diff --git "a/last-checkpoint/trainer_state.json" "b/last-checkpoint/trainer_state.json"
--- "a/last-checkpoint/trainer_state.json"
+++ "b/last-checkpoint/trainer_state.json"
@@ -1,9 +1,9 @@
 {
   "best_metric": 0.30425724387168884,
   "best_model_checkpoint": "./w2v-bert-2.0-luo_cv_fleurs_19h-v2/checkpoint-2000",
-  "epoch": 32.467532467532465,
+  "epoch": 38.96103896103896,
   "eval_steps": 1000,
-  "global_step": 5000,
+  "global_step": 6000,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -35057,6 +35057,7016 @@
       "eval_steps_per_second": 0.541,
       "eval_wer": 0.3558201058201058,
       "step": 5000
+    },
+    {
+      "epoch": 32.47402597402598,
+      "grad_norm": 0.7455947399139404,
+      "learning_rate": 2.9231384615384617e-05,
+      "loss": 0.0142,
+      "step": 5001
+    },
+    {
+      "epoch": 32.48051948051948,
+      "grad_norm": 0.595359206199646,
+      "learning_rate": 2.9231076923076922e-05,
+      "loss": 0.0105,
+      "step": 5002
+    },
+    {
+      "epoch": 32.48701298701299,
+      "grad_norm": 0.23928628861904144,
+      "learning_rate": 2.923076923076923e-05,
+      "loss": 0.0029,
+      "step": 5003
+    },
+    {
+      "epoch": 32.493506493506494,
+      "grad_norm": 1.417263388633728,
+      "learning_rate": 2.923046153846154e-05,
+      "loss": 0.0566,
+      "step": 5004
+    },
+    {
+      "epoch": 32.5,
+      "grad_norm": 0.9653884768486023,
+      "learning_rate": 2.9230153846153844e-05,
+      "loss": 0.1181,
+      "step": 5005
+    },
+    {
+      "epoch": 32.506493506493506,
+      "grad_norm": 0.5853801965713501,
+      "learning_rate": 2.9229846153846156e-05,
+      "loss": 0.0873,
+      "step": 5006
+    },
+    {
+      "epoch": 32.51298701298701,
+      "grad_norm": 0.4240158796310425,
+      "learning_rate": 2.9229538461538464e-05,
+      "loss": 0.0576,
+      "step": 5007
+    },
+    {
+      "epoch": 32.51948051948052,
+      "grad_norm": 0.4580501616001129,
+      "learning_rate": 2.922923076923077e-05,
+      "loss": 0.0544,
+      "step": 5008
+    },
+    {
+      "epoch": 32.52597402597402,
+      "grad_norm": 0.5052347183227539,
+      "learning_rate": 2.9228923076923077e-05,
+      "loss": 0.0341,
+      "step": 5009
+    },
+    {
+      "epoch": 32.532467532467535,
+      "grad_norm": 0.4432257115840912,
+      "learning_rate": 2.9228615384615385e-05,
+      "loss": 0.0338,
+      "step": 5010
+    },
+    {
+      "epoch": 32.53896103896104,
+      "grad_norm": 0.7629462480545044,
+      "learning_rate": 2.922830769230769e-05,
+      "loss": 0.0539,
+      "step": 5011
+    },
+    {
+      "epoch": 32.54545454545455,
+      "grad_norm": 0.4871881604194641,
+      "learning_rate": 2.9228e-05,
+      "loss": 0.0385,
+      "step": 5012
+    },
+    {
+      "epoch": 32.55194805194805,
+      "grad_norm": 0.3923622965812683,
+      "learning_rate": 2.922769230769231e-05,
+      "loss": 0.027,
+      "step": 5013
+    },
+    {
+      "epoch": 32.55844155844156,
+      "grad_norm": 0.6062350869178772,
+      "learning_rate": 2.9227384615384615e-05,
+      "loss": 0.0365,
+      "step": 5014
+    },
+    {
+      "epoch": 32.564935064935064,
+      "grad_norm": 0.34979578852653503,
+      "learning_rate": 2.9227076923076924e-05,
+      "loss": 0.0178,
+      "step": 5015
+    },
+    {
+      "epoch": 32.57142857142857,
+      "grad_norm": 0.3995114862918854,
+      "learning_rate": 2.9226769230769232e-05,
+      "loss": 0.0194,
+      "step": 5016
+    },
+    {
+      "epoch": 32.577922077922075,
+      "grad_norm": 0.3907703161239624,
+      "learning_rate": 2.9226461538461537e-05,
+      "loss": 0.0183,
+      "step": 5017
+    },
+    {
+      "epoch": 32.58441558441559,
+      "grad_norm": 0.539358377456665,
+      "learning_rate": 2.9226153846153845e-05,
+      "loss": 0.0212,
+      "step": 5018
+    },
+    {
+      "epoch": 32.59090909090909,
+      "grad_norm": 0.4485572278499603,
+      "learning_rate": 2.9225846153846154e-05,
+      "loss": 0.0207,
+      "step": 5019
+    },
+    {
+      "epoch": 32.5974025974026,
+      "grad_norm": 1.1342281103134155,
+      "learning_rate": 2.9225538461538465e-05,
+      "loss": 0.0144,
+      "step": 5020
+    },
+    {
+      "epoch": 32.603896103896105,
+      "grad_norm": 0.44675344228744507,
+      "learning_rate": 2.922523076923077e-05,
+      "loss": 0.0226,
+      "step": 5021
+    },
+    {
+      "epoch": 32.61038961038961,
+      "grad_norm": 0.8697386384010315,
+      "learning_rate": 2.922492307692308e-05,
+      "loss": 0.0206,
+      "step": 5022
+    },
+    {
+      "epoch": 32.616883116883116,
+      "grad_norm": 0.36490869522094727,
+      "learning_rate": 2.9224615384615387e-05,
+      "loss": 0.0135,
+      "step": 5023
+    },
+    {
+      "epoch": 32.62337662337662,
+      "grad_norm": 0.7864395976066589,
+      "learning_rate": 2.9224307692307692e-05,
+      "loss": 0.0176,
+      "step": 5024
+    },
+    {
+      "epoch": 32.62987012987013,
+      "grad_norm": 0.49690279364585876,
+      "learning_rate": 2.9224e-05,
+      "loss": 0.0286,
+      "step": 5025
+    },
+    {
+      "epoch": 32.63636363636363,
+      "grad_norm": 1.0221036672592163,
+      "learning_rate": 2.922369230769231e-05,
+      "loss": 0.0165,
+      "step": 5026
+    },
+    {
+      "epoch": 32.642857142857146,
+      "grad_norm": 0.5382166504859924,
+      "learning_rate": 2.9223384615384617e-05,
+      "loss": 0.0082,
+      "step": 5027
+    },
+    {
+      "epoch": 32.64935064935065,
+      "grad_norm": 0.6461249589920044,
+      "learning_rate": 2.9223076923076925e-05,
+      "loss": 0.011,
+      "step": 5028
+    },
+    {
+      "epoch": 32.65584415584416,
+      "grad_norm": 0.5846178531646729,
+      "learning_rate": 2.9222769230769233e-05,
+      "loss": 0.0136,
+      "step": 5029
+    },
+    {
+      "epoch": 32.66233766233766,
+      "grad_norm": 0.6709246039390564,
+      "learning_rate": 2.9222461538461538e-05,
+      "loss": 0.0154,
+      "step": 5030
+    },
+    {
+      "epoch": 32.66883116883117,
+      "grad_norm": 0.30806440114974976,
+      "learning_rate": 2.9222153846153847e-05,
+      "loss": 0.0053,
+      "step": 5031
+    },
+    {
+      "epoch": 32.675324675324674,
+      "grad_norm": 0.8055159449577332,
+      "learning_rate": 2.9221846153846155e-05,
+      "loss": 0.0099,
+      "step": 5032
+    },
+    {
+      "epoch": 32.68181818181818,
+      "grad_norm": 1.1743619441986084,
+      "learning_rate": 2.922153846153846e-05,
+      "loss": 0.0411,
+      "step": 5033
+    },
+    {
+      "epoch": 32.688311688311686,
+      "grad_norm": 0.1986626237630844,
+      "learning_rate": 2.922123076923077e-05,
+      "loss": 0.0031,
+      "step": 5034
+    },
+    {
+      "epoch": 32.6948051948052,
+      "grad_norm": 0.9511582851409912,
+      "learning_rate": 2.922092307692308e-05,
+      "loss": 0.0166,
+      "step": 5035
+    },
+    {
+      "epoch": 32.701298701298704,
+      "grad_norm": 0.5934857726097107,
+      "learning_rate": 2.9220615384615385e-05,
+      "loss": 0.012,
+      "step": 5036
+    },
+    {
+      "epoch": 32.70779220779221,
+      "grad_norm": 0.7282407283782959,
+      "learning_rate": 2.9220307692307693e-05,
+      "loss": 0.0188,
+      "step": 5037
+    },
+    {
+      "epoch": 32.714285714285715,
+      "grad_norm": 0.4278605282306671,
+      "learning_rate": 2.922e-05,
+      "loss": 0.0062,
+      "step": 5038
+    },
+    {
+      "epoch": 32.72077922077922,
+      "grad_norm": 0.5733248591423035,
+      "learning_rate": 2.9219692307692306e-05,
+      "loss": 0.0076,
+      "step": 5039
+    },
+    {
+      "epoch": 32.72727272727273,
+      "grad_norm": 1.7734631299972534,
+      "learning_rate": 2.9219384615384615e-05,
+      "loss": 0.0403,
+      "step": 5040
+    },
+    {
+      "epoch": 32.73376623376623,
+      "grad_norm": 0.2512581944465637,
+      "learning_rate": 2.9219076923076926e-05,
+      "loss": 0.0087,
+      "step": 5041
+    },
+    {
+      "epoch": 32.74025974025974,
+      "grad_norm": 0.7879403233528137,
+      "learning_rate": 2.921876923076923e-05,
+      "loss": 0.0106,
+      "step": 5042
+    },
+    {
+      "epoch": 32.746753246753244,
+      "grad_norm": 0.7762377858161926,
+      "learning_rate": 2.921846153846154e-05,
+      "loss": 0.1138,
+      "step": 5043
+    },
+    {
+      "epoch": 32.753246753246756,
+      "grad_norm": 0.6951083540916443,
+      "learning_rate": 2.9218153846153848e-05,
+      "loss": 0.0988,
+      "step": 5044
+    },
+    {
+      "epoch": 32.75974025974026,
+      "grad_norm": 0.4834299087524414,
+      "learning_rate": 2.9217846153846153e-05,
+      "loss": 0.0673,
+      "step": 5045
+    },
+    {
+      "epoch": 32.76623376623377,
+      "grad_norm": 0.4215543866157532,
+      "learning_rate": 2.921753846153846e-05,
+      "loss": 0.0537,
+      "step": 5046
+    },
+    {
+      "epoch": 32.77272727272727,
+      "grad_norm": 0.38221612572669983,
+      "learning_rate": 2.9217230769230773e-05,
+      "loss": 0.0475,
+      "step": 5047
+    },
+    {
+      "epoch": 32.77922077922078,
+      "grad_norm": 0.3835408687591553,
+      "learning_rate": 2.9216923076923078e-05,
+      "loss": 0.0276,
+      "step": 5048
+    },
+    {
+      "epoch": 32.785714285714285,
+      "grad_norm": 0.5472840070724487,
+      "learning_rate": 2.9216615384615386e-05,
+      "loss": 0.0318,
+      "step": 5049
+    },
+    {
+      "epoch": 32.79220779220779,
+      "grad_norm": 0.7388037443161011,
+      "learning_rate": 2.9216307692307695e-05,
+      "loss": 0.0408,
+      "step": 5050
+    },
+    {
+      "epoch": 32.798701298701296,
+      "grad_norm": 0.6615577936172485,
+      "learning_rate": 2.9216e-05,
+      "loss": 0.0607,
+      "step": 5051
+    },
+    {
+      "epoch": 32.8051948051948,
+      "grad_norm": 0.548747181892395,
+      "learning_rate": 2.9215692307692308e-05,
+      "loss": 0.0304,
+      "step": 5052
+    },
+    {
+      "epoch": 32.811688311688314,
+      "grad_norm": 0.38156527280807495,
+      "learning_rate": 2.9215384615384616e-05,
+      "loss": 0.0229,
+      "step": 5053
+    },
+    {
+      "epoch": 32.81818181818182,
+      "grad_norm": 0.6100485324859619,
+      "learning_rate": 2.9215076923076924e-05,
+      "loss": 0.0282,
+      "step": 5054
+    },
+    {
+      "epoch": 32.824675324675326,
+      "grad_norm": 0.5645389556884766,
+      "learning_rate": 2.9214769230769233e-05,
+      "loss": 0.0267,
+      "step": 5055
+    },
+    {
+      "epoch": 32.83116883116883,
+      "grad_norm": 0.7538372874259949,
+      "learning_rate": 2.921446153846154e-05,
+      "loss": 0.0214,
+      "step": 5056
+    },
+    {
+      "epoch": 32.83766233766234,
+      "grad_norm": 0.41211986541748047,
+      "learning_rate": 2.9214153846153846e-05,
+      "loss": 0.0205,
+      "step": 5057
+    },
+    {
+      "epoch": 32.84415584415584,
+      "grad_norm": 0.993869960308075,
+      "learning_rate": 2.9213846153846154e-05,
+      "loss": 0.0504,
+      "step": 5058
+    },
+    {
+      "epoch": 32.85064935064935,
+      "grad_norm": 0.5521100163459778,
+      "learning_rate": 2.9213538461538463e-05,
+      "loss": 0.0171,
+      "step": 5059
+    },
+    {
+      "epoch": 32.857142857142854,
+      "grad_norm": 0.49331873655319214,
+      "learning_rate": 2.9213230769230768e-05,
+      "loss": 0.0194,
+      "step": 5060
+    },
+    {
+      "epoch": 32.86363636363637,
+      "grad_norm": 0.4689992368221283,
+      "learning_rate": 2.921292307692308e-05,
+      "loss": 0.0138,
+      "step": 5061
+    },
+    {
+      "epoch": 32.87012987012987,
+      "grad_norm": 0.6267738342285156,
+      "learning_rate": 2.9212615384615388e-05,
+      "loss": 0.0147,
+      "step": 5062
+    },
+    {
+      "epoch": 32.87662337662338,
+      "grad_norm": 0.4092702865600586,
+      "learning_rate": 2.9212307692307693e-05,
+      "loss": 0.0121,
+      "step": 5063
+    },
+    {
+      "epoch": 32.883116883116884,
+      "grad_norm": 1.5005707740783691,
+      "learning_rate": 2.9212e-05,
+      "loss": 0.0232,
+      "step": 5064
+    },
+    {
+      "epoch": 32.88961038961039,
+      "grad_norm": 0.4179195165634155,
+      "learning_rate": 2.921169230769231e-05,
+      "loss": 0.0116,
+      "step": 5065
+    },
+    {
+      "epoch": 32.896103896103895,
+      "grad_norm": 0.9752940535545349,
+      "learning_rate": 2.9211384615384614e-05,
+      "loss": 0.0587,
+      "step": 5066
+    },
+    {
+      "epoch": 32.9025974025974,
+      "grad_norm": 0.7442163228988647,
+      "learning_rate": 2.9211076923076922e-05,
+      "loss": 0.0212,
+      "step": 5067
+    },
+    {
+      "epoch": 32.90909090909091,
+      "grad_norm": 0.6272023916244507,
+      "learning_rate": 2.9210769230769234e-05,
+      "loss": 0.0131,
+      "step": 5068
+    },
+    {
+      "epoch": 32.91558441558441,
+      "grad_norm": 0.5838685035705566,
+      "learning_rate": 2.921046153846154e-05,
+      "loss": 0.0153,
+      "step": 5069
+    },
+    {
+      "epoch": 32.922077922077925,
+      "grad_norm": 1.2788761854171753,
+      "learning_rate": 2.9210153846153847e-05,
+      "loss": 0.0132,
+      "step": 5070
+    },
+    {
+      "epoch": 32.92857142857143,
+      "grad_norm": 1.2704384326934814,
+      "learning_rate": 2.9209846153846156e-05,
+      "loss": 0.0072,
+      "step": 5071
+    },
+    {
+      "epoch": 32.935064935064936,
+      "grad_norm": 0.36931920051574707,
+      "learning_rate": 2.920953846153846e-05,
+      "loss": 0.0073,
+      "step": 5072
+    },
+    {
+      "epoch": 32.94155844155844,
+      "grad_norm": 0.43888309597969055,
+      "learning_rate": 2.920923076923077e-05,
+      "loss": 0.0107,
+      "step": 5073
+    },
+    {
+      "epoch": 32.94805194805195,
+      "grad_norm": 0.7088915705680847,
+      "learning_rate": 2.9208923076923077e-05,
+      "loss": 0.0527,
+      "step": 5074
+    },
+    {
+      "epoch": 32.95454545454545,
+      "grad_norm": 0.6745349764823914,
+      "learning_rate": 2.9208615384615386e-05,
+      "loss": 0.0059,
+      "step": 5075
+    },
+    {
+      "epoch": 32.96103896103896,
+      "grad_norm": 2.0108890533447266,
+      "learning_rate": 2.9208307692307694e-05,
+      "loss": 0.0225,
+      "step": 5076
+    },
+    {
+      "epoch": 32.967532467532465,
+      "grad_norm": 0.7361552715301514,
+      "learning_rate": 2.9208000000000002e-05,
+      "loss": 0.007,
+      "step": 5077
+    },
+    {
+      "epoch": 32.97402597402598,
+      "grad_norm": 0.6981684565544128,
+      "learning_rate": 2.9207692307692307e-05,
+      "loss": 0.0305,
+      "step": 5078
+    },
+    {
+      "epoch": 32.98051948051948,
+      "grad_norm": 0.6083940267562866,
+      "learning_rate": 2.9207384615384615e-05,
+      "loss": 0.0061,
+      "step": 5079
+    },
+    {
+      "epoch": 32.98701298701299,
