0.8683,
+      "step": 124
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.753200529590611,
+      "learning_rate": 9.991337266383056e-06,
+      "loss": 0.7512,
+      "step": 125
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.693412664995114,
+      "learning_rate": 9.990964830098246e-06,
+      "loss": 0.7306,
+      "step": 126
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.082364863469143,
+      "learning_rate": 9.990584562598679e-06,
+      "loss": 0.7836,
+      "step": 127
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.491549029464382,
+      "learning_rate": 9.990196464481025e-06,
+      "loss": 0.5206,
+      "step": 128
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.579409745338709,
+      "learning_rate": 9.989800536354243e-06,
+      "loss": 0.629,
+      "step": 129
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.202005969283118,
+      "learning_rate": 9.989396778839572e-06,
+      "loss": 0.428,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.742828175807453,
+      "learning_rate": 9.988985192570541e-06,
+      "loss": 0.7628,
+      "step": 131
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.611548082050718,
+      "learning_rate": 9.98856577819296e-06,
+      "loss": 0.4257,
+      "step": 132
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.783167273856261,
+      "learning_rate": 9.988138536364922e-06,
+      "loss": 0.3008,
+      "step": 133
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.184412983765748,
+      "learning_rate": 9.987703467756807e-06,
+      "loss": 0.7117,
+      "step": 134
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.054224609301897,
+      "learning_rate": 9.987260573051268e-06,
+      "loss": 0.4991,
+      "step": 135
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.54045338023512,
+      "learning_rate": 9.986809852943244e-06,
+      "loss": 0.3961,
+      "step": 136
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.999202345977584,
+      "learning_rate": 9.986351308139948e-06,
+      "loss": 0.4433,
+      "step": 137
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.657041187137718,
+      "learning_rate": 9.985884939360873e-06,
+      "loss": 0.7117,
+      "step": 138
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.736255352268028,
+      "learning_rate": 9.985410747337788e-06,
+      "loss": 0.5681,
+      "step": 139
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.4160416582186235,
+      "learning_rate": 9.984928732814738e-06,
+      "loss": 0.7026,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.551085654208427,
+      "learning_rate": 9.984438896548043e-06,
+      "loss": 0.8086,
+      "step": 141
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.585721693914806,
+      "learning_rate": 9.983941239306291e-06,
+      "loss": 0.5382,
+      "step": 142
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.509174347972325,
+      "learning_rate": 9.98343576187035e-06,
+      "loss": 0.5778,
+      "step": 143
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.064880941306638,
+      "learning_rate": 9.98292246503335e-06,
+      "loss": 0.2871,
+      "step": 144
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.19792899129946,
+      "learning_rate": 9.982401349600695e-06,
+      "loss": 1.0037,
+      "step": 145
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.544744033275711,
+      "learning_rate": 9.981872416390055e-06,
+      "loss": 0.505,
+      "step": 146
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.765816943542017,
+      "learning_rate": 9.981335666231369e-06,
+      "loss": 0.6552,
+      "step": 147
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.94756159633514,
+      "learning_rate": 9.98079109996684e-06,
+      "loss": 0.8352,
+      "step": 148
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.327165984682571,
+      "learning_rate": 9.980238718450934e-06,
+      "loss": 0.6213,
+      "step": 149
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.680930771815867,
+      "learning_rate": 9.979678522550382e-06,
+      "loss": 0.561,
+      "step": 150
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.015659229577425,
+      "learning_rate": 9.979110513144175e-06,
+      "loss": 0.8584,
+      "step": 151
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.95101306790891,
+      "learning_rate": 9.978534691123563e-06,
+      "loss": 0.9032,
+      "step": 152
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.458806183904773,
+      "learning_rate": 9.977951057392057e-06,
+      "loss": 0.7042,
+      "step": 153
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.914539407204849,
+      "learning_rate": 9.977359612865424e-06,
+      "loss": 0.6532,
+      "step": 154
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.14762479409611,
+      "learning_rate": 9.976760358471687e-06,
+      "loss": 0.5867,
+      "step": 155
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.060015023394262,
+      "learning_rate": 9.976153295151123e-06,
+      "loss": 0.5223,
+      "step": 156
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.214670267078661,
+      "learning_rate": 9.975538423856261e-06,
+      "loss": 0.7876,
+      "step": 157
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.430010349609652,
+      "learning_rate": 9.974915745551882e-06,
+      "loss": 0.3879,
+      "step": 158
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.079436310446846,
+      "learning_rate": 9.97428526121502e-06,
+      "loss": 0.9568,
+      "step": 159
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.1641882247915,
+      "learning_rate": 9.97364697183495e-06,
+      "loss": 0.8483,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.245660186154265,
+      "learning_rate": 9.973000878413203e-06,
+      "loss": 0.5145,
+      "step": 161
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 10.003192641435247,
+      "learning_rate": 9.972346981963546e-06,
+      "loss": 0.7765,
+      "step": 162
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.705352774502806,
+      "learning_rate": 9.971685283511996e-06,
+      "loss": 0.2866,
+      "step": 163
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.376060609181173,
+      "learning_rate": 9.971015784096808e-06,
+      "loss": 0.3711,
+      "step": 164
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.075769514876911,
+      "learning_rate": 9.970338484768481e-06,
+      "loss": 0.4238,
+      "step": 165
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.949963812119086,
+      "learning_rate": 9.969653386589749e-06,
+      "loss": 0.6609,
+      "step": 166
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.650487883668113,
+      "learning_rate": 9.968960490635584e-06,
+      "loss": 0.7281,
+      "step": 167
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.03687742377937,
+      "learning_rate": 9.968259797993197e-06,
+      "loss": 0.2753,
+      "step": 168
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.007716350627047,
+      "learning_rate": 9.967551309762028e-06,
+      "loss": 0.5426,
+      "step": 169
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.353735627412365,
+      "learning_rate": 9.96683502705375e-06,
+      "loss": 0.7178,
+      "step": 170
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.649303693881235,
+      "learning_rate": 9.966110950992267e-06,
+      "loss": 0.4387,
+      "step": 171
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.989185317191312,
+      "learning_rate": 9.965379082713711e-06,
+      "loss": 0.3955,
+      "step": 172
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.616226128228174,
+      "learning_rate": 9.964639423366442e-06,
+      "loss": 0.6876,
+      "step": 173
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.0596544890992305,
+      "learning_rate": 9.963891974111042e-06,
+      "loss": 0.5467,
+      "step": 174
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.981627770800435,
+      "learning_rate": 9.963136736120318e-06,
+      "loss": 0.6744,
+      "step": 175
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.219503583871042,
+      "learning_rate": 9.962373710579296e-06,
+      "loss": 0.5867,
+      "step": 176
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.75735079747755,
+      "learning_rate": 9.961602898685225e-06,
+      "loss": 0.3495,
+      "step": 177
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.1393224951719985,
+      "learning_rate": 9.960824301647569e-06,
+      "loss": 0.459,
+      "step": 178
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.521809955844176,
+      "learning_rate": 9.960037920688007e-06,
+      "loss": 0.6667,
+      "step": 179
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.849867961965746,
+      "learning_rate": 9.959243757040434e-06,
+      "loss": 1.0293,
+      "step": 180
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.035407322874203,
+      "learning_rate": 9.958441811950952e-06,
+      "loss": 0.4224,
+      "step": 181
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.48672035787583,
+      "learning_rate": 9.957632086677876e-06,
+      "loss": 0.4148,
+      "step": 182
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.95586573225404,
+      "learning_rate": 9.956814582491731e-06,
+      "loss": 0.4398,
+      "step": 183
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.175368456311246,
+      "learning_rate": 9.955989300675243e-06,
+      "loss": 0.6539,
+      "step": 184
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.349612594988935,
+      "learning_rate": 9.955156242523345e-06,
+      "loss": 0.404,
+      "step": 185
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.043658422156373,
+      "learning_rate": 9.95431540934317e-06,
+      "loss": 0.8843,
+      "step": 186
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.611942060388665,
+      "learning_rate": 9.95346680245405e-06,
+      "loss": 0.2997,
+      "step": 187
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.327339649145689,
+      "learning_rate": 9.952610423187516e-06,
+      "loss": 0.4541,
+      "step": 188
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.636493879004018,
+      "learning_rate": 9.951746272887298e-06,
+      "loss": 0.601,
+      "step": 189
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.154432748014298,
+      "learning_rate": 9.95087435290931e-06,
+      "loss": 0.6419,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.807187869480093,
+      "learning_rate": 9.949994664621664e-06,
+      "loss": 0.7136,
+      "step": 191
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.89695756697466,
+      "learning_rate": 9.949107209404664e-06,
+      "loss": 0.6911,
+      "step": 192
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.63703021104669,
+      "learning_rate": 9.948211988650794e-06,
+      "loss": 0.513,
+      "step": 193
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.303030646490855,
+      "learning_rate": 9.947309003764723e-06,
+      "loss": 0.63,
+      "step": 194
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.4080767213233685,
+      "learning_rate": 9.946398256163307e-06,
+      "loss": 0.8587,
+      "step": 195
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.178629015840178,
+      "learning_rate": 9.94547974727558e-06,
+      "loss": 0.3381,
+      "step": 196
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.192474431550429,
+      "learning_rate": 9.944553478542757e-06,
+      "loss": 0.7226,
+      "step": 197
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.254971047700981,
+      "learning_rate": 9.943619451418225e-06,
+      "loss": 0.562,
+      "step": 198
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.066029622620954,
+      "learning_rate": 9.942677667367541e-06,
+      "loss": 0.4646,
+      "step": 199
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.633734882218207,
+      "learning_rate": 9.941728127868446e-06,
+      "loss": 0.4237,
+      "step": 200
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.8942365470381946,
+      "learning_rate": 9.940770834410836e-06,
+      "loss": 0.6326,
+      "step": 201
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.496701523436003,
+      "learning_rate": 9.939805788496778e-06,
+      "loss": 0.6076,
+      "step": 202
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.523333549756454,
+      "learning_rate": 9.938832991640512e-06,
+      "loss": 0.6535,
+      "step": 203
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.855352384023266,
+      "learning_rate": 9.937852445368427e-06,
+      "loss": 0.3365,
+      "step": 204
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.558584636294461,
+      "learning_rate": 9.936864151219077e-06,
+      "loss": 0.5154,
+      "step": 205
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.366101209111038,
+      "learning_rate": 9.935868110743175e-06,
+      "loss": 0.8519,
+      "step": 206
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.322957013995513,
+      "learning_rate": 9.934864325503584e-06,
+      "loss": 0.5198,
+      "step": 207
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.426973359648164,
+      "learning_rate": 9.933852797075325e-06,
+      "loss": 0.5909,
+      "step": 208
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.3776468368540336,
+      "learning_rate": 9.932833527045563e-06,
+      "loss": 0.7068,
+      "step": 209
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.203666798541823,
+      "learning_rate": 9.931806517013612e-06,
+      "loss": 0.2836,
+      "step": 210
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.448120072789972,
+      "learning_rate": 9.930771768590934e-06,
+      "loss": 0.6617,
+      "step": 211
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.140414903559803,
+      "learning_rate": 9.929729283401127e-06,
+      "loss": 1.1004,
+      "step": 212
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.471338878333157,
+      "learning_rate": 9.928679063079934e-06,
+      "loss": 1.0521,
+      "step": 213
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.674920645877626,
+      "learning_rate": 9.927621109275233e-06,
+      "loss": 0.667,
+      "step": 214
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.34475789270995,
+      "learning_rate": 9.926555423647035e-06,
+      "loss": 0.561,
+      "step": 215
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.257567861186612,
+      "learning_rate": 9.925482007867485e-06,
+      "loss": 0.5936,
+      "step": 216
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.560031013337673,
+      "learning_rate": 9.924400863620857e-06,
+      "loss": 0.4272,
+      "step": 217
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.224823857790595,
+      "learning_rate": 9.92331199260355e-06,
+      "loss": 0.3991,
+      "step": 218
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.517561386732611,
+      "learning_rate": 9.922215396524089e-06,
+      "loss": 0.8067,
+      "step": 219
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.32107399129161,
+      "learning_rate": 9.921111077103118e-06,
+      "loss": 0.8056,
+      "step": 220
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.723309673859593,
+      "learning_rate": 9.9199990360734e-06,
+      "loss": 0.4147,
+      "step": 221
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.235724496677339,
+      "learning_rate": 9.918879275179819e-06,
+      "loss": 0.7469,
+      "step": 222
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.6575004399317415,
+      "learning_rate": 9.91775179617936e-06,
+      "loss": 0.3459,
+      "step": 223
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.460870124299543,
+      "learning_rate": 9.916616600841133e-06,
+      "loss": 0.9322,
+      "step": 224
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.10853460812157,
+      "learning_rate": 9.915473690946345e-06,
+      "loss": 0.7995,
+      "step": 225
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.695166873286982,
+      "learning_rate": 9.914323068288312e-06,
+      "loss": 0.6969,
+      "step": 226
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.022883524436682,
+      "learning_rate": 9.91316473467245e-06,
+      "loss": 0.2709,
+      "step": 227
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.906177789166639,
+      "learning_rate": 9.911998691916275e-06,
+      "loss": 0.7719,
+      "step": 228
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.201805471092074,
+      "learning_rate": 9.910824941849401e-06,
+      "loss": 0.4027,
+      "step": 229
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.6378632826228,
+      "learning_rate": 9.909643486313533e-06,
+      "loss": 0.9008,
+      "step": 230
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.619090369092227,
+      "learning_rate": 9.908454327162469e-06,
+      "loss": 0.7992,
+      "step": 231
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.267889278364657,
+      "learning_rate": 9.90725746626209e-06,
+      "loss": 0.5345,
+      "step": 232
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.113464436748009,
+      "learning_rate": 9.906052905490366e-06,
+      "loss": 0.5715,
+      "step": 233
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.371927237132624,
+      "learning_rate": 9.904840646737346e-06,
+      "loss": 0.501,
+      "step": 234
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.5624418707969285,
+      "learning_rate": 9.90362069190516e-06,
+      "loss": 0.5762,
+      "step": 235
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.488590672245439,
+      "learning_rate": 9.902393042908015e-06,
+      "loss": 0.4949,
+      "step": 236
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.720823201186881,
+      "learning_rate": 9.901157701672183e-06,
+      "loss": 0.4122,
+      "step": 237
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.787613369924378,
+      "learning_rate": 9.899914670136016e-06,
+      "loss": 0.7468,
+      "step": 238
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.214927476209819,
+      "learning_rate": 9.898663950249925e-06,
+      "loss": 0.3431,
+      "step": 239
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.727400564621301,
+      "learning_rate": 9.89740554397639e-06,
+      "loss": 0.5952,
+      "step": 240
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.402217775362403,
+      "learning_rate": 9.896139453289946e-06,
+      "loss": 0.6448,
+      "step": 241
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.585103963636099,
+      "learning_rate": 9.89486568017719e-06,
+      "loss": 0.4152,
+      "step": 242
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.647479694264918,
+      "learning_rate": 9.893584226636773e-06,
+      "loss": 0.8351,
+      "step": 243
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.632200926220455,
+      "learning_rate": 9.892295094679394e-06,
+      "loss": 0.9451,
+      "step": 244
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.414039713352258,
+      "learning_rate": 9.890998286327802e-06,
+      "loss": 0.6671,
+      "step": 245
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.900483823935903,
+      "learning_rate": 9.889693803616793e-06,
+      "loss": 0.55,
+      "step": 246
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.593605468779421,
+      "learning_rate": 9.8883816485932e-06,
+      "loss": 0.6476,
+      "step": 247
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.609562634073377,
+      "learning_rate": 9.8870618233159e-06,
+      "loss": 0.6162,
+      "step": 248
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.011815477328181,
+      "learning_rate": 9.885734329855798e-06,
+      "loss": 0.7807,
+      "step": 249
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.26524254560341,
+      "learning_rate": 9.884399170295839e-06,
+      "loss": 0.2472,
+      "step": 250
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.335370910528567,
+      "learning_rate": 9.883056346730992e-06,
+      "loss": 0.6568,
+      "step": 251
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.997695191369208,
+      "learning_rate": 9.881705861268252e-06,
+      "loss": 0.7309,
+      "step": 252
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.126963578159049,
+      "learning_rate": 9.880347716026635e-06,
+      "loss": 0.4084,
+      "step": 253
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.079237445427333,
+      "learning_rate": 9.878981913137178e-06,
+      "loss": 0.788,
+      "step": 254
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.049948294026413,
+      "learning_rate": 9.877608454742935e-06,
+      "loss": 0.6085,
+      "step": 255
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.257499830934716,
+      "learning_rate": 9.87622734299897e-06,
+      "loss": 0.6666,
+      "step": 256
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.651887244077257,
+      "learning_rate": 9.87483858007235e-06,
+      "loss": 0.4793,
+      "step": 257
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.043785471550064,
+      "learning_rate": 9.873442168142158e-06,
+      "loss": 0.6279,
+      "step": 258
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.158571435308082,
+      "learning_rate": 9.872038109399474e-06,
+      "loss": 0.734,
+      "step": 259
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.798279585084529,
+      "learning_rate": 9.870626406047372e-06,
+      "loss": 0.7056,
+      "step": 260
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.80425883777596,
+      "learning_rate": 9.869207060300929e-06,
+      "loss": 0.5749,
+      "step": 261
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.590455161475363,
+      "learning_rate": 9.867780074387207e-06,
+      "loss": 0.2604,
+      "step": 262
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.525151836428044,
+      "learning_rate": 9.86634545054526e-06,
+      "loss": 0.267,
+      "step": 263
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.708055135034483,
+      "learning_rate": 9.864903191026125e-06,
+      "loss": 0.2953,
+      "step": 264
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.3983965002891905,
+      "learning_rate": 9.86345329809282e-06,
+      "loss": 0.5049,
+      "step": 265
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.781667712208913,
+      "learning_rate": 9.861995774020341e-06,
+      "loss": 0.656,
+      "step": 266
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.289483360206045,
+      "learning_rate": 9.860530621095659e-06,
+      "loss": 0.6584,
+      "step": 267
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.0602707296292895,
+      "learning_rate": 9.859057841617709e-06,
+      "loss": 0.4042,
+      "step": 268
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 20.867771447040084,
+      "learning_rate": 9.857577437897401e-06,
+      "loss": 0.7279,
+      "step": 269
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.68379750893328,
+      "learning_rate": 9.856089412257605e-06,
+      "loss": 0.7704,
+      "step": 270
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.956331836477761,
+      "learning_rate": 9.854593767033147e-06,
+      "loss": 0.746,
+      "step": 271
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.383533768646651,
+      "learning_rate": 9.853090504570813e-06,
+      "loss": 0.5509,
+      "step": 272
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.46597403560279,
+      "learning_rate": 9.85157962722934e-06,
+      "loss": 0.6701,
+      "step": 273
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.613120765745608,
+      "learning_rate": 9.850061137379414e-06,
+      "loss": 0.7618,
+      "step": 274
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.300109956666041,
+      "learning_rate": 9.84853503740366e-06,
+      "loss": 0.4096,
+      "step": 275
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.885673288118975,
+      "learning_rate": 9.847001329696653e-06,
+      "loss": 0.4864,
+      "step": 276
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.191105323565091,
+      "learning_rate": 9.845460016664899e-06,
+      "loss": 0.3362,
+      "step": 277
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.775653473869663,
+      "learning_rate": 9.843911100726838e-06,
+      "loss": 0.6733,
+      "step": 278
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.646672616812669,
+      "learning_rate": 9.842354584312841e-06,
+      "loss": 1.1503,
+      "step": 279
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.893398851266994,
+      "learning_rate": 9.840790469865205e-06,
+      "loss": 0.5574,
+      "step": 280
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.8364563178155295,
+      "learning_rate": 9.839218759838146e-06,
+      "loss": 0.541,
+      "step": 281
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.942051863781865,
+      "learning_rate": 9.837639456697802e-06,
+      "loss": 0.3568,
+      "step": 282
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.876918111726609,
+      "learning_rate": 9.836052562922224e-06,
+      "loss": 0.3939,
+      "step": 283
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.939547804165729,
+      "learning_rate": 9.83445808100137e-06,
+      "loss": 0.2633,
+      "step": 284
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 9.043850201881881,
+      "learning_rate": 9.83285601343711e-06,
+      "loss": 0.3511,
+      "step": 285
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.61884688616227,
+      "learning_rate": 9.83124636274321e-06,
+      "loss": 0.6302,
+      "step": 286
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.039344836134989,
+      "learning_rate": 9.829629131445342e-06,
+      "loss": 0.3386,
+      "step": 287
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.956129723796397,
+      "learning_rate": 9.828004322081067e-06,
+      "loss": 0.4755,
+      "step": 288
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.5635099147781375,
+      "learning_rate": 9.826371937199837e-06,
+      "loss": 0.6223,
+      "step": 289
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.435171788762695,
+      "learning_rate": 9.824731979362991e-06,
+      "loss": 0.4679,
+      "step": 290
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.412793395992515,
+      "learning_rate": 9.823084451143755e-06,
+      "loss": 0.5032,
+      "step": 291
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.2313885489579075,
+      "learning_rate": 9.821429355127225e-06,
+      "loss": 0.4112,
+      "step": 292
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.831319321123592,
+      "learning_rate": 9.819766693910382e-06,
+      "loss": 0.5151,
+      "step": 293
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.879006356515538,
+      "learning_rate": 9.818096470102067e-06,
+      "loss": 0.4978,
+      "step": 294
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.035423996978485,
+      "learning_rate": 9.816418686322995e-06,
+      "loss": 0.4944,
+      "step": 295
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.641312697525722,
+      "learning_rate": 9.814733345205736e-06,
+      "loss": 0.8333,
+      "step": 296
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.845084661500621,
+      "learning_rate": 9.813040449394728e-06,
+      "loss": 0.3966,
+      "step": 297
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.171572463981029,
+      "learning_rate": 9.811340001546252e-06,
+      "loss": 0.5276,
+      "step": 298
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.921858936764608,
+      "learning_rate": 9.809632004328448e-06,
+      "loss": 0.5579,
+      "step": 299
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.145090683302827,
+      "learning_rate": 9.807916460421294e-06,
+      "loss": 0.5193,
+      "step": 300
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.836341525658862,
+      "learning_rate": 9.806193372516615e-06,
+      "loss": 0.4026,
+      "step": 301
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.581487511831342,
+      "learning_rate": 9.80446274331807e-06,
+      "loss": 0.4996,
+      "step": 302
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.255141502498269,
+      "learning_rate": 9.802724575541152e-06,
+      "loss": 0.6947,
+      "step": 303
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.386386426049397,
+      "learning_rate": 9.800978871913181e-06,
+      "loss": 0.3353,
+      "step": 304
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.538415251290061,
+      "learning_rate": 9.799225635173305e-06,
+      "loss": 0.5751,
+      "step": 305
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.326703067594579,
+      "learning_rate": 9.797464868072489e-06,
+      "loss": 0.6409,
+      "step": 306
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.840338885309457,
+      "learning_rate": 9.795696573373511e-06,
+      "loss": 0.4939,
+      "step": 307
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.812226879320205,
+      "learning_rate": 9.793920753850972e-06,
+      "loss": 0.5296,
+      "step": 308
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.284231927305214,
+      "learning_rate": 9.792137412291265e-06,
+      "loss": 0.3649,
+      "step": 309
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.326199089079676,
+      "learning_rate": 9.790346551492594e-06,
+      "loss": 0.5225,
+      "step": 310
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.60346439100645,
+      "learning_rate": 9.788548174264961e-06,
+      "loss": 0.3588,
+      "step": 311
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.089527981344691,
+      "learning_rate": 9.78674228343016e-06,
+      "loss": 0.4737,
+      "step": 312
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.651164249429784,
+      "learning_rate": 9.784928881821776e-06,
+      "loss": 0.5429,
+      "step": 313
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.42810315949908,
+      "learning_rate": 9.783107972285177e-06,
+      "loss": 0.5282,
+      "step": 314
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.4585265485361045,
+      "learning_rate": 9.781279557677514e-06,
+      "loss": 0.4829,
+      "step": 315
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.873939546653713,
+      "learning_rate": 9.779443640867712e-06,
+      "loss": 0.3218,
+      "step": 316
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.271643730021547,
+      "learning_rate": 9.777600224736468e-06,
+      "loss": 0.7987,
+      "step": 317
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.862633180212283,
+      "learning_rate": 9.775749312176249e-06,
+      "loss": 0.8568,
+      "step": 318
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.214642683741037,
+      "learning_rate": 9.773890906091275e-06,
+      "loss": 0.4015,
+      "step": 319
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.97107359956487,
+      "learning_rate": 9.772025009397538e-06,
+      "loss": 0.3971,
+      "step": 320
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.433346044875373,
+      "learning_rate": 9.770151625022772e-06,
+      "loss": 0.3841,
+      "step": 321
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.943137012235509,
+      "learning_rate": 9.768270755906467e-06,
+      "loss": 0.5485,
+      "step": 322
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.649512258026864,
+      "learning_rate": 9.76638240499985e-06,
+      "loss": 0.3617,
+      "step": 323
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.23323791908471,
+      "learning_rate": 9.764486575265893e-06,
+      "loss": 0.7007,
+      "step": 324
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.567604255024964,
+      "learning_rate": 9.762583269679304e-06,
+      "loss": 0.6144,
+      "step": 325
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.892714103609171,
+      "learning_rate": 9.760672491226515e-06,
+      "loss": 0.2174,
+      "step": 326
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.774746583839811,
+      "learning_rate": 9.758754242905688e-06,
+      "loss": 0.7154,
+      "step": 327
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.714795038248178,
+      "learning_rate": 9.756828527726706e-06,
+      "loss": 0.5746,
+      "step": 328
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.471646883651928,
+      "learning_rate": 9.754895348711167e-06,
+      "loss": 0.54,
+      "step": 329
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.989112148176543,
+      "learning_rate": 9.752954708892379e-06,
+      "loss": 0.7058,
+      "step": 330
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.928347523538836,
+      "learning_rate": 9.751006611315357e-06,
+      "loss": 0.7443,
+      "step": 331
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.585908113982128,
+      "learning_rate": 9.749051059036821e-06,
+      "loss": 0.456,
+      "step": 332
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.117243531401913,
+      "learning_rate": 9.747088055125186e-06,
+      "loss": 0.5869,
+      "step": 333
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.6263618774567155,
+      "learning_rate": 9.745117602660556e-06,
+      "loss": 0.3162,
+      "step": 334
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.707229023801819,
+      "learning_rate": 9.743139704734729e-06,
+      "loss": 0.6052,
+      "step": 335
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.237336641601359,
+      "learning_rate": 9.741154364451179e-06,
+      "loss": 0.5417,
+      "step": 336
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.41298988622075,
+      "learning_rate": 9.739161584925061e-06,
+      "loss": 0.2447,
+      "step": 337
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.849373058789376,
+      "learning_rate": 9.737161369283201e-06,
+      "loss": 0.4595,
+      "step": 338
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.35682340061978,
+      "learning_rate": 9.735153720664096e-06,
+      "loss": 0.6239,
+      "step": 339
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.969091935864224,
+      "learning_rate": 9.733138642217905e-06,
+      "loss": 0.7563,
+      "step": 340
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.45712541102906,
+      "learning_rate": 9.731116137106441e-06,
+      "loss": 0.7126,
+      "step": 341
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.377006008503654,
+      "learning_rate": 9.729086208503174e-06,
+      "loss": 0.3643,
+      "step": 342
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.066581448477598,
+      "learning_rate": 9.727048859593223e-06,
+      "loss": 0.6599,
+      "step": 343
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 8.254306487939065,
+      "learning_rate": 9.725004093573343e-06,
+      "loss": 0.6487,
+      "step": 344
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.392083079398413,
+      "learning_rate": 9.722951913651937e-06,
+      "loss": 0.3145,
+      "step": 345
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.235957620735297,
+      "learning_rate": 9.720892323049034e-06,
+      "loss": 0.4961,
+      "step": 346
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.345313997047523,
+      "learning_rate": 9.718825324996294e-06,
+      "loss": 0.4739,
+      "step": 347
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.745936546272732,
+      "learning_rate": 9.716750922736998e-06,
+      "loss": 0.6252,
+      "step": 348
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.734311635581271,
+      "learning_rate": 9.714669119526043e-06,
+      "loss": 0.2562,
+      "step": 349
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.704455470536539,
+      "learning_rate": 9.712579918629947e-06,
+      "loss": 0.5932,
+      "step": 350
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.321438502428611,
+      "learning_rate": 9.710483323326826e-06,
+      "loss": 0.6632,
+      "step": 351
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.984339774346785,
+      "learning_rate": 9.708379336906404e-06,
+      "loss": 0.3536,
+      "step": 352
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.11816595055312,
+      "learning_rate": 9.706267962669999e-06,
+      "loss": 0.2847,
+      "step": 353
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.958987770737341,
+      "learning_rate": 9.704149203930522e-06,
+      "loss": 0.4137,
+      "step": 354
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.50824457960139,
+      "learning_rate": 9.702023064012473e-06,
+      "loss": 0.5092,
+      "step": 355
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.5318192556780845,
+      "learning_rate": 9.699889546251934e-06,
+      "loss": 0.506,
+      "step": 356
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.665264867994612,
+      "learning_rate": 9.697748653996558e-06,
+      "loss": 0.4056,
+      "step": 357
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.591022259272521,
+      "learning_rate": 9.695600390605573e-06,
+      "loss": 0.4226,
+      "step": 358
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.44892332271176,
+      "learning_rate": 9.693444759449772e-06,
+      "loss": 0.7158,
+      "step": 359
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.528700345746769,
+      "learning_rate": 9.691281763911513e-06,
+      "loss": 0.1974,
+      "step": 360
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.440396887673768,
+      "learning_rate": 9.689111407384703e-06,
+      "loss": 0.6358,
+      "step": 361
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.263669295314319,
+      "learning_rate": 9.686933693274801e-06,
+      "loss": 0.4228,
+      "step": 362
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.7863239124259795,
+      "learning_rate": 9.68474862499881e-06,
+      "loss": 0.7634,
+      "step": 363
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.544318757098548,
+      "learning_rate": 9.682556205985274e-06,
+      "loss": 0.7168,
+      "step": 364
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.915395816252159,
+      "learning_rate": 9.680356439674272e-06,
+      "loss": 0.6839,
+      "step": 365
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.734733387774487,
+      "learning_rate": 9.67814932951741e-06,
+      "loss": 0.2306,
+      "step": 366
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.8398546251233645,
+      "learning_rate": 9.675934878977814e-06,
+      "loss": 0.469,
+      "step": 367
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.854795901737554,
+      "learning_rate": 9.67371309153013e-06,
+      "loss": 0.9874,
+      "step": 368
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.637104635529683,
+      "learning_rate": 9.671483970660519e-06,
+      "loss": 0.514,
+      "step": 369
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.846763652901588,
+      "learning_rate": 9.669247519866645e-06,
+      "loss": 0.6784,
+      "step": 370
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.824369828623507,
+      "learning_rate": 9.667003742657674e-06,
+      "loss": 0.7653,
+      "step": 371
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.407640051164882,
+      "learning_rate": 9.664752642554272e-06,
+      "loss": 0.4574,
+      "step": 372
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.629447184503669,
+      "learning_rate": 9.662494223088586e-06,
+      "loss": 0.4313,
+      "step": 373
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.59917689662875,
+      "learning_rate": 9.660228487804254e-06,
+      "loss": 0.1838,
+      "step": 374
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.598790128111294,
+      "learning_rate": 9.657955440256396e-06,
+      "loss": 0.4146,
+      "step": 375
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.6925984270154135,
+      "learning_rate": 9.655675084011597e-06,
+      "loss": 0.6849,
+      "step": 376
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.231477605073852,
+      "learning_rate": 9.653387422647918e-06,
+      "loss": 0.3639,
+      "step": 377
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.381814285110513,
+      "learning_rate": 9.651092459754879e-06,
+      "loss": 0.6685,
+      "step": 378
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.552681803361334,
+      "learning_rate": 9.648790198933452e-06,
+      "loss": 0.2848,
+      "step": 379
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.34831473491917,
+      "learning_rate": 9.64648064379607e-06,
+      "loss": 0.5052,
+      "step": 380
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.151382078570767,
+      "learning_rate": 9.644163797966604e-06,
+      "loss": 0.6374,
+      "step": 381
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.673448235622143,
+      "learning_rate": 9.641839665080363e-06,
+      "loss": 0.3938,
+      "step": 382
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.451456369553214,
+      "learning_rate": 9.6395082487841e-06,
+      "loss": 0.2418,
+      "step": 383
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.835227594492041,
+      "learning_rate": 9.637169552735985e-06,
+      "loss": 0.646,
+      "step": 384
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.611919642055182,
+      "learning_rate": 9.634823580605616e-06,
+      "loss": 0.2649,
+      "step": 385
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.999553603810167,
+      "learning_rate": 9.632470336074009e-06,
+      "loss": 0.6146,
+      "step": 386
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.88211570975364,
+      "learning_rate": 9.63010982283359e-06,
+      "loss": 0.4935,
+      "step": 387
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.969216796160313,
+      "learning_rate": 9.627742044588183e-06,
+      "loss": 0.4057,
+      "step": 388
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.275824770459739,
+      "learning_rate": 9.625367005053023e-06,
+      "loss": 0.2308,
+      "step": 389
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.116035677820355,
+      "learning_rate": 9.622984707954732e-06,
+      "loss": 0.2201,
+      "step": 390
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.6867022195529024,
+      "learning_rate": 9.62059515703132e-06,
+      "loss": 0.3027,
+      "step": 391
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.22863003774383,
+      "learning_rate": 9.61819835603218e-06,
+      "loss": 0.4913,
+      "step": 392
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.565778187479567,
+      "learning_rate": 9.61579430871808e-06,
+      "loss": 0.4912,
+      "step": 393
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.457178472171472,
+      "learning_rate": 9.613383018861159e-06,
+      "loss": 0.4607,
+      "step": 394
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.041290869912107,
+      "learning_rate": 9.610964490244921e-06,
+      "loss": 0.349,
+      "step": 395
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.947518287356081,
+      "learning_rate": 9.608538726664224e-06,
+      "loss": 0.5991,
+      "step": 396
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.511793169400645,
+      "learning_rate": 9.606105731925284e-06,
+      "loss": 0.6116,
+      "step": 397
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.675740378847589,
+      "learning_rate": 9.603665509845657e-06,
+      "loss": 0.2027,
+      "step": 398
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.026549637984747,
+      "learning_rate": 9.601218064254245e-06,
+      "loss": 0.438,
+      "step": 399
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.778939427784003,
+      "learning_rate": 9.598763398991283e-06,
+      "loss": 0.768,
+      "step": 400
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.793963249204283,
+      "learning_rate": 9.596301517908329e-06,
+      "loss": 0.7986,
+      "step": 401
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.71313794103704,
+      "learning_rate": 9.593832424868271e-06,
+      "loss": 0.7055,
+      "step": 402
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.614542708882127,
+      "learning_rate": 9.591356123745307e-06,
+      "loss": 0.6463,
+      "step": 403
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.041038055336509,
+      "learning_rate": 9.588872618424949e-06,
+      "loss": 0.5673,
+      "step": 404
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.405384635242469,
+      "learning_rate": 9.58638191280401e-06,
+      "loss": 0.2453,
+      "step": 405
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.20641170842102,
+      "learning_rate": 9.583884010790605e-06,
+      "loss": 0.6098,
+      "step": 406
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.614248272427313,
+      "learning_rate": 9.581378916304134e-06,
+      "loss": 0.5737,
+      "step": 407
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.763469821060781,
+      "learning_rate": 9.578866633275289e-06,
+      "loss": 0.3714,
+      "step": 408
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.7575888291758375,
+      "learning_rate": 9.576347165646038e-06,
+      "loss": 0.7211,
+      "step": 409
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.33997986745691,
+      "learning_rate": 9.573820517369623e-06,
+      "loss": 0.3097,
+      "step": 410
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.274820287363962,
+      "learning_rate": 9.571286692410553e-06,
+      "loss": 0.276,
+      "step": 411
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.817705244436653,
+      "learning_rate": 9.568745694744599e-06,
+      "loss": 0.6004,
+      "step": 412
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.148630748957115,
+      "learning_rate": 9.566197528358785e-06,
+      "loss": 0.6065,
+      "step": 413
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.519936676810737,
+      "learning_rate": 9.563642197251382e-06,
+      "loss": 0.5844,
+      "step": 414
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.687868076343516,
+      "learning_rate": 9.561079705431907e-06,
+      "loss": 0.3832,
+      "step": 415
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.231937884394182,
+      "learning_rate": 9.55851005692111e-06,
+      "loss": 0.5117,
+      "step": 416
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.046340559624703,
+      "learning_rate": 9.555933255750968e-06,
+      "loss": 0.3931,
+      "step": 417
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.316621022594312,
+      "learning_rate": 9.553349305964687e-06,
+      "loss": 0.6415,
+      "step": 418
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.5185681614742395,
+      "learning_rate": 9.550758211616684e-06,
+      "loss": 0.6544,
+      "step": 419
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.106881598647304,
+      "learning_rate": 9.548159976772593e-06,
+      "loss": 0.2595,
+      "step": 420
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.923890852527746,
+      "learning_rate": 9.545554605509244e-06,
+      "loss": 0.5632,
+      "step": 421
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.8413519974174575,
+      "learning_rate": 9.54294210191467e-06,
+      "loss": 0.2704,
+      "step": 422
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.435662954282161,
+      "learning_rate": 9.540322470088094e-06,
+      "loss": 0.3175,
+      "step": 423
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.8171840997736455,
+      "learning_rate": 9.537695714139925e-06,
+      "loss": 0.5396,
+      "step": 424
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.937493762708229,
+      "learning_rate": 9.535061838191746e-06,
+      "loss": 0.387,
+      "step": 425
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.553249957249877,
+      "learning_rate": 9.532420846376316e-06,
+      "loss": 0.1648,
+      "step": 426
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.989119236935052,
+      "learning_rate": 9.529772742837557e-06,
+      "loss": 0.7881,
+      "step": 427
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.912295985527433,
+      "learning_rate": 9.527117531730552e-06,
+      "loss": 0.4188,
+      "step": 428
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.481371641086089,
+      "learning_rate": 9.524455217221537e-06,
+      "loss": 0.6939,
+      "step": 429
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.336352083185281,
+      "learning_rate": 9.521785803487888e-06,
+      "loss": 0.4077,
+      "step": 430
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.42636333137998,
+      "learning_rate": 9.519109294718127e-06,
+      "loss": 0.5134,
+      "step": 431
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.5944462019201575,
+      "learning_rate": 9.516425695111906e-06,
+      "loss": 0.3965,
+      "step": 432
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.282364249353646,
+      "learning_rate": 9.513735008880001e-06,
+      "loss": 0.4672,
+      "step": 433
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.4145928824884,
+      "learning_rate": 9.51103724024431e-06,
+      "loss": 0.7779,
+      "step": 434