+      "grad_norm": 1.8485974073410034,
+      "learning_rate": 2.9207076923076924e-05,
+      "loss": 0.0192,
+      "step": 5080
+    },
+    {
+      "epoch": 32.993506493506494,
+      "grad_norm": 0.7409580945968628,
+      "learning_rate": 2.920676923076923e-05,
+      "loss": 0.05,
+      "step": 5081
+    },
+    {
+      "epoch": 33.0,
+      "grad_norm": 0.5991113185882568,
+      "learning_rate": 2.920646153846154e-05,
+      "loss": 0.0249,
+      "step": 5082
+    },
+    {
+      "epoch": 33.006493506493506,
+      "grad_norm": 0.6193212866783142,
+      "learning_rate": 2.920615384615385e-05,
+      "loss": 0.0957,
+      "step": 5083
+    },
+    {
+      "epoch": 33.01298701298701,
+      "grad_norm": 0.5013349652290344,
+      "learning_rate": 2.9205846153846154e-05,
+      "loss": 0.0582,
+      "step": 5084
+    },
+    {
+      "epoch": 33.01948051948052,
+      "grad_norm": 0.4513569474220276,
+      "learning_rate": 2.9205538461538462e-05,
+      "loss": 0.0527,
+      "step": 5085
+    },
+    {
+      "epoch": 33.02597402597402,
+      "grad_norm": 0.36499112844467163,
+      "learning_rate": 2.920523076923077e-05,
+      "loss": 0.0495,
+      "step": 5086
+    },
+    {
+      "epoch": 33.032467532467535,
+      "grad_norm": 0.4244421422481537,
+      "learning_rate": 2.9204923076923075e-05,
+      "loss": 0.0411,
+      "step": 5087
+    },
+    {
+      "epoch": 33.03896103896104,
+      "grad_norm": 0.3226945698261261,
+      "learning_rate": 2.9204615384615384e-05,
+      "loss": 0.0255,
+      "step": 5088
+    },
+    {
+      "epoch": 33.04545454545455,
+      "grad_norm": 0.38724368810653687,
+      "learning_rate": 2.9204307692307695e-05,
+      "loss": 0.027,
+      "step": 5089
+    },
+    {
+      "epoch": 33.05194805194805,
+      "grad_norm": 0.49415165185928345,
+      "learning_rate": 2.9204e-05,
+      "loss": 0.0276,
+      "step": 5090
+    },
+    {
+      "epoch": 33.05844155844156,
+      "grad_norm": 0.41548120975494385,
+      "learning_rate": 2.920369230769231e-05,
+      "loss": 0.02,
+      "step": 5091
+    },
+    {
+      "epoch": 33.064935064935064,
+      "grad_norm": 0.6897282004356384,
+      "learning_rate": 2.9203384615384617e-05,
+      "loss": 0.0215,
+      "step": 5092
+    },
+    {
+      "epoch": 33.07142857142857,
+      "grad_norm": 0.4937240481376648,
+      "learning_rate": 2.9203076923076922e-05,
+      "loss": 0.0169,
+      "step": 5093
+    },
+    {
+      "epoch": 33.077922077922075,
+      "grad_norm": 0.49796339869499207,
+      "learning_rate": 2.920276923076923e-05,
+      "loss": 0.0172,
+      "step": 5094
+    },
+    {
+      "epoch": 33.08441558441559,
+      "grad_norm": 1.144960880279541,
+      "learning_rate": 2.920246153846154e-05,
+      "loss": 0.0228,
+      "step": 5095
+    },
+    {
+      "epoch": 33.09090909090909,
+      "grad_norm": 0.44828659296035767,
+      "learning_rate": 2.9202153846153847e-05,
+      "loss": 0.0095,
+      "step": 5096
+    },
+    {
+      "epoch": 33.0974025974026,
+      "grad_norm": 0.4165479838848114,
+      "learning_rate": 2.9201846153846155e-05,
+      "loss": 0.015,
+      "step": 5097
+    },
+    {
+      "epoch": 33.103896103896105,
+      "grad_norm": 0.4386371672153473,
+      "learning_rate": 2.9201538461538463e-05,
+      "loss": 0.0133,
+      "step": 5098
+    },
+    {
+      "epoch": 33.11038961038961,
+      "grad_norm": 0.3190487325191498,
+      "learning_rate": 2.9201230769230772e-05,
+      "loss": 0.0086,
+      "step": 5099
+    },
+    {
+      "epoch": 33.116883116883116,
+      "grad_norm": 0.6171437501907349,
+      "learning_rate": 2.9200923076923077e-05,
+      "loss": 0.0379,
+      "step": 5100
+    },
+    {
+      "epoch": 33.12337662337662,
+      "grad_norm": 0.5920683741569519,
+      "learning_rate": 2.9200615384615385e-05,
+      "loss": 0.0111,
+      "step": 5101
+    },
+    {
+      "epoch": 33.12987012987013,
+      "grad_norm": 0.40865978598594666,
+      "learning_rate": 2.9200307692307693e-05,
+      "loss": 0.0088,
+      "step": 5102
+    },
+    {
+      "epoch": 33.13636363636363,
+      "grad_norm": 0.3535599112510681,
+      "learning_rate": 2.92e-05,
+      "loss": 0.0072,
+      "step": 5103
+    },
+    {
+      "epoch": 33.142857142857146,
+      "grad_norm": 0.577937662601471,
+      "learning_rate": 2.919969230769231e-05,
+      "loss": 0.0268,
+      "step": 5104
+    },
+    {
+      "epoch": 33.14935064935065,
+      "grad_norm": 0.5617281198501587,
+      "learning_rate": 2.9199384615384618e-05,
+      "loss": 0.0088,
+      "step": 5105
+    },
+    {
+      "epoch": 33.15584415584416,
+      "grad_norm": 0.9031304717063904,
+      "learning_rate": 2.9199076923076923e-05,
+      "loss": 0.0129,
+      "step": 5106
+    },
+    {
+      "epoch": 33.16233766233766,
+      "grad_norm": 1.5352048873901367,
+      "learning_rate": 2.919876923076923e-05,
+      "loss": 0.015,
+      "step": 5107
+    },
+    {
+      "epoch": 33.16883116883117,
+      "grad_norm": 0.34195709228515625,
+      "learning_rate": 2.919846153846154e-05,
+      "loss": 0.0047,
+      "step": 5108
+    },
+    {
+      "epoch": 33.175324675324674,
+      "grad_norm": 0.4741317331790924,
+      "learning_rate": 2.9198153846153845e-05,
+      "loss": 0.0083,
+      "step": 5109
+    },
+    {
+      "epoch": 33.18181818181818,
+      "grad_norm": 0.6579654812812805,
+      "learning_rate": 2.9197846153846157e-05,
+      "loss": 0.013,
+      "step": 5110
+    },
+    {
+      "epoch": 33.188311688311686,
+      "grad_norm": 0.8151121139526367,
+      "learning_rate": 2.9197538461538465e-05,
+      "loss": 0.0158,
+      "step": 5111
+    },
+    {
+      "epoch": 33.1948051948052,
+      "grad_norm": 1.0251507759094238,
+      "learning_rate": 2.919723076923077e-05,
+      "loss": 0.0173,
+      "step": 5112
+    },
+    {
+      "epoch": 33.201298701298704,
+      "grad_norm": 0.5701705813407898,
+      "learning_rate": 2.9196923076923078e-05,
+      "loss": 0.0104,
+      "step": 5113
+    },
+    {
+      "epoch": 33.20779220779221,
+      "grad_norm": 0.7816833257675171,
+      "learning_rate": 2.9196615384615386e-05,
+      "loss": 0.0139,
+      "step": 5114
+    },
+    {
+      "epoch": 33.214285714285715,
+      "grad_norm": 0.4763857126235962,
+      "learning_rate": 2.919630769230769e-05,
+      "loss": 0.0077,
+      "step": 5115
+    },
+    {
+      "epoch": 33.22077922077922,
+      "grad_norm": 1.8317736387252808,
+      "learning_rate": 2.9196e-05,
+      "loss": 0.0161,
+      "step": 5116
+    },
+    {
+      "epoch": 33.22727272727273,
+      "grad_norm": 3.5923259258270264,
+      "learning_rate": 2.919569230769231e-05,
+      "loss": 0.0258,
+      "step": 5117
+    },
+    {
+      "epoch": 33.23376623376623,
+      "grad_norm": 0.6837280988693237,
+      "learning_rate": 2.9195384615384616e-05,
+      "loss": 0.0198,
+      "step": 5118
+    },
+    {
+      "epoch": 33.24025974025974,
+      "grad_norm": 0.8108289837837219,
+      "learning_rate": 2.9195076923076925e-05,
+      "loss": 0.0117,
+      "step": 5119
+    },
+    {
+      "epoch": 33.246753246753244,
+      "grad_norm": 0.9223237037658691,
+      "learning_rate": 2.9194769230769233e-05,
+      "loss": 0.0143,
+      "step": 5120
+    },
+    {
+      "epoch": 33.253246753246756,
+      "grad_norm": 0.6978357434272766,
+      "learning_rate": 2.9194461538461538e-05,
+      "loss": 0.105,
+      "step": 5121
+    },
+    {
+      "epoch": 33.25974025974026,
+      "grad_norm": 0.6224868297576904,
+      "learning_rate": 2.9194153846153846e-05,
+      "loss": 0.0686,
+      "step": 5122
+    },
+    {
+      "epoch": 33.26623376623377,
+      "grad_norm": 0.4567446708679199,
+      "learning_rate": 2.9193846153846155e-05,
+      "loss": 0.0441,
+      "step": 5123
+    },
+    {
+      "epoch": 33.27272727272727,
+      "grad_norm": 0.5249010920524597,
+      "learning_rate": 2.9193538461538463e-05,
+      "loss": 0.0485,
+      "step": 5124
+    },
+    {
+      "epoch": 33.27922077922078,
+      "grad_norm": 0.45523807406425476,
+      "learning_rate": 2.919323076923077e-05,
+      "loss": 0.0297,
+      "step": 5125
+    },
+    {
+      "epoch": 33.285714285714285,
+      "grad_norm": 0.31809040904045105,
+      "learning_rate": 2.919292307692308e-05,
+      "loss": 0.0234,
+      "step": 5126
+    },
+    {
+      "epoch": 33.29220779220779,
+      "grad_norm": 0.45942091941833496,
+      "learning_rate": 2.9192615384615384e-05,
+      "loss": 0.0294,
+      "step": 5127
+    },
+    {
+      "epoch": 33.298701298701296,
+      "grad_norm": 0.4964737892150879,
+      "learning_rate": 2.9192307692307693e-05,
+      "loss": 0.024,
+      "step": 5128
+    },
+    {
+      "epoch": 33.3051948051948,
+      "grad_norm": 0.6479532122612,
+      "learning_rate": 2.9192e-05,
+      "loss": 0.0563,
+      "step": 5129
+    },
+    {
+      "epoch": 33.311688311688314,
+      "grad_norm": 0.3270736634731293,
+      "learning_rate": 2.9191692307692306e-05,
+      "loss": 0.019,
+      "step": 5130
+    },
+    {
+      "epoch": 33.31818181818182,
+      "grad_norm": 0.35060492157936096,
+      "learning_rate": 2.9191384615384618e-05,
+      "loss": 0.0178,
+      "step": 5131
+    },
+    {
+      "epoch": 33.324675324675326,
+      "grad_norm": 0.8634567260742188,
+      "learning_rate": 2.9191076923076926e-05,
+      "loss": 0.0338,
+      "step": 5132
+    },
+    {
+      "epoch": 33.33116883116883,
+      "grad_norm": 0.4603855013847351,
+      "learning_rate": 2.919076923076923e-05,
+      "loss": 0.0202,
+      "step": 5133
+    },
+    {
+      "epoch": 33.33766233766234,
+      "grad_norm": 0.4173492193222046,
+      "learning_rate": 2.919046153846154e-05,
+      "loss": 0.0183,
+      "step": 5134
+    },
+    {
+      "epoch": 33.34415584415584,
+      "grad_norm": 0.34388092160224915,
+      "learning_rate": 2.9190153846153848e-05,
+      "loss": 0.009,
+      "step": 5135
+    },
+    {
+      "epoch": 33.35064935064935,
+      "grad_norm": 0.6773208975791931,
+      "learning_rate": 2.9189846153846153e-05,
+      "loss": 0.0389,
+      "step": 5136
+    },
+    {
+      "epoch": 33.357142857142854,
+      "grad_norm": 0.4459824562072754,
+      "learning_rate": 2.918953846153846e-05,
+      "loss": 0.0207,
+      "step": 5137
+    },
+    {
+      "epoch": 33.36363636363637,
+      "grad_norm": 0.47827303409576416,
+      "learning_rate": 2.9189230769230773e-05,
+      "loss": 0.0166,
+      "step": 5138
+    },
+    {
+      "epoch": 33.37012987012987,
+      "grad_norm": 0.592857301235199,
+      "learning_rate": 2.9188923076923077e-05,
+      "loss": 0.0133,
+      "step": 5139
+    },
+    {
+      "epoch": 33.37662337662338,
+      "grad_norm": 0.9384477734565735,
+      "learning_rate": 2.9188615384615386e-05,
+      "loss": 0.025,
+      "step": 5140
+    },
+    {
+      "epoch": 33.383116883116884,
+      "grad_norm": 0.3302897810935974,
+      "learning_rate": 2.9188307692307694e-05,
+      "loss": 0.007,
+      "step": 5141
+    },
+    {
+      "epoch": 33.38961038961039,
+      "grad_norm": 0.6045815348625183,
+      "learning_rate": 2.9188e-05,
+      "loss": 0.0134,
+      "step": 5142
+    },
+    {
+      "epoch": 33.396103896103895,
+      "grad_norm": 0.4486161768436432,
+      "learning_rate": 2.9187692307692307e-05,
+      "loss": 0.0072,
+      "step": 5143
+    },
+    {
+      "epoch": 33.4025974025974,
+      "grad_norm": 0.4483949542045593,
+      "learning_rate": 2.9187384615384616e-05,
+      "loss": 0.031,
+      "step": 5144
+    },
+    {
+      "epoch": 33.40909090909091,
+      "grad_norm": 0.522419810295105,
+      "learning_rate": 2.9187076923076924e-05,
+      "loss": 0.0122,
+      "step": 5145
+    },
+    {
+      "epoch": 33.41558441558441,
+      "grad_norm": 0.9663325548171997,
+      "learning_rate": 2.9186769230769232e-05,
+      "loss": 0.0206,
+      "step": 5146
+    },
+    {
+      "epoch": 33.422077922077925,
+      "grad_norm": 1.6213914155960083,
+      "learning_rate": 2.918646153846154e-05,
+      "loss": 0.0104,
+      "step": 5147
+    },
+    {
+      "epoch": 33.42857142857143,
+      "grad_norm": 1.0196571350097656,
+      "learning_rate": 2.9186153846153846e-05,
+      "loss": 0.0489,
+      "step": 5148
+    },
+    {
+      "epoch": 33.435064935064936,
+      "grad_norm": 0.45447394251823425,
+      "learning_rate": 2.9185846153846154e-05,
+      "loss": 0.0104,
+      "step": 5149
+    },
+    {
+      "epoch": 33.44155844155844,
+      "grad_norm": 0.46332013607025146,
+      "learning_rate": 2.9185538461538462e-05,
+      "loss": 0.0085,
+      "step": 5150
+    },
+    {
+      "epoch": 33.44805194805195,
+      "grad_norm": 0.2794162929058075,
+      "learning_rate": 2.9185230769230767e-05,
+      "loss": 0.0034,
+      "step": 5151
+    },
+    {
+      "epoch": 33.45454545454545,
+      "grad_norm": 1.2567662000656128,
+      "learning_rate": 2.918492307692308e-05,
+      "loss": 0.0438,
+      "step": 5152
+    },
+    {
+      "epoch": 33.46103896103896,
+      "grad_norm": 0.4975087344646454,
+      "learning_rate": 2.9184615384615387e-05,
+      "loss": 0.0046,
+      "step": 5153
+    },
+    {
+      "epoch": 33.467532467532465,
+      "grad_norm": 0.7620040774345398,
+      "learning_rate": 2.9184307692307692e-05,
+      "loss": 0.0128,
+      "step": 5154
+    },
+    {
+      "epoch": 33.47402597402598,
+      "grad_norm": 0.7590813636779785,
+      "learning_rate": 2.9184e-05,
+      "loss": 0.0075,
+      "step": 5155
+    },
+    {
+      "epoch": 33.48051948051948,
+      "grad_norm": 0.5330340266227722,
+      "learning_rate": 2.918369230769231e-05,
+      "loss": 0.005,
+      "step": 5156
+    },
+    {
+      "epoch": 33.48701298701299,
+      "grad_norm": 0.24444550275802612,
+      "learning_rate": 2.9183384615384614e-05,
+      "loss": 0.0033,
+      "step": 5157
+    },
+    {
+      "epoch": 33.493506493506494,
+      "grad_norm": 0.6891660690307617,
+      "learning_rate": 2.9183076923076922e-05,
+      "loss": 0.0116,
+      "step": 5158
+    },
+    {
+      "epoch": 33.5,
+      "grad_norm": 1.2010726928710938,
+      "learning_rate": 2.9182769230769234e-05,
+      "loss": 0.0934,
+      "step": 5159
+    },
+    {
+      "epoch": 33.506493506493506,
+      "grad_norm": 0.590378999710083,
+      "learning_rate": 2.918246153846154e-05,
+      "loss": 0.08,
+      "step": 5160
+    },
+    {
+      "epoch": 33.51298701298701,
+      "grad_norm": 0.9820928573608398,
+      "learning_rate": 2.9182153846153847e-05,
+      "loss": 0.0655,
+      "step": 5161
+    },
+    {
+      "epoch": 33.51948051948052,
+      "grad_norm": 0.4438777267932892,
+      "learning_rate": 2.9181846153846155e-05,
+      "loss": 0.0445,
+      "step": 5162
+    },
+    {
+      "epoch": 33.52597402597402,
+      "grad_norm": 0.5293325781822205,
+      "learning_rate": 2.918153846153846e-05,
+      "loss": 0.0502,
+      "step": 5163
+    },
+    {
+      "epoch": 33.532467532467535,
+      "grad_norm": 1.6670782566070557,
+      "learning_rate": 2.918123076923077e-05,
+      "loss": 0.0291,
+      "step": 5164
+    },
+    {
+      "epoch": 33.53896103896104,
+      "grad_norm": 0.37328842282295227,
+      "learning_rate": 2.9180923076923077e-05,
+      "loss": 0.0285,
+      "step": 5165
+    },
+    {
+      "epoch": 33.54545454545455,
+      "grad_norm": 0.4555578827857971,
+      "learning_rate": 2.9180615384615385e-05,
+      "loss": 0.0352,
+      "step": 5166
+    },
+    {
+      "epoch": 33.55194805194805,
+      "grad_norm": 0.5837788581848145,
+      "learning_rate": 2.9180307692307694e-05,
+      "loss": 0.0301,
+      "step": 5167
+    },
+    {
+      "epoch": 33.55844155844156,
+      "grad_norm": 0.43086379766464233,
+      "learning_rate": 2.9180000000000002e-05,
+      "loss": 0.0286,
+      "step": 5168
+    },
+    {
+      "epoch": 33.564935064935064,
+      "grad_norm": 0.3732292056083679,
+      "learning_rate": 2.9179692307692307e-05,
+      "loss": 0.0172,
+      "step": 5169
+    },
+    {
+      "epoch": 33.57142857142857,
+      "grad_norm": 0.3461165428161621,
+      "learning_rate": 2.9179384615384615e-05,
+      "loss": 0.0146,
+      "step": 5170
+    },
+    {
+      "epoch": 33.577922077922075,
+      "grad_norm": 0.4424782991409302,
+      "learning_rate": 2.9179076923076923e-05,
+      "loss": 0.0177,
+      "step": 5171
+    },
+    {
+      "epoch": 33.58441558441559,
+      "grad_norm": 0.5220151543617249,
+      "learning_rate": 2.917876923076923e-05,
+      "loss": 0.0206,
+      "step": 5172
+    },
+    {
+      "epoch": 33.59090909090909,
+      "grad_norm": 0.3175686001777649,
+      "learning_rate": 2.917846153846154e-05,
+      "loss": 0.0105,
+      "step": 5173
+    },
+    {
+      "epoch": 33.5974025974026,
+      "grad_norm": 0.6249086856842041,
+      "learning_rate": 2.917815384615385e-05,
+      "loss": 0.0199,
+      "step": 5174
+    },
+    {
+      "epoch": 33.603896103896105,
+      "grad_norm": 0.4141503572463989,
+      "learning_rate": 2.9177846153846153e-05,
+      "loss": 0.0144,
+      "step": 5175
+    },
+    {
+      "epoch": 33.61038961038961,
+      "grad_norm": 1.178198218345642,
+      "learning_rate": 2.917753846153846e-05,
+      "loss": 0.0204,
+      "step": 5176
+    },
+    {
+      "epoch": 33.616883116883116,
+      "grad_norm": 0.4076842963695526,
+      "learning_rate": 2.917723076923077e-05,
+      "loss": 0.0109,
+      "step": 5177
+    },
+    {
+      "epoch": 33.62337662337662,
+      "grad_norm": 0.4344927966594696,
+      "learning_rate": 2.9176923076923078e-05,
+      "loss": 0.013,
+      "step": 5178
+    },
+    {
+      "epoch": 33.62987012987013,
+      "grad_norm": 0.6804765462875366,
+      "learning_rate": 2.9176615384615383e-05,
+      "loss": 0.0176,
+      "step": 5179
+    },
+    {
+      "epoch": 33.63636363636363,
+      "grad_norm": 0.28606274724006653,
+      "learning_rate": 2.9176307692307695e-05,
+      "loss": 0.0036,
+      "step": 5180
+    },
+    {
+      "epoch": 33.642857142857146,
+      "grad_norm": 0.6881821155548096,
+      "learning_rate": 2.9176000000000003e-05,
+      "loss": 0.0242,
+      "step": 5181
+    },
+    {
+      "epoch": 33.64935064935065,
+      "grad_norm": 0.5474647879600525,
+      "learning_rate": 2.9175692307692308e-05,
+      "loss": 0.0108,
+      "step": 5182
+    },
+    {
+      "epoch": 33.65584415584416,
+      "grad_norm": 0.6660938858985901,
+      "learning_rate": 2.9175384615384616e-05,
+      "loss": 0.0139,
+      "step": 5183
+    },
+    {
+      "epoch": 33.66233766233766,
+      "grad_norm": 1.0489383935928345,
+      "learning_rate": 2.9175076923076925e-05,
+      "loss": 0.0155,
+      "step": 5184
+    },
+    {
+      "epoch": 33.66883116883117,
+      "grad_norm": 3.0114641189575195,
+      "learning_rate": 2.917476923076923e-05,
+      "loss": 0.012,
+      "step": 5185
+    },
+    {
+      "epoch": 33.675324675324674,
+      "grad_norm": 2.148083209991455,
+      "learning_rate": 2.9174461538461538e-05,
+      "loss": 0.0402,
+      "step": 5186
+    },
+    {
+      "epoch": 33.68181818181818,
+      "grad_norm": 0.9920523166656494,
+      "learning_rate": 2.917415384615385e-05,
+      "loss": 0.0179,
+      "step": 5187
+    },
+    {
+      "epoch": 33.688311688311686,
+      "grad_norm": 0.17630691826343536,
+      "learning_rate": 2.9173846153846155e-05,
+      "loss": 0.0032,
+      "step": 5188
+    },
+    {
+      "epoch": 33.6948051948052,
+      "grad_norm": 0.8171472549438477,
+      "learning_rate": 2.9173538461538463e-05,
+      "loss": 0.0143,
+      "step": 5189
+    },
+    {
+      "epoch": 33.701298701298704,
+      "grad_norm": 1.099873661994934,
+      "learning_rate": 2.917323076923077e-05,
+      "loss": 0.0133,
+      "step": 5190
+    },
+    {
+      "epoch": 33.70779220779221,
+      "grad_norm": 0.36660346388816833,
+      "learning_rate": 2.9172923076923076e-05,
+      "loss": 0.0056,
+      "step": 5191
+    },
+    {
+      "epoch": 33.714285714285715,
+      "grad_norm": 0.8074890375137329,
+      "learning_rate": 2.9172615384615385e-05,
+      "loss": 0.0117,
+      "step": 5192
+    },
+    {
+      "epoch": 33.72077922077922,
+      "grad_norm": 2.2832932472229004,
+      "learning_rate": 2.9172307692307693e-05,
+      "loss": 0.0322,
+      "step": 5193
+    },
+    {
+      "epoch": 33.72727272727273,
+      "grad_norm": 7.517980575561523,
+      "learning_rate": 2.9172e-05,
+      "loss": 0.0527,
+      "step": 5194
+    },
+    {
+      "epoch": 33.73376623376623,
+      "grad_norm": 0.6317906975746155,
+      "learning_rate": 2.917169230769231e-05,
+      "loss": 0.0106,
+      "step": 5195
+    },
+    {
+      "epoch": 33.74025974025974,
+      "grad_norm": 4.595775127410889,
+      "learning_rate": 2.9171384615384618e-05,
+      "loss": 0.1125,
+      "step": 5196
+    },
+    {
+      "epoch": 33.746753246753244,
+      "grad_norm": 0.6804779171943665,
+      "learning_rate": 2.9171076923076923e-05,
+      "loss": 0.106,
+      "step": 5197
+    },
+    {
+      "epoch": 33.753246753246756,
+      "grad_norm": 0.47375553846359253,
+      "learning_rate": 2.917076923076923e-05,
+      "loss": 0.0698,
+      "step": 5198
+    },
+    {
+      "epoch": 33.75974025974026,
+      "grad_norm": 0.39842429757118225,
+      "learning_rate": 2.917046153846154e-05,
+      "loss": 0.0505,
+      "step": 5199
+    },
+    {
+      "epoch": 33.76623376623377,
+      "grad_norm": 0.46608805656433105,
+      "learning_rate": 2.9170153846153844e-05,
+      "loss": 0.06,
+      "step": 5200
+    },
+    {
+      "epoch": 33.77272727272727,
+      "grad_norm": 0.4317260682582855,
+      "learning_rate": 2.9169846153846156e-05,
+      "loss": 0.0378,
+      "step": 5201
+    },
+    {
+      "epoch": 33.77922077922078,
+      "grad_norm": 0.43160080909729004,
+      "learning_rate": 2.9169538461538464e-05,
+      "loss": 0.0333,
+      "step": 5202
+    },
+    {
+      "epoch": 33.785714285714285,
+      "grad_norm": 0.5686500072479248,
+      "learning_rate": 2.916923076923077e-05,
+      "loss": 0.0332,
+      "step": 5203
+    },
+    {
+      "epoch": 33.79220779220779,
+      "grad_norm": 0.6219180226325989,
+      "learning_rate": 2.9168923076923078e-05,
+      "loss": 0.0534,
+      "step": 5204
+    },
+    {
+      "epoch": 33.798701298701296,
+      "grad_norm": 0.42976194620132446,
+      "learning_rate": 2.9168615384615386e-05,
+      "loss": 0.0217,
+      "step": 5205
+    },
+    {
+      "epoch": 33.8051948051948,
+      "grad_norm": 0.42062732577323914,
+      "learning_rate": 2.916830769230769e-05,
+      "loss": 0.0203,
+      "step": 5206
+    },
+    {
+      "epoch": 33.811688311688314,
+      "grad_norm": 0.33051377534866333,
+      "learning_rate": 2.9168e-05,
+      "loss": 0.0205,
+      "step": 5207
+    },
+    {
+      "epoch": 33.81818181818182,
+      "grad_norm": 0.40757471323013306,
+      "learning_rate": 2.916769230769231e-05,
+      "loss": 0.0239,
+      "step": 5208
+    },
+    {
+      "epoch": 33.824675324675326,
+      "grad_norm": 0.6034359335899353,
+      "learning_rate": 2.9167384615384616e-05,
+      "loss": 0.0274,
+      "step": 5209
+    },
+    {
+      "epoch": 33.83116883116883,
+      "grad_norm": 0.5356073379516602,
+      "learning_rate": 2.9167076923076924e-05,
+      "loss": 0.0202,
+      "step": 5210
+    },
+    {
+      "epoch": 33.83766233766234,
+      "grad_norm": 0.41984131932258606,
+      "learning_rate": 2.9166769230769233e-05,
+      "loss": 0.0147,
+      "step": 5211
+    },
+    {
+      "epoch": 33.84415584415584,
+      "grad_norm": 0.3775591254234314,
+      "learning_rate": 2.9166461538461537e-05,
+      "loss": 0.0125,
+      "step": 5212
+    },
+    {
+      "epoch": 33.85064935064935,
+      "grad_norm": 0.7033487558364868,
+      "learning_rate": 2.9166153846153846e-05,
+      "loss": 0.0216,
+      "step": 5213
+    },
+    {
+      "epoch": 33.857142857142854,
+      "grad_norm": 0.3927634358406067,
+      "learning_rate": 2.9165846153846154e-05,
+      "loss": 0.011,
+      "step": 5214
+    },
+    {
+      "epoch": 33.86363636363637,
+      "grad_norm": 0.7569650411605835,
+      "learning_rate": 2.9165538461538462e-05,
+      "loss": 0.0231,
+      "step": 5215
+    },
+    {
+      "epoch": 33.87012987012987,
+      "grad_norm": 0.4745166003704071,
+      "learning_rate": 2.916523076923077e-05,
+      "loss": 0.012,
+      "step": 5216
+    },
+    {
+      "epoch": 33.87662337662338,
+      "grad_norm": 0.3904447555541992,
+      "learning_rate": 2.916492307692308e-05,
+      "loss": 0.0082,
+      "step": 5217
+    },
+    {
+      "epoch": 33.883116883116884,
+      "grad_norm": 1.1203515529632568,
+      "learning_rate": 2.9164615384615384e-05,
+      "loss": 0.0299,
+      "step": 5218
+    },
+    {
+      "epoch": 33.88961038961039,
+      "grad_norm": 0.3816543519496918,
+      "learning_rate": 2.9164307692307692e-05,
+      "loss": 0.0064,
+      "step": 5219
+    },
+    {
+      "epoch": 33.896103896103895,
+      "grad_norm": 0.3861207067966461,
+      "learning_rate": 2.9164e-05,
+      "loss": 0.009,
+      "step": 5220
+    },
+    {
+      "epoch": 33.9025974025974,
+      "grad_norm": 1.6413307189941406,
+      "learning_rate": 2.9163692307692306e-05,
+      "loss": 0.0211,
+      "step": 5221
+    },
+    {
+      "epoch": 33.90909090909091,
+      "grad_norm": 0.36325445771217346,
+      "learning_rate": 2.9163384615384617e-05,
+      "loss": 0.0074,
+      "step": 5222
+    },
+    {
+      "epoch": 33.91558441558441,
+      "grad_norm": 1.0768922567367554,
+      "learning_rate": 2.9163076923076926e-05,
+      "loss": 0.0164,
+      "step": 5223
+    },
+    {
+      "epoch": 33.922077922077925,
+      "grad_norm": 0.779722273349762,
+      "learning_rate": 2.916276923076923e-05,
+      "loss": 0.0171,
+      "step": 5224
+    },
+    {
+      "epoch": 33.92857142857143,
+      "grad_norm": 1.2277405261993408,
+      "learning_rate": 2.916246153846154e-05,
+      "loss": 0.0506,
+      "step": 5225
+    },
+    {
+      "epoch": 33.935064935064936,
+      "grad_norm": 0.6585797071456909,
+      "learning_rate": 2.9162153846153847e-05,
+      "loss": 0.0156,
+      "step": 5226
+    },
+    {
+      "epoch": 33.94155844155844,
+      "grad_norm": 0.23704443871974945,
+      "learning_rate": 2.9161846153846152e-05,
+      "loss": 0.0035,
+      "step": 5227
+    },
+    {
+      "epoch": 33.94805194805195,
+      "grad_norm": 0.6500463485717773,
+      "learning_rate": 2.9161538461538464e-05,
+      "loss": 0.014,
+      "step": 5228
+    },
+    {
+      "epoch": 33.95454545454545,
+      "grad_norm": 0.13933050632476807,
+      "learning_rate": 2.9161230769230772e-05,
+      "loss": 0.0018,
+      "step": 5229
+    },
+    {
+      "epoch": 33.96103896103896,
+      "grad_norm": 0.36695221066474915,
+      "learning_rate": 2.9160923076923077e-05,
+      "loss": 0.0051,
+      "step": 5230
+    },
+    {
+      "epoch": 33.967532467532465,
+      "grad_norm": 2.7713279724121094,
+      "learning_rate": 2.9160615384615385e-05,
+      "loss": 0.0064,
+      "step": 5231
+    },
+    {
+      "epoch": 33.97402597402598,
+      "grad_norm": 0.7228372693061829,
+      "learning_rate": 2.9160307692307694e-05,
+      "loss": 0.0071,
+      "step": 5232
+    },
+    {
+      "epoch": 33.98051948051948,
+      "grad_norm": 1.5581496953964233,
+      "learning_rate": 2.916e-05,
+      "loss": 0.0288,
+      "step": 5233
+    },
+    {
+      "epoch": 33.98701298701299,
+      "grad_norm": 0.8226633667945862,
+      "learning_rate": 2.9159692307692307e-05,
+      "loss": 0.016,
+      "step": 5234
+    },
+    {
+      "epoch": 33.993506493506494,
+      "grad_norm": 0.46550416946411133,
+      "learning_rate": 2.915938461538462e-05,
+      "loss": 0.0324,
+      "step": 5235
+    },
+    {
+      "epoch": 34.0,
+      "grad_norm": 0.998394787311554,
+      "learning_rate": 2.9159076923076924e-05,
+      "loss": 0.0125,
+      "step": 5236
+    },
+    {
+      "epoch": 34.006493506493506,
+      "grad_norm": 0.6651492118835449,
+      "learning_rate": 2.9158769230769232e-05,
+      "loss": 0.0943,
+      "step": 5237
+    },
+    {
+      "epoch": 34.01298701298701,
+      "grad_norm": 0.46950453519821167,
+      "learning_rate": 2.915846153846154e-05,
+      "loss": 0.0543,
+      "step": 5238
+    },
+    {
+      "epoch": 34.01948051948052,
+      "grad_norm": 0.4727919399738312,
+      "learning_rate": 2.9158153846153845e-05,
+      "loss": 0.0381,
+      "step": 5239
+    },
+    {
+      "epoch": 34.02597402597402,
+      "grad_norm": 0.5461315512657166,
+      "learning_rate": 2.9157846153846153e-05,
+      "loss": 0.043,
+      "step": 5240
+    },
+    {
+      "epoch": 34.032467532467535,
+      "grad_norm": 0.47372451424598694,
+      "learning_rate": 2.9157538461538462e-05,
+      "loss": 0.0339,
+      "step": 5241
+    },
+    {
+      "epoch": 34.03896103896104,
+      "grad_norm": 0.39935654401779175,
+      "learning_rate": 2.915723076923077e-05,
+      "loss": 0.0299,
+      "step": 5242
+    },
+    {
+      "epoch": 34.04545454545455,
+      "grad_norm": 0.4123666286468506,
+      "learning_rate": 2.915692307692308e-05,
+      "loss": 0.0372,
+      "step": 5243
+    },
+    {
+      "epoch": 34.05194805194805,
+      "grad_norm": 0.3065956234931946,
+      "learning_rate": 2.9156615384615387e-05,
+      "loss": 0.0207,
+      "step": 5244
+    },
+    {
+      "epoch": 34.05844155844156,
+      "grad_norm": 0.7890008687973022,
+      "learning_rate": 2.9156307692307692e-05,