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.533429081064421,
+      "learning_rate": 9.508332393437845e-06,
+      "loss": 0.4267,
+      "step": 435
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.71155794896511,
+      "learning_rate": 9.505620472704721e-06,
+      "loss": 0.5358,
+      "step": 436
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.648989084603511,
+      "learning_rate": 9.502901482300155e-06,
+      "loss": 0.7728,
+      "step": 437
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.161754705078291,
+      "learning_rate": 9.500175426490455e-06,
+      "loss": 0.5708,
+      "step": 438
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.095891733875877,
+      "learning_rate": 9.497442309553017e-06,
+      "loss": 0.4843,
+      "step": 439
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.181342270774325,
+      "learning_rate": 9.494702135776315e-06,
+      "loss": 0.7018,
+      "step": 440
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.086856274717312,
+      "learning_rate": 9.491954909459895e-06,
+      "loss": 0.155,
+      "step": 441
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.586954161640706,
+      "learning_rate": 9.489200634914373e-06,
+      "loss": 0.591,
+      "step": 442
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.640634780202634,
+      "learning_rate": 9.48643931646142e-06,
+      "loss": 0.6774,
+      "step": 443
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.675825617162014,
+      "learning_rate": 9.48367095843376e-06,
+      "loss": 0.2631,
+      "step": 444
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.066578740033582,
+      "learning_rate": 9.480895565175166e-06,
+      "loss": 0.3844,
+      "step": 445
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.009829294775996,
+      "learning_rate": 9.478113141040444e-06,
+      "loss": 0.3959,
+      "step": 446
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.229930106917378,
+      "learning_rate": 9.475323690395439e-06,
+      "loss": 0.3022,
+      "step": 447
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.788266587913981,
+      "learning_rate": 9.472527217617016e-06,
+      "loss": 0.5625,
+      "step": 448
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.720053712994818,
+      "learning_rate": 9.469723727093061e-06,
+      "loss": 0.4241,
+      "step": 449
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.114279692875901,
+      "learning_rate": 9.466913223222467e-06,
+      "loss": 0.7787,
+      "step": 450
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.488837694516971,
+      "learning_rate": 9.464095710415138e-06,
+      "loss": 0.5442,
+      "step": 451
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.884916870489228,
+      "learning_rate": 9.461271193091971e-06,
+      "loss": 0.4661,
+      "step": 452
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.123702868706546,
+      "learning_rate": 9.458439675684854e-06,
+      "loss": 0.5812,
+      "step": 453
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.776381518764753,
+      "learning_rate": 9.455601162636662e-06,
+      "loss": 0.4924,
+      "step": 454
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.250665275980305,
+      "learning_rate": 9.452755658401243e-06,
+      "loss": 0.5599,
+      "step": 455
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.813101213507731,
+      "learning_rate": 9.449903167443415e-06,
+      "loss": 0.8663,
+      "step": 456
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.017585706041487,
+      "learning_rate": 9.447043694238961e-06,
+      "loss": 0.4124,
+      "step": 457
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.5162848791395644,
+      "learning_rate": 9.444177243274619e-06,
+      "loss": 0.4879,
+      "step": 458
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.799692304435781,
+      "learning_rate": 9.441303819048073e-06,
+      "loss": 0.1965,
+      "step": 459
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.663076169983753,
+      "learning_rate": 9.438423426067953e-06,
+      "loss": 0.4772,
+      "step": 460
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.7057492095885465,
+      "learning_rate": 9.435536068853819e-06,
+      "loss": 0.6837,
+      "step": 461
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.576537087092552,
+      "learning_rate": 9.432641751936162e-06,
+      "loss": 0.5122,
+      "step": 462
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.802315042476752,
+      "learning_rate": 9.42974047985639e-06,
+      "loss": 0.6463,
+      "step": 463
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.965963334451556,
+      "learning_rate": 9.426832257166831e-06,
+      "loss": 0.2304,
+      "step": 464
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.49349567152424,
+      "learning_rate": 9.42391708843071e-06,
+      "loss": 0.6584,
+      "step": 465
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.956976424505751,
+      "learning_rate": 9.420994978222156e-06,
+      "loss": 0.2796,
+      "step": 466
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.895187589079934,
+      "learning_rate": 9.418065931126188e-06,
+      "loss": 0.7422,
+      "step": 467
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.162009865187619,
+      "learning_rate": 9.415129951738713e-06,
+      "loss": 0.7592,
+      "step": 468
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.074943467654599,
+      "learning_rate": 9.41218704466651e-06,
+      "loss": 0.6322,
+      "step": 469
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.489214335013715,
+      "learning_rate": 9.40923721452723e-06,
+      "loss": 0.6547,
+      "step": 470
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.262757783995714,
+      "learning_rate": 9.406280465949394e-06,
+      "loss": 0.3637,
+      "step": 471
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.0907325669614325,
+      "learning_rate": 9.403316803572363e-06,
+      "loss": 0.5202,
+      "step": 472
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.885826552650618,
+      "learning_rate": 9.400346232046361e-06,
+      "loss": 0.7693,
+      "step": 473
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.757160363287021,
+      "learning_rate": 9.397368756032445e-06,
+      "loss": 0.3856,
+      "step": 474
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.178620343935924,
+      "learning_rate": 9.394384380202511e-06,
+      "loss": 0.4919,
+      "step": 475
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.044643698327802,
+      "learning_rate": 9.391393109239277e-06,
+      "loss": 0.3934,
+      "step": 476
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.03917926402934,
+      "learning_rate": 9.388394947836278e-06,
+      "loss": 0.4205,
+      "step": 477
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.6081217690733896,
+      "learning_rate": 9.38538990069787e-06,
+      "loss": 0.3395,
+      "step": 478
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.431102669647372,
+      "learning_rate": 9.382377972539202e-06,
+      "loss": 0.3721,
+      "step": 479
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.359552945387467,
+      "learning_rate": 9.379359168086231e-06,
+      "loss": 0.7253,
+      "step": 480
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.68792187512187,
+      "learning_rate": 9.376333492075692e-06,
+      "loss": 0.7079,
+      "step": 481
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.167717896122742,
+      "learning_rate": 9.373300949255112e-06,
+      "loss": 0.7042,
+      "step": 482
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.831407302067921,
+      "learning_rate": 9.370261544382784e-06,
+      "loss": 0.5675,
+      "step": 483
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.095153001002084,
+      "learning_rate": 9.367215282227775e-06,
+      "loss": 0.3116,
+      "step": 484
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.897250530963122,
+      "learning_rate": 9.364162167569907e-06,
+      "loss": 0.4143,
+      "step": 485
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.53624010008374,
+      "learning_rate": 9.361102205199762e-06,
+      "loss": 0.4532,
+      "step": 486
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.5500232887782,
+      "learning_rate": 9.358035399918652e-06,
+      "loss": 0.3851,
+      "step": 487
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.024281018095578,
+      "learning_rate": 9.35496175653864e-06,
+      "loss": 0.7731,
+      "step": 488
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.298614201489692,
+      "learning_rate": 9.351881279882512e-06,
+      "loss": 0.4516,
+      "step": 489
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.102037939841859,
+      "learning_rate": 9.348793974783778e-06,
+      "loss": 0.5464,
+      "step": 490
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.810046538133617,
+      "learning_rate": 9.34569984608666e-06,
+      "loss": 0.5021,
+      "step": 491
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.09677561170902,
+      "learning_rate": 9.34259889864609e-06,
+      "loss": 0.3154,
+      "step": 492
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.462968183574109,
+      "learning_rate": 9.339491137327696e-06,
+      "loss": 0.3365,
+      "step": 493
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.076862513783476,
+      "learning_rate": 9.336376567007799e-06,
+      "loss": 0.2389,
+      "step": 494
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.472436045042487,
+      "learning_rate": 9.333255192573404e-06,
+      "loss": 0.4488,
+      "step": 495
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.458578628390376,
+      "learning_rate": 9.330127018922195e-06,
+      "loss": 0.5609,
+      "step": 496
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.035193804293063,
+      "learning_rate": 9.326992050962515e-06,
+      "loss": 0.4033,
+      "step": 497
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.278576512322674,
+      "learning_rate": 9.32385029361338e-06,
+      "loss": 0.4599,
+      "step": 498
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.9099550443679,
+      "learning_rate": 9.320701751804451e-06,
+      "loss": 0.2679,
+      "step": 499
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.160666606684506,
+      "learning_rate": 9.317546430476036e-06,
+      "loss": 0.345,
+      "step": 500
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.390224524129515,
+      "learning_rate": 9.314384334579085e-06,
+      "loss": 0.7871,
+      "step": 501
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.169113614638828,
+      "learning_rate": 9.311215469075168e-06,
+      "loss": 0.5507,
+      "step": 502
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.505867716803198,
+      "learning_rate": 9.30803983893649e-06,
+      "loss": 0.4485,
+      "step": 503
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.997989343457168,
+      "learning_rate": 9.304857449145858e-06,
+      "loss": 0.224,
+      "step": 504
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.9889268179994275,
+      "learning_rate": 9.301668304696694e-06,
+      "loss": 0.4886,
+      "step": 505
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.650010453484125,
+      "learning_rate": 9.298472410593013e-06,
+      "loss": 0.2831,
+      "step": 506
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.519307312442743,
+      "learning_rate": 9.295269771849426e-06,
+      "loss": 0.4835,
+      "step": 507
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.999356502951759,
+      "learning_rate": 9.292060393491122e-06,
+      "loss": 0.4484,
+      "step": 508
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.313726471258957,
+      "learning_rate": 9.288844280553869e-06,
+      "loss": 0.1948,
+      "step": 509
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.714006463761147,
+      "learning_rate": 9.285621438083997e-06,
+      "loss": 0.5978,
+      "step": 510
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.379507615700409,
+      "learning_rate": 9.282391871138404e-06,
+      "loss": 0.4984,
+      "step": 511
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.658570973017829,
+      "learning_rate": 9.279155584784524e-06,
+      "loss": 0.4057,
+      "step": 512
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.500958756680587,
+      "learning_rate": 9.275912584100353e-06,
+      "loss": 0.6567,
+      "step": 513
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.310159285161232,
+      "learning_rate": 9.27266287417441e-06,
+      "loss": 0.3668,
+      "step": 514
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.831937675268848,
+      "learning_rate": 9.269406460105742e-06,
+      "loss": 0.5193,
+      "step": 515
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.099515238410736,
+      "learning_rate": 9.26614334700392e-06,
+      "loss": 0.8194,
+      "step": 516
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.3368859479351745,
+      "learning_rate": 9.262873539989023e-06,
+      "loss": 0.5705,
+      "step": 517
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.284847397211024,
+      "learning_rate": 9.259597044191635e-06,
+      "loss": 0.2042,
+      "step": 518
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.735916668945992,
+      "learning_rate": 9.256313864752838e-06,
+      "loss": 0.4556,
+      "step": 519
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.0754002428268885,
+      "learning_rate": 9.25302400682419e-06,
+      "loss": 0.6233,
+      "step": 520
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.856899048001975,
+      "learning_rate": 9.249727475567742e-06,
+      "loss": 0.1953,
+      "step": 521
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.643048611312609,
+      "learning_rate": 9.246424276156008e-06,
+      "loss": 0.3284,
+      "step": 522
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.959454034089729,
+      "learning_rate": 9.243114413771966e-06,
+      "loss": 0.4909,
+      "step": 523
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.911583548587121,
+      "learning_rate": 9.23979789360905e-06,
+      "loss": 0.552,
+      "step": 524
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.161282668470586,
+      "learning_rate": 9.23647472087114e-06,
+      "loss": 0.7181,
+      "step": 525
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.465134368035077,
+      "learning_rate": 9.233144900772553e-06,
+      "loss": 0.5358,
+      "step": 526
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.286736045944329,
+      "learning_rate": 9.22980843853804e-06,
+      "loss": 0.4559,
+      "step": 527
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.364453867411805,
+      "learning_rate": 9.226465339402768e-06,
+      "loss": 0.5925,
+      "step": 528
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.097684020659741,
+      "learning_rate": 9.223115608612325e-06,
+      "loss": 0.2044,
+      "step": 529
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.097457814626292,
+      "learning_rate": 9.2197592514227e-06,
+      "loss": 0.5472,
+      "step": 530
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.004579623578877,
+      "learning_rate": 9.21639627310028e-06,
+      "loss": 0.6041,
+      "step": 531
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.450798269783079,
+      "learning_rate": 9.21302667892184e-06,
+      "loss": 0.2549,
+      "step": 532
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.67869819466794,
+      "learning_rate": 9.209650474174539e-06,
+      "loss": 0.4604,
+      "step": 533
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.801778106016594,
+      "learning_rate": 9.206267664155906e-06,
+      "loss": 0.316,
+      "step": 534
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.933585701980946,
+      "learning_rate": 9.202878254173836e-06,
+      "loss": 0.531,
+      "step": 535
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.78864179174119,
+      "learning_rate": 9.199482249546577e-06,
+      "loss": 0.2779,
+      "step": 536
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.615827418976137,
+      "learning_rate": 9.196079655602728e-06,
+      "loss": 0.2057,
+      "step": 537
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.815998281078128,
+      "learning_rate": 9.192670477681224e-06,
+      "loss": 0.5645,
+      "step": 538
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.335007672245006,
+      "learning_rate": 9.189254721131333e-06,
+      "loss": 0.7056,
+      "step": 539
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.796413724050597,
+      "learning_rate": 9.185832391312644e-06,
+      "loss": 0.2466,
+      "step": 540
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.5791002306653334,
+      "learning_rate": 9.18240349359506e-06,
+      "loss": 0.2612,
+      "step": 541
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.328801436376093,
+      "learning_rate": 9.178968033358792e-06,
+      "loss": 0.6004,
+      "step": 542
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.747678480571594,
+      "learning_rate": 9.175526015994345e-06,
+      "loss": 0.2985,
+      "step": 543
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.291349934659031,
+      "learning_rate": 9.172077446902515e-06,
+      "loss": 0.3563,
+      "step": 544
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.699165337729385,
+      "learning_rate": 9.168622331494375e-06,
+      "loss": 0.6731,
+      "step": 545
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.760838927131119,
+      "learning_rate": 9.165160675191272e-06,
+      "loss": 0.3648,
+      "step": 546
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.886185767469909,
+      "learning_rate": 9.161692483424817e-06,
+      "loss": 0.3649,
+      "step": 547
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.240767179044931,
+      "learning_rate": 9.158217761636876e-06,
+      "loss": 0.3731,
+      "step": 548
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.291531961236362,
+      "learning_rate": 9.154736515279557e-06,
+      "loss": 0.4663,
+      "step": 549
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.2866096817589865,
+      "learning_rate": 9.151248749815208e-06,
+      "loss": 0.2821,
+      "step": 550
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.013726750760507,
+      "learning_rate": 9.147754470716407e-06,
+      "loss": 0.3941,
+      "step": 551
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.876090507590155,
+      "learning_rate": 9.144253683465953e-06,
+      "loss": 0.4349,
+      "step": 552
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.415423062861582,
+      "learning_rate": 9.140746393556853e-06,
+      "loss": 0.4424,
+      "step": 553
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.881403743928506,
+      "learning_rate": 9.137232606492323e-06,
+      "loss": 0.2772,
+      "step": 554
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.761249797477214,
+      "learning_rate": 9.133712327785769e-06,
+      "loss": 0.2298,
+      "step": 555
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.054664391701198,
+      "learning_rate": 9.13018556296078e-06,
+      "loss": 0.4511,
+      "step": 556
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.502707759038403,
+      "learning_rate": 9.12665231755113e-06,
+      "loss": 0.4485,
+      "step": 557
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.85306690787738,
+      "learning_rate": 9.123112597100759e-06,
+      "loss": 0.6611,
+      "step": 558
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.943386447646357,
+      "learning_rate": 9.11956640716376e-06,
+      "loss": 0.6677,
+      "step": 559
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.947155732146871,
+      "learning_rate": 9.11601375330439e-06,
+      "loss": 0.4153,
+      "step": 560
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.823155051539645,
+      "learning_rate": 9.112454641097037e-06,
+      "loss": 0.5055,
+      "step": 561
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.88174364914201,
+      "learning_rate": 9.108889076126226e-06,
+      "loss": 0.3548,
+      "step": 562
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.219007935618746,
+      "learning_rate": 9.105317063986613e-06,
+      "loss": 0.4683,
+      "step": 563
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.081347804242448,
+      "learning_rate": 9.101738610282956e-06,
+      "loss": 0.479,
+      "step": 564
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.720049851954457,
+      "learning_rate": 9.098153720630138e-06,
+      "loss": 0.3496,
+      "step": 565
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.302519555948863,
+      "learning_rate": 9.094562400653127e-06,
+      "loss": 0.4195,
+      "step": 566
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.908405903770226,
+      "learning_rate": 9.090964655986985e-06,
+      "loss": 0.398,
+      "step": 567
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.689078199906083,
+      "learning_rate": 9.087360492276858e-06,
+      "loss": 0.5135,
+      "step": 568
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.765767677497044,
+      "learning_rate": 9.083749915177959e-06,
+      "loss": 0.4673,
+      "step": 569
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.98511312955719,
+      "learning_rate": 9.080132930355567e-06,
+      "loss": 0.8216,
+      "step": 570
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.8405831251850655,
+      "learning_rate": 9.076509543485014e-06,
+      "loss": 0.5163,
+      "step": 571
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.3180793693619846,
+      "learning_rate": 9.07287976025168e-06,
+      "loss": 0.3032,
+      "step": 572
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.223416951598585,
+      "learning_rate": 9.069243586350976e-06,
+      "loss": 0.2761,
+      "step": 573
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.461059520813073,
+      "learning_rate": 9.065601027488345e-06,
+      "loss": 0.414,
+      "step": 574
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.988423555806269,
+      "learning_rate": 9.061952089379248e-06,
+      "loss": 0.2759,
+      "step": 575
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.741573008092071,
+      "learning_rate": 9.058296777749154e-06,
+      "loss": 0.2549,
+      "step": 576
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.085477411513353,
+      "learning_rate": 9.054635098333532e-06,
+      "loss": 0.309,
+      "step": 577
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.374244270142311,
+      "learning_rate": 9.050967056877846e-06,
+      "loss": 0.4485,
+      "step": 578
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.1420596895146895,
+      "learning_rate": 9.047292659137542e-06,
+      "loss": 0.399,
+      "step": 579
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.39580420579575,
+      "learning_rate": 9.043611910878031e-06,
+      "loss": 0.2966,
+      "step": 580
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.542793329823031,
+      "learning_rate": 9.0399248178747e-06,
+      "loss": 0.2776,
+      "step": 581
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.7841474890677125,
+      "learning_rate": 9.03623138591289e-06,
+      "loss": 0.1636,
+      "step": 582
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.106091244850837,
+      "learning_rate": 9.032531620787879e-06,
+      "loss": 0.597,
+      "step": 583
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.9611394452724635,
+      "learning_rate": 9.028825528304892e-06,
+      "loss": 0.2254,
+      "step": 584
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.799929143801151,
+      "learning_rate": 9.025113114279076e-06,
+      "loss": 0.5035,
+      "step": 585
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.9930514568363513,
+      "learning_rate": 9.0213943845355e-06,
+      "loss": 0.3169,
+      "step": 586
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.394127381774008,
+      "learning_rate": 9.017669344909143e-06,
+      "loss": 0.3502,
+      "step": 587
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.6361982202751015,
+      "learning_rate": 9.013938001244885e-06,
+      "loss": 0.1503,
+      "step": 588
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.883508947746172,
+      "learning_rate": 9.010200359397495e-06,
+      "loss": 0.733,
+      "step": 589
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.2364534272751335,
+      "learning_rate": 9.006456425231624e-06,
+      "loss": 0.5015,
+      "step": 590
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.147546378113793,
+      "learning_rate": 9.002706204621802e-06,
+      "loss": 0.1173,
+      "step": 591
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.456674295259403,
+      "learning_rate": 8.998949703452418e-06,
+      "loss": 0.6133,
+      "step": 592
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.7048223122125234,
+      "learning_rate": 8.995186927617714e-06,
+      "loss": 0.5821,
+      "step": 593
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.221477340291791,
+      "learning_rate": 8.99141788302178e-06,
+      "loss": 0.5367,
+      "step": 594
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.853876499204625,
+      "learning_rate": 8.987642575578546e-06,
+      "loss": 0.2262,
+      "step": 595
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.851521493110232,
+      "learning_rate": 8.98386101121176e-06,
+      "loss": 0.3634,
+      "step": 596
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.793360927811436,
+      "learning_rate": 8.980073195854999e-06,
+      "loss": 0.3085,
+      "step": 597
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.684086790765379,
+      "learning_rate": 8.976279135451636e-06,
+      "loss": 0.4492,
+      "step": 598
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.406048209770403,
+      "learning_rate": 8.972478835954851e-06,
+      "loss": 0.4793,
+      "step": 599
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.532470906160693,
+      "learning_rate": 8.968672303327614e-06,
+      "loss": 0.3609,
+      "step": 600
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.321851260805017,
+      "learning_rate": 8.96485954354267e-06,
+      "loss": 0.4429,
+      "step": 601
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.585618498670008,
+      "learning_rate": 8.96104056258254e-06,
+      "loss": 0.6076,
+      "step": 602
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.350156432156571,
+      "learning_rate": 8.9572153664395e-06,
+      "loss": 0.3776,
+      "step": 603
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.554467816095734,
+      "learning_rate": 8.953383961115586e-06,
+      "loss": 0.3746,
+      "step": 604
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.778674710491944,
+      "learning_rate": 8.949546352622573e-06,
+      "loss": 0.3309,
+      "step": 605
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.875920570672478,
+      "learning_rate": 8.94570254698197e-06,
+      "loss": 0.5326,
+      "step": 606
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.69369852448707,
+      "learning_rate": 8.941852550225005e-06,
+      "loss": 0.4415,
+      "step": 607
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.720349321793076,
+      "learning_rate": 8.937996368392629e-06,
+      "loss": 0.1353,
+      "step": 608
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.1578710796940594,
+      "learning_rate": 8.93413400753549e-06,
+      "loss": 0.4838,
+      "step": 609
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.3784366525516445,
+      "learning_rate": 8.930265473713939e-06,
+      "loss": 0.4246,
+      "step": 610
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.811956853801444,
+      "learning_rate": 8.926390772998004e-06,
+      "loss": 0.332,
+      "step": 611
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.116776780546333,
+      "learning_rate": 8.922509911467395e-06,
+      "loss": 0.503,
+      "step": 612
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.233705981168302,
+      "learning_rate": 8.91862289521149e-06,
+      "loss": 0.5296,
+      "step": 613
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.83131844562604,
+      "learning_rate": 8.914729730329321e-06,
+      "loss": 0.5964,
+      "step": 614
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5487472129396322,
+      "learning_rate": 8.910830422929566e-06,
+      "loss": 0.1257,
+      "step": 615
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.976462171343212,
+      "learning_rate": 8.906924979130548e-06,
+      "loss": 0.2292,
+      "step": 616
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.125082454484557,
+      "learning_rate": 8.903013405060212e-06,
+      "loss": 0.2502,
+      "step": 617
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.2103922468222414,
+      "learning_rate": 8.899095706856122e-06,
+      "loss": 0.6488,
+      "step": 618
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.563943064276356,
+      "learning_rate": 8.895171890665457e-06,
+      "loss": 0.3041,
+      "step": 619
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.90549608756185,
+      "learning_rate": 8.891241962644992e-06,
+      "loss": 0.3908,
+      "step": 620
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.5647758747705645,
+      "learning_rate": 8.887305928961088e-06,
+      "loss": 0.2606,
+      "step": 621
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.338540588701125,
+      "learning_rate": 8.883363795789694e-06,
+      "loss": 0.4628,
+      "step": 622
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.989041661991694,
+      "learning_rate": 8.879415569316324e-06,
+      "loss": 0.3614,
+      "step": 623
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.72503531576978,
+      "learning_rate": 8.875461255736055e-06,
+      "loss": 0.6349,
+      "step": 624
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.99429370355517,
+      "learning_rate": 8.871500861253515e-06,
+      "loss": 0.6988,
+      "step": 625
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.741269406283492,
+      "learning_rate": 8.867534392082873e-06,
+      "loss": 0.3059,
+      "step": 626
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.082198357748838,
+      "learning_rate": 8.863561854447829e-06,
+      "loss": 0.3409,
+      "step": 627
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.886963998552751,
+      "learning_rate": 8.859583254581604e-06,
+      "loss": 0.2079,
+      "step": 628
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.777052226164107,
+      "learning_rate": 8.85559859872694e-06,
+      "loss": 0.6578,
+      "step": 629
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.477167421114731,
+      "learning_rate": 8.851607893136065e-06,
+      "loss": 0.691,
+      "step": 630
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.719875740447138,
+      "learning_rate": 8.847611144070716e-06,
+      "loss": 0.6848,
+      "step": 631
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.034115251265882,
+      "learning_rate": 8.8436083578021e-06,
+      "loss": 0.4325,
+      "step": 632
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.372608770075029,
+      "learning_rate": 8.839599540610906e-06,
+      "loss": 0.3621,
+      "step": 633
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.232505799557821,
+      "learning_rate": 8.83558469878728e-06,
+      "loss": 0.2291,
+      "step": 634
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.288205315432176,
+      "learning_rate": 8.831563838630825e-06,
+      "loss": 0.5354,
+      "step": 635
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.384007929077674,
+      "learning_rate": 8.827536966450584e-06,
+      "loss": 0.5507,
+      "step": 636
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.510150381590252,
+      "learning_rate": 8.823504088565035e-06,
+      "loss": 0.2524,
+      "step": 637
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.303322110036294,
+      "learning_rate": 8.819465211302081e-06,
+      "loss": 0.674,
+      "step": 638
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.999983164193584,
+      "learning_rate": 8.815420340999034e-06,
+      "loss": 0.2924,
+      "step": 639
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.12093234982364,
+      "learning_rate": 8.811369484002614e-06,
+      "loss": 0.3968,
+      "step": 640
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.619460408812439,
+      "learning_rate": 8.807312646668932e-06,
+      "loss": 0.4993,
+      "step": 641
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.934053692341061,
+      "learning_rate": 8.803249835363486e-06,
+      "loss": 0.2572,
+      "step": 642
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.30826448851593,
+      "learning_rate": 8.799181056461143e-06,
+      "loss": 0.4038,
+      "step": 643
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.47423916678203,
+      "learning_rate": 8.795106316346136e-06,
+      "loss": 0.4849,
+      "step": 644
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.466568199114376,
+      "learning_rate": 8.791025621412052e-06,
+      "loss": 0.332,
+      "step": 645
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.243633106944956,
+      "learning_rate": 8.78693897806182e-06,
+      "loss": 0.3374,
+      "step": 646
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.424150262851797,
+      "learning_rate": 8.782846392707704e-06,
+      "loss": 0.3594,
+      "step": 647
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.45245317318313,
+      "learning_rate": 8.778747871771293e-06,
+      "loss": 0.2824,
+      "step": 648
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.324510289911242,
+      "learning_rate": 8.774643421683484e-06,
+      "loss": 0.3548,
+      "step": 649
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.708580457516981,
+      "learning_rate": 8.770533048884483e-06,
+      "loss": 0.326,
+      "step": 650
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.074061664096671,
+      "learning_rate": 8.766416759823785e-06,
+      "loss": 0.3699,
+      "step": 651
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.7163699528315157,
+      "learning_rate": 8.762294560960173e-06,
+      "loss": 0.1294,
+      "step": 652
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.540201543268772,
+      "learning_rate": 8.758166458761695e-06,
+      "loss": 0.3724,
+      "step": 653
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.112779353599681,
+      "learning_rate": 8.754032459705672e-06,
+      "loss": 0.2772,
+      "step": 654
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.368645623986805,
+      "learning_rate": 8.74989257027867e-06,
+      "loss": 0.2362,
+      "step": 655
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.123496368634227,
+      "learning_rate": 8.7457467969765e-06,
+      "loss": 0.2468,
+      "step": 656
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.020379051578081,
+      "learning_rate": 8.741595146304202e-06,
+      "loss": 0.3748,
+      "step": 657
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2442462447122233,
+      "learning_rate": 8.737437624776047e-06,
+      "loss": 0.1612,
+      "step": 658
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.990264060435372,
+      "learning_rate": 8.733274238915508e-06,
+      "loss": 0.4543,
+      "step": 659
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.249144903142453,
+      "learning_rate": 8.729104995255265e-06,
+      "loss": 0.1474,
+      "step": 660
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.613450551881159,
+      "learning_rate": 8.724929900337186e-06,
+      "loss": 0.3388,
+      "step": 661
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.822050202241509,
+      "learning_rate": 8.720748960712323e-06,
+      "loss": 0.6221,
+      "step": 662
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.740105180638977,
+      "learning_rate": 8.7165621829409e-06,
+      "loss": 0.2057,
+      "step": 663
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.9460178659084795,
+      "learning_rate": 8.712369573592296e-06,
+      "loss": 0.3938,
+      "step": 664
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.31040494345963,
+      "learning_rate": 8.708171139245045e-06,
+      "loss": 0.2213,
+      "step": 665
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.788913480644609,
+      "learning_rate": 8.703966886486819e-06,
+      "loss": 0.3471,
+      "step": 666
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.545026373357742,
+      "learning_rate": 8.69975682191442e-06,
+      "loss": 0.6506,
+      "step": 667
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.344671011113987,
+      "learning_rate": 8.695540952133769e-06,
+      "loss": 0.3016,
+      "step": 668
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.303114281176954,
+      "learning_rate": 8.691319283759896e-06,
+      "loss": 0.5347,
+      "step": 669
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.723666597161375,
+      "learning_rate": 8.68709182341693e-06,
+      "loss": 0.5101,
+      "step": 670
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.085242528833489,
+      "learning_rate": 8.682858577738086e-06,
+      "loss": 0.3768,
+      "step": 671
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.779620333726095,
+      "learning_rate": 8.67861955336566e-06,
+      "loss": 0.5043,
+      "step": 672
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.687776970224553,
+      "learning_rate": 8.67437475695101e-06,
+      "loss": 0.4646,
+      "step": 673
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.345313862016524,
+      "learning_rate": 8.670124195154557e-06,
+      "loss": 0.2227,
+      "step": 674
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.169312047845613,
+      "learning_rate": 8.665867874645767e-06,
+      "loss": 0.3778,
+      "step": 675
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.425793391537271,
+      "learning_rate": 8.661605802103134e-06,
+      "loss": 0.7285,
+      "step": 676
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.659876500909084,
+      "learning_rate": 8.657337984214189e-06,
+      "loss": 0.2796,
+      "step": 677
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.009208184545386,
+      "learning_rate": 8.65306442767547e-06,
+      "loss": 0.1256,
+      "step": 678
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.0990976733828735,
+      "learning_rate": 8.648785139192526e-06,
+      "loss": 0.4689,
+      "step": 679
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.224789766419377,
+      "learning_rate": 8.644500125479891e-06,
+      "loss": 0.5291,
+      "step": 680
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.694503032236168,
+      "learning_rate": 8.640209393261087e-06,
+      "loss": 0.4087,
+      "step": 681
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.260352448355556,
+      "learning_rate": 8.635912949268614e-06,
+      "loss": 0.5022,
+      "step": 682
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.0785051869218645,
+      "learning_rate": 8.631610800243926e-06,
+      "loss": 0.5318,
+      "step": 683
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.480147091940225,
+      "learning_rate": 8.627302952937431e-06,
+      "loss": 0.3168,
+      "step": 684
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.3855043647303855,
+      "learning_rate": 8.622989414108479e-06,
+      "loss": 0.4281,
+      "step": 685
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.772660443031573,
+      "learning_rate": 8.61867019052535e-06,
+      "loss": 0.3886,
+      "step": 686
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.552568074182524,
+      "learning_rate": 8.614345288965249e-06,
+      "loss": 0.6878,
+      "step": 687
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.074780148567725,
+      "learning_rate": 8.61001471621428e-06,
+      "loss": 0.381,
+      "step": 688
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.885979991047838,
+      "learning_rate": 8.60567847906745e-06,
+      "loss": 0.1815,
+      "step": 689
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.654217815521764,
+      "learning_rate": 8.601336584328659e-06,
+      "loss": 0.7389,
+      "step": 690
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.137268516797368,
+      "learning_rate": 8.596989038810678e-06,
+      "loss": 0.572,
+      "step": 691
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.729037352933048,
+      "learning_rate": 8.592635849335148e-06,
+      "loss": 0.4381,
+      "step": 692
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.449555865086533,
+      "learning_rate": 8.58827702273256e-06,
+      "loss": 0.3368,
+      "step": 693
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.562912791658195,
+      "learning_rate": 8.583912565842258e-06,
+      "loss": 0.5002,
+      "step": 694
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.649673286196971,
+      "learning_rate": 8.579542485512416e-06,
+      "loss": 0.4945,
+      "step": 695
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.18833215378799,
+      "learning_rate": 8.575166788600031e-06,
+      "loss": 0.4533,
+      "step": 696
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.547724219921325,
+      "learning_rate": 8.570785481970915e-06,
+      "loss": 0.262,
+      "step": 697
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.29980581561246,
+      "learning_rate": 8.566398572499685e-06,
+      "loss": 0.4902,
+      "step": 698
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.632492170796724,
+      "learning_rate": 8.56200606706974e-06,
+      "loss": 0.4435,
+      "step": 699
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.737027857054075,
+      "learning_rate": 8.557607972573267e-06,
+      "loss": 0.4521,
+      "step": 700
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.080692048952784,
+      "learning_rate": 8.553204295911222e-06,
+      "loss": 0.4275,
+      "step": 701
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.3992850488391335,
+      "learning_rate": 8.548795043993316e-06,
+      "loss": 0.1881,
+      "step": 702
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.39630771421547,
+      "learning_rate": 8.544380223738015e-06,
+      "loss": 0.4196,
+      "step": 703
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.406526853578896,
+      "learning_rate": 8.539959842072513e-06,
+      "loss": 0.4037,
+      "step": 704
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.71538778598701,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 0.522,
+      "step": 705
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.148550809802702,
+      "learning_rate": 8.53110242226333e-06,
+      "loss": 0.2433,
+      "step": 706
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.835256793033029,
+      "learning_rate": 8.526665398017633e-06,
+      "loss": 0.2493,
+      "step": 707
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.771566456752833,
+      "learning_rate": 8.522222840157687e-06,
+      "loss": 0.462,
+      "step": 708
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.4112013405040225,
+      "learning_rate": 8.517774755654212e-06,
+      "loss": 0.5426,
+      "step": 709
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.476116214038771,
+      "learning_rate": 8.513321151486602e-06,
+      "loss": 0.3139,
+      "step": 710
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.105784418844728,
+      "learning_rate": 8.508862034642909e-06,
+      "loss": 0.4809,
+      "step": 711
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.956245030040778,
+      "learning_rate": 8.504397412119838e-06,
+      "loss": 0.4856,
+      "step": 712
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.353146817162132,
+      "learning_rate": 8.499927290922734e-06,
+      "loss": 0.2731,
+      "step": 713
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.387553795360139,
+      "learning_rate": 8.495451678065563e-06,
+      "loss": 0.3675,
+      "step": 714
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.9716862870401,
+      "learning_rate": 8.490970580570912e-06,
+      "loss": 0.3146,
+      "step": 715
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.465889783287375,
+      "learning_rate": 8.486484005469977e-06,
+      "loss": 0.3151,
+      "step": 716
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.457237970712992,
+      "learning_rate": 8.481991959802546e-06,
+      "loss": 0.5893,
+      "step": 717
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.08714400356136,
+      "learning_rate": 8.477494450616988e-06,
+      "loss": 0.317,
+      "step": 718
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.453397786116955,
+      "learning_rate": 8.472991484970247e-06,
+      "loss": 0.3618,
+      "step": 719
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.930592776162307,
+      "learning_rate": 8.468483069927832e-06,
+      "loss": 0.4635,
+      "step": 720
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.92336479181062,
+      "learning_rate": 8.463969212563796e-06,
+      "loss": 0.3627,
+      "step": 721
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.583733544579457,
+      "learning_rate": 8.459449919960737e-06,
+      "loss": 0.4629,
+      "step": 722
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.239532845260946,
+      "learning_rate": 8.454925199209778e-06,
+      "loss": 0.4101,
+      "step": 723
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3101666129775786,
+      "learning_rate": 8.450395057410561e-06,
+      "loss": 0.2588,
+      "step": 724
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.694791777010421,
+      "learning_rate": 8.445859501671232e-06,
+      "loss": 0.3593,
+      "step": 725
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.294140648416576,
+      "learning_rate": 8.441318539108433e-06,
+      "loss": 0.1859,
+      "step": 726
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.5982421619702745,
+      "learning_rate": 8.436772176847295e-06,
+      "loss": 0.5006,
+      "step": 727
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.32551203191427,
+      "learning_rate": 8.432220422021408e-06,
+      "loss": 0.3533,
+      "step": 728
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.7729572569482865,
+      "learning_rate": 8.42766328177284e-06,
+      "loss": 0.2034,
+      "step": 729
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.60106757513709,
+      "learning_rate": 8.423100763252094e-06,
+      "loss": 0.2384,
+      "step": 730
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.341121812752496,
+      "learning_rate": 8.418532873618125e-06,
+      "loss": 0.3294,
+      "step": 731
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.390056947750818,
+      "learning_rate": 8.413959620038306e-06,
+      "loss": 0.4776,
+      "step": 732
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.716304747185911,
+      "learning_rate": 8.409381009688431e-06,
+      "loss": 0.4473,
+      "step": 733
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.077923171897502,
+      "learning_rate": 8.404797049752697e-06,
+      "loss": 0.525,
+      "step": 734
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.90524035318987,
+      "learning_rate": 8.4002077474237e-06,
+      "loss": 0.3571,
+      "step": 735
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.353718092364014,
+      "learning_rate": 8.39561310990241e-06,
+      "loss": 0.6481,
+      "step": 736
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.2212785232108425,
+      "learning_rate": 8.39101314439818e-06,
+      "loss": 0.8407,
+      "step": 737
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.632940916977337,
+      "learning_rate": 8.386407858128707e-06,
+      "loss": 0.3278,
+      "step": 738
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.421823200834819,
+      "learning_rate": 8.38179725832005e-06,
+      "loss": 0.4727,
+      "step": 739
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.251013698044102,
+      "learning_rate": 8.377181352206604e-06,
+      "loss": 0.4168,
+      "step": 740
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.5407855751458355,
+      "learning_rate": 8.372560147031087e-06,
+      "loss": 0.4087,
+      "step": 741
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.221369280283807,
+      "learning_rate": 8.367933650044526e-06,
+      "loss": 0.2892,
+      "step": 742
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.308213092755246,
+      "learning_rate": 8.363301868506264e-06,
+      "loss": 0.2703,
+      "step": 743
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.980934617217376,
+      "learning_rate": 8.358664809683926e-06,
+      "loss": 0.5673,
+      "step": 744