+      "loss": 0.0284,
+      "step": 5245
+    },
+    {
+      "epoch": 34.064935064935064,
+      "grad_norm": 0.31738463044166565,
+      "learning_rate": 2.9156e-05,
+      "loss": 0.0159,
+      "step": 5246
+    },
+    {
+      "epoch": 34.07142857142857,
+      "grad_norm": 0.4964287281036377,
+      "learning_rate": 2.915569230769231e-05,
+      "loss": 0.0137,
+      "step": 5247
+    },
+    {
+      "epoch": 34.077922077922075,
+      "grad_norm": 0.38780248165130615,
+      "learning_rate": 2.9155384615384613e-05,
+      "loss": 0.0238,
+      "step": 5248
+    },
+    {
+      "epoch": 34.08441558441559,
+      "grad_norm": 0.3760470151901245,
+      "learning_rate": 2.9155076923076925e-05,
+      "loss": 0.0147,
+      "step": 5249
+    },
+    {
+      "epoch": 34.09090909090909,
+      "grad_norm": 0.5653063654899597,
+      "learning_rate": 2.9154769230769233e-05,
+      "loss": 0.0126,
+      "step": 5250
+    },
+    {
+      "epoch": 34.0974025974026,
+      "grad_norm": 0.43375396728515625,
+      "learning_rate": 2.9154461538461538e-05,
+      "loss": 0.0147,
+      "step": 5251
+    },
+    {
+      "epoch": 34.103896103896105,
+      "grad_norm": 0.3634866178035736,
+      "learning_rate": 2.9154153846153847e-05,
+      "loss": 0.0116,
+      "step": 5252
+    },
+    {
+      "epoch": 34.11038961038961,
+      "grad_norm": 0.4312509298324585,
+      "learning_rate": 2.9153846153846155e-05,
+      "loss": 0.018,
+      "step": 5253
+    },
+    {
+      "epoch": 34.116883116883116,
+      "grad_norm": 0.5749011635780334,
+      "learning_rate": 2.915353846153846e-05,
+      "loss": 0.0118,
+      "step": 5254
+    },
+    {
+      "epoch": 34.12337662337662,
+      "grad_norm": 0.31237083673477173,
+      "learning_rate": 2.9153230769230768e-05,
+      "loss": 0.0087,
+      "step": 5255
+    },
+    {
+      "epoch": 34.12987012987013,
+      "grad_norm": 0.4380163550376892,
+      "learning_rate": 2.915292307692308e-05,
+      "loss": 0.0067,
+      "step": 5256
+    },
+    {
+      "epoch": 34.13636363636363,
+      "grad_norm": 0.3805018663406372,
+      "learning_rate": 2.9152615384615385e-05,
+      "loss": 0.0106,
+      "step": 5257
+    },
+    {
+      "epoch": 34.142857142857146,
+      "grad_norm": 0.3207034766674042,
+      "learning_rate": 2.9152307692307693e-05,
+      "loss": 0.0078,
+      "step": 5258
+    },
+    {
+      "epoch": 34.14935064935065,
+      "grad_norm": 1.258542776107788,
+      "learning_rate": 2.9152e-05,
+      "loss": 0.0154,
+      "step": 5259
+    },
+    {
+      "epoch": 34.15584415584416,
+      "grad_norm": 0.3034595847129822,
+      "learning_rate": 2.915169230769231e-05,
+      "loss": 0.0168,
+      "step": 5260
+    },
+    {
+      "epoch": 34.16233766233766,
+      "grad_norm": 0.4040125012397766,
+      "learning_rate": 2.9151384615384615e-05,
+      "loss": 0.006,
+      "step": 5261
+    },
+    {
+      "epoch": 34.16883116883117,
+      "grad_norm": 0.31128889322280884,
+      "learning_rate": 2.9151076923076923e-05,
+      "loss": 0.0033,
+      "step": 5262
+    },
+    {
+      "epoch": 34.175324675324674,
+      "grad_norm": 0.4169544279575348,
+      "learning_rate": 2.9150769230769235e-05,
+      "loss": 0.0066,
+      "step": 5263
+    },
+    {
+      "epoch": 34.18181818181818,
+      "grad_norm": 0.6150069832801819,
+      "learning_rate": 2.915046153846154e-05,
+      "loss": 0.0267,
+      "step": 5264
+    },
+    {
+      "epoch": 34.188311688311686,
+      "grad_norm": 0.4691615104675293,
+      "learning_rate": 2.9150153846153848e-05,
+      "loss": 0.0078,
+      "step": 5265
+    },
+    {
+      "epoch": 34.1948051948052,
+      "grad_norm": 0.6778148412704468,
+      "learning_rate": 2.9149846153846156e-05,
+      "loss": 0.007,
+      "step": 5266
+    },
+    {
+      "epoch": 34.201298701298704,
+      "grad_norm": 2.118201494216919,
+      "learning_rate": 2.914953846153846e-05,
+      "loss": 0.0118,
+      "step": 5267
+    },
+    {
+      "epoch": 34.20779220779221,
+      "grad_norm": 0.4020325243473053,
+      "learning_rate": 2.914923076923077e-05,
+      "loss": 0.0115,
+      "step": 5268
+    },
+    {
+      "epoch": 34.214285714285715,
+      "grad_norm": 1.1559373140335083,
+      "learning_rate": 2.9148923076923078e-05,
+      "loss": 0.0153,
+      "step": 5269
+    },
+    {
+      "epoch": 34.22077922077922,
+      "grad_norm": 0.6937928199768066,
+      "learning_rate": 2.9148615384615386e-05,
+      "loss": 0.0072,
+      "step": 5270
+    },
+    {
+      "epoch": 34.22727272727273,
+      "grad_norm": 2.118891954421997,
+      "learning_rate": 2.9148307692307694e-05,
+      "loss": 0.0251,
+      "step": 5271
+    },
+    {
+      "epoch": 34.23376623376623,
+      "grad_norm": 0.7974960207939148,
+      "learning_rate": 2.9148000000000003e-05,
+      "loss": 0.0154,
+      "step": 5272
+    },
+    {
+      "epoch": 34.24025974025974,
+      "grad_norm": 1.1007604598999023,
+      "learning_rate": 2.9147692307692308e-05,
+      "loss": 0.0225,
+      "step": 5273
+    },
+    {
+      "epoch": 34.246753246753244,
+      "grad_norm": 0.738494873046875,
+      "learning_rate": 2.9147384615384616e-05,
+      "loss": 0.011,
+      "step": 5274
+    },
+    {
+      "epoch": 34.253246753246756,
+      "grad_norm": 0.9713312387466431,
+      "learning_rate": 2.9147076923076924e-05,
+      "loss": 0.1203,
+      "step": 5275
+    },
+    {
+      "epoch": 34.25974025974026,
+      "grad_norm": 0.6073378920555115,
+      "learning_rate": 2.914676923076923e-05,
+      "loss": 0.0688,
+      "step": 5276
+    },
+    {
+      "epoch": 34.26623376623377,
+      "grad_norm": 0.4947046935558319,
+      "learning_rate": 2.914646153846154e-05,
+      "loss": 0.0502,
+      "step": 5277
+    },
+    {
+      "epoch": 34.27272727272727,
+      "grad_norm": 0.6308859586715698,
+      "learning_rate": 2.914615384615385e-05,
+      "loss": 0.0436,
+      "step": 5278
+    },
+    {
+      "epoch": 34.27922077922078,
+      "grad_norm": 0.5347867012023926,
+      "learning_rate": 2.9145846153846154e-05,
+      "loss": 0.0448,
+      "step": 5279
+    },
+    {
+      "epoch": 34.285714285714285,
+      "grad_norm": 0.5856087803840637,
+      "learning_rate": 2.9145538461538463e-05,
+      "loss": 0.0383,
+      "step": 5280
+    },
+    {
+      "epoch": 34.29220779220779,
+      "grad_norm": 0.738502025604248,
+      "learning_rate": 2.914523076923077e-05,
+      "loss": 0.0346,
+      "step": 5281
+    },
+    {
+      "epoch": 34.298701298701296,
+      "grad_norm": 0.5556287169456482,
+      "learning_rate": 2.9144923076923076e-05,
+      "loss": 0.043,
+      "step": 5282
+    },
+    {
+      "epoch": 34.3051948051948,
+      "grad_norm": 0.49851131439208984,
+      "learning_rate": 2.9144615384615384e-05,
+      "loss": 0.0451,
+      "step": 5283
+    },
+    {
+      "epoch": 34.311688311688314,
+      "grad_norm": 0.41321220993995667,
+      "learning_rate": 2.9144307692307696e-05,
+      "loss": 0.0217,
+      "step": 5284
+    },
+    {
+      "epoch": 34.31818181818182,
+      "grad_norm": 0.5214315056800842,
+      "learning_rate": 2.9144e-05,
+      "loss": 0.0206,
+      "step": 5285
+    },
+    {
+      "epoch": 34.324675324675326,
+      "grad_norm": 0.5181014537811279,
+      "learning_rate": 2.914369230769231e-05,
+      "loss": 0.0236,
+      "step": 5286
+    },
+    {
+      "epoch": 34.33116883116883,
+      "grad_norm": 0.8338339328765869,
+      "learning_rate": 2.9143384615384617e-05,
+      "loss": 0.0177,
+      "step": 5287
+    },
+    {
+      "epoch": 34.33766233766234,
+      "grad_norm": 0.3325013816356659,
+      "learning_rate": 2.9143076923076922e-05,
+      "loss": 0.0176,
+      "step": 5288
+    },
+    {
+      "epoch": 34.34415584415584,
+      "grad_norm": 0.39470797777175903,
+      "learning_rate": 2.914276923076923e-05,
+      "loss": 0.0153,
+      "step": 5289
+    },
+    {
+      "epoch": 34.35064935064935,
+      "grad_norm": 0.7602513432502747,
+      "learning_rate": 2.914246153846154e-05,
+      "loss": 0.0463,
+      "step": 5290
+    },
+    {
+      "epoch": 34.357142857142854,
+      "grad_norm": 0.6509610414505005,
+      "learning_rate": 2.9142153846153847e-05,
+      "loss": 0.0179,
+      "step": 5291
+    },
+    {
+      "epoch": 34.36363636363637,
+      "grad_norm": 0.2593560814857483,
+      "learning_rate": 2.9141846153846156e-05,
+      "loss": 0.0103,
+      "step": 5292
+    },
+    {
+      "epoch": 34.37012987012987,
+      "grad_norm": 0.6391077041625977,
+      "learning_rate": 2.9141538461538464e-05,
+      "loss": 0.0128,
+      "step": 5293
+    },
+    {
+      "epoch": 34.37662337662338,
+      "grad_norm": 0.27730414271354675,
+      "learning_rate": 2.914123076923077e-05,
+      "loss": 0.0069,
+      "step": 5294
+    },
+    {
+      "epoch": 34.383116883116884,
+      "grad_norm": 0.9945710301399231,
+      "learning_rate": 2.9140923076923077e-05,
+      "loss": 0.0208,
+      "step": 5295
+    },
+    {
+      "epoch": 34.38961038961039,
+      "grad_norm": 0.37668702006340027,
+      "learning_rate": 2.9140615384615386e-05,
+      "loss": 0.0103,
+      "step": 5296
+    },
+    {
+      "epoch": 34.396103896103895,
+      "grad_norm": 0.4021940529346466,
+      "learning_rate": 2.914030769230769e-05,
+      "loss": 0.009,
+      "step": 5297
+    },
+    {
+      "epoch": 34.4025974025974,
+      "grad_norm": 2.7613797187805176,
+      "learning_rate": 2.9140000000000002e-05,
+      "loss": 0.0199,
+      "step": 5298
+    },
+    {
+      "epoch": 34.40909090909091,
+      "grad_norm": 0.5716807842254639,
+      "learning_rate": 2.913969230769231e-05,
+      "loss": 0.0085,
+      "step": 5299
+    },
+    {
+      "epoch": 34.41558441558441,
+      "grad_norm": 1.032850980758667,
+      "learning_rate": 2.9139384615384615e-05,
+      "loss": 0.0168,
+      "step": 5300
+    },
+    {
+      "epoch": 34.422077922077925,
+      "grad_norm": 1.0856531858444214,
+      "learning_rate": 2.9139076923076924e-05,
+      "loss": 0.0158,
+      "step": 5301
+    },
+    {
+      "epoch": 34.42857142857143,
+      "grad_norm": 0.9900875687599182,
+      "learning_rate": 2.9138769230769232e-05,
+      "loss": 0.0166,
+      "step": 5302
+    },
+    {
+      "epoch": 34.435064935064936,
+      "grad_norm": 1.2374485731124878,
+      "learning_rate": 2.9138461538461537e-05,
+      "loss": 0.0177,
+      "step": 5303
+    },
+    {
+      "epoch": 34.44155844155844,
+      "grad_norm": 0.9744414687156677,
+      "learning_rate": 2.9138153846153845e-05,
+      "loss": 0.0197,
+      "step": 5304
+    },
+    {
+      "epoch": 34.44805194805195,
+      "grad_norm": 0.5119012594223022,
+      "learning_rate": 2.9137846153846157e-05,
+      "loss": 0.0248,
+      "step": 5305
+    },
+    {
+      "epoch": 34.45454545454545,
+      "grad_norm": 0.8543151617050171,
+      "learning_rate": 2.9137538461538462e-05,
+      "loss": 0.0175,
+      "step": 5306
+    },
+    {
+      "epoch": 34.46103896103896,
+      "grad_norm": 0.604468822479248,
+      "learning_rate": 2.913723076923077e-05,
+      "loss": 0.0076,
+      "step": 5307
+    },
+    {
+      "epoch": 34.467532467532465,
+      "grad_norm": 1.545064926147461,
+      "learning_rate": 2.913692307692308e-05,
+      "loss": 0.0263,
+      "step": 5308
+    },
+    {
+      "epoch": 34.47402597402598,
+      "grad_norm": 0.7411176562309265,
+      "learning_rate": 2.9136615384615384e-05,
+      "loss": 0.0097,
+      "step": 5309
+    },
+    {
+      "epoch": 34.48051948051948,
+      "grad_norm": 1.1311073303222656,
+      "learning_rate": 2.9136307692307692e-05,
+      "loss": 0.0079,
+      "step": 5310
+    },
+    {
+      "epoch": 34.48701298701299,
+      "grad_norm": 0.7269066572189331,
+      "learning_rate": 2.9136e-05,
+      "loss": 0.0101,
+      "step": 5311
+    },
+    {
+      "epoch": 34.493506493506494,
+      "grad_norm": 2.4829859733581543,
+      "learning_rate": 2.913569230769231e-05,
+      "loss": 0.0583,
+      "step": 5312
+    },
+    {
+      "epoch": 34.5,
+      "grad_norm": 0.8113126754760742,
+      "learning_rate": 2.9135384615384617e-05,
+      "loss": 0.1129,
+      "step": 5313
+    },
+    {
+      "epoch": 34.506493506493506,
+      "grad_norm": 0.629552960395813,
+      "learning_rate": 2.9135076923076925e-05,
+      "loss": 0.0795,
+      "step": 5314
+    },
+    {
+      "epoch": 34.51298701298701,
+      "grad_norm": 0.5981005430221558,
+      "learning_rate": 2.913476923076923e-05,
+      "loss": 0.0641,
+      "step": 5315
+    },
+    {
+      "epoch": 34.51948051948052,
+      "grad_norm": 0.49797946214675903,
+      "learning_rate": 2.913446153846154e-05,
+      "loss": 0.0532,
+      "step": 5316
+    },
+    {
+      "epoch": 34.52597402597402,
+      "grad_norm": 0.343521386384964,
+      "learning_rate": 2.9134153846153847e-05,
+      "loss": 0.0343,
+      "step": 5317
+    },
+    {
+      "epoch": 34.532467532467535,
+      "grad_norm": 0.444717139005661,
+      "learning_rate": 2.913384615384615e-05,
+      "loss": 0.0259,
+      "step": 5318
+    },
+    {
+      "epoch": 34.53896103896104,
+      "grad_norm": 0.5592092275619507,
+      "learning_rate": 2.9133538461538463e-05,
+      "loss": 0.0272,
+      "step": 5319
+    },
+    {
+      "epoch": 34.54545454545455,
+      "grad_norm": 0.3717966675758362,
+      "learning_rate": 2.9133230769230772e-05,
+      "loss": 0.0221,
+      "step": 5320
+    },
+    {
+      "epoch": 34.55194805194805,
+      "grad_norm": 0.44754713773727417,
+      "learning_rate": 2.9132923076923077e-05,
+      "loss": 0.0266,
+      "step": 5321
+    },
+    {
+      "epoch": 34.55844155844156,
+      "grad_norm": 0.5954673886299133,
+      "learning_rate": 2.9132615384615385e-05,
+      "loss": 0.0316,
+      "step": 5322
+    },
+    {
+      "epoch": 34.564935064935064,
+      "grad_norm": 0.48587605357170105,
+      "learning_rate": 2.9132307692307693e-05,
+      "loss": 0.0191,
+      "step": 5323
+    },
+    {
+      "epoch": 34.57142857142857,
+      "grad_norm": 0.4678362011909485,
+      "learning_rate": 2.9131999999999998e-05,
+      "loss": 0.0217,
+      "step": 5324
+    },
+    {
+      "epoch": 34.577922077922075,
+      "grad_norm": 0.432365745306015,
+      "learning_rate": 2.9131692307692307e-05,
+      "loss": 0.0195,
+      "step": 5325
+    },
+    {
+      "epoch": 34.58441558441559,
+      "grad_norm": 0.8224279284477234,
+      "learning_rate": 2.9131384615384618e-05,
+      "loss": 0.0181,
+      "step": 5326
+    },
+    {
+      "epoch": 34.59090909090909,
+      "grad_norm": 0.43323376774787903,
+      "learning_rate": 2.9131076923076923e-05,
+      "loss": 0.0289,
+      "step": 5327
+    },
+    {
+      "epoch": 34.5974025974026,
+      "grad_norm": 0.6326972246170044,
+      "learning_rate": 2.913076923076923e-05,
+      "loss": 0.0132,
+      "step": 5328
+    },
+    {
+      "epoch": 34.603896103896105,
+      "grad_norm": 0.557214081287384,
+      "learning_rate": 2.913046153846154e-05,
+      "loss": 0.0189,
+      "step": 5329
+    },
+    {
+      "epoch": 34.61038961038961,
+      "grad_norm": 0.5050615668296814,
+      "learning_rate": 2.9130153846153845e-05,
+      "loss": 0.0162,
+      "step": 5330
+    },
+    {
+      "epoch": 34.616883116883116,
+      "grad_norm": 0.27772578597068787,
+      "learning_rate": 2.9129846153846153e-05,
+      "loss": 0.0078,
+      "step": 5331
+    },
+    {
+      "epoch": 34.62337662337662,
+      "grad_norm": 1.832878828048706,
+      "learning_rate": 2.912953846153846e-05,
+      "loss": 0.0253,
+      "step": 5332
+    },
+    {
+      "epoch": 34.62987012987013,
+      "grad_norm": 0.5369575023651123,
+      "learning_rate": 2.912923076923077e-05,
+      "loss": 0.0113,
+      "step": 5333
+    },
+    {
+      "epoch": 34.63636363636363,
+      "grad_norm": 0.5768107771873474,
+      "learning_rate": 2.9128923076923078e-05,
+      "loss": 0.0155,
+      "step": 5334
+    },
+    {
+      "epoch": 34.642857142857146,
+      "grad_norm": 0.7064384818077087,
+      "learning_rate": 2.9128615384615386e-05,
+      "loss": 0.0141,
+      "step": 5335
+    },
+    {
+      "epoch": 34.64935064935065,
+      "grad_norm": 0.5787583589553833,
+      "learning_rate": 2.912830769230769e-05,
+      "loss": 0.0215,
+      "step": 5336
+    },
+    {
+      "epoch": 34.65584415584416,
+      "grad_norm": 0.33400586247444153,
+      "learning_rate": 2.9128e-05,
+      "loss": 0.0124,
+      "step": 5337
+    },
+    {
+      "epoch": 34.66233766233766,
+      "grad_norm": 0.31130489706993103,
+      "learning_rate": 2.9127692307692308e-05,
+      "loss": 0.0054,
+      "step": 5338
+    },
+    {
+      "epoch": 34.66883116883117,
+      "grad_norm": 0.5748124122619629,
+      "learning_rate": 2.9127384615384616e-05,
+      "loss": 0.0483,
+      "step": 5339
+    },
+    {
+      "epoch": 34.675324675324674,
+      "grad_norm": 0.3011142909526825,
+      "learning_rate": 2.9127076923076925e-05,
+      "loss": 0.0038,
+      "step": 5340
+    },
+    {
+      "epoch": 34.68181818181818,
+      "grad_norm": 1.5520853996276855,
+      "learning_rate": 2.9126769230769233e-05,
+      "loss": 0.0281,
+      "step": 5341
+    },
+    {
+      "epoch": 34.688311688311686,
+      "grad_norm": 0.6826903223991394,
+      "learning_rate": 2.912646153846154e-05,
+      "loss": 0.0054,
+      "step": 5342
+    },
+    {
+      "epoch": 34.6948051948052,
+      "grad_norm": 0.41030851006507874,
+      "learning_rate": 2.9126153846153846e-05,
+      "loss": 0.0091,
+      "step": 5343
+    },
+    {
+      "epoch": 34.701298701298704,
+      "grad_norm": 0.5148544907569885,
+      "learning_rate": 2.9125846153846154e-05,
+      "loss": 0.008,
+      "step": 5344
+    },
+    {
+      "epoch": 34.70779220779221,
+      "grad_norm": 0.8189192414283752,
+      "learning_rate": 2.9125538461538463e-05,
+      "loss": 0.0154,
+      "step": 5345
+    },
+    {
+      "epoch": 34.714285714285715,
+      "grad_norm": 0.5207547545433044,
+      "learning_rate": 2.9125230769230768e-05,
+      "loss": 0.0116,
+      "step": 5346
+    },
+    {
+      "epoch": 34.72077922077922,
+      "grad_norm": 0.8722161650657654,
+      "learning_rate": 2.912492307692308e-05,
+      "loss": 0.0145,
+      "step": 5347
+    },
+    {
+      "epoch": 34.72727272727273,
+      "grad_norm": 1.0792169570922852,
+      "learning_rate": 2.9124615384615388e-05,
+      "loss": 0.0093,
+      "step": 5348
+    },
+    {
+      "epoch": 34.73376623376623,
+      "grad_norm": 0.5235753059387207,
+      "learning_rate": 2.9124307692307693e-05,
+      "loss": 0.0189,
+      "step": 5349
+    },
+    {
+      "epoch": 34.74025974025974,
+      "grad_norm": 0.940521776676178,
+      "learning_rate": 2.9124e-05,
+      "loss": 0.0161,
+      "step": 5350
+    },
+    {
+      "epoch": 34.746753246753244,
+      "grad_norm": 0.7171229720115662,
+      "learning_rate": 2.912369230769231e-05,
+      "loss": 0.0966,
+      "step": 5351
+    },
+    {
+      "epoch": 34.753246753246756,
+      "grad_norm": 0.508734941482544,
+      "learning_rate": 2.9123384615384614e-05,
+      "loss": 0.0641,
+      "step": 5352
+    },
+    {
+      "epoch": 34.75974025974026,
+      "grad_norm": 0.49755552411079407,
+      "learning_rate": 2.9123076923076923e-05,
+      "loss": 0.0603,
+      "step": 5353
+    },
+    {
+      "epoch": 34.76623376623377,
+      "grad_norm": 0.417592853307724,
+      "learning_rate": 2.9122769230769234e-05,
+      "loss": 0.0433,
+      "step": 5354
+    },
+    {
+      "epoch": 34.77272727272727,
+      "grad_norm": 0.5044029951095581,
+      "learning_rate": 2.912246153846154e-05,
+      "loss": 0.051,
+      "step": 5355
+    },
+    {
+      "epoch": 34.77922077922078,
+      "grad_norm": 0.5635175704956055,
+      "learning_rate": 2.9122153846153848e-05,
+      "loss": 0.0323,
+      "step": 5356
+    },
+    {
+      "epoch": 34.785714285714285,
+      "grad_norm": 0.6393215656280518,
+      "learning_rate": 2.9121846153846156e-05,
+      "loss": 0.0404,
+      "step": 5357
+    },
+    {
+      "epoch": 34.79220779220779,
+      "grad_norm": 0.7523524761199951,
+      "learning_rate": 2.912153846153846e-05,
+      "loss": 0.033,
+      "step": 5358
+    },
+    {
+      "epoch": 34.798701298701296,
+      "grad_norm": 0.38730788230895996,
+      "learning_rate": 2.912123076923077e-05,
+      "loss": 0.0224,
+      "step": 5359
+    },
+    {
+      "epoch": 34.8051948051948,
+      "grad_norm": 0.45913487672805786,
+      "learning_rate": 2.9120923076923077e-05,
+      "loss": 0.0222,
+      "step": 5360
+    },
+    {
+      "epoch": 34.811688311688314,
+      "grad_norm": 0.3380180597305298,
+      "learning_rate": 2.9120615384615386e-05,
+      "loss": 0.0195,
+      "step": 5361
+    },
+    {
+      "epoch": 34.81818181818182,
+      "grad_norm": 0.2975977957248688,
+      "learning_rate": 2.9120307692307694e-05,
+      "loss": 0.0157,
+      "step": 5362
+    },
+    {
+      "epoch": 34.824675324675326,
+      "grad_norm": 0.3227158188819885,
+      "learning_rate": 2.9120000000000002e-05,
+      "loss": 0.018,
+      "step": 5363
+    },
+    {
+      "epoch": 34.83116883116883,
+      "grad_norm": 0.5855149626731873,
+      "learning_rate": 2.9119692307692307e-05,
+      "loss": 0.0173,
+      "step": 5364
+    },
+    {
+      "epoch": 34.83766233766234,
+      "grad_norm": 0.46049290895462036,
+      "learning_rate": 2.9119384615384616e-05,
+      "loss": 0.0155,
+      "step": 5365
+    },
+    {
+      "epoch": 34.84415584415584,
+      "grad_norm": 0.7055251598358154,
+      "learning_rate": 2.9119076923076924e-05,
+      "loss": 0.0163,
+      "step": 5366
+    },
+    {
+      "epoch": 34.85064935064935,
+      "grad_norm": 0.6814199090003967,
+      "learning_rate": 2.911876923076923e-05,
+      "loss": 0.0232,
+      "step": 5367
+    },
+    {
+      "epoch": 34.857142857142854,
+      "grad_norm": 0.5258548259735107,
+      "learning_rate": 2.911846153846154e-05,
+      "loss": 0.0164,
+      "step": 5368
+    },
+    {
+      "epoch": 34.86363636363637,
+      "grad_norm": 0.9523411989212036,
+      "learning_rate": 2.911815384615385e-05,
+      "loss": 0.0211,
+      "step": 5369
+    },
+    {
+      "epoch": 34.87012987012987,
+      "grad_norm": 0.5174683928489685,
+      "learning_rate": 2.9117846153846154e-05,
+      "loss": 0.0129,
+      "step": 5370
+    },
+    {
+      "epoch": 34.87662337662338,
+      "grad_norm": 0.561558187007904,
+      "learning_rate": 2.9117538461538462e-05,
+      "loss": 0.0147,
+      "step": 5371
+    },
+    {
+      "epoch": 34.883116883116884,
+      "grad_norm": 2.3458762168884277,
+      "learning_rate": 2.911723076923077e-05,
+      "loss": 0.0175,
+      "step": 5372
+    },
+    {
+      "epoch": 34.88961038961039,
+      "grad_norm": 0.3211030960083008,
+      "learning_rate": 2.9116923076923075e-05,
+      "loss": 0.0056,
+      "step": 5373
+    },
+    {
+      "epoch": 34.896103896103895,
+      "grad_norm": 0.8096930980682373,
+      "learning_rate": 2.9116615384615384e-05,
+      "loss": 0.0132,
+      "step": 5374
+    },
+    {
+      "epoch": 34.9025974025974,
+      "grad_norm": 0.5573617815971375,
+      "learning_rate": 2.9116307692307695e-05,
+      "loss": 0.0292,
+      "step": 5375
+    },
+    {
+      "epoch": 34.90909090909091,
+      "grad_norm": 0.47696778178215027,
+      "learning_rate": 2.9116e-05,
+      "loss": 0.0063,
+      "step": 5376
+    },
+    {
+      "epoch": 34.91558441558441,
+      "grad_norm": 1.0395174026489258,
+      "learning_rate": 2.911569230769231e-05,
+      "loss": 0.0169,
+      "step": 5377
+    },
+    {
+      "epoch": 34.922077922077925,
+      "grad_norm": 0.4949539005756378,
+      "learning_rate": 2.9115384615384617e-05,
+      "loss": 0.0058,
+      "step": 5378
+    },
+    {
+      "epoch": 34.92857142857143,
+      "grad_norm": 0.8717275857925415,
+      "learning_rate": 2.9115076923076922e-05,
+      "loss": 0.0383,
+      "step": 5379
+    },
+    {
+      "epoch": 34.935064935064936,
+      "grad_norm": 0.3513164818286896,
+      "learning_rate": 2.911476923076923e-05,
+      "loss": 0.0057,
+      "step": 5380
+    },
+    {
+      "epoch": 34.94155844155844,
+      "grad_norm": 0.31253522634506226,
+      "learning_rate": 2.911446153846154e-05,
+      "loss": 0.0042,
+      "step": 5381
+    },
+    {
+      "epoch": 34.94805194805195,
+      "grad_norm": 0.5153536796569824,
+      "learning_rate": 2.9114153846153847e-05,
+      "loss": 0.0461,
+      "step": 5382
+    },
+    {
+      "epoch": 34.95454545454545,
+      "grad_norm": 1.0771937370300293,
+      "learning_rate": 2.9113846153846155e-05,
+      "loss": 0.0393,
+      "step": 5383
+    },
+    {
+      "epoch": 34.96103896103896,
+      "grad_norm": 0.6515333652496338,
+      "learning_rate": 2.9113538461538464e-05,
+      "loss": 0.0106,
+      "step": 5384
+    },
+    {
+      "epoch": 34.967532467532465,
+      "grad_norm": 1.4070911407470703,
+      "learning_rate": 2.911323076923077e-05,
+      "loss": 0.0234,
+      "step": 5385
+    },
+    {
+      "epoch": 34.97402597402598,
+      "grad_norm": 1.3474302291870117,
+      "learning_rate": 2.9112923076923077e-05,
+      "loss": 0.0124,
+      "step": 5386
+    },
+    {
+      "epoch": 34.98051948051948,
+      "grad_norm": 1.4982504844665527,
+      "learning_rate": 2.9112615384615385e-05,
+      "loss": 0.0242,
+      "step": 5387
+    },
+    {
+      "epoch": 34.98701298701299,
+      "grad_norm": 0.6688902378082275,
+      "learning_rate": 2.911230769230769e-05,
+      "loss": 0.0087,
+      "step": 5388
+    },
+    {
+      "epoch": 34.993506493506494,
+      "grad_norm": 1.5412135124206543,
+      "learning_rate": 2.9112000000000002e-05,
+      "loss": 0.0358,
+      "step": 5389
+    },
+    {
+      "epoch": 35.0,
+      "grad_norm": 0.770210325717926,
+      "learning_rate": 2.911169230769231e-05,
+      "loss": 0.0153,
+      "step": 5390
+    },
+    {
+      "epoch": 35.006493506493506,
+      "grad_norm": 0.6208969354629517,
+      "learning_rate": 2.9111384615384615e-05,
+      "loss": 0.0743,
+      "step": 5391
+    },
+    {
+      "epoch": 35.01298701298701,
+      "grad_norm": 0.46510785818099976,
+      "learning_rate": 2.9111076923076923e-05,
+      "loss": 0.0492,
+      "step": 5392
+    },
+    {
+      "epoch": 35.01948051948052,
+      "grad_norm": 0.4651740789413452,
+      "learning_rate": 2.9110769230769232e-05,
+      "loss": 0.0589,
+      "step": 5393
+    },
+    {
+      "epoch": 35.02597402597402,
+      "grad_norm": 0.3621070981025696,
+      "learning_rate": 2.9110461538461537e-05,
+      "loss": 0.0424,
+      "step": 5394
+    },
+    {
+      "epoch": 35.032467532467535,
+      "grad_norm": 0.511978268623352,
+      "learning_rate": 2.9110153846153845e-05,
+      "loss": 0.0281,
+      "step": 5395
+    },
+    {
+      "epoch": 35.03896103896104,
+      "grad_norm": 0.3234733045101166,
+      "learning_rate": 2.9109846153846157e-05,
+      "loss": 0.024,
+      "step": 5396
+    },
+    {
+      "epoch": 35.04545454545455,
+      "grad_norm": 0.4431886076927185,
+      "learning_rate": 2.910953846153846e-05,
+      "loss": 0.0244,
+      "step": 5397
+    },
+    {
+      "epoch": 35.05194805194805,
+      "grad_norm": 0.544076144695282,
+      "learning_rate": 2.910923076923077e-05,
+      "loss": 0.0217,
+      "step": 5398
+    },
+    {
+      "epoch": 35.05844155844156,
+      "grad_norm": 0.5601357221603394,
+      "learning_rate": 2.9108923076923078e-05,
+      "loss": 0.0253,
+      "step": 5399
+    },
+    {
+      "epoch": 35.064935064935064,
+      "grad_norm": 0.6342061161994934,
+      "learning_rate": 2.9108615384615383e-05,
+      "loss": 0.0189,
+      "step": 5400
+    },
+    {
+      "epoch": 35.07142857142857,
+      "grad_norm": 0.5895953178405762,
+      "learning_rate": 2.910830769230769e-05,
+      "loss": 0.0215,
+      "step": 5401
+    },
+    {
+      "epoch": 35.077922077922075,
+      "grad_norm": 0.6751048564910889,
+      "learning_rate": 2.9108000000000003e-05,
+      "loss": 0.0158,
+      "step": 5402
+    },
+    {
+      "epoch": 35.08441558441559,
+      "grad_norm": 0.27848485112190247,
+      "learning_rate": 2.9107692307692308e-05,
+      "loss": 0.0107,
+      "step": 5403
+    },
+    {
+      "epoch": 35.09090909090909,
+      "grad_norm": 0.2510458827018738,
+      "learning_rate": 2.9107384615384616e-05,
+      "loss": 0.0096,
+      "step": 5404
+    },
+    {
+      "epoch": 35.0974025974026,
+      "grad_norm": 0.5650168061256409,
+      "learning_rate": 2.9107076923076925e-05,
+      "loss": 0.0135,
+      "step": 5405
+    },
+    {
+      "epoch": 35.103896103896105,
+      "grad_norm": 0.565434992313385,
+      "learning_rate": 2.910676923076923e-05,
+      "loss": 0.0219,
+      "step": 5406
+    },
+    {
+      "epoch": 35.11038961038961,
+      "grad_norm": 0.4377562701702118,
+      "learning_rate": 2.9106461538461538e-05,
+      "loss": 0.0101,
+      "step": 5407
+    },
+    {
+      "epoch": 35.116883116883116,
+      "grad_norm": 0.5619506239891052,
+      "learning_rate": 2.9106153846153846e-05,
+      "loss": 0.0117,
+      "step": 5408
+    },
+    {
+      "epoch": 35.12337662337662,
+      "grad_norm": 0.8575261831283569,
+      "learning_rate": 2.9105846153846155e-05,
+      "loss": 0.0224,
+      "step": 5409
+    },
+    {
+      "epoch": 35.12987012987013,
+      "grad_norm": 0.3889850080013275,
+      "learning_rate": 2.9105538461538463e-05,
+      "loss": 0.0048,
+      "step": 5410
+    },
+    {