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.1931510638054945,
+      "learning_rate": 8.354022480853418e-06,
+      "loss": 0.5605,
+      "step": 745
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.117437936432802,
+      "learning_rate": 8.349374889298923e-06,
+      "loss": 0.1584,
+      "step": 746
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.136085182252913,
+      "learning_rate": 8.344722042312872e-06,
+      "loss": 0.4435,
+      "step": 747
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.550301029816411,
+      "learning_rate": 8.340063947195947e-06,
+      "loss": 0.2945,
+      "step": 748
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.800424006474591,
+      "learning_rate": 8.335400611257067e-06,
+      "loss": 0.4362,
+      "step": 749
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.160866290111288,
+      "learning_rate": 8.330732041813367e-06,
+      "loss": 0.5082,
+      "step": 750
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.143961468455159,
+      "learning_rate": 8.326058246190202e-06,
+      "loss": 0.3298,
+      "step": 751
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.360607497775467,
+      "learning_rate": 8.321379231721123e-06,
+      "loss": 0.336,
+      "step": 752
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.934394227353047,
+      "learning_rate": 8.316695005747866e-06,
+      "loss": 0.5735,
+      "step": 753
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.161434531278339,
+      "learning_rate": 8.312005575620355e-06,
+      "loss": 0.3328,
+      "step": 754
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.371996663426846,
+      "learning_rate": 8.307310948696667e-06,
+      "loss": 0.426,
+      "step": 755
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.347178597445798,
+      "learning_rate": 8.302611132343042e-06,
+      "loss": 0.3719,
+      "step": 756
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.302882867714401,
+      "learning_rate": 8.297906133933861e-06,
+      "loss": 0.1944,
+      "step": 757
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.77277910734547,
+      "learning_rate": 8.293195960851634e-06,
+      "loss": 0.5761,
+      "step": 758
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.644844069972676,
+      "learning_rate": 8.288480620486991e-06,
+      "loss": 0.5338,
+      "step": 759
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6321830270740136,
+      "learning_rate": 8.283760120238672e-06,
+      "loss": 0.1059,
+      "step": 760
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.461153624018655,
+      "learning_rate": 8.27903446751351e-06,
+      "loss": 0.1658,
+      "step": 761
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 11.673090789730939,
+      "learning_rate": 8.274303669726427e-06,
+      "loss": 0.3966,
+      "step": 762
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.756068482704599,
+      "learning_rate": 8.26956773430041e-06,
+      "loss": 0.2664,
+      "step": 763
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.00523516311952,
+      "learning_rate": 8.264826668666516e-06,
+      "loss": 0.3951,
+      "step": 764
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5683192166863535,
+      "learning_rate": 8.26008048026385e-06,
+      "loss": 0.1858,
+      "step": 765
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.365757277166145,
+      "learning_rate": 8.255329176539552e-06,
+      "loss": 0.3197,
+      "step": 766
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.54615438593096,
+      "learning_rate": 8.250572764948787e-06,
+      "loss": 0.6379,
+      "step": 767
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.341950067856082,
+      "learning_rate": 8.245811252954741e-06,
+      "loss": 0.4991,
+      "step": 768
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.890676412538074,
+      "learning_rate": 8.241044648028597e-06,
+      "loss": 0.6154,
+      "step": 769
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.398870850919496,
+      "learning_rate": 8.236272957649534e-06,
+      "loss": 0.1698,
+      "step": 770
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.32526464985607,
+      "learning_rate": 8.231496189304704e-06,
+      "loss": 0.2087,
+      "step": 771
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.44216431315945,
+      "learning_rate": 8.226714350489235e-06,
+      "loss": 0.4175,
+      "step": 772
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.473168102716619,
+      "learning_rate": 8.221927448706204e-06,
+      "loss": 0.4919,
+      "step": 773
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.995285658195084,
+      "learning_rate": 8.217135491466636e-06,
+      "loss": 0.3075,
+      "step": 774
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.124387228426363,
+      "learning_rate": 8.212338486289486e-06,
+      "loss": 0.4427,
+      "step": 775
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.0730998605401,
+      "learning_rate": 8.207536440701633e-06,
+      "loss": 0.2747,
+      "step": 776
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.207579874395538,
+      "learning_rate": 8.20272936223786e-06,
+      "loss": 0.4562,
+      "step": 777
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.409820639904443,
+      "learning_rate": 8.197917258440851e-06,
+      "loss": 0.4363,
+      "step": 778
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.147185631193565,
+      "learning_rate": 8.193100136861174e-06,
+      "loss": 0.2087,
+      "step": 779
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.644403115592335,
+      "learning_rate": 8.18827800505727e-06,
+      "loss": 0.4722,
+      "step": 780
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.593727345567647,
+      "learning_rate": 8.183450870595443e-06,
+      "loss": 0.4737,
+      "step": 781
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.5883233493303393,
+      "learning_rate": 8.178618741049841e-06,
+      "loss": 0.1089,
+      "step": 782
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.565888525765652,
+      "learning_rate": 8.173781624002456e-06,
+      "loss": 0.2804,
+      "step": 783
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.90704359502963,
+      "learning_rate": 8.168939527043104e-06,
+      "loss": 0.2742,
+      "step": 784
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.9667965149128985,
+      "learning_rate": 8.164092457769415e-06,
+      "loss": 0.2393,
+      "step": 785
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.73666596849711,
+      "learning_rate": 8.15924042378682e-06,
+      "loss": 0.4671,
+      "step": 786
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.498735388385353,
+      "learning_rate": 8.15438343270854e-06,
+      "loss": 0.4297,
+      "step": 787
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.297444749293851,
+      "learning_rate": 8.149521492155573e-06,
+      "loss": 0.4278,
+      "step": 788
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.887902501920645,
+      "learning_rate": 8.144654609756685e-06,
+      "loss": 0.386,
+      "step": 789
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.53146198817337,
+      "learning_rate": 8.1397827931484e-06,
+      "loss": 0.4026,
+      "step": 790
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.816985787516275,
+      "learning_rate": 8.134906049974975e-06,
+      "loss": 0.4611,
+      "step": 791
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.7708813505955527,
+      "learning_rate": 8.130024387888402e-06,
+      "loss": 0.2199,
+      "step": 792
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.9273358705222225,
+      "learning_rate": 8.125137814548394e-06,
+      "loss": 0.6214,
+      "step": 793
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.907725612922181,
+      "learning_rate": 8.120246337622364e-06,
+      "loss": 0.3915,
+      "step": 794
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.279782883297365,
+      "learning_rate": 8.115349964785425e-06,
+      "loss": 0.1899,
+      "step": 795
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.328073155849916,
+      "learning_rate": 8.110448703720372e-06,
+      "loss": 0.4714,
+      "step": 796
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.867309032110139,
+      "learning_rate": 8.105542562117663e-06,
+      "loss": 0.3014,
+      "step": 797
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.471102341195892,
+      "learning_rate": 8.100631547675417e-06,
+      "loss": 0.2651,
+      "step": 798
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.173289710599521,
+      "learning_rate": 8.095715668099406e-06,
+      "loss": 0.2958,
+      "step": 799
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.1909467558791595,
+      "learning_rate": 8.090794931103026e-06,
+      "loss": 0.333,
+      "step": 800
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.606462455050537,
+      "learning_rate": 8.085869344407301e-06,
+      "loss": 0.1894,
+      "step": 801
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.061123196068857,
+      "learning_rate": 8.080938915740863e-06,
+      "loss": 0.4263,
+      "step": 802
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.795244573922029,
+      "learning_rate": 8.076003652839936e-06,
+      "loss": 0.37,
+      "step": 803
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.990154637814864,
+      "learning_rate": 8.071063563448341e-06,
+      "loss": 0.3033,
+      "step": 804
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.179204271459913,
+      "learning_rate": 8.066118655317458e-06,
+      "loss": 0.1661,
+      "step": 805
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.254996210451282,
+      "learning_rate": 8.06116893620624e-06,
+      "loss": 0.7173,
+      "step": 806
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.06120092010497,
+      "learning_rate": 8.056214413881183e-06,
+      "loss": 0.577,
+      "step": 807
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.733355509595694,
+      "learning_rate": 8.051255096116322e-06,
+      "loss": 0.4846,
+      "step": 808
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.984245285971206,
+      "learning_rate": 8.04629099069321e-06,
+      "loss": 0.4047,
+      "step": 809
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.7598327807264855,
+      "learning_rate": 8.041322105400923e-06,
+      "loss": 0.5301,
+      "step": 810
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.195908506506143,
+      "learning_rate": 8.036348448036029e-06,
+      "loss": 0.6084,
+      "step": 811
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.35207554435034,
+      "learning_rate": 8.031370026402585e-06,
+      "loss": 0.3607,
+      "step": 812
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.32056746367122,
+      "learning_rate": 8.026386848312125e-06,
+      "loss": 0.3335,
+      "step": 813
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.267898004484538,
+      "learning_rate": 8.021398921583644e-06,
+      "loss": 0.1832,
+      "step": 814
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.80773704394529,
+      "learning_rate": 8.016406254043595e-06,
+      "loss": 0.269,
+      "step": 815
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.322922906891823,
+      "learning_rate": 8.01140885352586e-06,
+      "loss": 0.2891,
+      "step": 816
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.229550855302739,
+      "learning_rate": 8.006406727871754e-06,
+      "loss": 0.4966,
+      "step": 817
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.314039777508532,
+      "learning_rate": 8.001399884930004e-06,
+      "loss": 0.6837,
+      "step": 818
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.00525793622508,
+      "learning_rate": 7.996388332556735e-06,
+      "loss": 0.227,
+      "step": 819
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.772809921299112,
+      "learning_rate": 7.991372078615469e-06,
+      "loss": 0.292,
+      "step": 820
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.66819830861659,
+      "learning_rate": 7.9863511309771e-06,
+      "loss": 0.1952,
+      "step": 821
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.0325243013892385,
+      "learning_rate": 7.981325497519892e-06,
+      "loss": 0.2871,
+      "step": 822
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.253237938165674,
+      "learning_rate": 7.97629518612945e-06,
+      "loss": 0.6696,
+      "step": 823
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.175000357164085,
+      "learning_rate": 7.971260204698732e-06,
+      "loss": 0.4533,
+      "step": 824
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.069034528454901,
+      "learning_rate": 7.966220561128018e-06,
+      "loss": 0.3306,
+      "step": 825
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.373203481491902,
+      "learning_rate": 7.961176263324902e-06,
+      "loss": 0.5284,
+      "step": 826
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.871795914928086,
+      "learning_rate": 7.95612731920428e-06,
+      "loss": 0.3533,
+      "step": 827
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.248578895471283,
+      "learning_rate": 7.951073736688348e-06,
+      "loss": 0.6994,
+      "step": 828
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.9304222967760385,
+      "learning_rate": 7.946015523706566e-06,
+      "loss": 0.2685,
+      "step": 829
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.010468902967375,
+      "learning_rate": 7.940952688195668e-06,
+      "loss": 0.6237,
+      "step": 830
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.516919378035556,
+      "learning_rate": 7.93588523809964e-06,
+      "loss": 0.5792,
+      "step": 831
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.710684931910942,
+      "learning_rate": 7.930813181369713e-06,
+      "loss": 0.7422,
+      "step": 832
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.979364689556802,
+      "learning_rate": 7.925736525964332e-06,
+      "loss": 0.2954,
+      "step": 833
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.24363576008588,
+      "learning_rate": 7.920655279849173e-06,
+      "loss": 0.3,
+      "step": 834
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.45551648778925,
+      "learning_rate": 7.915569450997107e-06,
+      "loss": 0.4452,
+      "step": 835
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.837107505748551,
+      "learning_rate": 7.9104790473882e-06,
+      "loss": 0.372,
+      "step": 836
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.885385168899783,
+      "learning_rate": 7.905384077009693e-06,
+      "loss": 0.261,
+      "step": 837
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.728421661944464,
+      "learning_rate": 7.900284547855992e-06,
+      "loss": 0.2118,
+      "step": 838
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.25899147490279,
+      "learning_rate": 7.895180467928658e-06,
+      "loss": 0.6256,
+      "step": 839
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.764744182282839,
+      "learning_rate": 7.890071845236395e-06,
+      "loss": 0.7904,
+      "step": 840
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3448007090140215,
+      "learning_rate": 7.88495868779503e-06,
+      "loss": 0.318,
+      "step": 841
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.7238728497394575,
+      "learning_rate": 7.87984100362751e-06,
+      "loss": 0.5187,
+      "step": 842
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.648105123380729,
+      "learning_rate": 7.874718800763876e-06,
+      "loss": 0.3736,
+      "step": 843
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.992995102650532,
+      "learning_rate": 7.869592087241271e-06,
+      "loss": 0.7633,
+      "step": 844
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.697566530596448,
+      "learning_rate": 7.86446087110391e-06,
+      "loss": 0.3344,
+      "step": 845
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.027726262201498,
+      "learning_rate": 7.859325160403073e-06,
+      "loss": 0.3193,
+      "step": 846
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.297653433057747,
+      "learning_rate": 7.854184963197088e-06,
+      "loss": 0.567,
+      "step": 847
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.9724829344373047,
+      "learning_rate": 7.849040287551331e-06,
+      "loss": 0.2911,
+      "step": 848
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.316916475097543,
+      "learning_rate": 7.843891141538201e-06,
+      "loss": 0.3931,
+      "step": 849
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.1997865766796885,
+      "learning_rate": 7.838737533237111e-06,
+      "loss": 0.3999,
+      "step": 850
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.399989189666105,
+      "learning_rate": 7.833579470734476e-06,
+      "loss": 0.3233,
+      "step": 851
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.0342483467760175,
+      "learning_rate": 7.8284169621237e-06,
+      "loss": 0.219,
+      "step": 852
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.715006791459248,
+      "learning_rate": 7.823250015505162e-06,
+      "loss": 0.4017,
+      "step": 853
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.6502726800310175,
+      "learning_rate": 7.818078638986208e-06,
+      "loss": 0.3295,
+      "step": 854
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.845699780687884,
+      "learning_rate": 7.812902840681132e-06,
+      "loss": 0.2155,
+      "step": 855
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.466549450449988,
+      "learning_rate": 7.807722628711167e-06,
+      "loss": 0.3823,
+      "step": 856
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6928919746660114,
+      "learning_rate": 7.80253801120447e-06,
+      "loss": 0.1823,
+      "step": 857
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.882294075493104,
+      "learning_rate": 7.797348996296116e-06,
+      "loss": 0.3771,
+      "step": 858
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.407601911957696,
+      "learning_rate": 7.792155592128072e-06,
+      "loss": 0.456,
+      "step": 859
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.3998668046998,
+      "learning_rate": 7.786957806849197e-06,
+      "loss": 0.5148,
+      "step": 860
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6673883324077163,
+      "learning_rate": 7.781755648615223e-06,
+      "loss": 0.148,
+      "step": 861
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.145714133004689,
+      "learning_rate": 7.776549125588743e-06,
+      "loss": 0.346,
+      "step": 862
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.278468737428145,
+      "learning_rate": 7.771338245939205e-06,
+      "loss": 0.424,
+      "step": 863
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.292056531078291,
+      "learning_rate": 7.766123017842877e-06,
+      "loss": 0.4587,
+      "step": 864
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.458109108457931,
+      "learning_rate": 7.76090344948287e-06,
+      "loss": 0.6041,
+      "step": 865
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.349536451305507,
+      "learning_rate": 7.755679549049093e-06,
+      "loss": 0.528,
+      "step": 866
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3563335803250114,
+      "learning_rate": 7.75045132473825e-06,
+      "loss": 0.0775,
+      "step": 867
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.506228819584935,
+      "learning_rate": 7.745218784753841e-06,
+      "loss": 0.1533,
+      "step": 868
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.195522191451391,
+      "learning_rate": 7.739981937306127e-06,
+      "loss": 0.5374,
+      "step": 869
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.016228061521943,
+      "learning_rate": 7.734740790612137e-06,
+      "loss": 0.4524,
+      "step": 870
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.609372709352629,
+      "learning_rate": 7.729495352895633e-06,
+      "loss": 0.4124,
+      "step": 871
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.531520674537432,
+      "learning_rate": 7.724245632387122e-06,
+      "loss": 0.3826,
+      "step": 872
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.794990068253417,
+      "learning_rate": 7.718991637323828e-06,
+      "loss": 0.5657,
+      "step": 873
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.98381321495084,
+      "learning_rate": 7.713733375949677e-06,
+      "loss": 0.427,
+      "step": 874
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.08443591266254,
+      "learning_rate": 7.708470856515295e-06,
+      "loss": 0.1837,
+      "step": 875
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.750278414313818,
+      "learning_rate": 7.703204087277989e-06,
+      "loss": 0.2697,
+      "step": 876
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.06479166683791,
+      "learning_rate": 7.69793307650173e-06,
+      "loss": 0.375,
+      "step": 877
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.511833008406359,
+      "learning_rate": 7.692657832457146e-06,
+      "loss": 0.5867,
+      "step": 878
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.270173470987993,
+      "learning_rate": 7.687378363421512e-06,
+      "loss": 0.3481,
+      "step": 879
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.964743391352324,
+      "learning_rate": 7.682094677678726e-06,
+      "loss": 0.2531,
+      "step": 880
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.330434416723052,
+      "learning_rate": 7.676806783519304e-06,
+      "loss": 0.3912,
+      "step": 881
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.75484369586985,
+      "learning_rate": 7.671514689240366e-06,
+      "loss": 0.3605,
+      "step": 882
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.5040681375813305,
+      "learning_rate": 7.666218403145625e-06,
+      "loss": 0.3065,
+      "step": 883
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.425655568614444,
+      "learning_rate": 7.660917933545367e-06,
+      "loss": 0.3633,
+      "step": 884
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.2692885007461765,
+      "learning_rate": 7.655613288756443e-06,
+      "loss": 0.2331,
+      "step": 885
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.8933744104639296,
+      "learning_rate": 7.650304477102258e-06,
+      "loss": 0.272,
+      "step": 886
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.999363750590228,
+      "learning_rate": 7.644991506912753e-06,
+      "loss": 0.163,
+      "step": 887
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.689783040358317,
+      "learning_rate": 7.639674386524395e-06,
+      "loss": 0.5043,
+      "step": 888
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.595954396481512,
+      "learning_rate": 7.634353124280161e-06,
+      "loss": 0.3451,
+      "step": 889
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.094488391215071,
+      "learning_rate": 7.629027728529527e-06,
+      "loss": 0.1512,
+      "step": 890
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.484424693580878,
+      "learning_rate": 7.623698207628458e-06,
+      "loss": 0.2746,
+      "step": 891
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.3526795832851395,
+      "learning_rate": 7.61836456993939e-06,
+      "loss": 0.2982,
+      "step": 892
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.764334167761146,
+      "learning_rate": 7.613026823831217e-06,
+      "loss": 0.3148,
+      "step": 893
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.504617406586926,
+      "learning_rate": 7.607684977679284e-06,
+      "loss": 0.1511,
+      "step": 894
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.453008004875096,
+      "learning_rate": 7.602339039865362e-06,
+      "loss": 0.3653,
+      "step": 895
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.738371851296883,
+      "learning_rate": 7.5969890187776474e-06,
+      "loss": 0.346,
+      "step": 896
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.630936210205719,
+      "learning_rate": 7.591634922810744e-06,
+      "loss": 0.2132,
+      "step": 897
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.379714333632608,
+      "learning_rate": 7.586276760365645e-06,
+      "loss": 0.3159,
+      "step": 898
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.547032749723221,
+      "learning_rate": 7.580914539849731e-06,
+      "loss": 0.506,
+      "step": 899
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.734574813134246,
+      "learning_rate": 7.575548269676741e-06,
+      "loss": 0.2792,
+      "step": 900
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.110732626869322,
+      "learning_rate": 7.570177958266775e-06,
+      "loss": 0.4091,
+      "step": 901
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.810898674658677,
+      "learning_rate": 7.564803614046276e-06,
+      "loss": 0.5199,
+      "step": 902
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.390388498220371,
+      "learning_rate": 7.559425245448006e-06,
+      "loss": 0.2347,
+      "step": 903
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.9401391012049816,
+      "learning_rate": 7.554042860911049e-06,
+      "loss": 0.3488,
+      "step": 904
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.375961686591688,
+      "learning_rate": 7.548656468880788e-06,
+      "loss": 0.2671,
+      "step": 905
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.652853364172935,
+      "learning_rate": 7.543266077808893e-06,
+      "loss": 0.3276,
+      "step": 906
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.292072425620882,
+      "learning_rate": 7.53787169615331e-06,
+      "loss": 0.1898,
+      "step": 907
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.401327363677469,
+      "learning_rate": 7.5324733323782465e-06,
+      "loss": 0.4271,
+      "step": 908
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.544508218544989,
+      "learning_rate": 7.52707099495416e-06,
+      "loss": 0.518,
+      "step": 909
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.844606060047626,
+      "learning_rate": 7.521664692357737e-06,
+      "loss": 0.6057,
+      "step": 910
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.842438410217087,
+      "learning_rate": 7.516254433071894e-06,
+      "loss": 0.3442,
+      "step": 911
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.680098754657996,
+      "learning_rate": 7.510840225585749e-06,
+      "loss": 0.1931,
+      "step": 912
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.902844697369808,
+      "learning_rate": 7.50542207839462e-06,
+      "loss": 0.2623,
+      "step": 913
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.8169096747986573,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.5817,
+      "step": 914
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.191379249441057,
+      "learning_rate": 7.49457399890956e-06,
+      "loss": 0.2416,
+      "step": 915
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.833328428113343,
+      "learning_rate": 7.489144083637117e-06,
+      "loss": 0.635,
+      "step": 916
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.489457809497737,
+      "learning_rate": 7.483710262702635e-06,
+      "loss": 0.3246,
+      "step": 917
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.64131197732772,
+      "learning_rate": 7.478272544632204e-06,
+      "loss": 0.2175,
+      "step": 918
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.812018202683798,
+      "learning_rate": 7.472830937958029e-06,
+      "loss": 0.2529,
+      "step": 919
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.718869633788661,
+      "learning_rate": 7.467385451218418e-06,
+      "loss": 0.2765,
+      "step": 920
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.063079789697448,
+      "learning_rate": 7.461936092957767e-06,
+      "loss": 0.4305,
+      "step": 921
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.79127333539024,
+      "learning_rate": 7.456482871726545e-06,
+      "loss": 0.5141,
+      "step": 922
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.728505755068348,
+      "learning_rate": 7.4510257960812824e-06,
+      "loss": 0.5281,
+      "step": 923
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.1025867041171304,
+      "learning_rate": 7.44556487458456e-06,
+      "loss": 0.0836,
+      "step": 924
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.678403047928699,
+      "learning_rate": 7.440100115804991e-06,
+      "loss": 0.3458,
+      "step": 925
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.967428953129573,
+      "learning_rate": 7.434631528317209e-06,
+      "loss": 0.4046,
+      "step": 926
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.232743773961248,
+      "learning_rate": 7.4291591207018565e-06,
+      "loss": 0.3373,
+      "step": 927
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.9836939042896895,
+      "learning_rate": 7.4236829015455725e-06,
+      "loss": 0.1616,
+      "step": 928
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.234073847373181,
+      "learning_rate": 7.418202879440969e-06,
+      "loss": 0.4425,
+      "step": 929
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.790231896267228,
+      "learning_rate": 7.412719062986632e-06,
+      "loss": 0.5823,
+      "step": 930
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.178475738206307,
+      "learning_rate": 7.407231460787099e-06,
+      "loss": 0.3375,
+      "step": 931
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.099374284642818,
+      "learning_rate": 7.401740081452848e-06,
+      "loss": 0.6621,
+      "step": 932
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.159467108092978,
+      "learning_rate": 7.396244933600285e-06,
+      "loss": 0.5217,
+      "step": 933
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.9454721846184,
+      "learning_rate": 7.390746025851725e-06,
+      "loss": 0.2959,
+      "step": 934
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.177230638443921,
+      "learning_rate": 7.385243366835385e-06,
+      "loss": 0.5034,
+      "step": 935
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.041459142057979,
+      "learning_rate": 7.379736965185369e-06,
+      "loss": 0.2509,
+      "step": 936
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.909045409772038,
+      "learning_rate": 7.374226829541652e-06,
+      "loss": 0.5253,
+      "step": 937
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.416554472438831,
+      "learning_rate": 7.368712968550068e-06,
+      "loss": 0.4668,
+      "step": 938
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.7027022064011295,
+      "learning_rate": 7.363195390862298e-06,
+      "loss": 0.2844,
+      "step": 939
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.388308131366065,
+      "learning_rate": 7.3576741051358525e-06,
+      "loss": 0.4073,
+      "step": 940
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.898365393008674,
+      "learning_rate": 7.352149120034062e-06,
+      "loss": 0.1619,
+      "step": 941
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.8824562479569185,
+      "learning_rate": 7.3466204442260605e-06,
+      "loss": 0.4562,
+      "step": 942
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.190760940230458,
+      "learning_rate": 7.341088086386775e-06,
+      "loss": 0.303,
+      "step": 943
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.689279155420741,
+      "learning_rate": 7.3355520551969055e-06,
+      "loss": 0.3656,
+      "step": 944
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.707317777738777,
+      "learning_rate": 7.330012359342919e-06,
+      "loss": 0.2931,
+      "step": 945
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.166281060531139,
+      "learning_rate": 7.324469007517035e-06,
+      "loss": 0.5711,
+      "step": 946
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.3640713862293055,
+      "learning_rate": 7.318922008417203e-06,
+      "loss": 0.6074,
+      "step": 947
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.97816965650459,
+      "learning_rate": 7.313371370747104e-06,
+      "loss": 0.3745,
+      "step": 948
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.139472376430774,
+      "learning_rate": 7.3078171032161175e-06,
+      "loss": 0.3448,
+      "step": 949
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.036326371064701,
+      "learning_rate": 7.302259214539327e-06,
+      "loss": 0.5587,
+      "step": 950
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.296169035536824,
+      "learning_rate": 7.296697713437495e-06,
+      "loss": 0.3618,
+      "step": 951
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.890391546892352,
+      "learning_rate": 7.291132608637053e-06,
+      "loss": 0.5157,
+      "step": 952
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.888042735854257,
+      "learning_rate": 7.285563908870086e-06,
+      "loss": 0.2886,
+      "step": 953
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.254305854815032,
+      "learning_rate": 7.279991622874319e-06,
+      "loss": 0.1928,
+      "step": 954
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.173601890042715,
+      "learning_rate": 7.274415759393107e-06,
+      "loss": 0.6018,
+      "step": 955
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.339315049781077,
+      "learning_rate": 7.268836327175417e-06,
+      "loss": 0.5289,
+      "step": 956
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.111630198110997,
+      "learning_rate": 7.263253334975812e-06,
+      "loss": 0.2405,
+      "step": 957
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.280261001417196,
+      "learning_rate": 7.257666791554448e-06,
+      "loss": 0.2744,
+      "step": 958
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.607057236377991,
+      "learning_rate": 7.252076705677046e-06,
+      "loss": 0.3588,
+      "step": 959
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.417022781462303,
+      "learning_rate": 7.24648308611489e-06,
+      "loss": 0.487,
+      "step": 960
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.384842219101503,
+      "learning_rate": 7.2408859416448065e-06,
+      "loss": 0.5305,
+      "step": 961
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.815721179272225,
+      "learning_rate": 7.235285281049154e-06,
+      "loss": 0.194,
+      "step": 962
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.1484704211368975,
+      "learning_rate": 7.22968111311581e-06,
+      "loss": 0.7335,
+      "step": 963
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.792626651193991,
+      "learning_rate": 7.224073446638149e-06,
+      "loss": 0.4214,
+      "step": 964
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.258760541652446,
+      "learning_rate": 7.21846229041504e-06,
+      "loss": 0.4073,
+      "step": 965
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.614835640589382,
+      "learning_rate": 7.212847653250828e-06,
+      "loss": 0.5402,
+      "step": 966
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.609410649170137,
+      "learning_rate": 7.207229543955319e-06,
+      "loss": 0.2112,
+      "step": 967
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.816974369567264,
+      "learning_rate": 7.201607971343763e-06,
+      "loss": 0.2038,
+      "step": 968
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.236157249748036,
+      "learning_rate": 7.195982944236853e-06,
+      "loss": 0.2521,
+      "step": 969
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.402175205571623,
+      "learning_rate": 7.190354471460692e-06,
+      "loss": 0.2382,
+      "step": 970
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.819611590718435,
+      "learning_rate": 7.1847225618467975e-06,
+      "loss": 0.3573,
+      "step": 971
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.1465471735663746,
+      "learning_rate": 7.1790872242320775e-06,
+      "loss": 0.1005,
+      "step": 972
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.640336574149923,
+      "learning_rate": 7.173448467458817e-06,
+      "loss": 0.3745,
+      "step": 973
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.500844710543615,
+      "learning_rate": 7.167806300374665e-06,
+      "loss": 0.6411,
+      "step": 974
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.689235896418885,
+      "learning_rate": 7.162160731832627e-06,
+      "loss": 0.3168,
+      "step": 975
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.09472144596906,
+      "learning_rate": 7.15651177069104e-06,
+      "loss": 0.5588,
+      "step": 976
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.4432814427661684,
+      "learning_rate": 7.1508594258135685e-06,
+      "loss": 0.4689,
+      "step": 977
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.543801171222169,
+      "learning_rate": 7.145203706069183e-06,
+      "loss": 0.3286,
+      "step": 978
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.6396114540245357,
+      "learning_rate": 7.139544620332151e-06,
+      "loss": 0.2385,
+      "step": 979
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.974055244651273,
+      "learning_rate": 7.133882177482019e-06,
+      "loss": 0.4335,
+      "step": 980
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.56871393982033,
+      "learning_rate": 7.128216386403608e-06,
+      "loss": 0.46,
+      "step": 981
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.489308899737765,
+      "learning_rate": 7.122547255986985e-06,
+      "loss": 0.4826,
+      "step": 982
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.607582520293363,
+      "learning_rate": 7.1168747951274596e-06,
+      "loss": 0.2561,
+      "step": 983
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.463694616722066,
+      "learning_rate": 7.1111990127255684e-06,
+      "loss": 0.4603,
+      "step": 984
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.945644671403082,
+      "learning_rate": 7.105519917687058e-06,
+      "loss": 0.6607,
+      "step": 985
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.750019937917883,
+      "learning_rate": 7.099837518922873e-06,
+      "loss": 0.1904,
+      "step": 986
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.482299071320674,
+      "learning_rate": 7.094151825349145e-06,
+      "loss": 0.3152,
+      "step": 987
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.8434907070965467,
+      "learning_rate": 7.088462845887168e-06,
+      "loss": 0.4441,
+      "step": 988
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.773628678898187,
+      "learning_rate": 7.082770589463398e-06,
+      "loss": 0.4988,
+      "step": 989
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.213841525985786,
+      "learning_rate": 7.0770750650094335e-06,
+      "loss": 0.4936,
+      "step": 990
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.432768016996358,
+      "learning_rate": 7.071376281461994e-06,
+      "loss": 0.4266,
+      "step": 991
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.157815564365318,
+      "learning_rate": 7.065674247762924e-06,
+      "loss": 0.5153,
+      "step": 992
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.029204598653521,
+      "learning_rate": 7.059968972859155e-06,
+      "loss": 0.4522,
+      "step": 993
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.046118804737403,
+      "learning_rate": 7.054260465702712e-06,
+      "loss": 0.4115,
+      "step": 994
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.399234145678143,
+      "learning_rate": 7.04854873525069e-06,
+      "loss": 0.3488,
+      "step": 995
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.811268916684834,
+      "learning_rate": 7.042833790465241e-06,
+      "loss": 0.442,
+      "step": 996
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.097682814805097,
+      "learning_rate": 7.037115640313563e-06,
+      "loss": 0.5786,
+      "step": 997
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.89406435386151,
+      "learning_rate": 7.031394293767879e-06,
+      "loss": 0.2893,
+      "step": 998
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.741921341840173,
+      "learning_rate": 7.025669759805431e-06,
+      "loss": 0.3645,
+      "step": 999
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.097084643366875,
+      "learning_rate": 7.019942047408461e-06,
+      "loss": 0.1835,
+      "step": 1000
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6627766749272523,
+      "learning_rate": 7.0142111655642e-06,
+      "loss": 0.2201,
+      "step": 1001
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.037169315513487,
+      "learning_rate": 7.008477123264849e-06,
+      "loss": 0.2294,
+      "step": 1002
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.723506514756312,
+      "learning_rate": 7.002739929507569e-06,
+      "loss": 0.2124,
+      "step": 1003
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.303370968502258,
+      "learning_rate": 6.996999593294466e-06,
+      "loss": 0.2925,
+      "step": 1004
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.279075537465586,
+      "learning_rate": 6.991256123632577e-06,
+      "loss": 0.7073,
+      "step": 1005
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.04058869264715,
+      "learning_rate": 6.985509529533859e-06,
+      "loss": 0.5377,
+      "step": 1006
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.8875963143709975,
+      "learning_rate": 6.979759820015166e-06,
+      "loss": 0.4554,
+      "step": 1007
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.406641061009185,
+      "learning_rate": 6.974007004098243e-06,
+      "loss": 0.4458,
+      "step": 1008
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.93528175971489,
+      "learning_rate": 6.968251090809708e-06,
+      "loss": 0.2491,
+      "step": 1009
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.046559605036485,
+      "learning_rate": 6.96249208918104e-06,
+      "loss": 0.2308,
+      "step": 1010
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.406269492678815,
+      "learning_rate": 6.956730008248565e-06,
+      "loss": 0.3685,
+      "step": 1011
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.898383037743084,
+      "learning_rate": 6.950964857053436e-06,
+      "loss": 0.2557,
+      "step": 1012
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.800167808069986,
+      "learning_rate": 6.945196644641631e-06,
+      "loss": 0.3821,
+      "step": 1013
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.1650493510014055,
+      "learning_rate": 6.939425380063924e-06,
+      "loss": 0.5418,
+      "step": 1014
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.478798815675141,
+      "learning_rate": 6.933651072375884e-06,
+      "loss": 0.2895,
+      "step": 1015
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7967882889163005,
+      "learning_rate": 6.927873730637849e-06,
+      "loss": 0.2873,
+      "step": 1016
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.968736109982194,
+      "learning_rate": 6.922093363914923e-06,
+      "loss": 0.3376,
+      "step": 1017
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.9904418841960014,
+      "learning_rate": 6.916309981276954e-06,
+      "loss": 0.4059,
+      "step": 1018
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.889256192891055,
+      "learning_rate": 6.9105235917985215e-06,
+      "loss": 0.3218,
+      "step": 1019
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.110778907024558,
+      "learning_rate": 6.9047342045589224e-06,
+      "loss": 0.4627,
+      "step": 1020
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.685777372986321,
+      "learning_rate": 6.898941828642163e-06,
+      "loss": 0.4065,
+      "step": 1021
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.331353037515637,
+      "learning_rate": 6.89314647313693e-06,
+      "loss": 0.342,
+      "step": 1022
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.765773716278677,
+      "learning_rate": 6.887348147136592e-06,
+      "loss": 0.2128,
+      "step": 1023
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.447548302533927,
+      "learning_rate": 6.8815468597391785e-06,
+      "loss": 0.4476,
+      "step": 1024
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.147038446259682,
+      "learning_rate": 6.875742620047361e-06,
+      "loss": 0.4365,
+      "step": 1025
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.0080880227394027,
+      "learning_rate": 6.869935437168449e-06,
+      "loss": 0.2937,
+      "step": 1026
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.713681508637163,
+      "learning_rate": 6.864125320214364e-06,
+      "loss": 0.311,
+      "step": 1027
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.9305866178493307,
+      "learning_rate": 6.858312278301638e-06,
+      "loss": 0.4492,
+      "step": 1028
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.132501267665453,
+      "learning_rate": 6.852496320551387e-06,
+      "loss": 0.2314,
+      "step": 1029
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.581024219407659,
+      "learning_rate": 6.846677456089305e-06,
+      "loss": 0.6607,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.6534322322023804,
+      "learning_rate": 6.840855694045647e-06,
+      "loss": 0.222,
+      "step": 1031
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.9230792561451957,
+      "learning_rate": 6.835031043555211e-06,
+      "loss": 0.4596,
+      "step": 1032
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.4441479989012125,
+      "learning_rate": 6.829203513757332e-06,
+      "loss": 0.4975,
+      "step": 1033
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.960178871281018,
+      "learning_rate": 6.82337311379586e-06,
+      "loss": 0.2867,
+      "step": 1034
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.6012987952419007,
+      "learning_rate": 6.817539852819149e-06,
+      "loss": 0.2068,
+      "step": 1035
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.873659106808283,
+      "learning_rate": 6.811703739980045e-06,
+      "loss": 0.4902,
+      "step": 1036
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.9016539751511976,
+      "learning_rate": 6.8058647844358625e-06,
+      "loss": 0.4564,
+      "step": 1037
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.615831624104007,
+      "learning_rate": 6.800022995348381e-06,
+      "loss": 0.2232,
+      "step": 1038
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.517887212480329,
+      "learning_rate": 6.794178381883827e-06,
+      "loss": 0.3346,
+      "step": 1039
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.874539223333172,
+      "learning_rate": 6.788330953212854e-06,
+      "loss": 0.4881,
+      "step": 1040
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.216710935786579,
+      "learning_rate": 6.782480718510538e-06,
+      "loss": 0.6149,
+      "step": 1041
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.037340989288139,
+      "learning_rate": 6.776627686956354e-06,
+      "loss": 0.5184,
+      "step": 1042
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.499040907639204,
+      "learning_rate": 6.770771867734169e-06,
+      "loss": 0.6088,
+      "step": 1043
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.994247873916255,
+      "learning_rate": 6.76491327003222e-06,
+      "loss": 0.6212,
+      "step": 1044
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.941613195271821,
+      "learning_rate": 6.7590519030431054e-06,
+      "loss": 0.4855,
+      "step": 1045
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.294732445961128,
+      "learning_rate": 6.753187775963773e-06,
+      "loss": 0.2688,
+      "step": 1046
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.768122707458756,
+      "learning_rate": 6.747320897995493e-06,
+      "loss": 0.5366,
+      "step": 1047
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.2196424193217545,
+      "learning_rate": 6.7414512783438575e-06,
+      "loss": 0.3854,
+      "step": 1048
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.3578808239763775,
+      "learning_rate": 6.73557892621876e-06,
+      "loss": 0.348,
+      "step": 1049
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.502428374052916,
+      "learning_rate": 6.729703850834381e-06,
+      "loss": 0.3777,
+      "step": 1050
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.9362328259032635,
+      "learning_rate": 6.723826061409176e-06,
+      "loss": 0.3594,
+      "step": 1051
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.30420303398428,
+      "learning_rate": 6.717945567165854e-06,
+      "loss": 0.4955,
+      "step": 1052
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.615544401286117,
+      "learning_rate": 6.712062377331371e-06,
+      "loss": 0.5446,
+      "step": 1053
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.902992375707539,
+      "learning_rate": 6.706176501136914e-06,
+      "loss": 0.3112,
+      "step": 1054
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.408063450853472,
+      "learning_rate": 6.700287947817885e-06,
+      "loss": 0.2684,
+      "step": 1055
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.036836225427641,
+      "learning_rate": 6.694396726613883e-06,
+      "loss": 0.2459,
+      "step": 1056
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.066047406609241,
+      "learning_rate": 6.688502846768697e-06,
+      "loss": 0.513,
+      "step": 1057
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.7869968785109425,
+      "learning_rate": 6.682606317530284e-06,
+      "loss": 0.172,
+      "step": 1058
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.359044285906457,
+      "learning_rate": 6.676707148150763e-06,
+      "loss": 0.4342,
+      "step": 1059
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.698820258121499,
+      "learning_rate": 6.670805347886392e-06,
+      "loss": 0.5618,
+      "step": 1060
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.425351971010927,
+      "learning_rate": 6.6649009259975585e-06,
+      "loss": 0.4421,
+      "step": 1061
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.285062575279868,
+      "learning_rate": 6.65899389174876e-06,
+      "loss": 0.2069,
+      "step": 1062
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.8461884048317,
+      "learning_rate": 6.653084254408599e-06,
+      "loss": 0.4529,
+      "step": 1063
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.527993871516133,
+      "learning_rate": 6.647172023249758e-06,
+      "loss": 0.5812,
+      "step": 1064
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.838554724261919,
+      "learning_rate": 6.641257207548992e-06,
+      "loss": 0.3948,
+      "step": 1065
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.8931808000422543,
+      "learning_rate": 6.635339816587109e-06,
+      "loss": 0.1686,
+      "step": 1066
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.92898728170045,
+      "learning_rate": 6.629419859648959e-06,
+      "loss": 0.3539,
+      "step": 1067
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.054418910841877,
+      "learning_rate": 6.6234973460234184e-06,
+      "loss": 0.6506,
+      "step": 1068
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.327327643023074,
+      "learning_rate": 6.6175722850033755e-06,
+      "loss": 0.237,
+      "step": 1069
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.088852064696572,
+      "learning_rate": 6.611644685885713e-06,
+      "loss": 0.3575,
+      "step": 1070
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.5379329445691985,
+      "learning_rate": 6.6057145579713015e-06,
+      "loss": 0.3103,
+      "step": 1071
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.238507620341176,
+      "learning_rate": 6.599781910564973e-06,
+      "loss": 0.5496,
+      "step": 1072
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.2361587413049415,
+      "learning_rate": 6.593846752975518e-06,
+      "loss": 0.3359,
+      "step": 1073
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.723970206443219,
+      "learning_rate": 6.587909094515663e-06,
+      "loss": 0.5703,
+      "step": 1074
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.1645926050266273,
+      "learning_rate": 6.58196894450206e-06,
+      "loss": 0.0898,
+      "step": 1075
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.90095699604733,
+      "learning_rate": 6.576026312255268e-06,
+      "loss": 0.4937,
+      "step": 1076
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.332468638885417,
+      "learning_rate": 6.570081207099744e-06,
+      "loss": 0.5269,
+      "step": 1077
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.85470005034806,
+      "learning_rate": 6.564133638363823e-06,
+      "loss": 0.2602,
+      "step": 1078
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.026178179195995,
+      "learning_rate": 6.558183615379708e-06,
+      "loss": 0.3851,
+      "step": 1079
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.683079412951481,
+      "learning_rate": 6.552231147483448e-06,
+      "loss": 0.537,
+      "step": 1080
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.4753014518903536,
+      "learning_rate": 6.546276244014933e-06,
+      "loss": 0.3069,
+      "step": 1081
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.759588871696704,
+      "learning_rate": 6.5403189143178725e-06,
+      "loss": 0.3842,
+      "step": 1082
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.658300705288632,
+      "learning_rate": 6.534359167739784e-06,
+      "loss": 0.5289,
+      "step": 1083
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.710464926969626,
+      "learning_rate": 6.528397013631975e-06,
+      "loss": 0.2284,
+      "step": 1084
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.333273685475453,
+      "learning_rate": 6.522432461349536e-06,
+      "loss": 0.2599,
+      "step": 1085
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.5792537507612883,
+      "learning_rate": 6.5164655202513135e-06,
+      "loss": 0.11,
+      "step": 1086
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.530875134448532,
+      "learning_rate": 6.510496199699906e-06,
+      "loss": 0.5215,
+      "step": 1087
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.8562640009376037,
+      "learning_rate": 6.504524509061646e-06,
+      "loss": 0.2141,
+      "step": 1088
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.487716847033595,
+      "learning_rate": 6.498550457706584e-06,
+      "loss": 0.2573,
+      "step": 1089
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.292321959679036,
+      "learning_rate": 6.492574055008474e-06,
+      "loss": 0.3859,
+      "step": 1090
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.250167288607771,
+      "learning_rate": 6.4865953103447585e-06,
+      "loss": 0.1826,
+      "step": 1091
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.002364415651722,
+      "learning_rate": 6.480614233096558e-06,
+      "loss": 0.2539,
+      "step": 1092
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.399024878313421,
+      "learning_rate": 6.474630832648651e-06,
+      "loss": 0.1673,
+      "step": 1093
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.575294574941341,
+      "learning_rate": 6.4686451183894604e-06,
+      "loss": 0.224,
+      "step": 1094
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.657883459327564,
+      "learning_rate": 6.462657099711044e-06,
+      "loss": 0.4167,
+      "step": 1095
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.6453216067272676,
+      "learning_rate": 6.4566667860090674e-06,
+      "loss": 0.3451,
+      "step": 1096
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.65412738800545,
+      "learning_rate": 6.4506741866828035e-06,
+      "loss": 0.4441,
+      "step": 1097
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.392961673927681,
+      "learning_rate": 6.444679311135112e-06,
+      "loss": 0.404,
+      "step": 1098
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.858981707508947,
+      "learning_rate": 6.438682168772421e-06,
+      "loss": 0.5141,
+      "step": 1099
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.983253209510925,
+      "learning_rate": 6.432682769004717e-06,
+      "loss": 0.301,
+      "step": 1100
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.103632831354062,
+      "learning_rate": 6.426681121245527e-06,
+      "loss": 0.0999,
+      "step": 1101
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.462511305934385,
+      "learning_rate": 6.420677234911908e-06,
+      "loss": 0.2751,
+      "step": 1102
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.241387633043713,
+      "learning_rate": 6.414671119424426e-06,
+      "loss": 0.2869,
+      "step": 1103
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.0268500807699805,
+      "learning_rate": 6.408662784207149e-06,
+      "loss": 0.4188,
+      "step": 1104
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.99883312346557,
+      "learning_rate": 6.402652238687624e-06,
+      "loss": 0.4171,
+      "step": 1105
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.513925241895601,
+      "learning_rate": 6.396639492296868e-06,
+      "loss": 0.103,
+      "step": 1106
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.692900850075641,
+      "learning_rate": 6.390624554469351e-06,
+      "loss": 0.333,
+      "step": 1107
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.17466899381888,
+      "learning_rate": 6.384607434642982e-06,
+      "loss": 0.6551,
+      "step": 1108
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.87571778209691,
+      "learning_rate": 6.378588142259094e-06,
+      "loss": 0.4127,
+      "step": 1109
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.647447251806186,
+      "learning_rate": 6.372566686762427e-06,
+      "loss": 0.2528,
+      "step": 1110
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.7169210061877,
+      "learning_rate": 6.366543077601116e-06,
+      "loss": 0.3252,
+      "step": 1111
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.381070555240646,
+      "learning_rate": 6.360517324226676e-06,
+      "loss": 0.2383,
+      "step": 1112
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.056682885353319,
+      "learning_rate": 6.354489436093987e-06,
+      "loss": 0.3579,
+      "step": 1113
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.728850574726878,
+      "learning_rate": 6.348459422661276e-06,
+      "loss": 0.5767,
+      "step": 1114
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.7650735524192065,
+      "learning_rate": 6.342427293390108e-06,
+      "loss": 0.4645,
+      "step": 1115
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.107658350302222,
+      "learning_rate": 6.336393057745365e-06,
+      "loss": 0.5181,
+      "step": 1116
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.050427103553093,
+      "learning_rate": 6.330356725195237e-06,
+      "loss": 0.4415,
+      "step": 1117
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.113583600744585,
+      "learning_rate": 6.324318305211201e-06,
+      "loss": 0.2275,
+      "step": 1118
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.177665868869402,
+      "learning_rate": 6.318277807268013e-06,
+      "loss": 0.4803,
+      "step": 1119
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.894687258632692,
+      "learning_rate": 6.312235240843685e-06,
+      "loss": 0.4936,
+      "step": 1120
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.75435248759306,
+      "learning_rate": 6.306190615419476e-06,
+      "loss": 0.22,
+      "step": 1121
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.892314258562756,
+      "learning_rate": 6.300143940479881e-06,
+      "loss": 0.6011,
+      "step": 1122
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.379855244418597,
+      "learning_rate": 6.294095225512604e-06,
+      "loss": 0.3588,
+      "step": 1123
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.884767438661871,
+      "learning_rate": 6.2880444800085535e-06,
+      "loss": 0.2015,
+      "step": 1124
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.551356667662048,
+      "learning_rate": 6.281991713461823e-06,
+      "loss": 0.2999,
+      "step": 1125
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.02565697507532,
+      "learning_rate": 6.275936935369675e-06,
+      "loss": 0.3088,
+      "step": 1126
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.259174749542063,
+      "learning_rate": 6.269880155232534e-06,
+      "loss": 0.2645,
+      "step": 1127
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.1260340427461255,
+      "learning_rate": 6.2638213825539595e-06,
+      "loss": 0.2912,
+      "step": 1128
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.3305769669563485,
+      "learning_rate": 6.257760626840644e-06,
+      "loss": 0.3689,
+      "step": 1129
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.018191721498685,
+      "learning_rate": 6.251697897602384e-06,
+      "loss": 0.4977,
+      "step": 1130
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.162623471632947,
+      "learning_rate": 6.245633204352078e-06,
+      "loss": 0.2183,
+      "step": 1131
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.104107461505769,
+      "learning_rate": 6.239566556605706e-06,
+      "loss": 0.411,
+      "step": 1132
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.719875066793219,
+      "learning_rate": 6.233497963882314e-06,
+      "loss": 0.5093,
+      "step": 1133
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.794565517083581,
+      "learning_rate": 6.227427435703997e-06,
+      "loss": 0.4455,
+      "step": 1134
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.032460892724786,
+      "learning_rate": 6.22135498159589e-06,
+      "loss": 0.4597,
+      "step": 1135
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.015209913146625,
+      "learning_rate": 6.215280611086149e-06,
+      "loss": 0.5067,
+      "step": 1136
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.075035112182945,
+      "learning_rate": 6.209204333705937e-06,
+      "loss": 0.4369,
+      "step": 1137
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.93502641479607,
+      "learning_rate": 6.203126158989411e-06,
+      "loss": 0.4362,
+      "step": 1138
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.815104553602145,
+      "learning_rate": 6.197046096473701e-06,
+      "loss": 0.4026,
+      "step": 1139
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.051113586771054,
+      "learning_rate": 6.190964155698903e-06,
+      "loss": 0.5172,
+      "step": 1140
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.723729964761336,
+      "learning_rate": 6.184880346208056e-06,
+      "loss": 0.3351,
+      "step": 1141
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1836934395419347,
+      "learning_rate": 6.178794677547138e-06,
+      "loss": 0.3614,
+      "step": 1142
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.2274658876489495,
+      "learning_rate": 6.1727071592650345e-06,
+      "loss": 0.3113,
+      "step": 1143
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.908194274279897,
+      "learning_rate": 6.1666178009135426e-06,
+      "loss": 0.4549,
+      "step": 1144
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.863167550276418,
+      "learning_rate": 6.160526612047339e-06,
+      "loss": 0.4472,
+      "step": 1145
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.81905721465025,
+      "learning_rate": 6.154433602223979e-06,
+      "loss": 0.2906,
+      "step": 1146
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.870031172516451,
+      "learning_rate": 6.148338781003873e-06,
+      "loss": 0.3564,
+      "step": 1147
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.634649363433993,
+      "learning_rate": 6.14224215795027e-06,
+      "loss": 0.3849,
+      "step": 1148
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.024500636140412,
+      "learning_rate": 6.136143742629252e-06,
+      "loss": 0.4495,
+      "step": 1149
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.668985557086553,
+      "learning_rate": 6.130043544609707e-06,
+      "loss": 0.4648,
+      "step": 1150
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.42891353501445,
+      "learning_rate": 6.123941573463327e-06,
+      "loss": 0.3441,
+      "step": 1151
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.857536388028443,
+      "learning_rate": 6.117837838764579e-06,
+      "loss": 0.5887,
+      "step": 1152
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2776577992229217,
+      "learning_rate": 6.111732350090703e-06,
+      "loss": 0.3666,
+      "step": 1153
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.096908639197986,
+      "learning_rate": 6.105625117021692e-06,
+      "loss": 0.4971,
+      "step": 1154
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.3840741300229147,
+      "learning_rate": 6.099516149140267e-06,
+      "loss": 0.1509,
+      "step": 1155
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.6619159705833555,
+      "learning_rate": 6.09340545603188e-06,
+      "loss": 0.4491,
+      "step": 1156
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.258172288915472,
+      "learning_rate": 6.087293047284687e-06,
+      "loss": 0.5557,
+      "step": 1157
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.236971319237751,
+      "learning_rate": 6.0811789324895365e-06,
+      "loss": 0.537,
+      "step": 1158
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.080538142556825,
+      "learning_rate": 6.075063121239954e-06,
+      "loss": 0.5135,
+      "step": 1159
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.884510279760993,
+      "learning_rate": 6.068945623132125e-06,
+      "loss": 0.3686,
+      "step": 1160
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.718368897824202,
+      "learning_rate": 6.062826447764883e-06,
+      "loss": 0.2197,
+      "step": 1161
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.142915658609104,
+      "learning_rate": 6.056705604739696e-06,
+      "loss": 0.302,
+      "step": 1162
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.38869337946155,
+      "learning_rate": 6.050583103660643e-06,
+      "loss": 0.3433,
+      "step": 1163
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.333723556530458,
+      "learning_rate": 6.044458954134411e-06,
+      "loss": 0.3373,
+      "step": 1164
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.1630218045058935,
+      "learning_rate": 6.038333165770267e-06,
+      "loss": 0.2824,
+      "step": 1165
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.964394593601022,
+      "learning_rate": 6.032205748180054e-06,
+      "loss": 0.3608,
+      "step": 1166
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.07047404818321,
+      "learning_rate": 6.026076710978172e-06,
+      "loss": 0.4182,
+      "step": 1167
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.965978987107587,
+      "learning_rate": 6.019946063781559e-06,
+      "loss": 0.1604,
+      "step": 1168
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.586994011253932,
+      "learning_rate": 6.01381381620968e-06,
+      "loss": 0.4027,
+      "step": 1169
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.133231119435957,
+      "learning_rate": 6.0076799778845105e-06,
+      "loss": 0.3024,
+      "step": 1170
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.02658686052961,
+      "learning_rate": 6.001544558430527e-06,
+      "loss": 0.1995,
+      "step": 1171
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.890567611341687,
+      "learning_rate": 5.99540756747468e-06,
+      "loss": 0.3806,
+      "step": 1172
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8223830234142473,
+      "learning_rate": 5.989269014646393e-06,
+      "loss": 0.2737,
+      "step": 1173
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.429044083910661,
+      "learning_rate": 5.983128909577532e-06,
+      "loss": 0.2405,
+      "step": 1174
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.261424803528859,
+      "learning_rate": 5.976987261902403e-06,
+      "loss": 0.247,
+      "step": 1175
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.973107952087142,
+      "learning_rate": 5.970844081257734e-06,
+      "loss": 0.4896,
+      "step": 1176
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.580671442780074,
+      "learning_rate": 5.964699377282657e-06,
+      "loss": 0.5314,
+      "step": 1177
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.4169154107952195,
+      "learning_rate": 5.958553159618693e-06,
+      "loss": 0.4399,
+      "step": 1178
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.0613887064501135,
+      "learning_rate": 5.952405437909738e-06,
+      "loss": 0.6028,
+      "step": 1179
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.006412042006303,
+      "learning_rate": 5.946256221802052e-06,
+      "loss": 0.3751,
+      "step": 1180
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.3575450429058185,
+      "learning_rate": 5.940105520944232e-06,
+      "loss": 0.4863,
+      "step": 1181
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.024201890669275,
+      "learning_rate": 5.933953344987215e-06,
+      "loss": 0.5701,
+      "step": 1182
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.93172089619764,
+      "learning_rate": 5.927799703584242e-06,
+      "loss": 0.5711,
+      "step": 1183
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.534927950024668,
+      "learning_rate": 5.921644606390862e-06,
+      "loss": 0.3593,
+      "step": 1184
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.060085078207586,
+      "learning_rate": 5.915488063064902e-06,
+      "loss": 0.2987,
+      "step": 1185
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.934505560678856,
+      "learning_rate": 5.9093300832664625e-06,
+      "loss": 0.4388,
+      "step": 1186
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.442592625568581,
+      "learning_rate": 5.903170676657894e-06,
+      "loss": 0.1431,
+      "step": 1187
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3185934657339384,
+      "learning_rate": 5.897009852903792e-06,
+      "loss": 0.2223,
+      "step": 1188
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.734216767611118,
+      "learning_rate": 5.890847621670966e-06,
+      "loss": 0.1683,
+      "step": 1189
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.865184566691955,
+      "learning_rate": 5.8846839926284435e-06,
+      "loss": 0.6131,
+      "step": 1190
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.565513874168808,
+      "learning_rate": 5.878518975447439e-06,
+      "loss": 0.5469,
+      "step": 1191
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.064162651208678,
+      "learning_rate": 5.872352579801351e-06,
+      "loss": 0.4186,
+      "step": 1192
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.0208368031220365,
+      "learning_rate": 5.866184815365733e-06,
+      "loss": 0.1921,
+      "step": 1193
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.248481097308518,
+      "learning_rate": 5.860015691818292e-06,
+      "loss": 0.5729,
+      "step": 1194
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.049651461983444,
+      "learning_rate": 5.853845218838868e-06,
+      "loss": 0.1473,
+      "step": 1195
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.609426453789442,
+      "learning_rate": 5.847673406109413e-06,
+      "loss": 0.1931,
+      "step": 1196
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.196204602326466,
+      "learning_rate": 5.841500263313987e-06,
+      "loss": 0.5721,
+      "step": 1197
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.011308483929049,
+      "learning_rate": 5.835325800138736e-06,
+      "loss": 0.4812,
+      "step": 1198
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.405669496225648,
+      "learning_rate": 5.829150026271871e-06,
+      "loss": 0.1026,
+      "step": 1199
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.433134817157713,
+      "learning_rate": 5.82297295140367e-06,
+      "loss": 0.4697,
+      "step": 1200
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.247143845938337,
+      "learning_rate": 5.816794585226445e-06,
+      "loss": 0.4352,
+      "step": 1201
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.019507920925523,
+      "learning_rate": 5.810614937434537e-06,
+      "loss": 0.2723,
+      "step": 1202
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.467886743770641,
+      "learning_rate": 5.8044340177242995e-06,
+      "loss": 0.2454,
+      "step": 1203
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.169152841005073,
+      "learning_rate": 5.7982518357940755e-06,
+      "loss": 0.4425,
+      "step": 1204
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.067046748413299,
+      "learning_rate": 5.792068401344195e-06,
+      "loss": 0.3379,
+      "step": 1205
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.89229424685211,
+      "learning_rate": 5.78588372407695e-06,
+      "loss": 0.4855,
+      "step": 1206
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.432031432757888,
+      "learning_rate": 5.779697813696587e-06,
+      "loss": 0.4653,
+      "step": 1207
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.166187334895502,
+      "learning_rate": 5.77351067990928e-06,
+      "loss": 0.3022,
+      "step": 1208
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.566813188099879,
+      "learning_rate": 5.767322332423128e-06,
+      "loss": 0.3696,
+      "step": 1209
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.873947643336162,
+      "learning_rate": 5.761132780948132e-06,
+      "loss": 0.4058,
+      "step": 1210
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.8649685765666915,
+      "learning_rate": 5.7549420351961845e-06,
+      "loss": 0.2907,
+      "step": 1211
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.972624031864407,
+      "learning_rate": 5.748750104881051e-06,
+      "loss": 0.0882,
+      "step": 1212
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.629809020290339,
+      "learning_rate": 5.742556999718353e-06,
+      "loss": 0.4197,
+      "step": 1213
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.019146007779661,
+      "learning_rate": 5.736362729425558e-06,
+      "loss": 0.4386,
+      "step": 1214
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.879325619503961,
+      "learning_rate": 5.730167303721963e-06,
+      "loss": 0.2226,
+      "step": 1215
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.302162364705677,
+      "learning_rate": 5.723970732328675e-06,
+      "loss": 0.3707,
+      "step": 1216
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.693996901628024,
+      "learning_rate": 5.717773024968602e-06,
+      "loss": 0.2543,
+      "step": 1217
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.45242990400126,
+      "learning_rate": 5.711574191366427e-06,
+      "loss": 0.2894,
+      "step": 1218
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.69524047261105,
+      "learning_rate": 5.70537424124861e-06,
+      "loss": 0.4739,
+      "step": 1219
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.42156011111522,
+      "learning_rate": 5.699173184343357e-06,
+      "loss": 0.3222,
+      "step": 1220
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.5591476661146224,
+      "learning_rate": 5.692971030380613e-06,
+      "loss": 0.5504,
+      "step": 1221
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.404897507050902,
+      "learning_rate": 5.686767789092041e-06,
+      "loss": 0.3497,
+      "step": 1222
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.140349965762885,
+      "learning_rate": 5.6805634702110155e-06,
+      "loss": 0.4362,
+      "step": 1223
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.6771882068304116,
+      "learning_rate": 5.674358083472598e-06,
+      "loss": 0.1782,
+      "step": 1224
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.289961142006053,
+      "learning_rate": 5.668151638613524e-06,
+      "loss": 0.0961,
+      "step": 1225
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.210728581639879,
+      "learning_rate": 5.661944145372193e-06,
+      "loss": 0.4107,
+      "step": 1226
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.538860034888927,
+      "learning_rate": 5.655735613488651e-06,
+      "loss": 0.6956,
+      "step": 1227
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.972687849652108,
+      "learning_rate": 5.649526052704567e-06,
+      "loss": 0.2778,
+      "step": 1228
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.286073873714111,
+      "learning_rate": 5.643315472763229e-06,
+      "loss": 0.6749,
+      "step": 1229
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.354439211123255,
+      "learning_rate": 5.637103883409525e-06,
+      "loss": 0.2811,
+      "step": 1230
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.169129954895993,
+      "learning_rate": 5.630891294389923e-06,
+      "loss": 0.1283,
+      "step": 1231
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.748502550405167,
+      "learning_rate": 5.624677715452465e-06,
+      "loss": 0.3549,
+      "step": 1232
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.0941336491437115,
+      "learning_rate": 5.61846315634674e-06,
+      "loss": 0.3485,
+      "step": 1233
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.531285950464142,
+      "learning_rate": 5.612247626823878e-06,
+      "loss": 0.3336,
+      "step": 1234
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.1672818038900274,
+      "learning_rate": 5.606031136636534e-06,
+      "loss": 0.4179,
+      "step": 1235
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.861177099264431,
+      "learning_rate": 5.599813695538866e-06,
+      "loss": 0.1578,
+      "step": 1236
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.969074014830787,
+      "learning_rate": 5.593595313286526e-06,
+      "loss": 0.3372,
+      "step": 1237
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.087763165405732,
+      "learning_rate": 5.587375999636645e-06,
+      "loss": 0.6821,
+      "step": 1238
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.68834589702857,
+      "learning_rate": 5.581155764347812e-06,
+      "loss": 0.4139,
+      "step": 1239
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.766295320914873,
+      "learning_rate": 5.574934617180063e-06,
+      "loss": 0.6889,
+      "step": 1240
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.109333746550744,
+      "learning_rate": 5.568712567894866e-06,
+      "loss": 0.3811,
+      "step": 1241
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.9414847431989393,
+      "learning_rate": 5.562489626255104e-06,
+      "loss": 0.1047,
+      "step": 1242
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.386540890896379,
+      "learning_rate": 5.5562658020250585e-06,
+      "loss": 0.2444,
+      "step": 1243
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.735938555005505,
+      "learning_rate": 5.550041104970398e-06,
+      "loss": 0.2727,
+      "step": 1244
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.925222500287141,
+      "learning_rate": 5.5438155448581585e-06,
+      "loss": 0.5884,
+      "step": 1245
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.740961407977184,
+      "learning_rate": 5.5375891314567335e-06,
+      "loss": 0.397,
+      "step": 1246
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.306381654811139,
+      "learning_rate": 5.531361874535853e-06,
+      "loss": 0.5472,
+      "step": 1247
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.947520287156723,
+      "learning_rate": 5.52513378386657e-06,
+      "loss": 0.2943,
+      "step": 1248
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.3047376384719405,
+      "learning_rate": 5.518904869221245e-06,
+      "loss": 0.5119,
+      "step": 1249
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.537957792999452,
+      "learning_rate": 5.512675140373537e-06,
+      "loss": 0.3736,
+      "step": 1250
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.697394404742845,
+      "learning_rate": 5.506444607098375e-06,
+      "loss": 0.2577,
+      "step": 1251
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.7575481500881733,
+      "learning_rate": 5.500213279171955e-06,
+      "loss": 0.2839,
+      "step": 1252
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.325080892169685,
+      "learning_rate": 5.49398116637172e-06,
+      "loss": 0.5195,
+      "step": 1253
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.310682333635457,
+      "learning_rate": 5.487748278476342e-06,
+      "loss": 0.3729,
+      "step": 1254
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.542306776762291,
+      "learning_rate": 5.481514625265709e-06,
+      "loss": 0.3621,
+      "step": 1255
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.240042011769809,
+      "learning_rate": 5.475280216520913e-06,
+      "loss": 0.2586,
+      "step": 1256
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3475441815800413,
+      "learning_rate": 5.469045062024231e-06,
+      "loss": 0.3936,
+      "step": 1257
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.444550567362665,
+      "learning_rate": 5.462809171559104e-06,
+      "loss": 0.4832,
+      "step": 1258
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.49277554364756,
+      "learning_rate": 5.456572554910137e-06,
+      "loss": 0.2727,
+      "step": 1259
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.964218374207385,
+      "learning_rate": 5.450335221863068e-06,
+      "loss": 0.3893,
+      "step": 1260
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.8200888851705495,
+      "learning_rate": 5.444097182204762e-06,
+      "loss": 0.251,
+      "step": 1261
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.843886546917367,
+      "learning_rate": 5.437858445723191e-06,
+      "loss": 0.6017,
+      "step": 1262
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.582756610616576,
+      "learning_rate": 5.431619022207422e-06,
+      "loss": 0.1755,
+      "step": 1263
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.8065586351134115,
+      "learning_rate": 5.4253789214475975e-06,
+      "loss": 0.4714,
+      "step": 1264
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.348854201133222,
+      "learning_rate": 5.4191381532349264e-06,
+      "loss": 0.0689,
+      "step": 1265
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.308690605055334,
+      "learning_rate": 5.412896727361663e-06,
+      "loss": 0.2572,
+      "step": 1266
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.8132512454305925,
+      "learning_rate": 5.406654653621092e-06,
+      "loss": 0.5671,
+      "step": 1267
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.7302006434461283,
+      "learning_rate": 5.400411941807516e-06,
+      "loss": 0.6295,
+      "step": 1268
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.79180495036863,
+      "learning_rate": 5.39416860171624e-06,
+      "loss": 0.3222,
+      "step": 1269
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.133468204234008,
+      "learning_rate": 5.387924643143553e-06,
+      "loss": 0.2101,
+      "step": 1270
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.7855426839085893,
+      "learning_rate": 5.381680075886716e-06,
+      "loss": 0.4163,
+      "step": 1271
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.7767304290598305,
+      "learning_rate": 5.375434909743942e-06,
+      "loss": 0.0507,
+      "step": 1272
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8935497578214058,
+      "learning_rate": 5.369189154514388e-06,
+      "loss": 0.0535,
+      "step": 1273
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1298562133641754,
+      "learning_rate": 5.362942819998131e-06,
+      "loss": 0.4245,
+      "step": 1274
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.7844454000946315,
+      "learning_rate": 5.356695915996162e-06,
+      "loss": 0.3826,
+      "step": 1275
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.0834238707287165,
+      "learning_rate": 5.350448452310362e-06,
+      "loss": 0.4891,
+      "step": 1276
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.304986818984701,
+      "learning_rate": 5.344200438743489e-06,
+      "loss": 0.5352,
+      "step": 1277
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.033241097326988,
+      "learning_rate": 5.337951885099167e-06,
+      "loss": 0.6091,
+      "step": 1278
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.301711417656752,
+      "learning_rate": 5.3317028011818635e-06,
+      "loss": 0.2551,
+      "step": 1279
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.215987643808079,
+      "learning_rate": 5.3254531967968845e-06,
+      "loss": 0.3335,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.966519649083018,
+      "learning_rate": 5.319203081750348e-06,
+      "loss": 0.295,
+      "step": 1281
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.46701738285164,
+      "learning_rate": 5.312952465849173e-06,
+      "loss": 0.4782,
+      "step": 1282
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.297541144411482,
+      "learning_rate": 5.306701358901066e-06,
+      "loss": 0.3318,
+      "step": 1283
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.569942596811918,
+      "learning_rate": 5.300449770714502e-06,
+      "loss": 0.1394,
+      "step": 1284
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.976107483882604,
+      "learning_rate": 5.294197711098716e-06,
+      "loss": 0.1441,
+      "step": 1285
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.264637322106505,
+      "learning_rate": 5.287945189863676e-06,
+      "loss": 0.2133,
+      "step": 1286
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6591815017274834,
+      "learning_rate": 5.281692216820078e-06,
+      "loss": 0.3583,
+      "step": 1287
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.76105833314619,
+      "learning_rate": 5.275438801779328e-06,
+      "loss": 0.4813,
+      "step": 1288
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.635897240645958,
+      "learning_rate": 5.269184954553522e-06,
+      "loss": 0.2132,
+      "step": 1289
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.8099601746613936,
+      "learning_rate": 5.262930684955439e-06,
+      "loss": 0.2209,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.100196409365574,
+      "learning_rate": 5.2566760027985166e-06,
+      "loss": 0.4084,
+      "step": 1291
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.165643243140668,
+      "learning_rate": 5.2504209178968395e-06,
+      "loss": 0.4816,
+      "step": 1292
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.703730212664933,
+      "learning_rate": 5.2441654400651255e-06,
+      "loss": 0.283,
+      "step": 1293
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.613521938704651,
+      "learning_rate": 5.237909579118713e-06,
+      "loss": 0.4435,
+      "step": 1294
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.998155482772728,
+      "learning_rate": 5.231653344873534e-06,
+      "loss": 0.4067,
+      "step": 1295
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.592365604412431,
+      "learning_rate": 5.225396747146112e-06,
+      "loss": 0.4533,
+      "step": 1296
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.671757389339714,
+      "learning_rate": 5.219139795753539e-06,
+      "loss": 0.2443,
+      "step": 1297
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.003199894047307,
+      "learning_rate": 5.212882500513462e-06,
+      "loss": 0.3811,
+      "step": 1298
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.798964858875185,
+      "learning_rate": 5.206624871244066e-06,
+      "loss": 0.4514,
+      "step": 1299
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.548843252426148,
+      "learning_rate": 5.200366917764062e-06,
+      "loss": 0.5154,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.1170905572408065,
+      "learning_rate": 5.194108649892672e-06,
+      "loss": 0.4039,
+      "step": 1301
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.527101598475088,
+      "learning_rate": 5.187850077449604e-06,
+      "loss": 0.3006,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.887831108593034,
+      "learning_rate": 5.181591210255051e-06,
+      "loss": 0.3493,
+      "step": 1303
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.4220653913357,
+      "learning_rate": 5.175332058129664e-06,
+      "loss": 0.5213,
+      "step": 1304
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.949112894197746,
+      "learning_rate": 5.169072630894545e-06,
+      "loss": 0.4652,
+      "step": 1305
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.5970766089444446,
+      "learning_rate": 5.162812938371226e-06,
+      "loss": 0.3941,
+      "step": 1306
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.600667656635013,
+      "learning_rate": 5.1565529903816525e-06,
+      "loss": 0.2357,
+      "step": 1307
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.603442867822164,
+      "learning_rate": 5.150292796748174e-06,
+      "loss": 0.5449,
+      "step": 1308
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.572907471592153,
+      "learning_rate": 5.144032367293525e-06,
+      "loss": 0.3549,
+      "step": 1309
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.918967329511568,
+      "learning_rate": 5.137771711840811e-06,
+      "loss": 0.1822,
+      "step": 1310
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.751416230937269,
+      "learning_rate": 5.131510840213488e-06,
+      "loss": 0.1396,
+      "step": 1311
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.127729907821657,
+      "learning_rate": 5.125249762235357e-06,
+      "loss": 0.3859,
+      "step": 1312
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.276359069115821,
+      "learning_rate": 5.118988487730537e-06,
+      "loss": 0.5037,
+      "step": 1313
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.559645832299882,
+      "learning_rate": 5.112727026523461e-06,
+      "loss": 0.3524,
+      "step": 1314
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.098503918173065,
+      "learning_rate": 5.1064653884388515e-06,
+      "loss": 0.5115,
+      "step": 1315
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.765279023013616,
+      "learning_rate": 5.100203583301706e-06,
+      "loss": 0.3172,
+      "step": 1316
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.68916982314476,
+      "learning_rate": 5.0939416209372905e-06,
+      "loss": 0.364,
+      "step": 1317
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.526477907946366,
+      "learning_rate": 5.087679511171113e-06,
+      "loss": 0.1592,
+      "step": 1318
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.455813802201591,
+      "learning_rate": 5.081417263828914e-06,
+      "loss": 0.3091,
+      "step": 1319
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.055962351478127,
+      "learning_rate": 5.075154888736653e-06,
+      "loss": 0.6518,
+      "step": 1320
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.341974535809478,
+      "learning_rate": 5.068892395720482e-06,
+      "loss": 0.348,
+      "step": 1321
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.614419414981455,
+      "learning_rate": 5.062629794606748e-06,
+      "loss": 0.5504,
+      "step": 1322
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.004453011585351,
+      "learning_rate": 5.056367095221959e-06,
+      "loss": 0.4447,
+      "step": 1323
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.619127552811418,
+      "learning_rate": 5.050104307392783e-06,
+      "loss": 0.2753,
+      "step": 1324
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.568949512139795,
+      "learning_rate": 5.0438414409460234e-06,
+      "loss": 0.1231,
+      "step": 1325
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.397955295331865,
+      "learning_rate": 5.03757850570861e-06,
+      "loss": 0.648,
+      "step": 1326
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.44598650287468,
+      "learning_rate": 5.031315511507577e-06,
+      "loss": 0.4283,
+      "step": 1327
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.832899160657488,
+      "learning_rate": 5.025052468170054e-06,
+      "loss": 0.3432,
+      "step": 1328
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.8931850210303343,
+      "learning_rate": 5.018789385523245e-06,
+      "loss": 0.4084,
+      "step": 1329
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.950788434997343,
+      "learning_rate": 5.01252627339442e-06,
+      "loss": 0.1616,
+      "step": 1330
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.319062326100509,
+      "learning_rate": 5.006263141610891e-06,
+      "loss": 0.3678,
+      "step": 1331
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.041192425356723,
+      "learning_rate": 5e-06,
+      "loss": 0.4499,
+      "step": 1332
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.452899060490361,
+      "learning_rate": 4.993736858389111e-06,
+      "loss": 0.415,
+      "step": 1333
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.8849610265370007,
+      "learning_rate": 4.987473726605581e-06,
+      "loss": 0.5768,
+      "step": 1334
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.901606558710477,
+      "learning_rate": 4.981210614476755e-06,
+      "loss": 0.4216,
+      "step": 1335
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.475995918212704,
+      "learning_rate": 4.974947531829947e-06,
+      "loss": 0.5771,
+      "step": 1336
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.27343656664111,
+      "learning_rate": 4.968684488492424e-06,
+      "loss": 0.4872,
+      "step": 1337
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.81995353853467,
+      "learning_rate": 4.9624214942913916e-06,
+      "loss": 0.3985,
+      "step": 1338
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.106657369146505,
+      "learning_rate": 4.956158559053977e-06,
+      "loss": 0.3555,
+      "step": 1339
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.9024337598370034,
+      "learning_rate": 4.949895692607218e-06,
+      "loss": 0.1435,
+      "step": 1340
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.493152091605872,
+      "learning_rate": 4.9436329047780424e-06,
+      "loss": 0.2258,
+      "step": 1341
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.034364669142676,
+      "learning_rate": 4.9373702053932534e-06,
+      "loss": 0.3212,
+      "step": 1342
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.750046182270505,
+      "learning_rate": 4.9311076042795185e-06,
+      "loss": 0.4621,
+      "step": 1343
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.365292555964028,
+      "learning_rate": 4.924845111263349e-06,
+      "loss": 0.1799,
+      "step": 1344
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.040471548662068,
+      "learning_rate": 4.918582736171086e-06,
+      "loss": 0.3409,
+      "step": 1345
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.13485029700616,
+      "learning_rate": 4.912320488828887e-06,
+      "loss": 0.1801,
+      "step": 1346
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.574729597843415,
+      "learning_rate": 4.906058379062712e-06,
+      "loss": 0.4022,
+      "step": 1347
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.276802780132034,
+      "learning_rate": 4.899796416698296e-06,
+      "loss": 0.6159,
+      "step": 1348
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.6475497368492125,
+      "learning_rate": 4.893534611561152e-06,
+      "loss": 0.5064,
+      "step": 1349
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.814418825204974,
+      "learning_rate": 4.88727297347654e-06,
+      "loss": 0.4989,
+      "step": 1350
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.002754678951272,
+      "learning_rate": 4.881011512269464e-06,
+      "loss": 0.4045,
+      "step": 1351
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.9827601091357865,
+      "learning_rate": 4.874750237764645e-06,
+      "loss": 0.2493,
+      "step": 1352
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.693486262287033,
+      "learning_rate": 4.868489159786513e-06,
+      "loss": 0.444,
+      "step": 1353
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.335688140712484,
+      "learning_rate": 4.862228288159191e-06,
+      "loss": 0.5152,
+      "step": 1354
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.71196650068979,
+      "learning_rate": 4.8559676327064755e-06,
+      "loss": 0.4351,
+      "step": 1355
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.674192796837757,
+      "learning_rate": 4.8497072032518274e-06,
+      "loss": 0.417,
+      "step": 1356
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.649732752060125,
+      "learning_rate": 4.843447009618351e-06,
+      "loss": 0.3274,
+      "step": 1357
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.3033211138913,
+      "learning_rate": 4.837187061628777e-06,
+      "loss": 0.1602,
+      "step": 1358
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.315875668613648,
+      "learning_rate": 4.830927369105457e-06,
+      "loss": 0.3796,
+      "step": 1359
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.685982332069527,
+      "learning_rate": 4.824667941870338e-06,
+      "loss": 0.4421,
+      "step": 1360
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.347686458179461,
+      "learning_rate": 4.818408789744951e-06,
+      "loss": 0.3121,
+      "step": 1361
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.657288387082717,
+      "learning_rate": 4.8121499225503974e-06,