+      "epoch": 35.13636363636363,
+      "grad_norm": 0.4632609188556671,
+      "learning_rate": 2.910523076923077e-05,
+      "loss": 0.0111,
+      "step": 5411
+    },
+    {
+      "epoch": 35.142857142857146,
+      "grad_norm": 0.3049633502960205,
+      "learning_rate": 2.9104923076923076e-05,
+      "loss": 0.0064,
+      "step": 5412
+    },
+    {
+      "epoch": 35.14935064935065,
+      "grad_norm": 0.44959554076194763,
+      "learning_rate": 2.9104615384615385e-05,
+      "loss": 0.0073,
+      "step": 5413
+    },
+    {
+      "epoch": 35.15584415584416,
+      "grad_norm": 0.7940086722373962,
+      "learning_rate": 2.9104307692307693e-05,
+      "loss": 0.0287,
+      "step": 5414
+    },
+    {
+      "epoch": 35.16233766233766,
+      "grad_norm": 0.401343435049057,
+      "learning_rate": 2.9103999999999998e-05,
+      "loss": 0.0206,
+      "step": 5415
+    },
+    {
+      "epoch": 35.16883116883117,
+      "grad_norm": 0.22474849224090576,
+      "learning_rate": 2.910369230769231e-05,
+      "loss": 0.0053,
+      "step": 5416
+    },
+    {
+      "epoch": 35.175324675324674,
+      "grad_norm": 1.3230173587799072,
+      "learning_rate": 2.9103384615384618e-05,
+      "loss": 0.048,
+      "step": 5417
+    },
+    {
+      "epoch": 35.18181818181818,
+      "grad_norm": 0.7631206512451172,
+      "learning_rate": 2.9103076923076923e-05,
+      "loss": 0.0121,
+      "step": 5418
+    },
+    {
+      "epoch": 35.188311688311686,
+      "grad_norm": 0.4835094213485718,
+      "learning_rate": 2.910276923076923e-05,
+      "loss": 0.0053,
+      "step": 5419
+    },
+    {
+      "epoch": 35.1948051948052,
+      "grad_norm": 0.7741308212280273,
+      "learning_rate": 2.910246153846154e-05,
+      "loss": 0.015,
+      "step": 5420
+    },
+    {
+      "epoch": 35.201298701298704,
+      "grad_norm": 0.5356241464614868,
+      "learning_rate": 2.9102153846153848e-05,
+      "loss": 0.0058,
+      "step": 5421
+    },
+    {
+      "epoch": 35.20779220779221,
+      "grad_norm": 0.994607150554657,
+      "learning_rate": 2.9101846153846153e-05,
+      "loss": 0.0225,
+      "step": 5422
+    },
+    {
+      "epoch": 35.214285714285715,
+      "grad_norm": 0.789943516254425,
+      "learning_rate": 2.9101538461538464e-05,
+      "loss": 0.0095,
+      "step": 5423
+    },
+    {
+      "epoch": 35.22077922077922,
+      "grad_norm": 1.6964573860168457,
+      "learning_rate": 2.9101230769230773e-05,
+      "loss": 0.0216,
+      "step": 5424
+    },
+    {
+      "epoch": 35.22727272727273,
+      "grad_norm": 21.47344207763672,
+      "learning_rate": 2.9100923076923078e-05,
+      "loss": 0.0557,
+      "step": 5425
+    },
+    {
+      "epoch": 35.23376623376623,
+      "grad_norm": 0.30064910650253296,
+      "learning_rate": 2.9100615384615386e-05,
+      "loss": 0.0032,
+      "step": 5426
+    },
+    {
+      "epoch": 35.24025974025974,
+      "grad_norm": 0.12497668713331223,
+      "learning_rate": 2.9100307692307694e-05,
+      "loss": 0.0017,
+      "step": 5427
+    },
+    {
+      "epoch": 35.246753246753244,
+      "grad_norm": 0.7868456840515137,
+      "learning_rate": 2.91e-05,
+      "loss": 0.0458,
+      "step": 5428
+    },
+    {
+      "epoch": 35.253246753246756,
+      "grad_norm": 1.1124745607376099,
+      "learning_rate": 2.9099692307692308e-05,
+      "loss": 0.1121,
+      "step": 5429
+    },
+    {
+      "epoch": 35.25974025974026,
+      "grad_norm": 0.5016067028045654,
+      "learning_rate": 2.909938461538462e-05,
+      "loss": 0.0568,
+      "step": 5430
+    },
+    {
+      "epoch": 35.26623376623377,
+      "grad_norm": 0.6188507676124573,
+      "learning_rate": 2.9099076923076924e-05,
+      "loss": 0.0366,
+      "step": 5431
+    },
+    {
+      "epoch": 35.27272727272727,
+      "grad_norm": 0.4792235493659973,
+      "learning_rate": 2.9098769230769232e-05,
+      "loss": 0.0424,
+      "step": 5432
+    },
+    {
+      "epoch": 35.27922077922078,
+      "grad_norm": 0.39298224449157715,
+      "learning_rate": 2.909846153846154e-05,
+      "loss": 0.0355,
+      "step": 5433
+    },
+    {
+      "epoch": 35.285714285714285,
+      "grad_norm": 0.3892764151096344,
+      "learning_rate": 2.9098153846153846e-05,
+      "loss": 0.0255,
+      "step": 5434
+    },
+    {
+      "epoch": 35.29220779220779,
+      "grad_norm": 0.39262932538986206,
+      "learning_rate": 2.9097846153846154e-05,
+      "loss": 0.0347,
+      "step": 5435
+    },
+    {
+      "epoch": 35.298701298701296,
+      "grad_norm": 0.4668377637863159,
+      "learning_rate": 2.9097538461538462e-05,
+      "loss": 0.0227,
+      "step": 5436
+    },
+    {
+      "epoch": 35.3051948051948,
+      "grad_norm": 0.5487108826637268,
+      "learning_rate": 2.909723076923077e-05,
+      "loss": 0.0213,
+      "step": 5437
+    },
+    {
+      "epoch": 35.311688311688314,
+      "grad_norm": 0.49824461340904236,
+      "learning_rate": 2.909692307692308e-05,
+      "loss": 0.0236,
+      "step": 5438
+    },
+    {
+      "epoch": 35.31818181818182,
+      "grad_norm": 0.7446942925453186,
+      "learning_rate": 2.9096615384615387e-05,
+      "loss": 0.0254,
+      "step": 5439
+    },
+    {
+      "epoch": 35.324675324675326,
+      "grad_norm": 0.4612240195274353,
+      "learning_rate": 2.9096307692307692e-05,
+      "loss": 0.0164,
+      "step": 5440
+    },
+    {
+      "epoch": 35.33116883116883,
+      "grad_norm": 0.514288604259491,
+      "learning_rate": 2.9096e-05,
+      "loss": 0.0197,
+      "step": 5441
+    },
+    {
+      "epoch": 35.33766233766234,
+      "grad_norm": 0.328241229057312,
+      "learning_rate": 2.909569230769231e-05,
+      "loss": 0.0089,
+      "step": 5442
+    },
+    {
+      "epoch": 35.34415584415584,
+      "grad_norm": 0.49577921628952026,
+      "learning_rate": 2.9095384615384614e-05,
+      "loss": 0.0097,
+      "step": 5443
+    },
+    {
+      "epoch": 35.35064935064935,
+      "grad_norm": 0.27970683574676514,
+      "learning_rate": 2.9095076923076926e-05,
+      "loss": 0.0095,
+      "step": 5444
+    },
+    {
+      "epoch": 35.357142857142854,
+      "grad_norm": 0.5652754306793213,
+      "learning_rate": 2.9094769230769234e-05,
+      "loss": 0.0114,
+      "step": 5445
+    },
+    {
+      "epoch": 35.36363636363637,
+      "grad_norm": 0.7900232672691345,
+      "learning_rate": 2.909446153846154e-05,
+      "loss": 0.0162,
+      "step": 5446
+    },
+    {
+      "epoch": 35.37012987012987,
+      "grad_norm": 0.5476013422012329,
+      "learning_rate": 2.9094153846153847e-05,
+      "loss": 0.0164,
+      "step": 5447
+    },
+    {
+      "epoch": 35.37662337662338,
+      "grad_norm": 0.32560667395591736,
+      "learning_rate": 2.9093846153846155e-05,
+      "loss": 0.0149,
+      "step": 5448
+    },
+    {
+      "epoch": 35.383116883116884,
+      "grad_norm": 0.2903243899345398,
+      "learning_rate": 2.909353846153846e-05,
+      "loss": 0.0053,
+      "step": 5449
+    },
+    {
+      "epoch": 35.38961038961039,
+      "grad_norm": 0.5862712264060974,
+      "learning_rate": 2.909323076923077e-05,
+      "loss": 0.0064,
+      "step": 5450
+    },
+    {
+      "epoch": 35.396103896103895,
+      "grad_norm": 0.44692492485046387,
+      "learning_rate": 2.909292307692308e-05,
+      "loss": 0.012,
+      "step": 5451
+    },
+    {
+      "epoch": 35.4025974025974,
+      "grad_norm": 0.35976046323776245,
+      "learning_rate": 2.9092615384615385e-05,
+      "loss": 0.0064,
+      "step": 5452
+    },
+    {
+      "epoch": 35.40909090909091,
+      "grad_norm": 0.5869482159614563,
+      "learning_rate": 2.9092307692307694e-05,
+      "loss": 0.0128,
+      "step": 5453
+    },
+    {
+      "epoch": 35.41558441558441,
+      "grad_norm": 0.48824265599250793,
+      "learning_rate": 2.9092000000000002e-05,
+      "loss": 0.0054,
+      "step": 5454
+    },
+    {
+      "epoch": 35.422077922077925,
+      "grad_norm": 0.7205779552459717,
+      "learning_rate": 2.9091692307692307e-05,
+      "loss": 0.0133,
+      "step": 5455
+    },
+    {
+      "epoch": 35.42857142857143,
+      "grad_norm": 0.9559537172317505,
+      "learning_rate": 2.9091384615384615e-05,
+      "loss": 0.0214,
+      "step": 5456
+    },
+    {
+      "epoch": 35.435064935064936,
+      "grad_norm": 0.31414738297462463,
+      "learning_rate": 2.9091076923076924e-05,
+      "loss": 0.0053,
+      "step": 5457
+    },
+    {
+      "epoch": 35.44155844155844,
+      "grad_norm": 0.20352095365524292,
+      "learning_rate": 2.9090769230769232e-05,
+      "loss": 0.0052,
+      "step": 5458
+    },
+    {
+      "epoch": 35.44805194805195,
+      "grad_norm": 0.8375682234764099,
+      "learning_rate": 2.909046153846154e-05,
+      "loss": 0.009,
+      "step": 5459
+    },
+    {
+      "epoch": 35.45454545454545,
+      "grad_norm": 0.6300340890884399,
+      "learning_rate": 2.909015384615385e-05,
+      "loss": 0.0037,
+      "step": 5460
+    },
+    {
+      "epoch": 35.46103896103896,
+      "grad_norm": 0.614241898059845,
+      "learning_rate": 2.9089846153846153e-05,
+      "loss": 0.0032,
+      "step": 5461
+    },
+    {
+      "epoch": 35.467532467532465,
+      "grad_norm": 0.464739590883255,
+      "learning_rate": 2.9089538461538462e-05,
+      "loss": 0.0052,
+      "step": 5462
+    },
+    {
+      "epoch": 35.47402597402598,
+      "grad_norm": 0.18681010603904724,
+      "learning_rate": 2.908923076923077e-05,
+      "loss": 0.0024,
+      "step": 5463
+    },
+    {
+      "epoch": 35.48051948051948,
+      "grad_norm": 0.9883227348327637,
+      "learning_rate": 2.9088923076923075e-05,
+      "loss": 0.029,
+      "step": 5464
+    },
+    {
+      "epoch": 35.48701298701299,
+      "grad_norm": 0.5163398385047913,
+      "learning_rate": 2.9088615384615387e-05,
+      "loss": 0.0048,
+      "step": 5465
+    },
+    {
+      "epoch": 35.493506493506494,
+      "grad_norm": 1.1780750751495361,
+      "learning_rate": 2.9088307692307695e-05,
+      "loss": 0.0088,
+      "step": 5466
+    },
+    {
+      "epoch": 35.5,
+      "grad_norm": 0.8101599812507629,
+      "learning_rate": 2.9088e-05,
+      "loss": 0.0873,
+      "step": 5467
+    },
+    {
+      "epoch": 35.506493506493506,
+      "grad_norm": 0.5987259745597839,
+      "learning_rate": 2.908769230769231e-05,
+      "loss": 0.0667,
+      "step": 5468
+    },
+    {
+      "epoch": 35.51298701298701,
+      "grad_norm": 0.44257867336273193,
+      "learning_rate": 2.9087384615384617e-05,
+      "loss": 0.0408,
+      "step": 5469
+    },
+    {
+      "epoch": 35.51948051948052,
+      "grad_norm": 0.38900884985923767,
+      "learning_rate": 2.908707692307692e-05,
+      "loss": 0.0362,
+      "step": 5470
+    },
+    {
+      "epoch": 35.52597402597402,
+      "grad_norm": 0.3826066851615906,
+      "learning_rate": 2.908676923076923e-05,
+      "loss": 0.0266,
+      "step": 5471
+    },
+    {
+      "epoch": 35.532467532467535,
+      "grad_norm": 0.3471587300300598,
+      "learning_rate": 2.908646153846154e-05,
+      "loss": 0.0224,
+      "step": 5472
+    },
+    {
+      "epoch": 35.53896103896104,
+      "grad_norm": 0.4912363588809967,
+      "learning_rate": 2.9086153846153847e-05,
+      "loss": 0.0274,
+      "step": 5473
+    },
+    {
+      "epoch": 35.54545454545455,
+      "grad_norm": 0.36949363350868225,
+      "learning_rate": 2.9085846153846155e-05,
+      "loss": 0.0218,
+      "step": 5474
+    },
+    {
+      "epoch": 35.55194805194805,
+      "grad_norm": 0.4779680371284485,
+      "learning_rate": 2.9085538461538463e-05,
+      "loss": 0.0205,
+      "step": 5475
+    },
+    {
+      "epoch": 35.55844155844156,
+      "grad_norm": 0.6375389695167542,
+      "learning_rate": 2.9085230769230768e-05,
+      "loss": 0.0166,
+      "step": 5476
+    },
+    {
+      "epoch": 35.564935064935064,
+      "grad_norm": 0.5356177687644958,
+      "learning_rate": 2.9084923076923076e-05,
+      "loss": 0.0197,
+      "step": 5477
+    },
+    {
+      "epoch": 35.57142857142857,
+      "grad_norm": 0.6952133178710938,
+      "learning_rate": 2.9084615384615385e-05,
+      "loss": 0.0178,
+      "step": 5478
+    },
+    {
+      "epoch": 35.577922077922075,
+      "grad_norm": 0.44713854789733887,
+      "learning_rate": 2.9084307692307693e-05,
+      "loss": 0.0203,
+      "step": 5479
+    },
+    {
+      "epoch": 35.58441558441559,
+      "grad_norm": 0.3580329418182373,
+      "learning_rate": 2.9084e-05,
+      "loss": 0.0058,
+      "step": 5480
+    },
+    {
+      "epoch": 35.59090909090909,
+      "grad_norm": 0.47033828496932983,
+      "learning_rate": 2.908369230769231e-05,
+      "loss": 0.0116,
+      "step": 5481
+    },
+    {
+      "epoch": 35.5974025974026,
+      "grad_norm": 0.6218892931938171,
+      "learning_rate": 2.9083384615384615e-05,
+      "loss": 0.035,
+      "step": 5482
+    },
+    {
+      "epoch": 35.603896103896105,
+      "grad_norm": 0.49907755851745605,
+      "learning_rate": 2.9083076923076923e-05,
+      "loss": 0.0155,
+      "step": 5483
+    },
+    {
+      "epoch": 35.61038961038961,
+      "grad_norm": 0.3479296565055847,
+      "learning_rate": 2.908276923076923e-05,
+      "loss": 0.0055,
+      "step": 5484
+    },
+    {
+      "epoch": 35.616883116883116,
+      "grad_norm": 0.3294694125652313,
+      "learning_rate": 2.9082461538461536e-05,
+      "loss": 0.0089,
+      "step": 5485
+    },
+    {
+      "epoch": 35.62337662337662,
+      "grad_norm": 0.5581619143486023,
+      "learning_rate": 2.9082153846153848e-05,
+      "loss": 0.0114,
+      "step": 5486
+    },
+    {
+      "epoch": 35.62987012987013,
+      "grad_norm": 0.15970663726329803,
+      "learning_rate": 2.9081846153846156e-05,
+      "loss": 0.003,
+      "step": 5487
+    },
+    {
+      "epoch": 35.63636363636363,
+      "grad_norm": 0.7730697989463806,
+      "learning_rate": 2.908153846153846e-05,
+      "loss": 0.0111,
+      "step": 5488
+    },
+    {
+      "epoch": 35.642857142857146,
+      "grad_norm": 0.6104624271392822,
+      "learning_rate": 2.908123076923077e-05,
+      "loss": 0.0113,
+      "step": 5489
+    },
+    {
+      "epoch": 35.64935064935065,
+      "grad_norm": 0.28909027576446533,
+      "learning_rate": 2.9080923076923078e-05,
+      "loss": 0.0163,
+      "step": 5490
+    },
+    {
+      "epoch": 35.65584415584416,
+      "grad_norm": 0.5803835988044739,
+      "learning_rate": 2.9080615384615383e-05,
+      "loss": 0.0087,
+      "step": 5491
+    },
+    {
+      "epoch": 35.66233766233766,
+      "grad_norm": 1.817765712738037,
+      "learning_rate": 2.908030769230769e-05,
+      "loss": 0.0166,
+      "step": 5492
+    },
+    {
+      "epoch": 35.66883116883117,
+      "grad_norm": 0.38326776027679443,
+      "learning_rate": 2.9080000000000003e-05,