+      "loss": 0.4413,
+      "step": 1362
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.748923724907403,
+      "learning_rate": 4.8058913501073294e-06,
+      "loss": 0.4128,
+      "step": 1363
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.405162615018604,
+      "learning_rate": 4.799633082235938e-06,
+      "loss": 0.3405,
+      "step": 1364
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.414455172353855,
+      "learning_rate": 4.793375128755934e-06,
+      "loss": 0.3546,
+      "step": 1365
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.736569171545471,
+      "learning_rate": 4.787117499486539e-06,
+      "loss": 0.3899,
+      "step": 1366
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.884147814922411,
+      "learning_rate": 4.7808602042464615e-06,
+      "loss": 0.6129,
+      "step": 1367
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.5321049022077795,
+      "learning_rate": 4.774603252853889e-06,
+      "loss": 0.1721,
+      "step": 1368
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.500281378442747,
+      "learning_rate": 4.768346655126467e-06,
+      "loss": 0.5033,
+      "step": 1369
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.718876105082406,
+      "learning_rate": 4.762090420881289e-06,
+      "loss": 0.1936,
+      "step": 1370
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.199762316955981,
+      "learning_rate": 4.755834559934875e-06,
+      "loss": 0.284,
+      "step": 1371
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8804372991631624,
+      "learning_rate": 4.749579082103163e-06,
+      "loss": 0.2758,
+      "step": 1372
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.9318027094172727,
+      "learning_rate": 4.743323997201485e-06,
+      "loss": 0.1626,
+      "step": 1373
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.5867167821567083,
+      "learning_rate": 4.737069315044562e-06,
+      "loss": 0.1868,
+      "step": 1374
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.514197990561089,
+      "learning_rate": 4.730815045446478e-06,
+      "loss": 0.3716,
+      "step": 1375
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.776928401989964,
+      "learning_rate": 4.724561198220672e-06,
+      "loss": 0.3965,
+      "step": 1376
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.9094633914677335,
+      "learning_rate": 4.718307783179924e-06,
+      "loss": 0.3339,
+      "step": 1377
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.215323208202909,
+      "learning_rate": 4.712054810136327e-06,
+      "loss": 0.1818,
+      "step": 1378
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.228300881993021,
+      "learning_rate": 4.705802288901286e-06,
+      "loss": 0.328,
+      "step": 1379
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.464179689618592,
+      "learning_rate": 4.699550229285499e-06,
+      "loss": 0.6309,
+      "step": 1380
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.678966528112276,
+      "learning_rate": 4.693298641098935e-06,
+      "loss": 0.1651,
+      "step": 1381
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.459322859292966,
+      "learning_rate": 4.687047534150829e-06,
+      "loss": 0.3151,
+      "step": 1382
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.2787024520816175,
+      "learning_rate": 4.680796918249653e-06,
+      "loss": 0.4516,
+      "step": 1383
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.023343782044322,
+      "learning_rate": 4.674546803203116e-06,
+      "loss": 0.223,
+      "step": 1384
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.09531271863508,
+      "learning_rate": 4.668297198818137e-06,
+      "loss": 0.1708,
+      "step": 1385
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.754513599916183,
+      "learning_rate": 4.662048114900837e-06,
+      "loss": 0.1628,
+      "step": 1386
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.220505886201745,
+      "learning_rate": 4.6557995612565146e-06,
+      "loss": 0.3616,
+      "step": 1387
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.033642894202756,
+      "learning_rate": 4.649551547689641e-06,
+      "loss": 0.2401,
+      "step": 1388
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.643153166140661,
+      "learning_rate": 4.643304084003839e-06,
+      "loss": 0.3787,
+      "step": 1389
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.560528244548518,
+      "learning_rate": 4.6370571800018695e-06,
+      "loss": 0.1877,
+      "step": 1390
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.388491608904832,
+      "learning_rate": 4.630810845485613e-06,
+      "loss": 0.1232,
+      "step": 1391
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.93637784624128,
+      "learning_rate": 4.624565090256059e-06,
+      "loss": 0.365,
+      "step": 1392
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.257432987702576,
+      "learning_rate": 4.618319924113286e-06,
+      "loss": 0.2224,
+      "step": 1393
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.786647388241515,
+      "learning_rate": 4.612075356856447e-06,
+      "loss": 0.5707,
+      "step": 1394
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.935423072133717,
+      "learning_rate": 4.60583139828376e-06,
+      "loss": 0.3992,
+      "step": 1395
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.511290981704092,
+      "learning_rate": 4.599588058192487e-06,
+      "loss": 0.3921,
+      "step": 1396
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.785240331651109,
+      "learning_rate": 4.5933453463789105e-06,
+      "loss": 0.6688,
+      "step": 1397
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.561205508605427,
+      "learning_rate": 4.587103272638339e-06,
+      "loss": 0.3536,
+      "step": 1398
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.352232431684483,
+      "learning_rate": 4.580861846765075e-06,
+      "loss": 0.6256,
+      "step": 1399
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.446901089704637,
+      "learning_rate": 4.574621078552403e-06,
+      "loss": 0.6396,
+      "step": 1400
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.3216933923128265,
+      "learning_rate": 4.568380977792581e-06,
+      "loss": 0.5249,
+      "step": 1401
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.826445694399736,
+      "learning_rate": 4.562141554276811e-06,
+      "loss": 0.4093,
+      "step": 1402
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.254819834764432,
+      "learning_rate": 4.555902817795239e-06,
+      "loss": 0.5174,
+      "step": 1403
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.466656980375848,
+      "learning_rate": 4.549664778136933e-06,
+      "loss": 0.2572,
+      "step": 1404
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.175437382426703,
+      "learning_rate": 4.543427445089863e-06,
+      "loss": 0.2028,
+      "step": 1405
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.8911882705272,
+      "learning_rate": 4.537190828440898e-06,
+      "loss": 0.4679,
+      "step": 1406
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.345626242449101,
+      "learning_rate": 4.5309549379757724e-06,
+      "loss": 0.2619,
+      "step": 1407
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.47212229781788,
+      "learning_rate": 4.524719783479088e-06,
+      "loss": 0.4092,
+      "step": 1408
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.3189583567005485,
+      "learning_rate": 4.5184853747342926e-06,
+      "loss": 0.3102,
+      "step": 1409
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.748587805024686,
+      "learning_rate": 4.512251721523659e-06,
+      "loss": 0.4089,
+      "step": 1410
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.137858866638523,
+      "learning_rate": 4.506018833628281e-06,
+      "loss": 0.104,
+      "step": 1411
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.694425481506557,
+      "learning_rate": 4.499786720828046e-06,
+      "loss": 0.3668,
+      "step": 1412
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.934987324623524,
+      "learning_rate": 4.493555392901626e-06,
+      "loss": 0.3352,
+      "step": 1413
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.09958435430272,
+      "learning_rate": 4.487324859626465e-06,
+      "loss": 0.3388,
+      "step": 1414
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.646619577266805,
+      "learning_rate": 4.481095130778756e-06,
+      "loss": 0.5508,
+      "step": 1415
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.9639420627608435,
+      "learning_rate": 4.4748662161334335e-06,
+      "loss": 0.2911,
+      "step": 1416
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.396224153912776,
+      "learning_rate": 4.4686381254641495e-06,
+      "loss": 0.0965,
+      "step": 1417
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.283643875944138,
+      "learning_rate": 4.462410868543268e-06,
+      "loss": 0.3829,
+      "step": 1418
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.688876838583003,
+      "learning_rate": 4.456184455141843e-06,
+      "loss": 0.1055,
+      "step": 1419
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.134505921249341,
+      "learning_rate": 4.449958895029604e-06,
+      "loss": 0.1662,
+      "step": 1420
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.357120548680949,
+      "learning_rate": 4.443734197974943e-06,
+      "loss": 0.5369,
+      "step": 1421
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.264477593354643,
+      "learning_rate": 4.437510373744897e-06,
+      "loss": 0.3237,
+      "step": 1422
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.6578273420833596,
+      "learning_rate": 4.431287432105134e-06,
+      "loss": 0.4314,
+      "step": 1423
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.444778095296808,
+      "learning_rate": 4.4250653828199366e-06,
+      "loss": 0.5048,
+      "step": 1424
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.096365587109272,
+      "learning_rate": 4.418844235652188e-06,
+      "loss": 0.1961,
+      "step": 1425
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.145851994118955,
+      "learning_rate": 4.4126240003633565e-06,
+      "loss": 0.5091,
+      "step": 1426
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.890069160290513,
+      "learning_rate": 4.4064046867134755e-06,
+      "loss": 0.573,
+      "step": 1427
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.471820633417499,
+      "learning_rate": 4.400186304461136e-06,
+      "loss": 0.3534,
+      "step": 1428
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.700061575505297,
+      "learning_rate": 4.393968863363468e-06,
+      "loss": 0.4357,
+      "step": 1429
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.549335046345982,
+      "learning_rate": 4.387752373176123e-06,
+      "loss": 0.5236,
+      "step": 1430
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.805485968388716,
+      "learning_rate": 4.381536843653262e-06,
+      "loss": 0.2868,
+      "step": 1431
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.01988251960787,
+      "learning_rate": 4.375322284547536e-06,
+      "loss": 0.5035,
+      "step": 1432
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.785660213357935,
+      "learning_rate": 4.369108705610077e-06,
+      "loss": 0.2794,
+      "step": 1433
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8903109542957437,
+      "learning_rate": 4.362896116590475e-06,
+      "loss": 0.3184,
+      "step": 1434
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.9537539325999314,
+      "learning_rate": 4.356684527236773e-06,
+      "loss": 0.2546,
+      "step": 1435
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.628936578239452,
+      "learning_rate": 4.3504739472954355e-06,
+      "loss": 0.3653,
+      "step": 1436
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.881018316430561,
+      "learning_rate": 4.344264386511351e-06,
+      "loss": 0.1687,
+      "step": 1437
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.9588487703461333,
+      "learning_rate": 4.3380558546278075e-06,
+      "loss": 0.3115,
+      "step": 1438
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.701950011501446,
+      "learning_rate": 4.331848361386478e-06,
+      "loss": 0.5479,
+      "step": 1439
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.710195317053987,
+      "learning_rate": 4.325641916527405e-06,
+      "loss": 0.2061,
+      "step": 1440
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.741740338842962,
+      "learning_rate": 4.319436529788985e-06,
+      "loss": 0.3259,
+      "step": 1441
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.145635596003466,
+      "learning_rate": 4.313232210907959e-06,
+      "loss": 0.2426,
+      "step": 1442
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.446204265135104,
+      "learning_rate": 4.3070289696193886e-06,
+      "loss": 0.3982,
+      "step": 1443
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.425547733704053,
+      "learning_rate": 4.3008268156566444e-06,
+      "loss": 0.5111,
+      "step": 1444
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.33781990925659,
+      "learning_rate": 4.294625758751392e-06,
+      "loss": 0.3787,
+      "step": 1445
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.103452942258401,
+      "learning_rate": 4.2884258086335755e-06,
+      "loss": 0.3662,
+      "step": 1446
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.0655623814372195,
+      "learning_rate": 4.282226975031402e-06,
+      "loss": 0.4652,
+      "step": 1447
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.766225343491909,
+      "learning_rate": 4.276029267671327e-06,
+      "loss": 0.525,
+      "step": 1448
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.472322802063309,
+      "learning_rate": 4.269832696278038e-06,
+      "loss": 0.131,
+      "step": 1449
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.2210082573560435,
+      "learning_rate": 4.2636372705744425e-06,
+      "loss": 0.381,
+      "step": 1450
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.068230476015941,
+      "learning_rate": 4.257443000281648e-06,
+      "loss": 0.1899,
+      "step": 1451
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.062133989529346,
+      "learning_rate": 4.25124989511895e-06,
+      "loss": 0.2387,
+      "step": 1452
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.253395716948605,
+      "learning_rate": 4.245057964803815e-06,
+      "loss": 0.3772,
+      "step": 1453
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.066243021091036,
+      "learning_rate": 4.238867219051868e-06,
+      "loss": 0.2109,
+      "step": 1454
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.31968234675491,
+      "learning_rate": 4.232677667576874e-06,
+      "loss": 0.3079,
+      "step": 1455
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.9903737266786115,
+      "learning_rate": 4.226489320090723e-06,
+      "loss": 0.2489,
+      "step": 1456
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.785627309087942,
+      "learning_rate": 4.220302186303416e-06,
+      "loss": 0.2637,
+      "step": 1457
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.6660891871944035,
+      "learning_rate": 4.214116275923051e-06,
+      "loss": 0.3304,
+      "step": 1458
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.552726695058103,
+      "learning_rate": 4.2079315986558056e-06,
+      "loss": 0.2224,
+      "step": 1459
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.255849264152336,
+      "learning_rate": 4.201748164205925e-06,
+      "loss": 0.3666,
+      "step": 1460
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.921409238431859,
+      "learning_rate": 4.195565982275701e-06,
+      "loss": 0.3434,
+      "step": 1461
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1791514628396063,
+      "learning_rate": 4.1893850625654626e-06,
+      "loss": 0.4433,
+      "step": 1462
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.688767849030837,
+      "learning_rate": 4.183205414773555e-06,
+      "loss": 0.1771,
+      "step": 1463
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.274330270883109,
+      "learning_rate": 4.17702704859633e-06,
+      "loss": 0.3412,
+      "step": 1464
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.276379403881767,
+      "learning_rate": 4.1708499737281305e-06,
+      "loss": 0.2252,
+      "step": 1465
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.4275714990157535,
+      "learning_rate": 4.1646741998612676e-06,
+      "loss": 0.3196,
+      "step": 1466
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.805569818336066,
+      "learning_rate": 4.1584997366860145e-06,
+      "loss": 0.3324,
+      "step": 1467
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.705265396538318,
+      "learning_rate": 4.1523265938905885e-06,
+      "loss": 0.213,
+      "step": 1468
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.1938800793681095,
+      "learning_rate": 4.146154781161133e-06,
+      "loss": 0.4945,
+      "step": 1469
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0377205928657487,
+      "learning_rate": 4.1399843081817085e-06,
+      "loss": 0.0463,
+      "step": 1470
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3144372452006325,
+      "learning_rate": 4.133815184634269e-06,
+      "loss": 0.393,
+      "step": 1471
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.017508208082044,
+      "learning_rate": 4.12764742019865e-06,
+      "loss": 0.2596,
+      "step": 1472
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7919045395811386,
+      "learning_rate": 4.121481024552561e-06,
+      "loss": 0.541,
+      "step": 1473
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.95882206036411,
+      "learning_rate": 4.115316007371557e-06,
+      "loss": 0.4398,
+      "step": 1474
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.66973771498853,
+      "learning_rate": 4.109152378329036e-06,
+      "loss": 0.3763,
+      "step": 1475
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.847762996262314,
+      "learning_rate": 4.1029901470962105e-06,
+      "loss": 0.2938,
+      "step": 1476
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.179071502240395,
+      "learning_rate": 4.0968293233421065e-06,
+      "loss": 0.1904,
+      "step": 1477
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.9752752193041143,
+      "learning_rate": 4.090669916733539e-06,
+      "loss": 0.1964,
+      "step": 1478
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.6355064008210185,
+      "learning_rate": 4.0845119369350995e-06,
+      "loss": 0.3628,
+      "step": 1479
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.509494639492551,
+      "learning_rate": 4.078355393609139e-06,
+      "loss": 0.3887,
+      "step": 1480
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.56076210034677,
+      "learning_rate": 4.0722002964157585e-06,
+      "loss": 0.3614,
+      "step": 1481
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.470680502720224,
+      "learning_rate": 4.066046655012786e-06,
+      "loss": 0.2317,
+      "step": 1482
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.997678909046692,
+      "learning_rate": 4.059894479055767e-06,
+      "loss": 0.3313,
+      "step": 1483
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.0015515719649355,
+      "learning_rate": 4.053743778197951e-06,
+      "loss": 0.1241,
+      "step": 1484
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.0533753005570246,
+      "learning_rate": 4.047594562090264e-06,
+      "loss": 0.2029,
+      "step": 1485
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.129446225514946,
+      "learning_rate": 4.041446840381309e-06,
+      "loss": 0.3577,
+      "step": 1486
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.693087869441793,
+      "learning_rate": 4.0353006227173455e-06,
+      "loss": 0.2564,
+      "step": 1487
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.2698696515898,
+      "learning_rate": 4.029155918742268e-06,
+      "loss": 0.3401,
+      "step": 1488
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.4125741124273725,
+      "learning_rate": 4.023012738097598e-06,
+      "loss": 0.4311,
+      "step": 1489
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.510306762824228,
+      "learning_rate": 4.016871090422471e-06,
+      "loss": 0.2413,
+      "step": 1490
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.219252553428604,
+      "learning_rate": 4.010730985353609e-06,
+      "loss": 0.3228,
+      "step": 1491
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.9801080840084544,
+      "learning_rate": 4.00459243252532e-06,
+      "loss": 0.304,
+      "step": 1492
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.5396212091755626,
+      "learning_rate": 3.998455441569473e-06,
+      "loss": 0.4879,
+      "step": 1493
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.7772318325190013,
+      "learning_rate": 3.992320022115492e-06,
+      "loss": 0.4769,
+      "step": 1494
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.0039394809010185,
+      "learning_rate": 3.986186183790323e-06,
+      "loss": 0.3172,
+      "step": 1495
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.204364028635148,
+      "learning_rate": 3.980053936218444e-06,
+      "loss": 0.247,
+      "step": 1496
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.317121569144144,
+      "learning_rate": 3.973923289021829e-06,
+      "loss": 0.3766,
+      "step": 1497
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.998904443826016,
+      "learning_rate": 3.9677942518199465e-06,
+      "loss": 0.441,
+      "step": 1498
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.433776021513012,
+      "learning_rate": 3.961666834229734e-06,
+      "loss": 0.3183,
+      "step": 1499
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.369736004956909,
+      "learning_rate": 3.955541045865591e-06,
+      "loss": 0.3855,
+      "step": 1500
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9916271547290005,
+      "learning_rate": 3.949416896339357e-06,
+      "loss": 0.3071,
+      "step": 1501
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.512835533817584,
+      "learning_rate": 3.943294395260305e-06,
+      "loss": 0.1732,
+      "step": 1502
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.328324861730488,
+      "learning_rate": 3.937173552235117e-06,
+      "loss": 0.3571,
+      "step": 1503
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.030903577474148,
+      "learning_rate": 3.931054376867877e-06,
+      "loss": 0.4652,
+      "step": 1504
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.344547597819089,
+      "learning_rate": 3.924936878760048e-06,
+      "loss": 0.3512,
+      "step": 1505
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.299986753165218,
+      "learning_rate": 3.918821067510464e-06,
+      "loss": 0.1322,
+      "step": 1506
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.356397527437635,
+      "learning_rate": 3.912706952715314e-06,
+      "loss": 0.2978,
+      "step": 1507
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.703673582358423,
+      "learning_rate": 3.906594543968122e-06,
+      "loss": 0.2282,
+      "step": 1508
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.503063276257467,
+      "learning_rate": 3.900483850859735e-06,
+      "loss": 0.1066,
+      "step": 1509
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.134104636570549,
+      "learning_rate": 3.89437488297831e-06,
+      "loss": 0.4441,
+      "step": 1510
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.756484523844902,
+      "learning_rate": 3.888267649909296e-06,
+      "loss": 0.3926,
+      "step": 1511
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.504780524924184,
+      "learning_rate": 3.882162161235421e-06,
+      "loss": 0.1757,
+      "step": 1512
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.126572009653953,
+      "learning_rate": 3.876058426536674e-06,
+      "loss": 0.606,
+      "step": 1513
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.663230747358417,
+      "learning_rate": 3.869956455390295e-06,
+      "loss": 0.4179,
+      "step": 1514
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.181315471156965,
+      "learning_rate": 3.86385625737075e-06,
+      "loss": 0.289,
+      "step": 1515
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.201227898458757,
+      "learning_rate": 3.857757842049732e-06,
+      "loss": 0.1346,
+      "step": 1516
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.926969503211117,
+      "learning_rate": 3.851661218996129e-06,
+      "loss": 0.4248,
+      "step": 1517
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.989317637241524,
+      "learning_rate": 3.845566397776022e-06,
+      "loss": 0.3161,
+      "step": 1518
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.874726914919087,
+      "learning_rate": 3.839473387952662e-06,
+      "loss": 0.2798,
+      "step": 1519
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.174614508967492,
+      "learning_rate": 3.833382199086459e-06,
+      "loss": 0.5569,
+      "step": 1520
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.4966931392537384,
+      "learning_rate": 3.827292840734966e-06,
+      "loss": 0.2482,
+      "step": 1521
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.9320863315443138,
+      "learning_rate": 3.821205322452863e-06,
+      "loss": 0.2117,
+      "step": 1522
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.440310700744331,
+      "learning_rate": 3.815119653791943e-06,
+      "loss": 0.2702,
+      "step": 1523
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.7478768889650524,
+      "learning_rate": 3.8090358443010993e-06,
+      "loss": 0.3356,
+      "step": 1524
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.008056794967351,
+      "learning_rate": 3.802953903526301e-06,
+      "loss": 0.1924,
+      "step": 1525
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.080002486180463,
+      "learning_rate": 3.796873841010591e-06,
+      "loss": 0.2048,
+      "step": 1526
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.217585136256839,
+      "learning_rate": 3.7907956662940644e-06,
+      "loss": 0.0916,
+      "step": 1527
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.8086105307397395,
+      "learning_rate": 3.784719388913853e-06,
+      "loss": 0.2909,
+      "step": 1528
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.5725664063240155,
+      "learning_rate": 3.778645018404112e-06,
+      "loss": 0.4978,
+      "step": 1529
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.3115117944969565,
+      "learning_rate": 3.7725725642960047e-06,
+      "loss": 0.4703,
+      "step": 1530
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.8455271216801306,
+      "learning_rate": 3.7665020361176872e-06,
+      "loss": 0.4041,
+      "step": 1531
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.957077460576616,
+      "learning_rate": 3.760433443394294e-06,
+      "loss": 0.4267,
+      "step": 1532
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.219325430622278,
+      "learning_rate": 3.754366795647923e-06,
+      "loss": 0.3815,
+      "step": 1533
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.767149721252658,
+      "learning_rate": 3.748302102397618e-06,
+      "loss": 0.4733,
+      "step": 1534
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.5153750892227036,
+      "learning_rate": 3.7422393731593586e-06,
+      "loss": 0.1929,
+      "step": 1535
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.182170655134872,
+      "learning_rate": 3.7361786174460414e-06,
+      "loss": 0.2514,
+      "step": 1536
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.401273986878207,
+      "learning_rate": 3.730119844767468e-06,
+      "loss": 0.1653,
+      "step": 1537
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.9891550305249703,
+      "learning_rate": 3.7240630646303262e-06,
+      "loss": 0.4162,
+      "step": 1538
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.475869932565687,
+      "learning_rate": 3.718008286538179e-06,
+      "loss": 0.3706,
+      "step": 1539
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.624030179882473,
+      "learning_rate": 3.711955519991447e-06,
+      "loss": 0.2352,
+      "step": 1540
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.4289749422225055,
+      "learning_rate": 3.705904774487396e-06,
+      "loss": 0.239,
+      "step": 1541
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.672470473292984,
+      "learning_rate": 3.6998560595201188e-06,
+      "loss": 0.1951,
+      "step": 1542
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.227877497623781,
+      "learning_rate": 3.6938093845805257e-06,
+      "loss": 0.4455,
+      "step": 1543
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.585073856751496,
+      "learning_rate": 3.687764759156318e-06,
+      "loss": 0.2735,
+      "step": 1544
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.731607106926894,
+      "learning_rate": 3.68172219273199e-06,
+      "loss": 0.3307,
+      "step": 1545
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.404167408537416,
+      "learning_rate": 3.675681694788801e-06,
+      "loss": 0.3731,
+      "step": 1546
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.515198173032717,
+      "learning_rate": 3.669643274804765e-06,
+      "loss": 0.188,
+      "step": 1547
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.495410609475275,
+      "learning_rate": 3.6636069422546363e-06,
+      "loss": 0.2857,
+      "step": 1548
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.37388058164263,
+      "learning_rate": 3.6575727066098936e-06,
+      "loss": 0.2787,
+      "step": 1549
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.324020488687119,
+      "learning_rate": 3.6515405773387257e-06,
+      "loss": 0.4444,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.0257406253270815,
+      "learning_rate": 3.645510563906014e-06,
+      "loss": 0.258,
+      "step": 1551
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.370561128774498,
+      "learning_rate": 3.639482675773324e-06,
+      "loss": 0.1959,
+      "step": 1552
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.2493296603566115,
+      "learning_rate": 3.633456922398887e-06,
+      "loss": 0.3543,
+      "step": 1553
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.335576674837506,
+      "learning_rate": 3.627433313237576e-06,
+      "loss": 0.5657,
+      "step": 1554
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.062024130226634,
+      "learning_rate": 3.621411857740908e-06,
+      "loss": 0.1434,
+      "step": 1555
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.979363096325817,
+      "learning_rate": 3.6153925653570186e-06,
+      "loss": 0.4033,
+      "step": 1556
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.71101568663758,
+      "learning_rate": 3.6093754455306495e-06,
+      "loss": 0.1745,
+      "step": 1557
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.8585032404200943,
+      "learning_rate": 3.603360507703133e-06,
+      "loss": 0.4782,
+      "step": 1558
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.784458526359172,
+      "learning_rate": 3.597347761312377e-06,
+      "loss": 0.5261,
+      "step": 1559
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.843226432822785,
+      "learning_rate": 3.5913372157928515e-06,
+      "loss": 0.3679,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.966301117391833,
+      "learning_rate": 3.585328880575574e-06,
+      "loss": 0.3992,
+      "step": 1561
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.367193452076943,
+      "learning_rate": 3.5793227650880928e-06,
+      "loss": 0.5085,
+      "step": 1562
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.586351823066316,
+      "learning_rate": 3.573318878754475e-06,
+      "loss": 0.4149,
+      "step": 1563
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.837885770381356,
+      "learning_rate": 3.5673172309952846e-06,
+      "loss": 0.3667,
+      "step": 1564
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.695094278265337,
+      "learning_rate": 3.5613178312275796e-06,
+      "loss": 0.2563,
+      "step": 1565
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.603957498733291,
+      "learning_rate": 3.555320688864889e-06,
+      "loss": 0.3894,
+      "step": 1566
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.365398294496251,
+      "learning_rate": 3.549325813317197e-06,
+      "loss": 0.3296,
+      "step": 1567
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.719689737508214,
+      "learning_rate": 3.5433332139909342e-06,
+      "loss": 0.2342,
+      "step": 1568
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.083072422306283,
+      "learning_rate": 3.5373429002889583e-06,
+      "loss": 0.4403,
+      "step": 1569
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.015949665717416,
+      "learning_rate": 3.531354881610539e-06,
+      "loss": 0.271,
+      "step": 1570
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.561547003390101,
+      "learning_rate": 3.525369167351349e-06,
+      "loss": 0.3076,
+      "step": 1571
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.618005155853049,
+      "learning_rate": 3.519385766903442e-06,
+      "loss": 0.2894,
+      "step": 1572
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.0852155842053905,
+      "learning_rate": 3.5134046896552436e-06,
+      "loss": 0.3645,
+      "step": 1573
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.439663902373668,
+      "learning_rate": 3.507425944991529e-06,
+      "loss": 0.4083,
+      "step": 1574
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.057067586030128,
+      "learning_rate": 3.501449542293418e-06,
+      "loss": 0.4067,
+      "step": 1575
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1617977677922733,
+      "learning_rate": 3.495475490938355e-06,
+      "loss": 0.1281,
+      "step": 1576
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.811102471895702,
+      "learning_rate": 3.4895038003000953e-06,
+      "loss": 0.5908,
+      "step": 1577
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.033044332940408,
+      "learning_rate": 3.483534479748688e-06,
+      "loss": 0.4497,
+      "step": 1578
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.015058932877953,
+      "learning_rate": 3.477567538650466e-06,
+      "loss": 0.4729,
+      "step": 1579
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.4444982631762167,
+      "learning_rate": 3.4716029863680256e-06,
+      "loss": 0.3643,
+      "step": 1580
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.638305557142188,
+      "learning_rate": 3.4656408322602168e-06,
+      "loss": 0.2885,
+      "step": 1581
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.255174305001011,
+      "learning_rate": 3.4596810856821304e-06,
+      "loss": 0.3581,
+      "step": 1582
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.523418614165512,
+      "learning_rate": 3.45372375598507e-06,
+      "loss": 0.3931,
+      "step": 1583
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.9112513072493846,
+      "learning_rate": 3.447768852516554e-06,
+      "loss": 0.3582,
+      "step": 1584
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.503925033573366,
+      "learning_rate": 3.4418163846202945e-06,
+      "loss": 0.2074,
+      "step": 1585
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.17076176202868,
+      "learning_rate": 3.4358663616361775e-06,
+      "loss": 0.4885,
+      "step": 1586
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.875859502345279,
+      "learning_rate": 3.429918792900257e-06,
+      "loss": 0.4738,
+      "step": 1587
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.124966067519853,
+      "learning_rate": 3.4239736877447327e-06,
+      "loss": 0.327,
+      "step": 1588
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.089392474409888,
+      "learning_rate": 3.4180310554979413e-06,
+      "loss": 0.3903,
+      "step": 1589
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.857269282488471,
+      "learning_rate": 3.4120909054843375e-06,
+      "loss": 0.3071,
+      "step": 1590
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.0690463549392355,
+      "learning_rate": 3.406153247024483e-06,
+      "loss": 0.2055,
+      "step": 1591
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4169111638971876,
+      "learning_rate": 3.400218089435029e-06,
+      "loss": 0.3195,
+      "step": 1592
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.506896436639098,
+      "learning_rate": 3.3942854420287006e-06,
+      "loss": 0.6153,
+      "step": 1593
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.323212175668684,
+      "learning_rate": 3.3883553141142884e-06,
+      "loss": 0.4096,
+      "step": 1594
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.571988379752501,
+      "learning_rate": 3.3824277149966265e-06,
+      "loss": 0.2163,
+      "step": 1595
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.925491898191029,
+      "learning_rate": 3.3765026539765832e-06,
+      "loss": 0.3347,
+      "step": 1596
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.810934803829099,
+      "learning_rate": 3.3705801403510417e-06,
+      "loss": 0.2624,
+      "step": 1597
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.13045526203993,
+      "learning_rate": 3.3646601834128924e-06,
+      "loss": 0.4272,
+      "step": 1598
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.679400999094655,
+      "learning_rate": 3.3587427924510086e-06,
+      "loss": 0.2084,
+      "step": 1599
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.671096588916482,
+      "learning_rate": 3.352827976750242e-06,
+      "loss": 0.6803,
+      "step": 1600
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.6355593379831377,
+      "learning_rate": 3.3469157455914013e-06,
+      "loss": 0.3139,
+      "step": 1601
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4716879424544707,
+      "learning_rate": 3.3410061082512422e-06,
+      "loss": 0.3129,
+      "step": 1602
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.275693478616472,
+      "learning_rate": 3.335099074002445e-06,
+      "loss": 0.2482,
+      "step": 1603
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.463750058107956,
+      "learning_rate": 3.32919465211361e-06,
+      "loss": 0.2431,
+      "step": 1604
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.420931104299418,
+      "learning_rate": 3.323292851849238e-06,
+      "loss": 0.32,
+      "step": 1605
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5310184393281894,
+      "learning_rate": 3.3173936824697174e-06,
+      "loss": 0.2225,
+      "step": 1606
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.898892610016629,
+      "learning_rate": 3.3114971532313058e-06,
+      "loss": 0.4428,
+      "step": 1607
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.6983351425803095,
+      "learning_rate": 3.3056032733861188e-06,
+      "loss": 0.4583,
+      "step": 1608
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2872913356956315,
+      "learning_rate": 3.2997120521821168e-06,
+      "loss": 0.4078,
+      "step": 1609
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.9667901873913753,
+      "learning_rate": 3.293823498863087e-06,
+      "loss": 0.1196,
+      "step": 1610
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.444362037455661,
+      "learning_rate": 3.28793762266863e-06,
+      "loss": 0.1999,
+      "step": 1611
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.1323845180596965,
+      "learning_rate": 3.2820544328341485e-06,
+      "loss": 0.2043,
+      "step": 1612
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.669166522622902,
+      "learning_rate": 3.2761739385908264e-06,
+      "loss": 0.1308,
+      "step": 1613
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1742174942176495,
+      "learning_rate": 3.2702961491656197e-06,
+      "loss": 0.0665,
+      "step": 1614
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9569157853804575,
+      "learning_rate": 3.264421073781241e-06,
+      "loss": 0.3077,
+      "step": 1615
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.188901513956453,
+      "learning_rate": 3.258548721656144e-06,
+      "loss": 0.1397,
+      "step": 1616
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.478623580194558,
+      "learning_rate": 3.252679102004509e-06,
+      "loss": 0.407,
+      "step": 1617
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.209800340962832,
+      "learning_rate": 3.2468122240362287e-06,
+      "loss": 0.522,
+      "step": 1618
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.611061350039146,
+      "learning_rate": 3.240948096956894e-06,
+      "loss": 0.3241,
+      "step": 1619
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.851662013093061,
+      "learning_rate": 3.2350867299677802e-06,
+      "loss": 0.4047,
+      "step": 1620
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.4414185335000855,
+      "learning_rate": 3.2292281322658315e-06,
+      "loss": 0.6093,
+      "step": 1621
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.256115339901157,
+      "learning_rate": 3.223372313043647e-06,
+      "loss": 0.4213,
+      "step": 1622
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.060735569137316,
+      "learning_rate": 3.2175192814894627e-06,
+      "loss": 0.323,
+      "step": 1623
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2550775250574673,
+      "learning_rate": 3.211669046787147e-06,
+      "loss": 0.1256,
+      "step": 1624
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.7634389316841625,
+      "learning_rate": 3.205821618116175e-06,
+      "loss": 0.3866,
+      "step": 1625
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.887011889531534,
+      "learning_rate": 3.1999770046516198e-06,
+      "loss": 0.2934,
+      "step": 1626
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.236055281555687,
+      "learning_rate": 3.194135215564139e-06,
+      "loss": 0.2454,
+      "step": 1627
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.298067101862836,
+      "learning_rate": 3.188296260019956e-06,
+      "loss": 0.4734,
+      "step": 1628
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.46461080053218,
+      "learning_rate": 3.1824601471808504e-06,
+      "loss": 0.1847,
+      "step": 1629
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6199644001707134,
+      "learning_rate": 3.1766268862041406e-06,
+      "loss": 0.1652,
+      "step": 1630
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.916631814306934,
+      "learning_rate": 3.17079648624267e-06,
+      "loss": 0.1468,
+      "step": 1631
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5932787909655586,
+      "learning_rate": 3.164968956444791e-06,
+      "loss": 0.1869,
+      "step": 1632
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.015353888738545,
+      "learning_rate": 3.159144305954356e-06,
+      "loss": 0.3621,
+      "step": 1633
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.9500143540112194,
+      "learning_rate": 3.1533225439106965e-06,
+      "loss": 0.1496,
+      "step": 1634
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.794321715793806,
+      "learning_rate": 3.1475036794486147e-06,
+      "loss": 0.3855,
+      "step": 1635
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.440922482063624,
+      "learning_rate": 3.141687721698363e-06,
+      "loss": 0.1055,
+      "step": 1636
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.6663959358882092,
+      "learning_rate": 3.1358746797856367e-06,
+      "loss": 0.422,
+      "step": 1637
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.932315484073457,
+      "learning_rate": 3.130064562831553e-06,
+      "loss": 0.1827,
+      "step": 1638
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.059351081158459,
+      "learning_rate": 3.1242573799526397e-06,
+      "loss": 0.3927,
+      "step": 1639
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.642735050617873,
+      "learning_rate": 3.118453140260823e-06,
+      "loss": 0.2246,
+      "step": 1640
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.914140625367253,
+      "learning_rate": 3.1126518528634096e-06,
+      "loss": 0.5902,
+      "step": 1641
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.27481680069469,
+      "learning_rate": 3.106853526863073e-06,
+      "loss": 0.3063,
+      "step": 1642
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.9534193599306784,
+      "learning_rate": 3.1010581713578403e-06,
+      "loss": 0.2117,
+      "step": 1643
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.403822031964082,
+      "learning_rate": 3.0952657954410792e-06,
+      "loss": 0.3243,
+      "step": 1644
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.17237295544702,
+      "learning_rate": 3.0894764082014805e-06,
+      "loss": 0.0755,
+      "step": 1645
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.143751849182849,
+      "learning_rate": 3.0836900187230475e-06,
+      "loss": 0.3995,
+      "step": 1646
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.080162008984293,
+      "learning_rate": 3.0779066360850774e-06,
+      "loss": 0.4505,
+      "step": 1647
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.31575469159342,
+      "learning_rate": 3.072126269362151e-06,
+      "loss": 0.3163,
+      "step": 1648
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.335292171405565,
+      "learning_rate": 3.0663489276241166e-06,
+      "loss": 0.5304,
+      "step": 1649
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.486175151787324,
+      "learning_rate": 3.0605746199360755e-06,
+      "loss": 0.359,
+      "step": 1650
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.27442481961856,
+      "learning_rate": 3.0548033553583707e-06,
+      "loss": 0.4114,
+      "step": 1651
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.766822274738467,
+      "learning_rate": 3.049035142946565e-06,
+      "loss": 0.2047,
+      "step": 1652
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.291066566094241,
+      "learning_rate": 3.0432699917514375e-06,
+      "loss": 0.1798,
+      "step": 1653
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.140959022245619,
+      "learning_rate": 3.0375079108189613e-06,
+      "loss": 0.2744,
+      "step": 1654
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.219552671471157,
+      "learning_rate": 3.0317489091902936e-06,
+      "loss": 0.2772,
+      "step": 1655
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2133288245929497,
+      "learning_rate": 3.0259929959017585e-06,
+      "loss": 0.0624,
+      "step": 1656
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.3653314556421225,
+      "learning_rate": 3.0202401799848347e-06,
+      "loss": 0.49,
+      "step": 1657
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2388325557812734,
+      "learning_rate": 3.0144904704661413e-06,
+      "loss": 0.1567,
+      "step": 1658
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.707238816625475,
+      "learning_rate": 3.0087438763674226e-06,
+      "loss": 0.2937,
+      "step": 1659
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.53005153553539,
+      "learning_rate": 3.003000406705535e-06,
+      "loss": 0.1577,
+      "step": 1660
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.058192745606313,
+      "learning_rate": 2.9972600704924348e-06,
+      "loss": 0.3803,
+      "step": 1661
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.365488213512427,
+      "learning_rate": 2.991522876735154e-06,
+      "loss": 0.323,
+      "step": 1662
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.502784046793897,
+      "learning_rate": 2.985788834435802e-06,
+      "loss": 0.6248,
+      "step": 1663
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.360244546359318,
+      "learning_rate": 2.9800579525915394e-06,
+      "loss": 0.5125,
+      "step": 1664
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.882626192669339,
+      "learning_rate": 2.9743302401945707e-06,
+      "loss": 0.4606,
+      "step": 1665
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.952718683598004,
+      "learning_rate": 2.9686057062321226e-06,
+      "loss": 0.2694,
+      "step": 1666
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2532355007595117,
+      "learning_rate": 2.9628843596864386e-06,
+      "loss": 0.3237,
+      "step": 1667
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.525233809899968,
+      "learning_rate": 2.95716620953476e-06,
+      "loss": 0.4447,
+      "step": 1668
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.30406452613882,
+      "learning_rate": 2.9514512647493123e-06,