+      "loss": 0.0056,
+      "step": 5493
+    },
+    {
+      "epoch": 35.675324675324674,
+      "grad_norm": 0.3740791976451874,
+      "learning_rate": 2.9079692307692308e-05,
+      "loss": 0.0037,
+      "step": 5494
+    },
+    {
+      "epoch": 35.68181818181818,
+      "grad_norm": 0.6539549231529236,
+      "learning_rate": 2.9079384615384616e-05,
+      "loss": 0.03,
+      "step": 5495
+    },
+    {
+      "epoch": 35.688311688311686,
+      "grad_norm": 1.2329083681106567,
+      "learning_rate": 2.9079076923076924e-05,
+      "loss": 0.0142,
+      "step": 5496
+    },
+    {
+      "epoch": 35.6948051948052,
+      "grad_norm": 0.6088599562644958,
+      "learning_rate": 2.9078769230769233e-05,
+      "loss": 0.0087,
+      "step": 5497
+    },
+    {
+      "epoch": 35.701298701298704,
+      "grad_norm": 0.40001776814460754,
+      "learning_rate": 2.9078461538461538e-05,
+      "loss": 0.007,
+      "step": 5498
+    },
+    {
+      "epoch": 35.70779220779221,
+      "grad_norm": 1.0654664039611816,
+      "learning_rate": 2.9078153846153846e-05,
+      "loss": 0.0232,
+      "step": 5499
+    },
+    {
+      "epoch": 35.714285714285715,
+      "grad_norm": 0.4179539084434509,
+      "learning_rate": 2.9077846153846158e-05,
+      "loss": 0.0076,
+      "step": 5500
+    },
+    {
+      "epoch": 35.72077922077922,
+      "grad_norm": 0.60594242811203,
+      "learning_rate": 2.9077538461538463e-05,
+      "loss": 0.0138,
+      "step": 5501
+    },
+    {
+      "epoch": 35.72727272727273,
+      "grad_norm": 0.6405591368675232,
+      "learning_rate": 2.907723076923077e-05,
+      "loss": 0.0122,
+      "step": 5502
+    },
+    {
+      "epoch": 35.73376623376623,
+      "grad_norm": 0.49801746010780334,
+      "learning_rate": 2.907692307692308e-05,
+      "loss": 0.0066,
+      "step": 5503
+    },
+    {
+      "epoch": 35.74025974025974,
+      "grad_norm": 1.2940011024475098,
+      "learning_rate": 2.9076615384615384e-05,
+      "loss": 0.0237,
+      "step": 5504
+    },
+    {
+      "epoch": 35.746753246753244,
+      "grad_norm": 0.5902006030082703,
+      "learning_rate": 2.9076307692307692e-05,
+      "loss": 0.0761,
+      "step": 5505
+    },
+    {
+      "epoch": 35.753246753246756,
+      "grad_norm": 0.5554302930831909,
+      "learning_rate": 2.9076e-05,
+      "loss": 0.075,
+      "step": 5506
+    },
+    {
+      "epoch": 35.75974025974026,
+      "grad_norm": 0.3603295087814331,
+      "learning_rate": 2.907569230769231e-05,
+      "loss": 0.0425,
+      "step": 5507
+    },
+    {
+      "epoch": 35.76623376623377,
+      "grad_norm": 0.6961526274681091,
+      "learning_rate": 2.9075384615384617e-05,
+      "loss": 0.0552,
+      "step": 5508
+    },
+    {
+      "epoch": 35.77272727272727,
+      "grad_norm": 0.3742905557155609,
+      "learning_rate": 2.9075076923076926e-05,
+      "loss": 0.0378,
+      "step": 5509
+    },
+    {
+      "epoch": 35.77922077922078,
+      "grad_norm": 0.3611226975917816,
+      "learning_rate": 2.907476923076923e-05,
+      "loss": 0.0259,
+      "step": 5510
+    },
+    {
+      "epoch": 35.785714285714285,
+      "grad_norm": 0.5611889958381653,
+      "learning_rate": 2.907446153846154e-05,
+      "loss": 0.0294,
+      "step": 5511
+    },
+    {
+      "epoch": 35.79220779220779,
+      "grad_norm": 0.6996146440505981,
+      "learning_rate": 2.9074153846153847e-05,
+      "loss": 0.0216,
+      "step": 5512
+    },
+    {
+      "epoch": 35.798701298701296,
+      "grad_norm": 0.4185541570186615,
+      "learning_rate": 2.9073846153846152e-05,
+      "loss": 0.0403,
+      "step": 5513
+    },
+    {
+      "epoch": 35.8051948051948,
+      "grad_norm": 0.40230315923690796,
+      "learning_rate": 2.9073538461538464e-05,
+      "loss": 0.0204,
+      "step": 5514
+    },
+    {
+      "epoch": 35.811688311688314,
+      "grad_norm": 0.7334201335906982,
+      "learning_rate": 2.9073230769230772e-05,
+      "loss": 0.0156,
+      "step": 5515
+    },
+    {
+      "epoch": 35.81818181818182,
+      "grad_norm": 0.35723817348480225,
+      "learning_rate": 2.9072923076923077e-05,
+      "loss": 0.0146,
+      "step": 5516
+    },
+    {
+      "epoch": 35.824675324675326,
+      "grad_norm": 0.25425267219543457,
+      "learning_rate": 2.9072615384615386e-05,
+      "loss": 0.0115,
+      "step": 5517
+    },
+    {
+      "epoch": 35.83116883116883,
+      "grad_norm": 0.3605209290981293,
+      "learning_rate": 2.9072307692307694e-05,
+      "loss": 0.0128,
+      "step": 5518
+    },
+    {
+      "epoch": 35.83766233766234,
+      "grad_norm": 0.25324591994285583,
+      "learning_rate": 2.9072e-05,
+      "loss": 0.0092,
+      "step": 5519
+    },
+    {
+      "epoch": 35.84415584415584,
+      "grad_norm": 0.31561002135276794,
+      "learning_rate": 2.9071692307692307e-05,
+      "loss": 0.0084,
+      "step": 5520
+    },
+    {
+      "epoch": 35.85064935064935,
+      "grad_norm": 0.800163209438324,
+      "learning_rate": 2.907138461538462e-05,
+      "loss": 0.018,
+      "step": 5521
+    },
+    {
+      "epoch": 35.857142857142854,
+      "grad_norm": 0.3143797516822815,
+      "learning_rate": 2.9071076923076924e-05,
+      "loss": 0.0061,
+      "step": 5522
+    },
+    {
+      "epoch": 35.86363636363637,
+      "grad_norm": 0.24487708508968353,
+      "learning_rate": 2.9070769230769232e-05,
+      "loss": 0.0049,
+      "step": 5523
+    },
+    {
+      "epoch": 35.87012987012987,
+      "grad_norm": 0.7393335103988647,
+      "learning_rate": 2.907046153846154e-05,
+      "loss": 0.0183,
+      "step": 5524
+    },
+    {
+      "epoch": 35.87662337662338,
+      "grad_norm": 0.8401263356208801,
+      "learning_rate": 2.9070153846153845e-05,
+      "loss": 0.0253,
+      "step": 5525
+    },
+    {
+      "epoch": 35.883116883116884,
+      "grad_norm": 0.506006121635437,
+      "learning_rate": 2.9069846153846154e-05,
+      "loss": 0.0067,
+      "step": 5526
+    },
+    {
+      "epoch": 35.88961038961039,
+      "grad_norm": 0.35310667753219604,
+      "learning_rate": 2.9069538461538462e-05,
+      "loss": 0.0097,
+      "step": 5527
+    },
+    {
+      "epoch": 35.896103896103895,
+      "grad_norm": 0.23659542202949524,
+      "learning_rate": 2.906923076923077e-05,
+      "loss": 0.0039,
+      "step": 5528
+    },
+    {
+      "epoch": 35.9025974025974,
+      "grad_norm": 0.3253742456436157,
+      "learning_rate": 2.906892307692308e-05,
+      "loss": 0.0087,
+      "step": 5529
+    },
+    {
+      "epoch": 35.90909090909091,
+      "grad_norm": 2.2575976848602295,
+      "learning_rate": 2.9068615384615387e-05,
+      "loss": 0.0206,
+      "step": 5530
+    },
+    {
+      "epoch": 35.91558441558441,
+      "grad_norm": 1.0390942096710205,
+      "learning_rate": 2.9068307692307692e-05,
+      "loss": 0.0126,
+      "step": 5531
+    },
+    {
+      "epoch": 35.922077922077925,
+      "grad_norm": 0.5909035801887512,
+      "learning_rate": 2.9068e-05,
+      "loss": 0.0133,
+      "step": 5532
+    },
+    {
+      "epoch": 35.92857142857143,
+      "grad_norm": 0.6804397106170654,
+      "learning_rate": 2.906769230769231e-05,
+      "loss": 0.0186,
+      "step": 5533
+    },
+    {
+      "epoch": 35.935064935064936,
+      "grad_norm": 1.6290465593338013,
+      "learning_rate": 2.9067384615384613e-05,
+      "loss": 0.0128,
+      "step": 5534
+    },
+    {
+      "epoch": 35.94155844155844,
+      "grad_norm": 1.1904802322387695,
+      "learning_rate": 2.9067076923076925e-05,
+      "loss": 0.0248,
+      "step": 5535
+    },
+    {
+      "epoch": 35.94805194805195,
+      "grad_norm": 0.44184446334838867,
+      "learning_rate": 2.9066769230769233e-05,
+      "loss": 0.0527,
+      "step": 5536
+    },
+    {
+      "epoch": 35.95454545454545,
+      "grad_norm": 0.5022152662277222,
+      "learning_rate": 2.906646153846154e-05,
+      "loss": 0.0157,
+      "step": 5537
+    },
+    {
+      "epoch": 35.96103896103896,
+      "grad_norm": 1.0548217296600342,
+      "learning_rate": 2.9066153846153847e-05,
+      "loss": 0.0117,
+      "step": 5538
+    },
+    {
+      "epoch": 35.967532467532465,
+      "grad_norm": 0.5167074203491211,
+      "learning_rate": 2.9065846153846155e-05,
+      "loss": 0.0069,
+      "step": 5539
+    },
+    {
+      "epoch": 35.97402597402598,
+      "grad_norm": 0.6413472294807434,
+      "learning_rate": 2.906553846153846e-05,
+      "loss": 0.0131,
+      "step": 5540
+    },
+    {
+      "epoch": 35.98051948051948,
+      "grad_norm": 1.2182412147521973,
+      "learning_rate": 2.906523076923077e-05,
+      "loss": 0.0132,
+      "step": 5541
+    },
+    {
+      "epoch": 35.98701298701299,
+      "grad_norm": 1.4319102764129639,
+      "learning_rate": 2.906492307692308e-05,
+      "loss": 0.0201,
+      "step": 5542
+    },
+    {
+      "epoch": 35.993506493506494,
+      "grad_norm": 0.5336865186691284,
+      "learning_rate": 2.9064615384615385e-05,
+      "loss": 0.0435,
+      "step": 5543
+    },
+    {
+      "epoch": 36.0,
+      "grad_norm": 0.4530264437198639,
+      "learning_rate": 2.9064307692307693e-05,
+      "loss": 0.0097,
+      "step": 5544
+    },
+    {
+      "epoch": 36.006493506493506,
+      "grad_norm": 0.5937870740890503,
+      "learning_rate": 2.9064e-05,
+      "loss": 0.071,
+      "step": 5545
+    },
+    {
+      "epoch": 36.01298701298701,
+      "grad_norm": 0.5256974697113037,
+      "learning_rate": 2.9063692307692307e-05,
+      "loss": 0.056,
+      "step": 5546
+    },
+    {
+      "epoch": 36.01948051948052,
+      "grad_norm": 0.3508014380931854,
+      "learning_rate": 2.9063384615384615e-05,
+      "loss": 0.0302,
+      "step": 5547
+    },
+    {
+      "epoch": 36.02597402597402,
+      "grad_norm": 0.34796005487442017,
+      "learning_rate": 2.9063076923076923e-05,
+      "loss": 0.0407,
+      "step": 5548
+    },
+    {
+      "epoch": 36.032467532467535,
+      "grad_norm": 0.380829393863678,
+      "learning_rate": 2.906276923076923e-05,
+      "loss": 0.0276,
+      "step": 5549
+    },
+    {
+      "epoch": 36.03896103896104,
+      "grad_norm": 0.3318963646888733,
+      "learning_rate": 2.906246153846154e-05,
+      "loss": 0.0161,
+      "step": 5550
+    },
+    {
+      "epoch": 36.04545454545455,
+      "grad_norm": 0.2507683336734772,
+      "learning_rate": 2.9062153846153848e-05,
+      "loss": 0.0171,
+      "step": 5551
+    },
+    {
+      "epoch": 36.05194805194805,
+      "grad_norm": 0.35036081075668335,
+      "learning_rate": 2.9061846153846153e-05,
+      "loss": 0.0201,
+      "step": 5552
+    },
+    {
+      "epoch": 36.05844155844156,
+      "grad_norm": 0.505987823009491,
+      "learning_rate": 2.906153846153846e-05,
+      "loss": 0.0217,
+      "step": 5553
+    },
+    {
+      "epoch": 36.064935064935064,
+      "grad_norm": 0.3023730218410492,
+      "learning_rate": 2.906123076923077e-05,
+      "loss": 0.012,
+      "step": 5554
+    },
+    {
+      "epoch": 36.07142857142857,
+      "grad_norm": 0.5266294479370117,
+      "learning_rate": 2.9060923076923075e-05,
+      "loss": 0.0156,
+      "step": 5555
+    },
+    {
+      "epoch": 36.077922077922075,
+      "grad_norm": 0.37526872754096985,
+      "learning_rate": 2.9060615384615386e-05,
+      "loss": 0.0146,
+      "step": 5556
+    },
+    {
+      "epoch": 36.08441558441559,
+      "grad_norm": 0.29122111201286316,
+      "learning_rate": 2.9060307692307695e-05,
+      "loss": 0.0101,
+      "step": 5557
+    },
+    {
+      "epoch": 36.09090909090909,
+      "grad_norm": 0.31415414810180664,
+      "learning_rate": 2.906e-05,
+      "loss": 0.0107,
+      "step": 5558
+    },
+    {
+      "epoch": 36.0974025974026,
+      "grad_norm": 0.6353932023048401,
+      "learning_rate": 2.9059692307692308e-05,
+      "loss": 0.0137,
+      "step": 5559
+    },
+    {
+      "epoch": 36.103896103896105,
+      "grad_norm": 0.5115739703178406,
+      "learning_rate": 2.9059384615384616e-05,
+      "loss": 0.0386,
+      "step": 5560
+    },
+    {
+      "epoch": 36.11038961038961,
+      "grad_norm": 0.42253628373146057,
+      "learning_rate": 2.905907692307692e-05,
+      "loss": 0.0111,
+      "step": 5561
+    },
+    {
+      "epoch": 36.116883116883116,
+      "grad_norm": 0.27850422263145447,
+      "learning_rate": 2.905876923076923e-05,
+      "loss": 0.0092,
+      "step": 5562
+    },
+    {
+      "epoch": 36.12337662337662,
+      "grad_norm": 0.20719462633132935,
+      "learning_rate": 2.905846153846154e-05,
+      "loss": 0.0054,
+      "step": 5563
+    },
+    {
+      "epoch": 36.12987012987013,
+      "grad_norm": 0.6023554801940918,
+      "learning_rate": 2.9058153846153846e-05,
+      "loss": 0.0092,
+      "step": 5564
+    },
+    {
+      "epoch": 36.13636363636363,
+      "grad_norm": 1.097813606262207,
+      "learning_rate": 2.9057846153846154e-05,
+      "loss": 0.0105,
+      "step": 5565
+    },
+    {
+      "epoch": 36.142857142857146,
+      "grad_norm": 0.5676377415657043,
+      "learning_rate": 2.9057538461538463e-05,
+      "loss": 0.0112,
+      "step": 5566
+    },
+    {
+      "epoch": 36.14935064935065,
+      "grad_norm": 0.5977246165275574,
+      "learning_rate": 2.9057230769230768e-05,
+      "loss": 0.0198,
+      "step": 5567
+    },
+    {
+      "epoch": 36.15584415584416,
+      "grad_norm": 0.29906946420669556,
+      "learning_rate": 2.9056923076923076e-05,
+      "loss": 0.0043,
+      "step": 5568
+    },
+    {
+      "epoch": 36.16233766233766,
+      "grad_norm": 0.433550089597702,
+      "learning_rate": 2.9056615384615384e-05,
+      "loss": 0.0094,
+      "step": 5569
+    },
+    {
+      "epoch": 36.16883116883117,
+      "grad_norm": 0.6595407128334045,
+      "learning_rate": 2.9056307692307693e-05,
+      "loss": 0.0079,
+      "step": 5570
+    },
+    {
+      "epoch": 36.175324675324674,
+      "grad_norm": 0.9656956791877747,
+      "learning_rate": 2.9056e-05,
+      "loss": 0.019,
+      "step": 5571
+    },
+    {
+      "epoch": 36.18181818181818,
+      "grad_norm": 0.16969792544841766,
+      "learning_rate": 2.905569230769231e-05,
+      "loss": 0.002,
+      "step": 5572
+    },
+    {
+      "epoch": 36.188311688311686,
+      "grad_norm": 2.333681344985962,
+      "learning_rate": 2.9055384615384614e-05,
+      "loss": 0.0094,
+      "step": 5573
+    },
+    {
+      "epoch": 36.1948051948052,
+      "grad_norm": 0.4919402599334717,
+      "learning_rate": 2.9055076923076923e-05,
+      "loss": 0.0072,
+      "step": 5574
+    },
+    {
+      "epoch": 36.201298701298704,
+      "grad_norm": 0.46083393692970276,
+      "learning_rate": 2.905476923076923e-05,
+      "loss": 0.005,
+      "step": 5575
+    },
+    {
+      "epoch": 36.20779220779221,
+      "grad_norm": 1.5113096237182617,