+      "loss": 0.1828,
+      "step": 1669
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.100129586502264,
+      "learning_rate": 2.9457395342972904e-06,
+      "loss": 0.2453,
+      "step": 1670
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.798501317618989,
+      "learning_rate": 2.940031027140848e-06,
+      "loss": 0.2783,
+      "step": 1671
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.239549366716743,
+      "learning_rate": 2.9343257522370784e-06,
+      "loss": 0.33,
+      "step": 1672
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.849134728746657,
+      "learning_rate": 2.928623718538006e-06,
+      "loss": 0.1592,
+      "step": 1673
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.467559854123593,
+      "learning_rate": 2.9229249349905686e-06,
+      "loss": 0.5214,
+      "step": 1674
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.066321792108032,
+      "learning_rate": 2.917229410536604e-06,
+      "loss": 0.5847,
+      "step": 1675
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.7684968096427176,
+      "learning_rate": 2.9115371541128344e-06,
+      "loss": 0.3402,
+      "step": 1676
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.770360438291747,
+      "learning_rate": 2.905848174650856e-06,
+      "loss": 0.1709,
+      "step": 1677
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0763589893400987,
+      "learning_rate": 2.900162481077126e-06,
+      "loss": 0.3201,
+      "step": 1678
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.6197372876556435,
+      "learning_rate": 2.8944800823129413e-06,
+      "loss": 0.2598,
+      "step": 1679
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.290460315443914,
+      "learning_rate": 2.8888009872744332e-06,
+      "loss": 0.5132,
+      "step": 1680
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.8173794768215568,
+      "learning_rate": 2.883125204872542e-06,
+      "loss": 0.2549,
+      "step": 1681
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.439447427484574,
+      "learning_rate": 2.8774527440130173e-06,
+      "loss": 0.2146,
+      "step": 1682
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.222232351099214,
+      "learning_rate": 2.8717836135963937e-06,
+      "loss": 0.5238,
+      "step": 1683
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.582334617398952,
+      "learning_rate": 2.866117822517982e-06,
+      "loss": 0.3758,
+      "step": 1684
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.959571024791742,
+      "learning_rate": 2.860455379667851e-06,
+      "loss": 0.156,
+      "step": 1685
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.931363949721377,
+      "learning_rate": 2.8547962939308187e-06,
+      "loss": 0.4552,
+      "step": 1686
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.3040823343772985,
+      "learning_rate": 2.8491405741864323e-06,
+      "loss": 0.3618,
+      "step": 1687
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.607659758841999,
+      "learning_rate": 2.8434882293089607e-06,
+      "loss": 0.3063,
+      "step": 1688
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.702741810993404,
+      "learning_rate": 2.837839268167373e-06,
+      "loss": 0.2151,
+      "step": 1689
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.211298705112521,
+      "learning_rate": 2.8321936996253368e-06,
+      "loss": 0.3437,
+      "step": 1690
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.345823636413663,
+      "learning_rate": 2.8265515325411856e-06,
+      "loss": 0.4581,
+      "step": 1691
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.443663189487989,
+      "learning_rate": 2.8209127757679246e-06,
+      "loss": 0.1207,
+      "step": 1692
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.209942661911642,
+      "learning_rate": 2.8152774381532033e-06,
+      "loss": 0.4597,
+      "step": 1693
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.399312147279553,
+      "learning_rate": 2.8096455285393094e-06,
+      "loss": 0.2113,
+      "step": 1694
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.174749198687062,
+      "learning_rate": 2.804017055763149e-06,
+      "loss": 0.4833,
+      "step": 1695
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.1444824944970495,
+      "learning_rate": 2.798392028656237e-06,
+      "loss": 0.2804,
+      "step": 1696
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.044934464655802,
+      "learning_rate": 2.792770456044683e-06,
+      "loss": 0.4196,
+      "step": 1697
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.341011730415659,
+      "learning_rate": 2.787152346749173e-06,
+      "loss": 0.27,
+      "step": 1698
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.1951400779268866,
+      "learning_rate": 2.7815377095849612e-06,
+      "loss": 0.1069,
+      "step": 1699
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.924086035872311,
+      "learning_rate": 2.775926553361855e-06,
+      "loss": 0.4745,
+      "step": 1700
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.3989896193178994,
+      "learning_rate": 2.7703188868841936e-06,
+      "loss": 0.4895,
+      "step": 1701
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.9757493138429996,
+      "learning_rate": 2.7647147189508485e-06,
+      "loss": 0.1893,
+      "step": 1702
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.039018022787459,
+      "learning_rate": 2.7591140583551956e-06,
+      "loss": 0.315,
+      "step": 1703
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6680430669769373,
+      "learning_rate": 2.7535169138851124e-06,
+      "loss": 0.1205,
+      "step": 1704
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.394568294103908,
+      "learning_rate": 2.7479232943229567e-06,
+      "loss": 0.1386,
+      "step": 1705
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.920795596494274,
+      "learning_rate": 2.7423332084455543e-06,
+      "loss": 0.1672,
+      "step": 1706
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.109653066753222,
+      "learning_rate": 2.7367466650241868e-06,
+      "loss": 0.586,
+      "step": 1707
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.636956032652642,
+      "learning_rate": 2.731163672824583e-06,
+      "loss": 0.2288,
+      "step": 1708
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.808929072946402,
+      "learning_rate": 2.7255842406068917e-06,
+      "loss": 0.521,
+      "step": 1709
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.075018120655451,
+      "learning_rate": 2.720008377125682e-06,
+      "loss": 0.363,
+      "step": 1710
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.459381989696827,
+      "learning_rate": 2.7144360911299158e-06,
+      "loss": 0.2035,
+      "step": 1711
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.3473002345601435,
+      "learning_rate": 2.708867391362948e-06,
+      "loss": 0.4227,
+      "step": 1712
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.9174475662573927,
+      "learning_rate": 2.703302286562506e-06,
+      "loss": 0.1723,
+      "step": 1713
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.7138861041297355,
+      "learning_rate": 2.697740785460675e-06,
+      "loss": 0.3219,
+      "step": 1714
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.143243281149177,
+      "learning_rate": 2.692182896783885e-06,
+      "loss": 0.6009,
+      "step": 1715
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.554380657981959,
+      "learning_rate": 2.686628629252899e-06,
+      "loss": 0.2703,
+      "step": 1716
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.018761179383061,
+      "learning_rate": 2.681077991582797e-06,
+      "loss": 0.3545,
+      "step": 1717
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.542355393780447,
+      "learning_rate": 2.6755309924829657e-06,
+      "loss": 0.1874,
+      "step": 1718
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.514922915379382,
+      "learning_rate": 2.6699876406570823e-06,
+      "loss": 0.1727,
+      "step": 1719
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0740736807638944,
+      "learning_rate": 2.664447944803097e-06,
+      "loss": 0.1471,
+      "step": 1720
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.20581139373578,
+      "learning_rate": 2.6589119136132273e-06,
+      "loss": 0.6312,
+      "step": 1721
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.256503390802347,
+      "learning_rate": 2.6533795557739407e-06,
+      "loss": 0.4009,
+      "step": 1722
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.387740355612719,
+      "learning_rate": 2.6478508799659393e-06,
+      "loss": 0.3199,
+      "step": 1723
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.968641510398305,
+      "learning_rate": 2.6423258948641483e-06,
+      "loss": 0.2967,
+      "step": 1724
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7168503542752256,
+      "learning_rate": 2.636804609137703e-06,
+      "loss": 0.322,
+      "step": 1725
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.447583109886695,
+      "learning_rate": 2.6312870314499335e-06,
+      "loss": 0.3068,
+      "step": 1726
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.9977352996698166,
+      "learning_rate": 2.62577317045835e-06,
+      "loss": 0.5861,
+      "step": 1727
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.047519665534339,
+      "learning_rate": 2.6202630348146323e-06,
+      "loss": 0.4104,
+      "step": 1728
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.99157274852523,
+      "learning_rate": 2.614756633164618e-06,
+      "loss": 0.2417,
+      "step": 1729
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.056993343008954,
+      "learning_rate": 2.609253974148278e-06,
+      "loss": 0.2285,
+      "step": 1730
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.676483893783997,
+      "learning_rate": 2.603755066399718e-06,
+      "loss": 0.4048,
+      "step": 1731
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.302491960324539,
+      "learning_rate": 2.5982599185471535e-06,
+      "loss": 0.4063,
+      "step": 1732
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.8631562321598905,
+      "learning_rate": 2.5927685392129033e-06,
+      "loss": 0.2333,
+      "step": 1733
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.929943829504728,
+      "learning_rate": 2.5872809370133704e-06,
+      "loss": 0.2795,
+      "step": 1734
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.844718932304175,
+      "learning_rate": 2.5817971205590343e-06,
+      "loss": 0.3706,
+      "step": 1735
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.804649506267364,
+      "learning_rate": 2.5763170984544304e-06,
+      "loss": 0.4786,
+      "step": 1736
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.680940953198759,
+      "learning_rate": 2.5708408792981443e-06,
+      "loss": 0.2343,
+      "step": 1737
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.491004565731436,
+      "learning_rate": 2.5653684716827904e-06,
+      "loss": 0.4233,
+      "step": 1738
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.71514538609898,
+      "learning_rate": 2.5598998841950105e-06,
+      "loss": 0.411,
+      "step": 1739
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6820129989056176,
+      "learning_rate": 2.5544351254154407e-06,
+      "loss": 0.1074,
+      "step": 1740
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.4278678511955714,
+      "learning_rate": 2.5489742039187184e-06,
+      "loss": 0.3178,
+      "step": 1741
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.991000421170609,
+      "learning_rate": 2.5435171282734563e-06,
+      "loss": 0.4246,
+      "step": 1742
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.116980081819,
+      "learning_rate": 2.5380639070422343e-06,
+      "loss": 0.1699,
+      "step": 1743
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.606339347586722,
+      "learning_rate": 2.5326145487815822e-06,
+      "loss": 0.2506,
+      "step": 1744
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.978536746939724,
+      "learning_rate": 2.527169062041972e-06,
+      "loss": 0.2521,
+      "step": 1745
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.542237963771189,
+      "learning_rate": 2.5217274553677975e-06,
+      "loss": 0.303,
+      "step": 1746
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.783993323145913,
+      "learning_rate": 2.516289737297366e-06,
+      "loss": 0.7893,
+      "step": 1747
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.1531787428384455,
+      "learning_rate": 2.510855916362884e-06,
+      "loss": 0.5545,
+      "step": 1748
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.025596900208484,
+      "learning_rate": 2.5054260010904423e-06,
+      "loss": 0.4583,
+      "step": 1749
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.534124333013831,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.4523,
+      "step": 1750
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.347062562066108,
+      "learning_rate": 2.494577921605382e-06,
+      "loss": 0.2637,
+      "step": 1751
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.5250474399781013,
+      "learning_rate": 2.489159774414252e-06,
+      "loss": 0.1547,
+      "step": 1752
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.045552337721784,
+      "learning_rate": 2.4837455669281075e-06,
+      "loss": 0.3852,
+      "step": 1753
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.580015984978388,
+      "learning_rate": 2.478335307642264e-06,
+      "loss": 0.1578,
+      "step": 1754
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.232880365049796,
+      "learning_rate": 2.472929005045842e-06,
+      "loss": 0.3332,
+      "step": 1755
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.046078565805272,
+      "learning_rate": 2.467526667621754e-06,
+      "loss": 0.461,
+      "step": 1756
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.851610044152447,
+      "learning_rate": 2.4621283038466916e-06,
+      "loss": 0.4053,
+      "step": 1757
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.970665341233688,
+      "learning_rate": 2.4567339221911086e-06,
+      "loss": 0.2675,
+      "step": 1758
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.396785887035718,
+      "learning_rate": 2.451343531119215e-06,
+      "loss": 0.2001,
+      "step": 1759
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.539376066279478,
+      "learning_rate": 2.4459571390889537e-06,
+      "loss": 0.3424,
+      "step": 1760
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.5926404292727785,
+      "learning_rate": 2.4405747545519966e-06,
+      "loss": 0.2022,
+      "step": 1761
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.980381363671957,
+      "learning_rate": 2.435196385953727e-06,
+      "loss": 0.2754,
+      "step": 1762
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.563763971441686,
+      "learning_rate": 2.4298220417332263e-06,
+      "loss": 0.1919,
+      "step": 1763
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.048892028409314,
+      "learning_rate": 2.424451730323261e-06,
+      "loss": 0.4567,
+      "step": 1764
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.328049914773882,
+      "learning_rate": 2.4190854601502717e-06,
+      "loss": 0.1598,
+      "step": 1765
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.898882776743654,
+      "learning_rate": 2.413723239634356e-06,
+      "loss": 0.4388,
+      "step": 1766
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.0984047535774355,
+      "learning_rate": 2.4083650771892576e-06,
+      "loss": 0.0878,
+      "step": 1767
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.127977143842788,
+      "learning_rate": 2.4030109812223534e-06,
+      "loss": 0.3876,
+      "step": 1768
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4286384315349805,
+      "learning_rate": 2.3976609601346395e-06,
+      "loss": 0.235,
+      "step": 1769
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.219348625459872,
+      "learning_rate": 2.3923150223207176e-06,
+      "loss": 0.2831,
+      "step": 1770
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.528009026452362,
+      "learning_rate": 2.386973176168783e-06,
+      "loss": 0.2909,
+      "step": 1771
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.993414814418132,
+      "learning_rate": 2.381635430060611e-06,
+      "loss": 0.2571,
+      "step": 1772
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.8509877796601737,
+      "learning_rate": 2.376301792371543e-06,
+      "loss": 0.2845,
+      "step": 1773
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.431228556677509,
+      "learning_rate": 2.370972271470475e-06,
+      "loss": 0.2373,
+      "step": 1774
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.903649421667131,
+      "learning_rate": 2.3656468757198414e-06,
+      "loss": 0.2355,
+      "step": 1775
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.842661793218539,
+      "learning_rate": 2.3603256134756066e-06,
+      "loss": 0.3632,
+      "step": 1776
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.8990150290189516,
+      "learning_rate": 2.3550084930872475e-06,
+      "loss": 0.1405,
+      "step": 1777
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.273142419554314,
+      "learning_rate": 2.3496955228977437e-06,
+      "loss": 0.3316,
+      "step": 1778
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.903402169816889,
+      "learning_rate": 2.3443867112435585e-06,
+      "loss": 0.1332,
+      "step": 1779
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.167027410847691,
+      "learning_rate": 2.3390820664546353e-06,
+      "loss": 0.3948,
+      "step": 1780
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.9550706014204544,
+      "learning_rate": 2.333781596854377e-06,
+      "loss": 0.1822,
+      "step": 1781
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.897982596771436,
+      "learning_rate": 2.328485310759635e-06,
+      "loss": 0.4817,
+      "step": 1782
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.641644997255952,
+      "learning_rate": 2.323193216480698e-06,
+      "loss": 0.1201,
+      "step": 1783
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.785292260897342,
+      "learning_rate": 2.317905322321276e-06,
+      "loss": 0.3171,
+      "step": 1784
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.1639291376943355,
+      "learning_rate": 2.3126216365784894e-06,
+      "loss": 0.3944,
+      "step": 1785
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.233736261489142,
+      "learning_rate": 2.307342167542854e-06,
+      "loss": 0.517,
+      "step": 1786
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8494868708678367,
+      "learning_rate": 2.3020669234982713e-06,
+      "loss": 0.0983,
+      "step": 1787
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9212668064989864,
+      "learning_rate": 2.296795912722014e-06,
+      "loss": 0.2703,
+      "step": 1788
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.893351719853449,
+      "learning_rate": 2.291529143484707e-06,
+      "loss": 0.2016,
+      "step": 1789
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.775310475610977,
+      "learning_rate": 2.286266624050326e-06,
+      "loss": 0.2352,
+      "step": 1790
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.861045659934974,
+      "learning_rate": 2.2810083626761757e-06,
+      "loss": 0.2019,
+      "step": 1791
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.620298672464006,
+      "learning_rate": 2.275754367612881e-06,
+      "loss": 0.5436,
+      "step": 1792
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.001508931792466,
+      "learning_rate": 2.2705046471043703e-06,
+      "loss": 0.3174,
+      "step": 1793
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.592378740037039,
+      "learning_rate": 2.265259209387867e-06,
+      "loss": 0.3409,
+      "step": 1794
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.571920468246785,
+      "learning_rate": 2.2600180626938738e-06,
+      "loss": 0.3798,
+      "step": 1795
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.489737273497598,
+      "learning_rate": 2.25478121524616e-06,
+      "loss": 0.3132,
+      "step": 1796
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2552839858457787,
+      "learning_rate": 2.249548675261751e-06,
+      "loss": 0.1245,
+      "step": 1797
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.697271869044669,
+      "learning_rate": 2.2443204509509094e-06,
+      "loss": 0.4708,
+      "step": 1798
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.0946877009284846,
+      "learning_rate": 2.2390965505171304e-06,
+      "loss": 0.1455,
+      "step": 1799
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.344739133195355,
+      "learning_rate": 2.2338769821571225e-06,
+      "loss": 0.17,
+      "step": 1800
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.536375158972799,
+      "learning_rate": 2.2286617540607974e-06,
+      "loss": 0.2821,
+      "step": 1801
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.6738225162461355,
+      "learning_rate": 2.2234508744112564e-06,
+      "loss": 0.2426,
+      "step": 1802
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.560354059448164,
+      "learning_rate": 2.2182443513847777e-06,
+      "loss": 0.4775,
+      "step": 1803
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2596417176638885,
+      "learning_rate": 2.213042193150804e-06,
+      "loss": 0.2052,
+      "step": 1804
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.9607589865283486,
+      "learning_rate": 2.207844407871929e-06,
+      "loss": 0.37,
+      "step": 1805
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.8086407477970563,
+      "learning_rate": 2.202651003703885e-06,
+      "loss": 0.4081,
+      "step": 1806
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.034000667010573,
+      "learning_rate": 2.1974619887955294e-06,
+      "loss": 0.2258,
+      "step": 1807
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4654356360972334,
+      "learning_rate": 2.1922773712888356e-06,
+      "loss": 0.5445,
+      "step": 1808
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.100062974354942,
+      "learning_rate": 2.1870971593188704e-06,
+      "loss": 0.2674,
+      "step": 1809
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.531630835810501,
+      "learning_rate": 2.181921361013794e-06,
+      "loss": 0.1823,
+      "step": 1810
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4791204312655166,
+      "learning_rate": 2.17674998449484e-06,
+      "loss": 0.4051,
+      "step": 1811
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.6143879526477787,
+      "learning_rate": 2.1715830378763025e-06,
+      "loss": 0.276,
+      "step": 1812
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.9283119165737888,
+      "learning_rate": 2.1664205292655253e-06,
+      "loss": 0.1797,
+      "step": 1813
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.558243741990926,
+      "learning_rate": 2.16126246676289e-06,
+      "loss": 0.397,
+      "step": 1814
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.279406464430176,
+      "learning_rate": 2.1561088584617995e-06,
+      "loss": 0.4484,
+      "step": 1815
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.822172198039966,
+      "learning_rate": 2.1509597124486693e-06,
+      "loss": 0.5161,
+      "step": 1816
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.8481469598497378,
+      "learning_rate": 2.1458150368029146e-06,
+      "loss": 0.4062,
+      "step": 1817
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.171123093227522,
+      "learning_rate": 2.140674839596931e-06,
+      "loss": 0.4381,
+      "step": 1818
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.7516300336535124,
+      "learning_rate": 2.135539128896092e-06,
+      "loss": 0.228,
+      "step": 1819
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.1320863200582485,
+      "learning_rate": 2.13040791275873e-06,
+      "loss": 0.4194,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.013216837242059,
+      "learning_rate": 2.1252811992361256e-06,
+      "loss": 0.2301,
+      "step": 1821
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4767324138964217,
+      "learning_rate": 2.1201589963724933e-06,
+      "loss": 0.2864,
+      "step": 1822
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.9590311234581295,
+      "learning_rate": 2.1150413122049714e-06,
+      "loss": 0.6389,
+      "step": 1823
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.719787599208209,
+      "learning_rate": 2.109928154763606e-06,
+      "loss": 0.5455,
+      "step": 1824
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.849412917078412,
+      "learning_rate": 2.104819532071343e-06,
+      "loss": 0.4647,
+      "step": 1825
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.977819357950538,
+      "learning_rate": 2.09971545214401e-06,
+      "loss": 0.4172,
+      "step": 1826
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.9708211287255444,
+      "learning_rate": 2.094615922990309e-06,
+      "loss": 0.3091,
+      "step": 1827
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.347703838314115,
+      "learning_rate": 2.0895209526118016e-06,
+      "loss": 0.3192,
+      "step": 1828
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.407786720546166,
+      "learning_rate": 2.084430549002894e-06,
+      "loss": 0.0593,
+      "step": 1829
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.352516959592958,
+      "learning_rate": 2.0793447201508288e-06,
+      "loss": 0.2893,
+      "step": 1830
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.9752958408104626,
+      "learning_rate": 2.0742634740356695e-06,
+      "loss": 0.3441,
+      "step": 1831
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.844520877279769,
+      "learning_rate": 2.0691868186302898e-06,
+      "loss": 0.1656,
+      "step": 1832
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.464660124764538,
+      "learning_rate": 2.064114761900359e-06,
+      "loss": 0.2321,
+      "step": 1833
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.5319744273333855,
+      "learning_rate": 2.0590473118043326e-06,
+      "loss": 0.1348,
+      "step": 1834
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.898414095415711,
+      "learning_rate": 2.0539844762934353e-06,
+      "loss": 0.4963,
+      "step": 1835
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.087696125078068,
+      "learning_rate": 2.0489262633116536e-06,
+      "loss": 0.3278,
+      "step": 1836
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.9744662973862352,
+      "learning_rate": 2.043872680795721e-06,
+      "loss": 0.2132,
+      "step": 1837
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2202779147594316,
+      "learning_rate": 2.0388237366751005e-06,
+      "loss": 0.0496,
+      "step": 1838
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6156922261128424,
+      "learning_rate": 2.0337794388719845e-06,
+      "loss": 0.1862,
+      "step": 1839
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.314353663515789,
+      "learning_rate": 2.0287397953012686e-06,
+      "loss": 0.2203,
+      "step": 1840
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.980969740133097,
+      "learning_rate": 2.023704813870551e-06,
+      "loss": 0.2546,
+      "step": 1841
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.3102029918766345,
+      "learning_rate": 2.01867450248011e-06,
+      "loss": 0.1992,
+      "step": 1842
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.546211276625146,
+      "learning_rate": 2.0136488690228993e-06,
+      "loss": 0.2234,
+      "step": 1843
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.448963702862516,
+      "learning_rate": 2.008627921384531e-06,
+      "loss": 0.328,
+      "step": 1844
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.607520018578633,
+      "learning_rate": 2.0036116674432653e-06,
+      "loss": 0.4105,
+      "step": 1845
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.146275622462938,
+      "learning_rate": 1.998600115069998e-06,
+      "loss": 0.3217,
+      "step": 1846
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.174020516804923,
+      "learning_rate": 1.993593272128248e-06,
+      "loss": 0.5677,
+      "step": 1847
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.154539002930115,
+      "learning_rate": 1.9885911464741413e-06,
+      "loss": 0.3293,
+      "step": 1848
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.6708980658065995,
+      "learning_rate": 1.9835937459564065e-06,
+      "loss": 0.1627,
+      "step": 1849
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2209070064120717,
+      "learning_rate": 1.978601078416357e-06,
+      "loss": 0.2464,
+      "step": 1850
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.051538378905791,
+      "learning_rate": 1.9736131516878776e-06,
+      "loss": 0.391,
+      "step": 1851
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.907065551145547,
+      "learning_rate": 1.9686299735974177e-06,
+      "loss": 0.4384,
+      "step": 1852
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.043205777487744,
+      "learning_rate": 1.9636515519639735e-06,
+      "loss": 0.2174,
+      "step": 1853
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.585680254109682,
+      "learning_rate": 1.9586778945990785e-06,
+      "loss": 0.46,
+      "step": 1854
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.1220159311342015,
+      "learning_rate": 1.9537090093067907e-06,
+      "loss": 0.1915,
+      "step": 1855
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7753553128253463,
+      "learning_rate": 1.9487449038836803e-06,
+      "loss": 0.2192,
+      "step": 1856
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.706950482884989,
+      "learning_rate": 1.9437855861188175e-06,
+      "loss": 0.2192,
+      "step": 1857
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.811551914070741,
+      "learning_rate": 1.9388310637937606e-06,
+      "loss": 0.3117,
+      "step": 1858
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.783095422388319,
+      "learning_rate": 1.933881344682543e-06,
+      "loss": 0.4881,
+      "step": 1859
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.461477591229335,
+      "learning_rate": 1.928936436551661e-06,
+      "loss": 0.3832,
+      "step": 1860
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.096882534563683,
+      "learning_rate": 1.9239963471600635e-06,
+      "loss": 0.1809,
+      "step": 1861
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.03141917029654,
+      "learning_rate": 1.9190610842591386e-06,
+      "loss": 0.3769,
+      "step": 1862
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.992746423129923,
+      "learning_rate": 1.9141306555926985e-06,
+      "loss": 0.1452,
+      "step": 1863
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.509861142515103,
+      "learning_rate": 1.9092050688969736e-06,
+      "loss": 0.2118,
+      "step": 1864
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.841888948721654,
+      "learning_rate": 1.9042843319005944e-06,
+      "loss": 0.2139,
+      "step": 1865
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.2507697273324085,
+      "learning_rate": 1.8993684523245842e-06,
+      "loss": 0.4808,
+      "step": 1866
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.091326455063292,
+      "learning_rate": 1.8944574378823406e-06,
+      "loss": 0.3611,
+      "step": 1867
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.193311837378678,
+      "learning_rate": 1.8895512962796304e-06,
+      "loss": 0.2952,
+      "step": 1868
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.128520130462512,
+      "learning_rate": 1.8846500352145753e-06,
+      "loss": 0.2216,
+      "step": 1869
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.402403936519249,
+      "learning_rate": 1.879753662377637e-06,
+      "loss": 0.2593,
+      "step": 1870
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.16036607149034,
+      "learning_rate": 1.874862185451608e-06,
+      "loss": 0.5864,
+      "step": 1871
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.703495245666888,
+      "learning_rate": 1.8699756121115997e-06,
+      "loss": 0.1701,
+      "step": 1872
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.227222988390471,
+      "learning_rate": 1.865093950025027e-06,
+      "loss": 0.4472,
+      "step": 1873
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.6940826204907418,
+      "learning_rate": 1.8602172068516011e-06,
+      "loss": 0.1333,
+      "step": 1874
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.087595824923308,
+      "learning_rate": 1.8553453902433144e-06,
+      "loss": 0.3781,
+      "step": 1875
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.1949383247133545,
+      "learning_rate": 1.8504785078444293e-06,
+      "loss": 0.2486,
+      "step": 1876
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.214626070351042,
+      "learning_rate": 1.8456165672914628e-06,
+      "loss": 0.4535,
+      "step": 1877
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.335984815989052,
+      "learning_rate": 1.8407595762131814e-06,
+      "loss": 0.0587,
+      "step": 1878
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.097901640087357,
+      "learning_rate": 1.8359075422305856e-06,
+      "loss": 0.2541,
+      "step": 1879
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4985719180072055,
+      "learning_rate": 1.8310604729568964e-06,
+      "loss": 0.4075,
+      "step": 1880
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.265585961033121,
+      "learning_rate": 1.826218375997545e-06,
+      "loss": 0.5293,
+      "step": 1881
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.017867853833034,
+      "learning_rate": 1.8213812589501611e-06,
+      "loss": 0.2587,
+      "step": 1882
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.181760996941523,
+      "learning_rate": 1.8165491294045596e-06,
+      "loss": 0.1503,
+      "step": 1883
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.4153449028242395,
+      "learning_rate": 1.811721994942731e-06,
+      "loss": 0.3288,
+      "step": 1884
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.861232817903692,
+      "learning_rate": 1.8068998631388268e-06,
+      "loss": 0.5081,
+      "step": 1885
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.140432536085061,
+      "learning_rate": 1.8020827415591496e-06,
+      "loss": 0.1753,
+      "step": 1886
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.403257328693415,
+      "learning_rate": 1.7972706377621412e-06,
+      "loss": 0.2535,
+      "step": 1887
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.702488954204706,
+      "learning_rate": 1.7924635592983687e-06,
+      "loss": 0.284,
+      "step": 1888
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.8967907167704707,
+      "learning_rate": 1.7876615137105142e-06,
+      "loss": 0.3759,
+      "step": 1889
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.9964616530740846,
+      "learning_rate": 1.7828645085333645e-06,
+      "loss": 0.1542,
+      "step": 1890
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.137944778948514,
+      "learning_rate": 1.7780725512937964e-06,
+      "loss": 0.3305,
+      "step": 1891
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.792807836796973,
+      "learning_rate": 1.7732856495107648e-06,
+      "loss": 0.3998,
+      "step": 1892
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.4646948533354776,
+      "learning_rate": 1.7685038106952952e-06,
+      "loss": 0.4072,
+      "step": 1893
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.590201368147436,
+      "learning_rate": 1.7637270423504664e-06,
+      "loss": 0.1166,
+      "step": 1894
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.864879188400069,
+      "learning_rate": 1.7589553519714019e-06,
+      "loss": 0.3662,
+      "step": 1895
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.0136210039162625,
+      "learning_rate": 1.7541887470452606e-06,
+      "loss": 0.194,
+      "step": 1896
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.837556349737367,
+      "learning_rate": 1.7494272350512137e-06,
+      "loss": 0.417,
+      "step": 1897
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.873319109693996,
+      "learning_rate": 1.7446708234604498e-06,
+      "loss": 0.395,
+      "step": 1898
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.255409082820054,
+      "learning_rate": 1.7399195197361507e-06,
+      "loss": 0.3846,
+      "step": 1899
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.184679040282454,
+      "learning_rate": 1.735173331333484e-06,
+      "loss": 0.206,
+      "step": 1900
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.217040552255799,
+      "learning_rate": 1.7304322656995908e-06,
+      "loss": 0.4931,
+      "step": 1901
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.9509138521280507,
+      "learning_rate": 1.7256963302735752e-06,
+      "loss": 0.131,
+      "step": 1902
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4422246860188808,
+      "learning_rate": 1.7209655324864898e-06,
+      "loss": 0.2605,
+      "step": 1903
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.732536136113016,
+      "learning_rate": 1.7162398797613284e-06,
+      "loss": 0.2366,
+      "step": 1904
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.398890019070956,
+      "learning_rate": 1.7115193795130086e-06,
+      "loss": 0.1436,
+      "step": 1905
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.896327349849701,
+      "learning_rate": 1.7068040391483676e-06,
+      "loss": 0.2583,
+      "step": 1906
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.892523718757571,
+      "learning_rate": 1.7020938660661407e-06,
+      "loss": 0.2705,
+      "step": 1907
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.584594296514275,
+      "learning_rate": 1.6973888676569594e-06,
+      "loss": 0.4623,
+      "step": 1908
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.3928091988912925,
+      "learning_rate": 1.692689051303335e-06,
+      "loss": 0.3382,
+      "step": 1909
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.985037371450277,
+      "learning_rate": 1.6879944243796477e-06,
+      "loss": 0.4469,
+      "step": 1910
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.995767937726994,
+      "learning_rate": 1.6833049942521346e-06,
+      "loss": 0.1979,
+      "step": 1911
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.344882901204299,
+      "learning_rate": 1.678620768278879e-06,
+      "loss": 0.3542,
+      "step": 1912
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.315501959638579,
+      "learning_rate": 1.6739417538097986e-06,
+      "loss": 0.2108,
+      "step": 1913
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.2631242630187645,
+      "learning_rate": 1.6692679581866334e-06,
+      "loss": 0.4056,
+      "step": 1914
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.067063018617635,
+      "learning_rate": 1.6645993887429345e-06,
+      "loss": 0.1566,
+      "step": 1915
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.8586377855324243,
+      "learning_rate": 1.6599360528040538e-06,
+      "loss": 0.3336,
+      "step": 1916
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.72609549725891,
+      "learning_rate": 1.6552779576871297e-06,
+      "loss": 0.1219,
+      "step": 1917
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.7225871897213025,
+      "learning_rate": 1.650625110701079e-06,
+      "loss": 0.3494,
+      "step": 1918
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.0000787349216385,
+      "learning_rate": 1.6459775191465827e-06,
+      "loss": 0.3741,
+      "step": 1919
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.2515474732936465,
+      "learning_rate": 1.6413351903160763e-06,
+      "loss": 0.2318,
+      "step": 1920
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.9321237728452942,
+      "learning_rate": 1.6366981314937374e-06,
+      "loss": 0.1037,
+      "step": 1921
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.494625429012883,
+      "learning_rate": 1.632066349955474e-06,
+      "loss": 0.463,
+      "step": 1922
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.93056708515331,
+      "learning_rate": 1.6274398529689144e-06,
+      "loss": 0.532,
+      "step": 1923
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.47652438721708,
+      "learning_rate": 1.6228186477933956e-06,
+      "loss": 0.2872,
+      "step": 1924
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.057854019071931,
+      "learning_rate": 1.6182027416799505e-06,
+      "loss": 0.5433,
+      "step": 1925
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.220721751604938,
+      "learning_rate": 1.6135921418712959e-06,
+      "loss": 0.2726,
+      "step": 1926
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.8912346045021518,
+      "learning_rate": 1.608986855601824e-06,
+      "loss": 0.238,
+      "step": 1927
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.904987966064022,
+      "learning_rate": 1.60438689009759e-06,
+      "loss": 0.1876,
+      "step": 1928
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.7473960548713343,
+      "learning_rate": 1.5997922525763015e-06,
+      "loss": 0.3482,
+      "step": 1929
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8209032042140376,
+      "learning_rate": 1.5952029502473032e-06,
+      "loss": 0.13,
+      "step": 1930
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.00413022039406,
+      "learning_rate": 1.5906189903115704e-06,
+      "loss": 0.1374,
+      "step": 1931
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.390396954765262,
+      "learning_rate": 1.5860403799616951e-06,
+      "loss": 0.2525,
+      "step": 1932
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.68223887860946,
+      "learning_rate": 1.581467126381876e-06,
+      "loss": 0.5401,
+      "step": 1933
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.9328192151159516,
+      "learning_rate": 1.5768992367479058e-06,
+      "loss": 0.3122,
+      "step": 1934
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.148297088018799,
+      "learning_rate": 1.5723367182271632e-06,
+      "loss": 0.1883,
+      "step": 1935
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.922163003108522,
+      "learning_rate": 1.5677795779785932e-06,
+      "loss": 0.0705,
+      "step": 1936
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.108390157771879,
+      "learning_rate": 1.5632278231527081e-06,
+      "loss": 0.2291,
+      "step": 1937
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.706029599468409,
+      "learning_rate": 1.5586814608915673e-06,
+      "loss": 0.3241,
+      "step": 1938
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.8920567592092015,
+      "learning_rate": 1.55414049832877e-06,
+      "loss": 0.1642,
+      "step": 1939
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.07836920347546,
+      "learning_rate": 1.549604942589441e-06,
+      "loss": 0.4499,
+      "step": 1940
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.099638437227029,
+      "learning_rate": 1.5450748007902234e-06,
+      "loss": 0.4009,
+      "step": 1941
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.7431169637360755,
+      "learning_rate": 1.5405500800392643e-06,
+      "loss": 0.209,
+      "step": 1942
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.916367575937835,
+      "learning_rate": 1.5360307874362052e-06,
+      "loss": 0.5596,
+      "step": 1943
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.878293599526504,
+      "learning_rate": 1.5315169300721694e-06,
+      "loss": 0.5261,
+      "step": 1944
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.632173888229477,
+      "learning_rate": 1.5270085150297538e-06,
+      "loss": 0.2211,
+      "step": 1945
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.180609088555902,
+      "learning_rate": 1.5225055493830132e-06,
+      "loss": 0.6547,
+      "step": 1946
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0562258716530493,
+      "learning_rate": 1.518008040197455e-06,
+      "loss": 0.3523,
+      "step": 1947
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.30719723193215,
+      "learning_rate": 1.5135159945300232e-06,
+      "loss": 0.4574,
+      "step": 1948
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.865967079964248,
+      "learning_rate": 1.5090294194290882e-06,
+      "loss": 0.1708,
+      "step": 1949
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.1514956790858735,
+      "learning_rate": 1.5045483219344387e-06,
+      "loss": 0.4878,
+      "step": 1950
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.948402002091091,
+      "learning_rate": 1.500072709077267e-06,
+      "loss": 0.1444,
+      "step": 1951
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.764960830166332,
+      "learning_rate": 1.4956025878801611e-06,
+      "loss": 0.1403,
+      "step": 1952
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.010936729544463,
+      "learning_rate": 1.4911379653570913e-06,
+      "loss": 0.4949,
+      "step": 1953
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.301235148641327,
+      "learning_rate": 1.4866788485133988e-06,
+      "loss": 0.3474,
+      "step": 1954
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.3562354621637662,
+      "learning_rate": 1.4822252443457896e-06,
+      "loss": 0.1339,
+      "step": 1955
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.760096093544358,
+      "learning_rate": 1.4777771598423147e-06,
+      "loss": 0.427,
+      "step": 1956
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1391426484601626,
+      "learning_rate": 1.473334601982368e-06,
+      "loss": 0.3402,
+      "step": 1957
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.193663684513153,
+      "learning_rate": 1.4688975777366716e-06,
+      "loss": 0.6715,
+      "step": 1958
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.611091736111432,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 0.4865,
+      "step": 1959
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.7290852229072966,
+      "learning_rate": 1.4600401579274876e-06,
+      "loss": 0.1898,
+      "step": 1960
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.830029871514128,
+      "learning_rate": 1.455619776261986e-06,
+      "loss": 0.2779,
+      "step": 1961
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.243605923130625,
+      "learning_rate": 1.4512049560066837e-06,
+      "loss": 0.5229,
+      "step": 1962
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.871388634951142,
+      "learning_rate": 1.4467957040887792e-06,
+      "loss": 0.4501,
+      "step": 1963
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.904867962592391,
+      "learning_rate": 1.442392027426735e-06,
+      "loss": 0.2435,
+      "step": 1964
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.856088282844029,
+      "learning_rate": 1.4379939329302629e-06,
+      "loss": 0.212,
+      "step": 1965
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.084729437993213,
+      "learning_rate": 1.433601427500318e-06,
+      "loss": 0.4676,
+      "step": 1966
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.7455140531670175,
+      "learning_rate": 1.4292145180290856e-06,
+      "loss": 0.6632,
+      "step": 1967
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.378019868076491,
+      "learning_rate": 1.4248332113999708e-06,
+      "loss": 0.5038,
+      "step": 1968
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.647863656216499,
+      "learning_rate": 1.4204575144875871e-06,
+      "loss": 0.2726,
+      "step": 1969
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.431152332928633,
+      "learning_rate": 1.4160874341577447e-06,
+      "loss": 0.2333,
+      "step": 1970
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.495704443756592,
+      "learning_rate": 1.4117229772674424e-06,
+      "loss": 0.4257,
+      "step": 1971
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.691984119348884,
+      "learning_rate": 1.4073641506648555e-06,
+      "loss": 0.1647,
+      "step": 1972
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.037385327982083,
+      "learning_rate": 1.4030109611893233e-06,
+      "loss": 0.418,
+      "step": 1973
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.718651261892519,
+      "learning_rate": 1.3986634156713418e-06,
+      "loss": 0.4115,
+      "step": 1974
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.653336887853578,
+      "learning_rate": 1.3943215209325505e-06,
+      "loss": 0.1272,
+      "step": 1975
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.427301242675204,
+      "learning_rate": 1.3899852837857219e-06,
+      "loss": 0.3621,
+      "step": 1976
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.912708632765017,
+      "learning_rate": 1.3856547110347523e-06,
+      "loss": 0.3707,
+      "step": 1977
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.75903615667389,
+      "learning_rate": 1.3813298094746491e-06,
+      "loss": 0.2424,
+      "step": 1978
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.9543030958828203,
+      "learning_rate": 1.3770105858915217e-06,
+      "loss": 0.1327,
+      "step": 1979
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.073839072996145,
+      "learning_rate": 1.3726970470625705e-06,
+      "loss": 0.2441,
+      "step": 1980
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.248038819366711,
+      "learning_rate": 1.368389199756075e-06,
+      "loss": 0.4179,
+      "step": 1981
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.706165976752562,
+      "learning_rate": 1.3640870507313859e-06,
+      "loss": 0.4898,
+      "step": 1982
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.6784957994753045,
+      "learning_rate": 1.3597906067389116e-06,
+      "loss": 0.2817,
+      "step": 1983
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.108801191012866,
+      "learning_rate": 1.3554998745201114e-06,
+      "loss": 0.189,
+      "step": 1984
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.2833266169990605,
+      "learning_rate": 1.351214860807476e-06,
+      "loss": 0.4302,
+      "step": 1985
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.046966896037209,
+      "learning_rate": 1.3469355723245303e-06,
+      "loss": 0.2545,
+      "step": 1986
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.681410091351572,
+      "learning_rate": 1.3426620157858122e-06,
+      "loss": 0.3114,
+      "step": 1987
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.520905347693073,
+      "learning_rate": 1.3383941978968673e-06,
+      "loss": 0.3039,
+      "step": 1988
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2925634956430687,
+      "learning_rate": 1.334132125354236e-06,
+      "loss": 0.0927,
+      "step": 1989
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.6608381582152525,
+      "learning_rate": 1.3298758048454436e-06,
+      "loss": 0.2809,
+      "step": 1990
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.598559928535005,
+      "learning_rate": 1.3256252430489907e-06,
+      "loss": 0.3549,
+      "step": 1991
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.750943366975744,
+      "learning_rate": 1.321380446634342e-06,
+      "loss": 0.4004,
+      "step": 1992
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.582402456253899,
+      "learning_rate": 1.317141422261915e-06,
+      "loss": 0.1714,
+      "step": 1993
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.9781433109450575,
+      "learning_rate": 1.3129081765830725e-06,
+      "loss": 0.2884,
+      "step": 1994
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.983333650655471,
+      "learning_rate": 1.308680716240106e-06,
+      "loss": 0.3647,
+      "step": 1995
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.037252926870133,
+      "learning_rate": 1.3044590478662328e-06,
+      "loss": 0.2001,
+      "step": 1996
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.7007787473476625,
+      "learning_rate": 1.3002431780855817e-06,
+      "loss": 0.4664,
+      "step": 1997
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.259652229021513,
+      "learning_rate": 1.2960331135131826e-06,
+      "loss": 0.5181,
+      "step": 1998
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.8749742006193864,
+      "learning_rate": 1.2918288607549568e-06,
+      "loss": 0.2406,
+      "step": 1999
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.79999946453577,
+      "learning_rate": 1.2876304264077056e-06,
+      "loss": 0.23,
+      "step": 2000
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.928603656545691,
+      "learning_rate": 1.2834378170591017e-06,
+      "loss": 0.4401,
+      "step": 2001
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.265612866438301,
+      "learning_rate": 1.2792510392876777e-06,
+      "loss": 0.397,
+      "step": 2002
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.888399383415932,
+      "learning_rate": 1.275070099662815e-06,
+      "loss": 0.4213,
+      "step": 2003
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.944330017371113,
+      "learning_rate": 1.270895004744737e-06,
+      "loss": 0.4076,
+      "step": 2004
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.800461573265687,
+      "learning_rate": 1.266725761084493e-06,
+      "loss": 0.2683,
+      "step": 2005
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.6299485297277014,
+      "learning_rate": 1.262562375223954e-06,
+      "loss": 0.1161,
+      "step": 2006
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.558917516160953,
+      "learning_rate": 1.2584048536957982e-06,
+      "loss": 0.3336,
+      "step": 2007
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.156952009933042,
+      "learning_rate": 1.2542532030235022e-06,
+      "loss": 0.1929,
+      "step": 2008
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.194404639752258,
+      "learning_rate": 1.2501074297213312e-06,
+      "loss": 0.1215,
+      "step": 2009
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.46505852031427,
+      "learning_rate": 1.245967540294329e-06,
+      "loss": 0.3978,
+      "step": 2010
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.307157765575073,
+      "learning_rate": 1.2418335412383048e-06,
+      "loss": 0.2229,
+      "step": 2011
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.316879722224256,
+      "learning_rate": 1.2377054390398285e-06,
+      "loss": 0.0755,
+      "step": 2012
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.837458208310748,
+      "learning_rate": 1.233583240176217e-06,
+      "loss": 0.0884,
+      "step": 2013
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.127736170224491,
+      "learning_rate": 1.2294669511155193e-06,
+      "loss": 0.4923,
+      "step": 2014
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.762526526414984,
+      "learning_rate": 1.2253565783165178e-06,
+      "loss": 0.1324,
+      "step": 2015
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.119209105814661,
+      "learning_rate": 1.2212521282287093e-06,
+      "loss": 0.3451,
+      "step": 2016
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.572573493797165,
+      "learning_rate": 1.2171536072922968e-06,
+      "loss": 0.4531,
+      "step": 2017
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.087635891798809,
+      "learning_rate": 1.2130610219381811e-06,
+      "loss": 0.3802,
+      "step": 2018
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.675084661173655,
+      "learning_rate": 1.2089743785879493e-06,
+      "loss": 0.2072,
+      "step": 2019
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.110603688136279,
+      "learning_rate": 1.204893683653865e-06,
+      "loss": 0.1613,
+      "step": 2020
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.7380721050272148,
+      "learning_rate": 1.2008189435388578e-06,
+      "loss": 0.3018,
+      "step": 2021
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.6167838386004427,
+      "learning_rate": 1.1967501646365147e-06,
+      "loss": 0.0758,
+      "step": 2022
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.207302312264849,
+      "learning_rate": 1.1926873533310696e-06,
+      "loss": 0.1603,
+      "step": 2023
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.399434592957552,
+      "learning_rate": 1.1886305159973883e-06,
+      "loss": 0.1417,
+      "step": 2024
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.914128828875052,
+      "learning_rate": 1.1845796590009684e-06,
+      "loss": 0.1855,
+      "step": 2025
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.8691992231252135,
+      "learning_rate": 1.1805347886979219e-06,
+      "loss": 0.4489,
+      "step": 2026
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.262352832761651,
+      "learning_rate": 1.1764959114349666e-06,
+      "loss": 0.1457,
+      "step": 2027
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.624405744820009,
+      "learning_rate": 1.172463033549418e-06,
+      "loss": 0.1777,
+      "step": 2028
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.025271662536933,
+      "learning_rate": 1.1684361613691769e-06,
+      "loss": 0.2803,
+      "step": 2029
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.9066979854150494,
+      "learning_rate": 1.1644153012127208e-06,
+      "loss": 0.2028,
+      "step": 2030
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.9564926271561855,
+      "learning_rate": 1.160400459389095e-06,
+      "loss": 0.372,
+      "step": 2031
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.479560878347843,
+      "learning_rate": 1.1563916421979004e-06,
+      "loss": 0.257,
+      "step": 2032
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.758125967710335,
+      "learning_rate": 1.1523888559292857e-06,
+      "loss": 0.1786,
+      "step": 2033
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1132504235175125,
+      "learning_rate": 1.1483921068639353e-06,
+      "loss": 0.2248,
+      "step": 2034
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.532879144747303,
+      "learning_rate": 1.144401401273062e-06,
+      "loss": 0.3273,
+      "step": 2035
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.182532557670968,
+      "learning_rate": 1.1404167454183957e-06,
+      "loss": 0.2285,
+      "step": 2036
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.350719304663705,
+      "learning_rate": 1.1364381455521728e-06,
+      "loss": 0.2246,
+      "step": 2037
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.9694165033268005,
+      "learning_rate": 1.1324656079171288e-06,
+      "loss": 0.1475,
+      "step": 2038
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.2326102236425305,
+      "learning_rate": 1.128499138746486e-06,
+      "loss": 0.4002,
+      "step": 2039
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.30206636616595,
+      "learning_rate": 1.1245387442639456e-06,
+      "loss": 0.2632,
+      "step": 2040
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.138805068569877,
+      "learning_rate": 1.120584430683676e-06,
+      "loss": 0.5556,
+      "step": 2041
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.664457648996525,
+      "learning_rate": 1.1166362042103056e-06,
+      "loss": 0.2151,
+      "step": 2042
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.481110043185694,
+      "learning_rate": 1.1126940710389128e-06,
+      "loss": 0.4308,
+      "step": 2043
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.916600677596624,
+      "learning_rate": 1.10875803735501e-06,
+      "loss": 0.3834,
+      "step": 2044
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.043080724308969,
+      "learning_rate": 1.1048281093345437e-06,
+      "loss": 0.1925,
+      "step": 2045
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.558594051183082,
+      "learning_rate": 1.1009042931438784e-06,
+      "loss": 0.3452,
+      "step": 2046
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.413240832614784,
+      "learning_rate": 1.0969865949397902e-06,
+      "loss": 0.2696,
+      "step": 2047
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.254044799732455,
+      "learning_rate": 1.0930750208694535e-06,
+      "loss": 0.2523,
+      "step": 2048
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.81778679253413,
+      "learning_rate": 1.0891695770704341e-06,
+      "loss": 0.1174,
+      "step": 2049
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.677530913602758,
+      "learning_rate": 1.0852702696706807e-06,
+      "loss": 0.5205,
+      "step": 2050
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.609905228090157,
+      "learning_rate": 1.0813771047885107e-06,
+      "loss": 0.3916,
+      "step": 2051
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.7234916972882646,
+      "learning_rate": 1.077490088532605e-06,
+      "loss": 0.2206,
+      "step": 2052
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.398685893251061,
+      "learning_rate": 1.0736092270019987e-06,
+      "loss": 0.3175,
+      "step": 2053
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5592097414637895,
+      "learning_rate": 1.0697345262860638e-06,
+      "loss": 0.1735,
+      "step": 2054
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.9940841160069587,
+      "learning_rate": 1.0658659924645109e-06,
+      "loss": 0.3928,
+      "step": 2055
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.511446606912383,
+      "learning_rate": 1.062003631607373e-06,
+      "loss": 0.2891,
+      "step": 2056
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.2704572363188085,
+      "learning_rate": 1.0581474497749965e-06,
+      "loss": 0.2182,
+      "step": 2057
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.939171355285332,
+      "learning_rate": 1.0542974530180327e-06,
+      "loss": 0.5411,
+      "step": 2058
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.91519392759276,
+      "learning_rate": 1.050453647377428e-06,
+      "loss": 0.2355,
+      "step": 2059
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2279221254710553,
+      "learning_rate": 1.0466160388844149e-06,
+      "loss": 0.2909,
+      "step": 2060
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.413789562276128,
+      "learning_rate": 1.0427846335605012e-06,
+      "loss": 0.381,
+      "step": 2061
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.092313236446898,
+      "learning_rate": 1.0389594374174628e-06,
+      "loss": 0.3005,
+      "step": 2062
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.197869963235752,
+      "learning_rate": 1.0351404564573314e-06,
+      "loss": 0.3983,
+      "step": 2063
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.8493024986407303,
+      "learning_rate": 1.0313276966723867e-06,
+      "loss": 0.4979,
+      "step": 2064
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.846811971577239,
+      "learning_rate": 1.0275211640451488e-06,
+      "loss": 0.2574,
+      "step": 2065
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.555778848514527,
+      "learning_rate": 1.0237208645483648e-06,
+      "loss": 0.4752,
+      "step": 2066
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.53702990137847,
+      "learning_rate": 1.0199268041450023e-06,
+      "loss": 0.4165,
+      "step": 2067
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.716237784894001,
+      "learning_rate": 1.0161389887882394e-06,
+      "loss": 0.2189,
+      "step": 2068
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.493212914863018,
+      "learning_rate": 1.0123574244214552e-06,
+      "loss": 0.38,
+      "step": 2069
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.627727168634854,
+      "learning_rate": 1.00858211697822e-06,
+      "loss": 0.4674,
+      "step": 2070
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.008450903032346,
+      "learning_rate": 1.0048130723822874e-06,
+      "loss": 0.2886,
+      "step": 2071
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2159768889456473,
+      "learning_rate": 1.0010502965475843e-06,
+      "loss": 0.1351,
+      "step": 2072
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.164295800423085,
+      "learning_rate": 9.972937953781985e-07,
+      "loss": 0.3149,
+      "step": 2073
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5698815129316523,
+      "learning_rate": 9.935435747683758e-07,
+      "loss": 0.3081,
+      "step": 2074
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.9989737239095793,
+      "learning_rate": 9.897996406025062e-07,
+      "loss": 0.0709,
+      "step": 2075
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.27243695334749,
+      "learning_rate": 9.860619987551157e-07,
+      "loss": 0.1984,
+      "step": 2076
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.137489306606834,
+      "learning_rate": 9.823306550908568e-07,
+      "loss": 0.4305,
+      "step": 2077
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.6600039547541012,
+      "learning_rate": 9.786056154645001e-07,
+      "loss": 0.1008,
+      "step": 2078
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5358824541027305,
+      "learning_rate": 9.74886885720925e-07,
+      "loss": 0.4141,
+      "step": 2079
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.37324014539792,
+      "learning_rate": 9.711744716951093e-07,
+      "loss": 0.2516,
+      "step": 2080
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.461854105449941,
+      "learning_rate": 9.674683792121214e-07,
+      "loss": 0.2222,
+      "step": 2081
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.9228980377943365,
+      "learning_rate": 9.637686140871121e-07,
+      "loss": 0.417,
+      "step": 2082
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0523698451974934,
+      "learning_rate": 9.600751821252996e-07,
+      "loss": 0.0772,
+      "step": 2083
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.519691539735653,
+      "learning_rate": 9.563880891219702e-07,
+      "loss": 0.2497,
+      "step": 2084
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1123399819194137,
+      "learning_rate": 9.52707340862461e-07,
+      "loss": 0.247,
+      "step": 2085
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0621168607130396,
+      "learning_rate": 9.490329431221545e-07,
+      "loss": 0.342,
+      "step": 2086
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.65955452278162,
+      "learning_rate": 9.45364901666469e-07,
+      "loss": 0.4608,
+      "step": 2087
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.044479388675468,
+      "learning_rate": 9.417032222508476e-07,
+      "loss": 0.3426,
+      "step": 2088
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.106807424334697,
+      "learning_rate": 9.380479106207535e-07,
+      "loss": 0.6374,
+      "step": 2089
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.095472272297561,
+      "learning_rate": 9.34398972511656e-07,
+      "loss": 0.4042,
+      "step": 2090
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.6407607405805673,
+      "learning_rate": 9.307564136490255e-07,
+      "loss": 0.4477,
+      "step": 2091
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.029536211055726,
+      "learning_rate": 9.271202397483214e-07,
+      "loss": 0.168,
+      "step": 2092
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1264121192362637,
+      "learning_rate": 9.234904565149866e-07,
+      "loss": 0.1865,
+      "step": 2093
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.600785134736331,
+      "learning_rate": 9.198670696444339e-07,
+      "loss": 0.569,
+      "step": 2094
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.354844649574498,
+      "learning_rate": 9.162500848220418e-07,
+      "loss": 0.3773,
+      "step": 2095
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.9763569877557283,
+      "learning_rate": 9.126395077231432e-07,
+      "loss": 0.5987,
+      "step": 2096
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.670891860269174,
+      "learning_rate": 9.090353440130156e-07,
+      "loss": 0.3412,
+      "step": 2097
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.9757048622207876,
+      "learning_rate": 9.054375993468745e-07,
+      "loss": 0.1836,
+      "step": 2098
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8169569598598896,
+      "learning_rate": 9.018462793698629e-07,
+      "loss": 0.2635,
+      "step": 2099
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5066396076847877,
+      "learning_rate": 8.982613897170439e-07,
+      "loss": 0.2857,
+      "step": 2100
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.574040814370678,
+      "learning_rate": 8.946829360133891e-07,
+      "loss": 0.1768,
+      "step": 2101
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.06970750749417,
+      "learning_rate": 8.911109238737748e-07,
+      "loss": 0.2801,
+      "step": 2102
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.478024686652142,
+      "learning_rate": 8.875453589029648e-07,
+      "loss": 0.1764,
+      "step": 2103
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.507724182930931,
+      "learning_rate": 8.839862466956112e-07,
+      "loss": 0.3428,
+      "step": 2104
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.46631075228589,
+      "learning_rate": 8.804335928362401e-07,
+      "loss": 0.3971,
+      "step": 2105
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.728797916190673,
+      "learning_rate": 8.768874028992431e-07,
+      "loss": 0.2106,
+      "step": 2106
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.01213483599429,
+      "learning_rate": 8.733476824488707e-07,
+      "loss": 0.3977,
+      "step": 2107
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8858479135518835,
+      "learning_rate": 8.698144370392209e-07,
+      "loss": 0.3053,
+      "step": 2108
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.721128699419567,
+      "learning_rate": 8.662876722142327e-07,
+      "loss": 0.3185,
+      "step": 2109
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.466868520310928,
+      "learning_rate": 8.627673935076769e-07,
+      "loss": 0.1758,
+      "step": 2110
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.433307332884818,
+      "learning_rate": 8.592536064431467e-07,
+      "loss": 0.4071,
+      "step": 2111
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.134998907112914,
+      "learning_rate": 8.557463165340479e-07,
+      "loss": 0.2446,
+      "step": 2112
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.00151766537157,
+      "learning_rate": 8.522455292835935e-07,
+      "loss": 0.1392,
+      "step": 2113
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.4654451968116975,
+      "learning_rate": 8.487512501847933e-07,
+      "loss": 0.4149,
+      "step": 2114
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.561830986099788,
+      "learning_rate": 8.452634847204449e-07,
+      "loss": 0.376,
+      "step": 2115
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.605604282870342,
+      "learning_rate": 8.417822383631258e-07,
+      "loss": 0.4221,
+      "step": 2116
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.305985198147468,
+      "learning_rate": 8.38307516575183e-07,
+      "loss": 0.4552,
+      "step": 2117
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.308173256421205,
+      "learning_rate": 8.348393248087289e-07,
+      "loss": 0.2849,
+      "step": 2118
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.649060332381415,
+      "learning_rate": 8.313776685056262e-07,
+      "loss": 0.3408,
+      "step": 2119
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.388896373470471,
+      "learning_rate": 8.279225530974866e-07,
+      "loss": 0.193,
+      "step": 2120
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7355215497481191,
+      "learning_rate": 8.244739840056554e-07,
+      "loss": 0.1914,
+      "step": 2121
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.410760774134362,
+      "learning_rate": 8.210319666412087e-07,
+      "loss": 0.3838,
+      "step": 2122
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.937757317673456,
+      "learning_rate": 8.175965064049401e-07,
+      "loss": 0.3069,
+      "step": 2123
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.188973863798863,
+      "learning_rate": 8.141676086873574e-07,
+      "loss": 0.4484,
+      "step": 2124
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1448707670417018,
+      "learning_rate": 8.107452788686681e-07,
+      "loss": 0.3897,
+      "step": 2125
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.79337077314214,
+      "learning_rate": 8.073295223187766e-07,
+      "loss": 0.1276,
+      "step": 2126
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.114734385974199,
+      "learning_rate": 8.039203443972726e-07,
+      "loss": 0.1251,
+      "step": 2127
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.8419358936248424,
+      "learning_rate": 8.005177504534228e-07,
+      "loss": 0.0652,
+      "step": 2128
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.163174007698943,
+      "learning_rate": 7.971217458261643e-07,
+      "loss": 0.3861,
+      "step": 2129
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.644778096748403,
+      "learning_rate": 7.937323358440935e-07,
+      "loss": 0.0956,
+      "step": 2130
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.9069050437062187,
+      "learning_rate": 7.903495258254624e-07,
+      "loss": 0.3901,
+      "step": 2131
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.17563498329954,
+      "learning_rate": 7.869733210781611e-07,
+      "loss": 0.396,
+      "step": 2132
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3817413649102677,
+      "learning_rate": 7.836037268997221e-07,
+      "loss": 0.1037,
+      "step": 2133
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.6837222956817044,
+      "learning_rate": 7.802407485773011e-07,
+      "loss": 0.2575,
+      "step": 2134
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2467913594081104,
+      "learning_rate": 7.768843913876756e-07,
+      "loss": 0.1036,
+      "step": 2135
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.164768780656091,
+      "learning_rate": 7.735346605972322e-07,
+      "loss": 0.3418,
+      "step": 2136
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.20933849417798,
+      "learning_rate": 7.701915614619615e-07,
+      "loss": 0.1194,
+      "step": 2137
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.441401383440821,
+      "learning_rate": 7.668550992274476e-07,
+      "loss": 0.3822,
+      "step": 2138
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.435654312501965,
+      "learning_rate": 7.635252791288611e-07,
+      "loss": 0.309,
+      "step": 2139
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.513225623672635,
+      "learning_rate": 7.60202106390951e-07,
+      "loss": 0.181,
+      "step": 2140
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.121618568580262,
+      "learning_rate": 7.568855862280356e-07,
+      "loss": 0.4322,
+      "step": 2141
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.069097365037233,
+      "learning_rate": 7.535757238439939e-07,
+      "loss": 0.2743,
+      "step": 2142
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.037148328924976,
+      "learning_rate": 7.502725244322596e-07,
+      "loss": 0.3722,
+      "step": 2143
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.769759623713564,
+      "learning_rate": 7.469759931758109e-07,
+      "loss": 0.1261,
+      "step": 2144
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.6513560243308065,
+      "learning_rate": 7.436861352471647e-07,
+      "loss": 0.3153,
+      "step": 2145
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.622277249400289,
+      "learning_rate": 7.404029558083653e-07,
+      "loss": 0.1517,
+      "step": 2146
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.282837150432579,
+      "learning_rate": 7.371264600109779e-07,
+      "loss": 0.4481,
+      "step": 2147
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.6116691472748395,
+      "learning_rate": 7.338566529960817e-07,
+      "loss": 0.1992,
+      "step": 2148
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.881261732047941,
+      "learning_rate": 7.305935398942598e-07,
+      "loss": 0.1339,
+      "step": 2149
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5362180276910373,
+      "learning_rate": 7.273371258255923e-07,
+      "loss": 0.281,
+      "step": 2150
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.231597373086418,
+      "learning_rate": 7.240874158996475e-07,
+      "loss": 0.5531,
+      "step": 2151
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.371128302847014,
+      "learning_rate": 7.208444152154759e-07,
+      "loss": 0.1399,
+      "step": 2152
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.358173010504196,
+      "learning_rate": 7.176081288615983e-07,
+      "loss": 0.135,
+      "step": 2153
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.562187126615553,
+      "learning_rate": 7.143785619160026e-07,
+      "loss": 0.2225,
+      "step": 2154
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.248814129365013,
+      "learning_rate": 7.11155719446131e-07,
+      "loss": 0.5118,
+      "step": 2155
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.7086853661453736,
+      "learning_rate": 7.079396065088773e-07,
+      "loss": 0.4558,
+      "step": 2156
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.570030705304649,
+      "learning_rate": 7.047302281505735e-07,
+      "loss": 0.426,
+      "step": 2157
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.286519515007753,
+      "learning_rate": 7.015275894069862e-07,
+      "loss": 0.1149,
+      "step": 2158
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0759124027897693,
+      "learning_rate": 6.983316953033064e-07,
+      "loss": 0.089,
+      "step": 2159
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.533310420852684,
+      "learning_rate": 6.951425508541432e-07,
+      "loss": 0.3715,
+      "step": 2160
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.966910003663425,
+      "learning_rate": 6.919601610635118e-07,
+      "loss": 0.2148,
+      "step": 2161
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.265341662817313,
+      "learning_rate": 6.887845309248326e-07,
+      "loss": 0.1649,
+      "step": 2162
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.2034762310858405,
+      "learning_rate": 6.856156654209173e-07,
+      "loss": 0.1545,
+      "step": 2163
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.164156647443338,
+      "learning_rate": 6.824535695239643e-07,
+      "loss": 0.2189,
+      "step": 2164
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.23657078796661,
+      "learning_rate": 6.792982481955502e-07,
+      "loss": 0.2903,
+      "step": 2165
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.343945320000238,
+      "learning_rate": 6.761497063866207e-07,
+      "loss": 0.3613,
+      "step": 2166
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.113784060762545,
+      "learning_rate": 6.730079490374852e-07,
+      "loss": 0.2078,
+      "step": 2167
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.238462587662076,
+      "learning_rate": 6.698729810778065e-07,
+      "loss": 0.4055,
+      "step": 2168
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.73984052317012,
+      "learning_rate": 6.667448074265954e-07,
+      "loss": 0.496,
+      "step": 2169
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.7840827126001346,
+      "learning_rate": 6.63623432992202e-07,
+      "loss": 0.4575,
+      "step": 2170
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.6539429361993143,
+      "learning_rate": 6.605088626723055e-07,
+      "loss": 0.1113,
+      "step": 2171
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.061510014960234,
+      "learning_rate": 6.574011013539111e-07,
+      "loss": 0.2968,
+      "step": 2172
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.691724520906602,
+      "learning_rate": 6.543001539133409e-07,
+      "loss": 0.4197,
+      "step": 2173
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0898603323675964,
+      "learning_rate": 6.512060252162228e-07,
+      "loss": 0.2099,
+      "step": 2174
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.893812609554106,
+      "learning_rate": 6.481187201174888e-07,
+      "loss": 0.3421,
+      "step": 2175
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.454004315665616,
+      "learning_rate": 6.450382434613612e-07,
+      "loss": 0.1934,
+      "step": 2176
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.0834905530575485,
+      "learning_rate": 6.41964600081349e-07,
+      "loss": 0.2604,
+      "step": 2177
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.631006931804685,
+      "learning_rate": 6.388977948002406e-07,
+      "loss": 0.2214,
+      "step": 2178
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.9515344707274664,
+      "learning_rate": 6.358378324300929e-07,
+      "loss": 0.0976,
+      "step": 2179
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8578028818205463,
+      "learning_rate": 6.327847177722274e-07,
+      "loss": 0.1415,
+      "step": 2180
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.250405384395825,
+      "learning_rate": 6.297384556172176e-07,
+      "loss": 0.2038,
+      "step": 2181
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.912978693166093,
+      "learning_rate": 6.2669905074489e-07,
+      "loss": 0.3054,
+      "step": 2182
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2949957962456864,
+      "learning_rate": 6.236665079243087e-07,
+      "loss": 0.2979,
+      "step": 2183
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.171347472760072,
+      "learning_rate": 6.206408319137703e-07,
+      "loss": 0.4694,
+      "step": 2184
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.075732236688332,
+      "learning_rate": 6.176220274607975e-07,
+      "loss": 0.4996,
+      "step": 2185
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.240542025206586,
+      "learning_rate": 6.146100993021308e-07,
+      "loss": 0.1871,
+      "step": 2186
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.0951669262417156,
+      "learning_rate": 6.116050521637218e-07,
+      "loss": 0.3241,
+      "step": 2187
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.55318070900083,
+      "learning_rate": 6.086068907607245e-07,
+      "loss": 0.1071,
+      "step": 2188
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6878309228777213,
+      "learning_rate": 6.056156197974888e-07,
+      "loss": 0.2196,
+      "step": 2189
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.165476001138745,
+      "learning_rate": 6.026312439675553e-07,
+      "loss": 0.2738,
+      "step": 2190
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.904441308239979,
+      "learning_rate": 5.996537679536401e-07,
+      "loss": 0.2987,
+      "step": 2191
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.166171824537586,
+      "learning_rate": 5.966831964276376e-07,
+      "loss": 0.4376,
+      "step": 2192
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.752733155157431,
+      "learning_rate": 5.93719534050608e-07,
+      "loss": 0.182,
+      "step": 2193
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.1621407767764165,
+      "learning_rate": 5.907627854727688e-07,
+      "loss": 0.4705,
+      "step": 2194
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.413922763072018,
+      "learning_rate": 5.878129553334905e-07,
+      "loss": 0.4403,
+      "step": 2195
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6748962070067934,
+      "learning_rate": 5.848700482612873e-07,
+      "loss": 0.1076,
+      "step": 2196
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.658284669907072,
+      "learning_rate": 5.819340688738118e-07,
+      "loss": 0.3573,
+      "step": 2197
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.244987066843069,
+      "learning_rate": 5.790050217778442e-07,
+      "loss": 0.292,
+      "step": 2198
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.610341401184943,
+      "learning_rate": 5.760829115692907e-07,
+      "loss": 0.1896,
+      "step": 2199
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.377775047650545,
+      "learning_rate": 5.7316774283317e-07,
+      "loss": 0.3236,
+      "step": 2200
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.779888792861332,
+      "learning_rate": 5.702595201436101e-07,
+      "loss": 0.3476,
+      "step": 2201
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.007938371187252,
+      "learning_rate": 5.673582480638395e-07,
+      "loss": 0.4435,
+      "step": 2202
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.151679689529081,
+      "learning_rate": 5.64463931146183e-07,
+      "loss": 0.2445,
+      "step": 2203
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.8556135935279046,
+      "learning_rate": 5.615765739320494e-07,
+      "loss": 0.1526,
+      "step": 2204
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.048550000357814,
+      "learning_rate": 5.586961809519287e-07,
+      "loss": 0.319,
+      "step": 2205
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6271400582880586,
+      "learning_rate": 5.558227567253832e-07,
+      "loss": 0.0709,
+      "step": 2206
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.081379093376282,
+      "learning_rate": 5.529563057610399e-07,
+      "loss": 0.1807,
+      "step": 2207
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6094950199211286,
+      "learning_rate": 5.500968325565859e-07,
+      "loss": 0.3581,
+      "step": 2208
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.397011302400871,
+      "learning_rate": 5.472443415987594e-07,
+      "loss": 0.404,
+      "step": 2209
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.449418872569559,
+      "learning_rate": 5.443988373633397e-07,
+      "loss": 0.3493,
+      "step": 2210
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.7212599348359543,
+      "learning_rate": 5.415603243151469e-07,
+      "loss": 0.1638,
+      "step": 2211
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.101584966107421,
+      "learning_rate": 5.387288069080298e-07,
+      "loss": 0.4121,
+      "step": 2212
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.726380754203162,
+      "learning_rate": 5.359042895848626e-07,
+      "loss": 0.28,
+      "step": 2213
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.52379869067051,
+      "learning_rate": 5.330867767775333e-07,
+      "loss": 0.1494,
+      "step": 2214
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.6137579601650396,
+      "learning_rate": 5.302762729069399e-07,
+      "loss": 0.1219,
+      "step": 2215
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.6887392408225614,
+      "learning_rate": 5.274727823829839e-07,
+      "loss": 0.0591,
+      "step": 2216
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.291731173000995,
+      "learning_rate": 5.246763096045604e-07,
+      "loss": 0.3837,
+      "step": 2217
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.494925188042238,
+      "learning_rate": 5.218868589595555e-07,
+      "loss": 0.3388,
+      "step": 2218
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.045719425080615,
+      "learning_rate": 5.191044348248358e-07,
+      "loss": 0.5384,
+      "step": 2219
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.583416393498012,
+      "learning_rate": 5.163290415662408e-07,
+      "loss": 0.1205,
+      "step": 2220
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.1126018660664005,
+      "learning_rate": 5.13560683538582e-07,
+      "loss": 0.3394,
+      "step": 2221
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.602219295464481,
+      "learning_rate": 5.107993650856285e-07,
+      "loss": 0.2336,
+      "step": 2222
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.4779864062241925,
+      "learning_rate": 5.080450905401057e-07,
+      "loss": 0.3208,
+      "step": 2223
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.221824269737294,
+      "learning_rate": 5.052978642236866e-07,
+      "loss": 0.1526,
+      "step": 2224
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.976218441578814,
+      "learning_rate": 5.025576904469842e-07,
+      "loss": 0.4964,
+      "step": 2225
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0440285075835165,
+      "learning_rate": 4.998245735095459e-07,
+      "loss": 0.5404,
+      "step": 2226
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.0172220978893245,
+      "learning_rate": 4.970985176998455e-07,
+      "loss": 0.2748,
+      "step": 2227
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.76160128353007,
+      "learning_rate": 4.943795272952795e-07,
+      "loss": 0.2733,
+      "step": 2228
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1256717890706294,
+      "learning_rate": 4.916676065621562e-07,
+      "loss": 0.1491,
+      "step": 2229
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3672764462319758,
+      "learning_rate": 4.889627597556911e-07,
+      "loss": 0.1067,
+      "step": 2230
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.6672614940569095,
+      "learning_rate": 4.862649911200007e-07,
+      "loss": 0.0883,
+      "step": 2231
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.8905968351187106,
+      "learning_rate": 4.835743048880959e-07,
+      "loss": 0.3253,
+      "step": 2232
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.977738056689484,
+      "learning_rate": 4.80890705281874e-07,
+      "loss": 0.3721,
+      "step": 2233
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.935860819344814,
+      "learning_rate": 4.782141965121129e-07,
+      "loss": 0.1413,
+      "step": 2234
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.77547836114163,
+      "learning_rate": 4.7554478277846427e-07,
+      "loss": 0.3436,
+      "step": 2235
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9662639339985204,
+      "learning_rate": 4.7288246826944826e-07,
+      "loss": 0.2761,
+      "step": 2236
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.088850430302421,
+      "learning_rate": 4.7022725716244387e-07,
+      "loss": 0.268,
+      "step": 2237
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.229226765216043,
+      "learning_rate": 4.6757915362368567e-07,
+      "loss": 0.1782,
+      "step": 2238
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.349486192130572,
+      "learning_rate": 4.64938161808256e-07,
+      "loss": 0.2291,
+      "step": 2239
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7510258273627577,
+      "learning_rate": 4.6230428586007734e-07,
+      "loss": 0.1063,
+      "step": 2240
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.317631021090621,
+      "learning_rate": 4.596775299119066e-07,
+      "loss": 0.2341,
+      "step": 2241
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.8447583158308825,
+      "learning_rate": 4.570578980853302e-07,
+      "loss": 0.4979,
+      "step": 2242
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.930941806279006,
+      "learning_rate": 4.5444539449075677e-07,
+      "loss": 0.2828,
+      "step": 2243
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.5638729630919426,
+      "learning_rate": 4.5184002322740784e-07,
+      "loss": 0.3164,
+      "step": 2244
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.792267002535738,
+      "learning_rate": 4.4924178838331554e-07,
+      "loss": 0.2149,
+      "step": 2245
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9765455635327998,
+      "learning_rate": 4.466506940353138e-07,
+      "loss": 0.1257,
+      "step": 2246
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.449574773706834,
+      "learning_rate": 4.4406674424903264e-07,
+      "loss": 0.4303,
+      "step": 2247
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.072806953412824,
+      "learning_rate": 4.414899430788916e-07,
+      "loss": 0.1642,
+      "step": 2248
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.209866210186225,
+      "learning_rate": 4.389202945680943e-07,
+      "loss": 0.3411,
+      "step": 2249
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.5137050490375983,
+      "learning_rate": 4.363578027486187e-07,
+      "loss": 0.3249,
+      "step": 2250
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.479277042576117,
+      "learning_rate": 4.338024716412165e-07,
+      "loss": 0.3783,
+      "step": 2251
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3536186914531054,
+      "learning_rate": 4.3125430525540193e-07,
+      "loss": 0.0922,
+      "step": 2252
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.45840467049957,
+      "learning_rate": 4.2871330758944786e-07,
+      "loss": 0.1824,
+      "step": 2253
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.8852710219387414,
+      "learning_rate": 4.261794826303783e-07,
+      "loss": 0.0863,
+      "step": 2254
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9884406942216013,
+      "learning_rate": 4.2365283435396366e-07,
+      "loss": 0.1911,
+      "step": 2255
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.900171809048658,
+      "learning_rate": 4.211333667247125e-07,
+      "loss": 0.221,
+      "step": 2256
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.6534575242811975,
+      "learning_rate": 4.186210836958676e-07,
+      "loss": 0.2694,
+      "step": 2257
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.7072879599529545,
+      "learning_rate": 4.16115989209398e-07,
+      "loss": 0.4097,
+      "step": 2258
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.1948718631671804,
+      "learning_rate": 4.1361808719599163e-07,
+      "loss": 0.1913,
+      "step": 2259
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3393465056721294,
+      "learning_rate": 4.111273815750527e-07,
+      "loss": 0.3053,
+      "step": 2260
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.035462572360308,
+      "learning_rate": 4.086438762546946e-07,
+      "loss": 0.3196,
+      "step": 2261
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.898366077230296,
+      "learning_rate": 4.0616757513173123e-07,
+      "loss": 0.3597,
+      "step": 2262
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2715825445215896,
+      "learning_rate": 4.036984820916723e-07,
+      "loss": 0.3562,
+      "step": 2263
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.595156067461667,
+      "learning_rate": 4.012366010087193e-07,
+      "loss": 0.2183,
+      "step": 2264
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.8984367186597173,
+      "learning_rate": 3.9878193574575566e-07,
+      "loss": 0.3473,
+      "step": 2265
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.033654283570532,
+      "learning_rate": 3.963344901543437e-07,
+      "loss": 0.173,
+      "step": 2266
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1131666765963493,
+      "learning_rate": 3.9389426807471764e-07,
+      "loss": 0.3853,
+      "step": 2267
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.1332637878964915,
+      "learning_rate": 3.9146127333577757e-07,
+      "loss": 0.2157,
+      "step": 2268
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.480222224954902,
+      "learning_rate": 3.8903550975508075e-07,
+      "loss": 0.2085,
+      "step": 2269
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.230038187245582,
+      "learning_rate": 3.866169811388415e-07,
+      "loss": 0.275,
+      "step": 2270
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.393893906232192,
+      "learning_rate": 3.8420569128192084e-07,
+      "loss": 0.4166,
+      "step": 2271
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.230483245488447,
+      "learning_rate": 3.8180164396782124e-07,
+      "loss": 0.3383,
+      "step": 2272
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.492645965140245,
+      "learning_rate": 3.794048429686803e-07,
+      "loss": 0.2577,
+      "step": 2273
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.665151567340731,
+      "learning_rate": 3.7701529204526856e-07,
+      "loss": 0.184,
+      "step": 2274
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.269273239781175,
+      "learning_rate": 3.7463299494697735e-07,
+      "loss": 0.2679,
+      "step": 2275
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.043086433472301,
+      "learning_rate": 3.722579554118172e-07,
+      "loss": 0.461,
+      "step": 2276
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.427482468742114,
+      "learning_rate": 3.6989017716641206e-07,
+      "loss": 0.4518,
+      "step": 2277
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.4356947674317664,
+      "learning_rate": 3.675296639259912e-07,
+      "loss": 0.3305,
+      "step": 2278
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.243042908459477,
+      "learning_rate": 3.6517641939438407e-07,
+      "loss": 0.284,
+      "step": 2279
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.7151734771965774,
+      "learning_rate": 3.6283044726401594e-07,
+      "loss": 0.2837,
+      "step": 2280
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9718825048206083,
+      "learning_rate": 3.604917512159012e-07,
+      "loss": 0.1858,
+      "step": 2281
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.9004321754841875,
+      "learning_rate": 3.581603349196372e-07,
+      "loss": 0.3112,
+      "step": 2282
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.767243144192718,
+      "learning_rate": 3.5583620203339774e-07,
+      "loss": 0.0786,
+      "step": 2283
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.902165427525528,
+      "learning_rate": 3.5351935620393075e-07,
+      "loss": 0.3765,
+      "step": 2284
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.9917697911824828,
+      "learning_rate": 3.5120980106654825e-07,
+      "loss": 0.1585,
+      "step": 2285
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2721323330322116,
+      "learning_rate": 3.4890754024512254e-07,
+      "loss": 0.094,
+      "step": 2286
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.610634166327765,
+      "learning_rate": 3.466125773520818e-07,
+      "loss": 0.1695,
+      "step": 2287
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.4325757687783907,
+      "learning_rate": 3.443249159884038e-07,
+      "loss": 0.1344,
+      "step": 2288
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7688963559167385,
+      "learning_rate": 3.420445597436056e-07,
+      "loss": 0.2178,
+      "step": 2289
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.02149252263074,
+      "learning_rate": 3.397715121957468e-07,
+      "loss": 0.2878,
+      "step": 2290
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.769142990340114,
+      "learning_rate": 3.3750577691141596e-07,
+      "loss": 0.31,
+      "step": 2291
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.607562051900548,
+      "learning_rate": 3.352473574457304e-07,
+      "loss": 0.3982,
+      "step": 2292
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.03675486612295,
+      "learning_rate": 3.329962573423262e-07,
+      "loss": 0.4732,
+      "step": 2293
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.444472483450235,
+      "learning_rate": 3.3075248013335614e-07,
+      "loss": 0.2885,
+      "step": 2294
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.842348385571747,
+      "learning_rate": 3.2851602933948236e-07,
+      "loss": 0.2496,
+      "step": 2295
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.7327071436136965,
+      "learning_rate": 3.262869084698711e-07,
+      "loss": 0.4226,
+      "step": 2296
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.625058061867976,
+      "learning_rate": 3.2406512102218903e-07,
+      "loss": 0.2804,
+      "step": 2297
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.03330180543428,
+      "learning_rate": 3.2185067048259245e-07,
+      "loss": 0.3524,
+      "step": 2298
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.780213531818724,
+      "learning_rate": 3.1964356032572866e-07,
+      "loss": 0.1921,
+      "step": 2299
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.947730158346107,
+      "learning_rate": 3.174437940147268e-07,
+      "loss": 0.3891,
+      "step": 2300
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.5808131036732567,
+      "learning_rate": 3.1525137500119207e-07,
+      "loss": 0.1723,
+      "step": 2301
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.71210086029987,
+      "learning_rate": 3.1306630672520153e-07,
+      "loss": 0.2112,
+      "step": 2302
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.460327447410836,
+      "learning_rate": 3.108885926152988e-07,
+      "loss": 0.2039,
+      "step": 2303
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.575085992901881,
+      "learning_rate": 3.087182360884872e-07,
+      "loss": 0.3629,
+      "step": 2304
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.369756650295015,
+      "learning_rate": 3.06555240550227e-07,
+      "loss": 0.1128,
+      "step": 2305
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.312354009353822,
+      "learning_rate": 3.0439960939442794e-07,
+      "loss": 0.3821,
+      "step": 2306
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.321709269718386,
+      "learning_rate": 3.0225134600344373e-07,
+      "loss": 0.5004,
+      "step": 2307
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.050755875630375,
+      "learning_rate": 3.001104537480676e-07,
+      "loss": 0.4332,
+      "step": 2308
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.694026462183988,
+      "learning_rate": 2.9797693598752673e-07,
+      "loss": 0.387,
+      "step": 2309
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.192090971482895,
+      "learning_rate": 2.9585079606947843e-07,
+      "loss": 0.5603,
+      "step": 2310
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.298875139962309,
+      "learning_rate": 2.9373203733000234e-07,
+      "loss": 0.1111,
+      "step": 2311
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.005963960691346,
+      "learning_rate": 2.916206630935969e-07,
+      "loss": 0.174,
+      "step": 2312
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.7574263191665196,
+      "learning_rate": 2.895166766731744e-07,
+      "loss": 0.0858,
+      "step": 2313
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.9971631541008374,
+      "learning_rate": 2.874200813700534e-07,
+      "loss": 0.0948,
+      "step": 2314
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.786237382549601,
+      "learning_rate": 2.8533088047395627e-07,
+      "loss": 0.2424,
+      "step": 2315
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.323124485530052,
+      "learning_rate": 2.8324907726300366e-07,
+      "loss": 0.223,
+      "step": 2316
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.728848173573977,
+      "learning_rate": 2.8117467500370756e-07,
+      "loss": 0.5376,
+      "step": 2317
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.599013605641036,
+      "learning_rate": 2.7910767695096707e-07,
+      "loss": 0.2773,
+      "step": 2318
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.751904231828526,
+      "learning_rate": 2.77048086348064e-07,
+      "loss": 0.351,
+      "step": 2319
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.779065873499481,
+      "learning_rate": 2.7499590642665773e-07,
+      "loss": 0.2772,
+      "step": 2320
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.677818272163841,
+      "learning_rate": 2.729511404067797e-07,
+      "loss": 0.2405,
+      "step": 2321
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.661325741592654,
+      "learning_rate": 2.7091379149682683e-07,
+      "loss": 0.5917,
+      "step": 2322
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.932608846333246,
+      "learning_rate": 2.688838628935603e-07,
+      "loss": 0.2891,
+      "step": 2323
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.041463500009093,
+      "learning_rate": 2.668613577820961e-07,
+      "loss": 0.5847,
+      "step": 2324
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.529300253653549,
+      "learning_rate": 2.6484627933590414e-07,
+      "loss": 0.3205,
+      "step": 2325
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.165638196233581,
+      "learning_rate": 2.628386307167996e-07,
+      "loss": 0.4662,
+      "step": 2326
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.946145270811574,
+      "learning_rate": 2.608384150749416e-07,
+      "loss": 0.3197,
+      "step": 2327
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.3117858943504945,
+      "learning_rate": 2.5884563554882336e-07,
+      "loss": 0.4974,
+      "step": 2328
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.303863712464286,
+      "learning_rate": 2.5686029526527266e-07,
+      "loss": 0.1986,
+      "step": 2329
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.002332008584412,
+      "learning_rate": 2.548823973394449e-07,
+      "loss": 0.2956,
+      "step": 2330
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.870626259973056,
+      "learning_rate": 2.5291194487481575e-07,
+      "loss": 0.1783,
+      "step": 2331
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.861901439263068,
+      "learning_rate": 2.509489409631799e-07,
+      "loss": 0.401,
+      "step": 2332
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.7764678082233942,
+      "learning_rate": 2.4899338868464404e-07,
+      "loss": 0.4056,
+      "step": 2333
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.863567510326271,
+      "learning_rate": 2.470452911076227e-07,
+      "loss": 0.4565,
+      "step": 2334
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.340754799737173,
+      "learning_rate": 2.4510465128883387e-07,
+      "loss": 0.2217,
+      "step": 2335
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.470125112029207,
+      "learning_rate": 2.431714722732942e-07,
+      "loss": 0.3163,
+      "step": 2336
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.874890794744219,
+      "learning_rate": 2.4124575709431276e-07,
+      "loss": 0.3827,
+      "step": 2337
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.626194552062349,
+      "learning_rate": 2.393275087734864e-07,
+      "loss": 0.2556,
+      "step": 2338
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.434333608965055,
+      "learning_rate": 2.3741673032069757e-07,
+      "loss": 0.6067,
+      "step": 2339
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.939447195699781,
+      "learning_rate": 2.355134247341073e-07,
+      "loss": 0.3322,
+      "step": 2340
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.5766429794284593,
+      "learning_rate": 2.3361759500015123e-07,
+      "loss": 0.2931,
+      "step": 2341
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.481706695120485,
+      "learning_rate": 2.317292440935348e-07,
+      "loss": 0.3289,
+      "step": 2342
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.365543528322596,
+      "learning_rate": 2.2984837497722844e-07,
+      "loss": 0.5175,
+      "step": 2343
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.433713418680976,
+      "learning_rate": 2.2797499060246253e-07,
+      "loss": 0.513,
+      "step": 2344
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.599160013829929,
+      "learning_rate": 2.2610909390872516e-07,
+      "loss": 0.0912,
+      "step": 2345
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.824188414041289,
+      "learning_rate": 2.242506878237538e-07,
+      "loss": 0.0429,
+      "step": 2346
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.646961121734687,
+      "learning_rate": 2.2239977526353263e-07,
+      "loss": 0.4166,
+      "step": 2347
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.720538238822282,
+      "learning_rate": 2.2055635913228845e-07,
+      "loss": 0.3376,
+      "step": 2348
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.976217160454024,
+      "learning_rate": 2.1872044232248646e-07,
+      "loss": 0.2981,
+      "step": 2349
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.9145300557247014,
+      "learning_rate": 2.1689202771482344e-07,
+      "loss": 0.1662,
+      "step": 2350
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.984116177511633,
+      "learning_rate": 2.1507111817822445e-07,
+      "loss": 0.3808,
+      "step": 2351
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.6122481167406004,
+      "learning_rate": 2.1325771656984075e-07,
+      "loss": 0.0804,
+      "step": 2352
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.7008062089929408,
+      "learning_rate": 2.1145182573504008e-07,
+      "loss": 0.1593,
+      "step": 2353
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.421804301547307,
+      "learning_rate": 2.0965344850740698e-07,
+      "loss": 0.1981,
+      "step": 2354
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.7565947537910307,
+      "learning_rate": 2.0786258770873647e-07,
+      "loss": 0.1585,
+      "step": 2355
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.2498597934224285,
+      "learning_rate": 2.060792461490302e-07,
+      "loss": 0.2037,
+      "step": 2356
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.155498728314941,
+      "learning_rate": 2.043034266264887e-07,
+      "loss": 0.2652,
+      "step": 2357
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.749570915651954,
+      "learning_rate": 2.0253513192751374e-07,
+      "loss": 0.4169,
+      "step": 2358
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3698109676375467,
+      "learning_rate": 2.0077436482669687e-07,
+      "loss": 0.3276,
+      "step": 2359
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.525804262974156,
+      "learning_rate": 1.9902112808682094e-07,
+      "loss": 0.3643,
+      "step": 2360
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.869879927514824,
+      "learning_rate": 1.972754244588504e-07,
+      "loss": 0.1856,
+      "step": 2361
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.699833572027133,
+      "learning_rate": 1.9553725668193192e-07,
+      "loss": 0.3203,
+      "step": 2362
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.6582337350552505,
+      "learning_rate": 1.9380662748338662e-07,
+      "loss": 0.3739,
+      "step": 2363
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.7386335227610403,
+      "learning_rate": 1.9208353957870684e-07,
+      "loss": 0.1664,
+      "step": 2364
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.55949737725435,
+      "learning_rate": 1.903679956715526e-07,
+      "loss": 0.354,
+      "step": 2365
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.413365297085287,
+      "learning_rate": 1.8865999845374794e-07,
+      "loss": 0.3282,
+      "step": 2366
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.203806054562433,
+      "learning_rate": 1.8695955060527292e-07,
+      "loss": 0.1678,
+      "step": 2367
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.435695808454198,
+      "learning_rate": 1.8526665479426386e-07,
+      "loss": 0.1676,
+      "step": 2368
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.9413738835117917,
+      "learning_rate": 1.835813136770065e-07,
+      "loss": 0.2952,
+      "step": 2369
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.9499772409701683,
+      "learning_rate": 1.8190352989793325e-07,
+      "loss": 0.4746,
+      "step": 2370
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.913242103759934,
+      "learning_rate": 1.8023330608961886e-07,
+      "loss": 0.2626,
+      "step": 2371
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.537523619393519,
+      "learning_rate": 1.7857064487277475e-07,
+      "loss": 0.5212,
+      "step": 2372
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.952200206049917,
+      "learning_rate": 1.7691554885624628e-07,
+      "loss": 0.2876,
+      "step": 2373
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.6971037175651738,
+      "learning_rate": 1.7526802063700943e-07,
+      "loss": 0.1865,
+      "step": 2374
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9720627918683027,
+      "learning_rate": 1.7362806280016464e-07,
+      "loss": 0.0704,
+      "step": 2375
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.935582362599151,
+      "learning_rate": 1.7199567791893524e-07,
+      "loss": 0.4445,
+      "step": 2376
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.341700486061497,
+      "learning_rate": 1.7037086855465902e-07,
+      "loss": 0.3082,
+      "step": 2377
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.243106213865727,
+      "learning_rate": 1.6875363725679052e-07,
+      "loss": 0.3016,
+      "step": 2378
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.07365651043146,
+      "learning_rate": 1.6714398656289154e-07,
+      "loss": 0.3474,
+      "step": 2379
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.9638957363571796,
+      "learning_rate": 1.655419189986307e-07,
+      "loss": 0.5941,
+      "step": 2380
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.250470121011741,
+      "learning_rate": 1.6394743707777772e-07,
+      "loss": 0.1473,
+      "step": 2381
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.567418416925741,
+      "learning_rate": 1.6236054330219853e-07,
+      "loss": 0.3394,
+      "step": 2382
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.834563403433531,
+      "learning_rate": 1.6078124016185525e-07,
+      "loss": 0.3445,
+      "step": 2383
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.333859975716066,
+      "learning_rate": 1.592095301347968e-07,
+      "loss": 0.4406,
+      "step": 2384
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.485342996585978,
+      "learning_rate": 1.576454156871604e-07,
+      "loss": 0.2695,
+      "step": 2385
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.540749889940476,
+      "learning_rate": 1.5608889927316407e-07,
+      "loss": 0.1982,
+      "step": 2386
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.284222162715035,
+      "learning_rate": 1.54539983335103e-07,
+      "loss": 0.1483,
+      "step": 2387
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.167960942317522,
+      "learning_rate": 1.5299867030334815e-07,
+      "loss": 0.3977,
+      "step": 2388
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.9460848430489204,
+      "learning_rate": 1.5146496259634103e-07,
+      "loss": 0.3962,
+      "step": 2389
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.877961840330807,
+      "learning_rate": 1.4993886262058833e-07,
+      "loss": 0.0831,
+      "step": 2390
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3243854065830085,
+      "learning_rate": 1.4842037277066012e-07,
+      "loss": 0.1368,
+      "step": 2391
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.12653640296523,
+      "learning_rate": 1.469094954291872e-07,
+      "loss": 0.3493,
+      "step": 2392
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.14409200126004,
+      "learning_rate": 1.4540623296685374e-07,
+      "loss": 0.328,
+      "step": 2393
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3420411567836426,
+      "learning_rate": 1.439105877423963e-07,
+      "loss": 0.331,
+      "step": 2394
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.5829924976567007,
+      "learning_rate": 1.424225621025993e-07,
+      "loss": 0.2863,
+      "step": 2395
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.5257665888028784,
+      "learning_rate": 1.4094215838229176e-07,
+      "loss": 0.1107,
+      "step": 2396
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6649501246247334,
+      "learning_rate": 1.3946937890434276e-07,
+      "loss": 0.3173,
+      "step": 2397
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.444779951760428,
+      "learning_rate": 1.3800422597965935e-07,
+      "loss": 0.4135,
+      "step": 2398
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.990551353390603,
+      "learning_rate": 1.3654670190718035e-07,
+      "loss": 0.1215,
+      "step": 2399
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3901286357647016,
+      "learning_rate": 1.350968089738758e-07,
+      "loss": 0.1025,
+      "step": 2400
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.888378452337556,
+      "learning_rate": 1.3365454945474088e-07,
+      "loss": 0.4471,
+      "step": 2401
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.942309701473477,
+      "learning_rate": 1.322199256127943e-07,
+      "loss": 0.4582,
+      "step": 2402
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.572422844933958,
+      "learning_rate": 1.3079293969907259e-07,
+      "loss": 0.337,
+      "step": 2403
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.299307252527783,
+      "learning_rate": 1.293735939526286e-07,
+      "loss": 0.1868,
+      "step": 2404
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.9020506965455546,
+      "learning_rate": 1.279618906005281e-07,
+      "loss": 0.2337,
+      "step": 2405
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.9987913798648007,
+      "learning_rate": 1.2655783185784253e-07,
+      "loss": 0.3397,
+      "step": 2406
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.697259423572701,
+      "learning_rate": 1.2516141992765074e-07,
+      "loss": 0.4186,
+      "step": 2407
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6738815340929847,
+      "learning_rate": 1.2377265700103225e-07,
+      "loss": 0.2155,
+      "step": 2408
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.590619004433068,
+      "learning_rate": 1.223915452570651e-07,
+      "loss": 0.1545,
+      "step": 2409
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.898850891496661,
+      "learning_rate": 1.210180868628219e-07,
+      "loss": 0.2885,
+      "step": 2410
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.5973904171805255,
+      "learning_rate": 1.196522839733666e-07,
+      "loss": 0.5292,
+      "step": 2411
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.26431185935259,
+      "learning_rate": 1.1829413873174988e-07,
+      "loss": 0.4353,
+      "step": 2412
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.371513798825584,
+      "learning_rate": 1.1694365326900936e-07,
+      "loss": 0.1542,
+      "step": 2413
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.806828363526848,
+      "learning_rate": 1.1560082970416164e-07,
+      "loss": 0.3467,
+      "step": 2414
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.5213477132783852,
+      "learning_rate": 1.1426567014420297e-07,
+      "loss": 0.3744,
+      "step": 2415
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.090420360408503,
+      "learning_rate": 1.1293817668410201e-07,
+      "loss": 0.2608,
+      "step": 2416
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.183683274675307,
+      "learning_rate": 1.1161835140680033e-07,
+      "loss": 0.2705,
+      "step": 2417
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.716722040421792,
+      "learning_rate": 1.1030619638320805e-07,
+      "loss": 0.1356,
+      "step": 2418
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.024852555730413,
+      "learning_rate": 1.0900171367219826e-07,
+      "loss": 0.2136,
+      "step": 2419
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.375596574605777,
+      "learning_rate": 1.0770490532060696e-07,
+      "loss": 0.1455,
+      "step": 2420
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1324080617824577,
+      "learning_rate": 1.0641577336322761e-07,
+      "loss": 0.1473,
+      "step": 2421
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.300359247540792,
+      "learning_rate": 1.0513431982280997e-07,
+      "loss": 0.5602,
+      "step": 2422
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9591857162850492,
+      "learning_rate": 1.0386054671005452e-07,
+      "loss": 0.0584,
+      "step": 2423
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.206503491737669,
+      "learning_rate": 1.0259445602361084e-07,
+      "loss": 0.4306,
+      "step": 2424
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.846524563135461,
+      "learning_rate": 1.0133604975007483e-07,
+      "loss": 0.241,
+      "step": 2425
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.04315192848501,
+      "learning_rate": 1.0008532986398422e-07,
+      "loss": 0.1092,
+      "step": 2426
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.62176548047698,
+      "learning_rate": 9.884229832781644e-08,
+      "loss": 0.4157,
+      "step": 2427
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.411617320395682,
+      "learning_rate": 9.760695709198575e-08,
+      "loss": 0.1242,
+      "step": 2428
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.147046863685668,
+      "learning_rate": 9.637930809483942e-08,
+      "loss": 0.4138,
+      "step": 2429
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.8360773978835905,
+      "learning_rate": 9.51593532626538e-08,
+      "loss": 0.3258,
+      "step": 2430
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.819721248087985,
+      "learning_rate": 9.39470945096349e-08,
+      "loss": 0.1967,
+      "step": 2431
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.946631832138262,
+      "learning_rate": 9.274253373791064e-08,
+      "loss": 0.3869,
+      "step": 2432
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.5429212520894335,
+      "learning_rate": 9.154567283753135e-08,
+      "loss": 0.1746,
+      "step": 2433
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.8105027567396035,
+      "learning_rate": 9.035651368646647e-08,
+      "loss": 0.3425,
+      "step": 2434
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.869488263166803,
+      "learning_rate": 8.917505815059902e-08,
+      "loss": 0.1844,
+      "step": 2435
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.178460812695998,
+      "learning_rate": 8.800130808372553e-08,
+      "loss": 0.5737,
+      "step": 2436
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.099213248905488,
+      "learning_rate": 8.683526532755171e-08,
+      "loss": 0.3112,
+      "step": 2437
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.4551360585327795,
+      "learning_rate": 8.567693171168956e-08,
+      "loss": 0.2497,
+      "step": 2438
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.149425301022077,
+      "learning_rate": 8.452630905365633e-08,
+      "loss": 0.2785,
+      "step": 2439
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.857104396726367,
+      "learning_rate": 8.338339915886784e-08,
+      "loss": 0.2252,
+      "step": 2440
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.267349455333625,
+      "learning_rate": 8.224820382064013e-08,
+      "loss": 0.3577,
+      "step": 2441
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.846666449441902,
+      "learning_rate": 8.11207248201834e-08,
+      "loss": 0.2175,
+      "step": 2442
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.739712327070508,
+      "learning_rate": 8.000096392660029e-08,
+      "loss": 0.2764,
+      "step": 2443
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.680967064212557,
+      "learning_rate": 7.888892289688366e-08,
+      "loss": 0.3105,
+      "step": 2444
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.497417055358871,
+      "learning_rate": 7.778460347591277e-08,
+      "loss": 0.3753,
+      "step": 2445
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.134835352692148,
+      "learning_rate": 7.6688007396451e-08,
+      "loss": 0.3449,
+      "step": 2446
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.954947804012956,
+      "learning_rate": 7.559913637914418e-08,
+      "loss": 0.225,
+      "step": 2447
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.680352421863636,
+      "learning_rate": 7.45179921325162e-08,
+      "loss": 0.3417,
+      "step": 2448
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.549047612322732,
+      "learning_rate": 7.34445763529662e-08,
+      "loss": 0.2182,
+      "step": 2449
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.698221199963706,
+      "learning_rate": 7.237889072476856e-08,
+      "loss": 0.2427,
+      "step": 2450
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.2043257289544425,
+      "learning_rate": 7.13209369200668e-08,
+      "loss": 0.422,
+      "step": 2451
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.506784247214024,
+      "learning_rate": 7.027071659887364e-08,
+      "loss": 0.6956,
+      "step": 2452
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.5338379473216435,
+      "learning_rate": 6.922823140906754e-08,
+      "loss": 0.39,
+      "step": 2453
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4635245942883532,
+      "learning_rate": 6.819348298638839e-08,
+      "loss": 0.2996,
+      "step": 2454
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.628592924851369,
+      "learning_rate": 6.716647295443801e-08,
+      "loss": 0.2383,
+      "step": 2455
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.346329942877061,
+      "learning_rate": 6.614720292467569e-08,
+      "loss": 0.4178,
+      "step": 2456
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.454209308730887,
+      "learning_rate": 6.5135674496416e-08,
+      "loss": 0.4222,
+      "step": 2457
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.897875667018516,
+      "learning_rate": 6.4131889256826e-08,
+      "loss": 0.411,
+      "step": 2458
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.920827659381575,
+      "learning_rate": 6.313584878092361e-08,
+      "loss": 0.2343,
+      "step": 2459
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6797078298491557,
+      "learning_rate": 6.214755463157417e-08,
+      "loss": 0.3456,
+      "step": 2460
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.7075003504881185,
+      "learning_rate": 6.116700835948842e-08,
+      "loss": 0.4986,
+      "step": 2461
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.9655977972924346,
+      "learning_rate": 6.019421150322114e-08,
+      "loss": 0.3024,
+      "step": 2462
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.809475860857254,
+      "learning_rate": 5.922916558916581e-08,
+      "loss": 0.1581,
+      "step": 2463
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.772941070339243,
+      "learning_rate": 5.8271872131555605e-08,
+      "loss": 0.1239,
+      "step": 2464
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.201401866483458,
+      "learning_rate": 5.7322332632458454e-08,
+      "loss": 0.2997,
+      "step": 2465
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6027730570681804,
+      "learning_rate": 5.638054858177644e-08,
+      "loss": 0.0951,
+      "step": 2466
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.533120472850452,
+      "learning_rate": 5.544652145724305e-08,
+      "loss": 0.2793,
+      "step": 2467
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.8829477134747488,
+      "learning_rate": 5.45202527244193e-08,
+      "loss": 0.2008,
+      "step": 2468
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.782765987982104,
+      "learning_rate": 5.36017438366937e-08,
+      "loss": 0.2675,
+      "step": 2469
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1394937379714922,
+      "learning_rate": 5.26909962352784e-08,
+      "loss": 0.2646,
+      "step": 2470
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.071604895324785,
+      "learning_rate": 5.178801134920808e-08,
+      "loss": 0.1296,
+      "step": 2471
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.046037995794092,
+      "learning_rate": 5.089279059533658e-08,
+      "loss": 0.5076,
+      "step": 2472
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.989306541734293,
+      "learning_rate": 5.0005335378335294e-08,
+      "loss": 0.4052,
+      "step": 2473
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.108135623557452,
+      "learning_rate": 4.91256470906909e-08,
+      "loss": 0.4083,
+      "step": 2474
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.125310792739269,
+      "learning_rate": 4.825372711270371e-08,
+      "loss": 0.2845,
+      "step": 2475
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.218625236456331,
+      "learning_rate": 4.73895768124838e-08,
+      "loss": 0.382,
+      "step": 2476
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.506600050375412,
+      "learning_rate": 4.653319754595098e-08,
+      "loss": 0.4095,
+      "step": 2477
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.462363796117681,
+      "learning_rate": 4.568459065683206e-08,
+      "loss": 0.1888,
+      "step": 2478
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.1031074445540283,
+      "learning_rate": 4.484375747665637e-08,
+      "loss": 0.2608,
+      "step": 2479
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.9194260512973553,
+      "learning_rate": 4.401069932475799e-08,
+      "loss": 0.3155,
+      "step": 2480
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9076216174540166,
+      "learning_rate": 4.318541750827021e-08,
+      "loss": 0.065,
+      "step": 2481
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.0033303580822226,
+      "learning_rate": 4.236791332212498e-08,
+      "loss": 0.4254,
+      "step": 2482
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.026903988839854,
+      "learning_rate": 4.155818804904954e-08,
+      "loss": 0.107,
+      "step": 2483
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.839210054593576,
+      "learning_rate": 4.0756242959567596e-08,
+      "loss": 0.5627,
+      "step": 2484
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.956139091148291,
+      "learning_rate": 3.996207931199314e-08,
+      "loss": 0.3366,
+      "step": 2485
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.582273864155257,
+      "learning_rate": 3.917569835243107e-08,
+      "loss": 0.2606,
+      "step": 2486
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.8876436065674,
+      "learning_rate": 3.839710131477492e-08,
+      "loss": 0.1378,
+      "step": 2487
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.563414680610009,
+      "learning_rate": 3.762628942070412e-08,
+      "loss": 0.192,
+      "step": 2488
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.570767032609257,
+      "learning_rate": 3.686326387968286e-08,
+      "loss": 0.1846,
+      "step": 2489
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.681878056667321,
+      "learning_rate": 3.610802588895845e-08,
+      "loss": 0.176,
+      "step": 2490
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.7111296544316255,
+      "learning_rate": 3.536057663355852e-08,
+      "loss": 0.246,
+      "step": 2491
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.429892003758981,
+      "learning_rate": 3.462091728628936e-08,
+      "loss": 0.5231,
+      "step": 2492
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.7128950589879866,
+      "learning_rate": 3.388904900773371e-08,
+      "loss": 0.3555,
+      "step": 2493
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.793644159392157,
+      "learning_rate": 3.316497294625132e-08,
+      "loss": 0.6867,
+      "step": 2494
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.180988923079452,
+      "learning_rate": 3.2448690237973366e-08,
+      "loss": 0.4084,
+      "step": 2495
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.287777924413648,
+      "learning_rate": 3.1740202006804166e-08,
+      "loss": 0.2953,
+      "step": 2496
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.329746316314543,
+      "learning_rate": 3.1039509364416686e-08,
+      "loss": 0.3636,
+      "step": 2497
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.338994924839404,
+      "learning_rate": 3.034661341025258e-08,
+      "loss": 0.3093,
+      "step": 2498
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.9125712203770338,
+      "learning_rate": 2.96615152315205e-08,
+      "loss": 0.3362,
+      "step": 2499
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.353841186113316,
+      "learning_rate": 2.898421590319278e-08,
+      "loss": 0.4441,
+      "step": 2500
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.726307880994095,
+      "learning_rate": 2.8314716488004878e-08,
+      "loss": 0.3498,
+      "step": 2501
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.993050474232988,
+      "learning_rate": 2.765301803645426e-08,
+      "loss": 0.2487,
+      "step": 2502
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.065567782858954,
+      "learning_rate": 2.6999121586797626e-08,
+      "loss": 0.3005,
+      "step": 2503
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.870193447538301,
+      "learning_rate": 2.6353028165049254e-08,
+      "loss": 0.315,
+      "step": 2504
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.270123183126187,
+      "learning_rate": 2.5714738784980986e-08,
+      "loss": 0.1365,
+      "step": 2505
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.571330197561437,
+      "learning_rate": 2.5084254448117794e-08,
+      "loss": 0.3508,
+      "step": 2506
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.005644196773933,
+      "learning_rate": 2.4461576143740007e-08,
+      "loss": 0.5137,
+      "step": 2507
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.374905997459383,
+      "learning_rate": 2.3846704848878298e-08,
+      "loss": 0.3205,
+      "step": 2508
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.337172424869481,
+      "learning_rate": 2.323964152831426e-08,
+      "loss": 0.2395,
+      "step": 2509
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.6707544297850334,
+      "learning_rate": 2.264038713457706e-08,
+      "loss": 0.1153,
+      "step": 2510
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.921124694953214,
+      "learning_rate": 2.204894260794399e-08,
+      "loss": 0.2637,
+      "step": 2511
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6874069579673088,
+      "learning_rate": 2.1465308876438275e-08,
+      "loss": 0.2582,
+      "step": 2512
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.7975784533444275,
+      "learning_rate": 2.0889486855826812e-08,
+      "loss": 0.2238,
+      "step": 2513
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.311751711747407,
+      "learning_rate": 2.0321477449619098e-08,
+      "loss": 0.3692,
+      "step": 2514
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.945762699148563,
+      "learning_rate": 1.9761281549067202e-08,
+      "loss": 0.3589,
+      "step": 2515
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.049157802613744,
+      "learning_rate": 1.9208900033161338e-08,
+      "loss": 0.06,
+      "step": 2516
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7649032589146585,
+      "learning_rate": 1.8664333768631526e-08,
+      "loss": 0.0726,
+      "step": 2517
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.3304364903166617,
+      "learning_rate": 1.8127583609945376e-08,
+      "loss": 0.4464,
+      "step": 2518
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.564667529502882,
+      "learning_rate": 1.7598650399305862e-08,
+      "loss": 0.1721,
+      "step": 2519
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.428294691529022,
+      "learning_rate": 1.7077534966650767e-08,
+      "loss": 0.3292,
+      "step": 2520
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.458748330778354,
+      "learning_rate": 1.6564238129650468e-08,
+      "loss": 0.1685,
+      "step": 2521
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.871700938049787,
+      "learning_rate": 1.6058760693708487e-08,
+      "loss": 0.3076,
+      "step": 2522
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.5337468242527925,
+      "learning_rate": 1.556110345195816e-08,
+      "loss": 0.4248,
+      "step": 2523
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.322905065071024,
+      "learning_rate": 1.5071267185262086e-08,
+      "loss": 0.1864,
+      "step": 2524
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.504052845237909,
+      "learning_rate": 1.4589252662213227e-08,
+      "loss": 0.3098,
+      "step": 2525
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.479172864918825,
+      "learning_rate": 1.411506063912882e-08,
+      "loss": 0.3844,
+      "step": 2526
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.091046952414351,
+      "learning_rate": 1.3648691860053686e-08,
+      "loss": 0.1502,
+      "step": 2527
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.0671040189744,
+      "learning_rate": 1.3190147056757474e-08,
+      "loss": 0.3661,
+      "step": 2528
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.367494494280107,
+      "learning_rate": 1.2739426948732426e-08,
+      "loss": 0.2815,
+      "step": 2529
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.5721072977285555,
+      "learning_rate": 1.2296532243193382e-08,
+      "loss": 0.3263,
+      "step": 2530
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.522497550698559,
+      "learning_rate": 1.1861463635077785e-08,
+      "loss": 0.3777,
+      "step": 2531
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.460847448476461,
+      "learning_rate": 1.1434221807041234e-08,
+      "loss": 0.2931,
+      "step": 2532
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0025198698598334,
+      "learning_rate": 1.1014807429460262e-08,
+      "loss": 0.1078,
+      "step": 2533
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.657526819003815,
+      "learning_rate": 1.0603221160429e-08,
+      "loss": 0.3756,
+      "step": 2534
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.4038560085940524,
+      "learning_rate": 1.019946364575808e-08,
+      "loss": 0.1278,
+      "step": 2535
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.22480764244677,
+      "learning_rate": 9.803535518975171e-09,
+      "loss": 0.1412,
+      "step": 2536
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.147722850177204,
+      "learning_rate": 9.41543740132167e-09,
+      "loss": 0.3178,
+      "step": 2537
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.132280351434606,
+      "learning_rate": 9.035169901754902e-09,
+      "loss": 0.3441,
+      "step": 2538
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.055217266405239,
+      "learning_rate": 8.662733616944253e-09,
+      "loss": 0.4877,
+      "step": 2539
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.9514117165189906,
+      "learning_rate": 8.298129131270594e-09,
+      "loss": 0.2908,
+      "step": 2540
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.862555042803328,
+      "learning_rate": 7.941357016827967e-09,
+      "loss": 0.3591,
+      "step": 2541
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.727874665401667,
+      "learning_rate": 7.59241783341913e-09,
+      "loss": 0.3701,
+      "step": 2542
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.120515532998743,
+      "learning_rate": 7.251312128556675e-09,
+      "loss": 0.3953,
+      "step": 2543
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.7310449788079407,
+      "learning_rate": 6.918040437463025e-09,
+      "loss": 0.29,
+      "step": 2544
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.307580188884742,
+      "learning_rate": 6.592603283067101e-09,
+      "loss": 0.1891,
+      "step": 2545
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.009717756133829,
+      "learning_rate": 6.2750011760054355e-09,
+      "loss": 0.15,
+      "step": 2546
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.801223121652817,
+      "learning_rate": 5.965234614620508e-09,
+      "loss": 0.2152,
+      "step": 2547
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.358510602122476,
+      "learning_rate": 5.6633040849601865e-09,
+      "loss": 0.1348,
+      "step": 2548
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2923506636351823,
+      "learning_rate": 5.369210060777175e-09,
+      "loss": 0.1001,
+      "step": 2549
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.512827829846558,
+      "learning_rate": 5.082953003528457e-09,
+      "loss": 0.231,
+      "step": 2550
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.179076900260911,
+      "learning_rate": 4.804533362373076e-09,
+      "loss": 0.4858,
+      "step": 2551
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.853661165753404,
+      "learning_rate": 4.533951574174356e-09,
+      "loss": 0.1578,
+      "step": 2552
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.590908201313937,
+      "learning_rate": 4.2712080634949024e-09,
+      "loss": 0.2932,
+      "step": 2553
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.542632474681511,
+      "learning_rate": 4.016303242600495e-09,
+      "loss": 0.355,
+      "step": 2554
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.927745661422115,
+      "learning_rate": 3.769237511457302e-09,
+      "loss": 0.3641,
+      "step": 2555
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3152458610500832,
+      "learning_rate": 3.530011257730226e-09,
+      "loss": 0.0891,
+      "step": 2556
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.026774623379328,
+      "learning_rate": 3.298624856784005e-09,
+      "loss": 0.347,
+      "step": 2557
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.817238944433984,
+      "learning_rate": 3.075078671682108e-09,
+      "loss": 0.3206,
+      "step": 2558
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.642502782905127,
+      "learning_rate": 2.8593730531861764e-09,
+      "loss": 0.3517,
+      "step": 2559
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.382834139043959,
+      "learning_rate": 2.6515083397549156e-09,
+      "loss": 0.3784,
+      "step": 2560
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.029589267164337,
+      "learning_rate": 2.4514848575446505e-09,
+      "loss": 0.2407,
+      "step": 2561
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.613691822112342,
+      "learning_rate": 2.2593029204076578e-09,
+      "loss": 0.4333,
+      "step": 2562
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.9969353002967742,
+      "learning_rate": 2.0749628298921688e-09,
+      "loss": 0.2206,
+      "step": 2563
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.782340417296486,
+      "learning_rate": 1.8984648752429222e-09,
+      "loss": 0.3964,
+      "step": 2564
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.638242369175846,
+      "learning_rate": 1.7298093333989463e-09,
+      "loss": 0.0925,
+      "step": 2565
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.926706281697563,
+      "learning_rate": 1.5689964689935555e-09,
+      "loss": 0.2707,
+      "step": 2566
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4660002503863585,
+      "learning_rate": 1.4160265343549084e-09,
+      "loss": 0.2439,
+      "step": 2567
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.72353386822579,
+      "learning_rate": 1.2708997695043412e-09,
+      "loss": 0.082,
+      "step": 2568
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.715665495542824,
+      "learning_rate": 1.133616402158033e-09,
+      "loss": 0.3755,
+      "step": 2569
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.597729541155232,
+      "learning_rate": 1.004176647724231e-09,
+      "loss": 0.4678,
+      "step": 2570
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.759024022744002,
+      "learning_rate": 8.82580709303249e-10,
+      "loss": 0.1463,
+      "step": 2571
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.7135158349399124,
+      "learning_rate": 7.688287776896897e-10,
+      "loss": 0.3065,
+      "step": 2572
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9021686062557372,
+      "learning_rate": 6.629210313680023e-10,
+      "loss": 0.123,
+      "step": 2573
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.661695973328677,
+      "learning_rate": 5.648576365169245e-10,
+      "loss": 0.2247,
+      "step": 2574
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.85994131895302,
+      "learning_rate": 4.746387470044855e-10,
+      "loss": 0.1752,
+      "step": 2575
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.889415920762853,
+      "learning_rate": 3.922645043924478e-10,
+      "loss": 0.379,
+      "step": 2576
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.786370660553833,
+      "learning_rate": 3.1773503793131043e-10,
+      "loss": 0.3649,
+      "step": 2577
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.692099772632725,
+      "learning_rate": 2.5105046456475047e-10,
+      "loss": 0.3483,
+      "step": 2578
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.9958814936777114,
+      "learning_rate": 1.9221088892518169e-10,
+      "loss": 0.4529,
+      "step": 2579
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.821969571566754,
+      "learning_rate": 1.4121640333653042e-10,
+      "loss": 0.4975,
+      "step": 2580
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.220564095926847,
+      "learning_rate": 9.806708781368025e-11,
+      "loss": 0.3491,
+      "step": 2581
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.874333232465943,
+      "learning_rate": 6.276301006080676e-11,
+      "loss": 0.2107,
+      "step": 2582
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.085812035622458,
+      "learning_rate": 3.530422547304291e-11,
+      "loss": 0.2492,
+      "step": 2583
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.090912860808705,
+      "learning_rate": 1.5690777135368707e-11,
+      "loss": 0.2154,
+      "step": 2584
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.693481016317709,
+      "learning_rate": 3.922695822611289e-12,
+      "loss": 0.2986,
+      "step": 2585
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.069385969005949,
+      "learning_rate": 0.0,
+      "loss": 0.5263,
+      "step": 2586
+    },
+    {
+      "epoch": 1.0,
+      "step": 2586,
+      "total_flos": 0.0,
+      "train_loss": 0.5355383884464229,
+      "train_runtime": 62473.0256,
+      "train_samples_per_second": 10.599,
+      "train_steps_per_second": 0.041
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2586,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "total_flos": 0.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}