+      "learning_rate": 2.905446153846154e-05,
+      "loss": 0.0686,
+      "step": 5576
+    },
+    {
+      "epoch": 36.214285714285715,
+      "grad_norm": 0.5647264122962952,
+      "learning_rate": 2.9054153846153848e-05,
+      "loss": 0.0139,
+      "step": 5577
+    },
+    {
+      "epoch": 36.22077922077922,
+      "grad_norm": 2.3536794185638428,
+      "learning_rate": 2.9053846153846156e-05,
+      "loss": 0.0303,
+      "step": 5578
+    },
+    {
+      "epoch": 36.22727272727273,
+      "grad_norm": 0.43088576197624207,
+      "learning_rate": 2.9053538461538464e-05,
+      "loss": 0.0136,
+      "step": 5579
+    },
+    {
+      "epoch": 36.23376623376623,
+      "grad_norm": 1.1980865001678467,
+      "learning_rate": 2.905323076923077e-05,
+      "loss": 0.0175,
+      "step": 5580
+    },
+    {
+      "epoch": 36.24025974025974,
+      "grad_norm": 0.4099942147731781,
+      "learning_rate": 2.9052923076923077e-05,
+      "loss": 0.0035,
+      "step": 5581
+    },
+    {
+      "epoch": 36.246753246753244,
+      "grad_norm": 1.6205854415893555,
+      "learning_rate": 2.9052615384615386e-05,
+      "loss": 0.048,
+      "step": 5582
+    },
+    {
+      "epoch": 36.253246753246756,
+      "grad_norm": 0.5835689306259155,
+      "learning_rate": 2.9052307692307694e-05,
+      "loss": 0.0781,
+      "step": 5583
+    },
+    {
+      "epoch": 36.25974025974026,
+      "grad_norm": 0.3758178651332855,
+      "learning_rate": 2.9052000000000002e-05,
+      "loss": 0.04,
+      "step": 5584
+    },
+    {
+      "epoch": 36.26623376623377,
+      "grad_norm": 0.6773163676261902,
+      "learning_rate": 2.905169230769231e-05,
+      "loss": 0.0431,
+      "step": 5585
+    },
+    {
+      "epoch": 36.27272727272727,
+      "grad_norm": 0.3295516073703766,
+      "learning_rate": 2.9051384615384616e-05,
+      "loss": 0.0303,
+      "step": 5586
+    },
+    {
+      "epoch": 36.27922077922078,
+      "grad_norm": 0.27221164107322693,
+      "learning_rate": 2.9051076923076924e-05,
+      "loss": 0.023,
+      "step": 5587
+    },
+    {
+      "epoch": 36.285714285714285,
+      "grad_norm": 0.3228011429309845,
+      "learning_rate": 2.9050769230769232e-05,
+      "loss": 0.0187,
+      "step": 5588
+    },
+    {
+      "epoch": 36.29220779220779,
+      "grad_norm": 0.3541872799396515,
+      "learning_rate": 2.9050461538461537e-05,
+      "loss": 0.0202,
+      "step": 5589
+    },
+    {
+      "epoch": 36.298701298701296,
+      "grad_norm": 0.5636196136474609,
+      "learning_rate": 2.905015384615385e-05,
+      "loss": 0.02,
+      "step": 5590
+    },
+    {
+      "epoch": 36.3051948051948,
+      "grad_norm": 0.8093762397766113,
+      "learning_rate": 2.9049846153846157e-05,
+      "loss": 0.0235,
+      "step": 5591
+    },
+    {
+      "epoch": 36.311688311688314,
+      "grad_norm": 0.35847240686416626,
+      "learning_rate": 2.9049538461538462e-05,
+      "loss": 0.0154,
+      "step": 5592
+    },
+    {
+      "epoch": 36.31818181818182,
+      "grad_norm": 0.4368458390235901,
+      "learning_rate": 2.904923076923077e-05,
+      "loss": 0.0152,
+      "step": 5593
+    },
+    {
+      "epoch": 36.324675324675326,
+      "grad_norm": 0.595810055732727,
+      "learning_rate": 2.904892307692308e-05,
+      "loss": 0.0131,
+      "step": 5594
+    },
+    {
+      "epoch": 36.33116883116883,
+      "grad_norm": 0.2645629644393921,
+      "learning_rate": 2.9048615384615384e-05,
+      "loss": 0.0077,
+      "step": 5595
+    },
+    {
+      "epoch": 36.33766233766234,
+      "grad_norm": 0.3524000942707062,
+      "learning_rate": 2.9048307692307692e-05,
+      "loss": 0.01,
+      "step": 5596
+    },
+    {
+      "epoch": 36.34415584415584,
+      "grad_norm": 0.42889848351478577,
+      "learning_rate": 2.9048000000000004e-05,
+      "loss": 0.0136,
+      "step": 5597
+    },
+    {
+      "epoch": 36.35064935064935,
+      "grad_norm": 0.6208891868591309,
+      "learning_rate": 2.904769230769231e-05,
+      "loss": 0.0175,
+      "step": 5598
+    },
+    {
+      "epoch": 36.357142857142854,
+      "grad_norm": 0.2720721960067749,
+      "learning_rate": 2.9047384615384617e-05,
+      "loss": 0.0113,
+      "step": 5599
+    },
+    {
+      "epoch": 36.36363636363637,
+      "grad_norm": 0.3387434184551239,
+      "learning_rate": 2.9047076923076925e-05,
+      "loss": 0.0071,
+      "step": 5600
+    },
+    {
+      "epoch": 36.37012987012987,
+      "grad_norm": 0.7956691384315491,
+      "learning_rate": 2.904676923076923e-05,
+      "loss": 0.0107,
+      "step": 5601
+    },
+    {
+      "epoch": 36.37662337662338,
+      "grad_norm": 0.5840033292770386,
+      "learning_rate": 2.904646153846154e-05,
+      "loss": 0.0161,
+      "step": 5602
+    },
+    {
+      "epoch": 36.383116883116884,
+      "grad_norm": 0.8667775988578796,
+      "learning_rate": 2.9046153846153847e-05,
+      "loss": 0.0095,
+      "step": 5603
+    },
+    {
+      "epoch": 36.38961038961039,
+      "grad_norm": 0.5208715796470642,
+      "learning_rate": 2.9045846153846155e-05,
+      "loss": 0.0115,
+      "step": 5604
+    },
+    {
+      "epoch": 36.396103896103895,
+      "grad_norm": 1.3019874095916748,
+      "learning_rate": 2.9045538461538464e-05,
+      "loss": 0.0109,
+      "step": 5605
+    },
+    {
+      "epoch": 36.4025974025974,
+      "grad_norm": 0.451679527759552,
+      "learning_rate": 2.9045230769230772e-05,
+      "loss": 0.0107,
+      "step": 5606
+    },
+    {
+      "epoch": 36.40909090909091,
+      "grad_norm": 0.4819625914096832,
+      "learning_rate": 2.9044923076923077e-05,
+      "loss": 0.0259,
+      "step": 5607
+    },
+    {
+      "epoch": 36.41558441558441,
+      "grad_norm": 0.27623212337493896,
+      "learning_rate": 2.9044615384615385e-05,
+      "loss": 0.006,
+      "step": 5608
+    },
+    {
+      "epoch": 36.422077922077925,
+      "grad_norm": 0.35889360308647156,
+      "learning_rate": 2.9044307692307693e-05,
+      "loss": 0.0051,
+      "step": 5609
+    },
+    {
+      "epoch": 36.42857142857143,
+      "grad_norm": 1.6308268308639526,
+      "learning_rate": 2.9044e-05,
+      "loss": 0.0233,
+      "step": 5610
+    },
+    {
+      "epoch": 36.435064935064936,
+      "grad_norm": 0.20552177727222443,
+      "learning_rate": 2.904369230769231e-05,
+      "loss": 0.0043,
+      "step": 5611
+    },
+    {
+      "epoch": 36.44155844155844,
+      "grad_norm": 0.3932984173297882,
+      "learning_rate": 2.904338461538462e-05,
+      "loss": 0.0053,
+      "step": 5612
+    },
+    {
+      "epoch": 36.44805194805195,
+      "grad_norm": 0.4277825355529785,
+      "learning_rate": 2.9043076923076923e-05,
+      "loss": 0.004,
+      "step": 5613
+    },
+    {
+      "epoch": 36.45454545454545,
+      "grad_norm": 0.3922305107116699,
+      "learning_rate": 2.904276923076923e-05,
+      "loss": 0.0055,
+      "step": 5614
+    },
+    {
+      "epoch": 36.46103896103896,
+      "grad_norm": 0.24312107264995575,
+      "learning_rate": 2.904246153846154e-05,
+      "loss": 0.003,
+      "step": 5615
+    },
+    {
+      "epoch": 36.467532467532465,
+      "grad_norm": 0.5969083309173584,
+      "learning_rate": 2.9042153846153845e-05,
+      "loss": 0.0132,
+      "step": 5616
+    },
+    {
+      "epoch": 36.47402597402598,
+      "grad_norm": 1.956734299659729,
+      "learning_rate": 2.9041846153846153e-05,
+      "loss": 0.0543,
+      "step": 5617
+    },
+    {
+      "epoch": 36.48051948051948,
+      "grad_norm": 0.8999709486961365,
+      "learning_rate": 2.9041538461538465e-05,
+      "loss": 0.0255,
+      "step": 5618
+    },
+    {
+      "epoch": 36.48701298701299,
+      "grad_norm": 1.2347708940505981,
+      "learning_rate": 2.904123076923077e-05,
+      "loss": 0.0149,
+      "step": 5619
+    },
+    {
+      "epoch": 36.493506493506494,
+      "grad_norm": 1.7901315689086914,
+      "learning_rate": 2.9040923076923078e-05,
+      "loss": 0.0445,
+      "step": 5620
+    },
+    {
+      "epoch": 36.5,
+      "grad_norm": 0.559679388999939,
+      "learning_rate": 2.9040615384615387e-05,
+      "loss": 0.0729,
+      "step": 5621
+    },
+    {
+      "epoch": 36.506493506493506,
+      "grad_norm": 0.5174139142036438,
+      "learning_rate": 2.904030769230769e-05,
+      "loss": 0.0572,
+      "step": 5622
+    },
+    {
+      "epoch": 36.51298701298701,
+      "grad_norm": 0.38822513818740845,
+      "learning_rate": 2.904e-05,
+      "loss": 0.0365,
+      "step": 5623
+    },
+    {
+      "epoch": 36.51948051948052,
+      "grad_norm": 0.39593538641929626,
+      "learning_rate": 2.9039692307692308e-05,
+      "loss": 0.0293,
+      "step": 5624
+    },
+    {
+      "epoch": 36.52597402597402,
+      "grad_norm": 0.3647960424423218,
+      "learning_rate": 2.9039384615384616e-05,
+      "loss": 0.0294,
+      "step": 5625
+    },
+    {
+      "epoch": 36.532467532467535,
+      "grad_norm": 0.38199394941329956,
+      "learning_rate": 2.9039076923076925e-05,
+      "loss": 0.0213,
+      "step": 5626
+    },
+    {
+      "epoch": 36.53896103896104,
+      "grad_norm": 0.3603741526603699,
+      "learning_rate": 2.9038769230769233e-05,
+      "loss": 0.0297,
+      "step": 5627
+    },
+    {
+      "epoch": 36.54545454545455,
+      "grad_norm": 0.4830622375011444,
+      "learning_rate": 2.9038461538461538e-05,
+      "loss": 0.0294,
+      "step": 5628
+    },
+    {
+      "epoch": 36.55194805194805,
+      "grad_norm": 0.6599977612495422,
+      "learning_rate": 2.9038153846153846e-05,
+      "loss": 0.0442,
+      "step": 5629
+    },
+    {
+      "epoch": 36.55844155844156,
+      "grad_norm": 0.6063061952590942,
+      "learning_rate": 2.9037846153846155e-05,
+      "loss": 0.0222,
+      "step": 5630
+    },
+    {
+      "epoch": 36.564935064935064,
+      "grad_norm": 0.29563385248184204,
+      "learning_rate": 2.903753846153846e-05,
+      "loss": 0.0116,
+      "step": 5631
+    },
+    {
+      "epoch": 36.57142857142857,
+      "grad_norm": 1.1445544958114624,
+      "learning_rate": 2.903723076923077e-05,
+      "loss": 0.0135,
+      "step": 5632
+    },
+    {
+      "epoch": 36.577922077922075,
+      "grad_norm": 0.4297947585582733,
+      "learning_rate": 2.903692307692308e-05,
+      "loss": 0.0181,
+      "step": 5633
+    },
+    {
+      "epoch": 36.58441558441559,
+      "grad_norm": 0.5096610188484192,
+      "learning_rate": 2.9036615384615385e-05,
+      "loss": 0.0122,
+      "step": 5634
+    },
+    {
+      "epoch": 36.59090909090909,
+      "grad_norm": 0.44934797286987305,
+      "learning_rate": 2.9036307692307693e-05,
+      "loss": 0.0149,
+      "step": 5635
+    },
+    {
+      "epoch": 36.5974025974026,
+      "grad_norm": 0.49636566638946533,
+      "learning_rate": 2.9036e-05,
+      "loss": 0.0121,
+      "step": 5636
+    },
+    {
+      "epoch": 36.603896103896105,
+      "grad_norm": 0.2698199450969696,
+      "learning_rate": 2.9035692307692306e-05,
+      "loss": 0.0086,
+      "step": 5637
+    },
+    {
+      "epoch": 36.61038961038961,
+      "grad_norm": 0.36220136284828186,
+      "learning_rate": 2.9035384615384614e-05,
+      "loss": 0.0123,
+      "step": 5638
+    },
+    {
+      "epoch": 36.616883116883116,
+      "grad_norm": 0.6276205778121948,
+      "learning_rate": 2.9035076923076926e-05,
+      "loss": 0.0086,
+      "step": 5639
+    },
+    {
+      "epoch": 36.62337662337662,
+      "grad_norm": 0.764651358127594,
+      "learning_rate": 2.903476923076923e-05,
+      "loss": 0.0167,
+      "step": 5640
+    },
+    {
+      "epoch": 36.62987012987013,
+      "grad_norm": 0.7851477265357971,
+      "learning_rate": 2.903446153846154e-05,
+      "loss": 0.0115,
+      "step": 5641
+    },
+    {
+      "epoch": 36.63636363636363,
+      "grad_norm": 1.075822114944458,
+      "learning_rate": 2.9034153846153848e-05,
+      "loss": 0.0172,
+      "step": 5642
+    },
+    {
+      "epoch": 36.642857142857146,
+      "grad_norm": 0.6303055882453918,
+      "learning_rate": 2.9033846153846153e-05,
+      "loss": 0.0052,
+      "step": 5643
+    },
+    {
+      "epoch": 36.64935064935065,
+      "grad_norm": 0.6038342714309692,
+      "learning_rate": 2.903353846153846e-05,
+      "loss": 0.0121,
+      "step": 5644
+    },
+    {
+      "epoch": 36.65584415584416,
+      "grad_norm": 0.4720795452594757,
+      "learning_rate": 2.903323076923077e-05,
+      "loss": 0.0113,
+      "step": 5645
+    },
+    {
+      "epoch": 36.66233766233766,
+      "grad_norm": 0.2050342559814453,
+      "learning_rate": 2.9032923076923078e-05,
+      "loss": 0.0026,
+      "step": 5646
+    },
+    {
+      "epoch": 36.66883116883117,
+      "grad_norm": 0.5154380202293396,
+      "learning_rate": 2.9032615384615386e-05,
+      "loss": 0.011,
+      "step": 5647
+    },
+    {
+      "epoch": 36.675324675324674,
+      "grad_norm": 0.7150618433952332,
+      "learning_rate": 2.9032307692307694e-05,
+      "loss": 0.0178,
+      "step": 5648
+    },
+    {
+      "epoch": 36.68181818181818,
+      "grad_norm": 1.0069153308868408,
+      "learning_rate": 2.9032e-05,
+      "loss": 0.0278,
+      "step": 5649
+    },
+    {
+      "epoch": 36.688311688311686,
+      "grad_norm": 0.29059624671936035,
+      "learning_rate": 2.9031692307692308e-05,
+      "loss": 0.0033,
+      "step": 5650
+    },
+    {
+      "epoch": 36.6948051948052,
+      "grad_norm": 1.0125490427017212,
+      "learning_rate": 2.9031384615384616e-05,
+      "loss": 0.0095,
+      "step": 5651
+    },
+    {
+      "epoch": 36.701298701298704,
+      "grad_norm": 0.28927430510520935,
+      "learning_rate": 2.903107692307692e-05,
+      "loss": 0.0035,
+      "step": 5652
+    },
+    {
+      "epoch": 36.70779220779221,
+      "grad_norm": 0.606968104839325,
+      "learning_rate": 2.9030769230769232e-05,
+      "loss": 0.0083,
+      "step": 5653
+    },
+    {
+      "epoch": 36.714285714285715,
+      "grad_norm": 0.8843885660171509,
+      "learning_rate": 2.903046153846154e-05,
+      "loss": 0.0048,
+      "step": 5654
+    },
+    {
+      "epoch": 36.72077922077922,
+      "grad_norm": 0.5820101499557495,
+      "learning_rate": 2.9030153846153846e-05,
+      "loss": 0.014,
+      "step": 5655
+    },
+    {
+      "epoch": 36.72727272727273,
+      "grad_norm": 1.4354671239852905,
+      "learning_rate": 2.9029846153846154e-05,
+      "loss": 0.0212,
+      "step": 5656
+    },
+    {
+      "epoch": 36.73376623376623,
+      "grad_norm": 0.6979734301567078,
+      "learning_rate": 2.9029538461538462e-05,
+      "loss": 0.0113,
+      "step": 5657
+    },
+    {
+      "epoch": 36.74025974025974,
+      "grad_norm": 1.1822783946990967,
+      "learning_rate": 2.902923076923077e-05,
+      "loss": 0.0163,
+      "step": 5658
+    },
+    {
+      "epoch": 36.746753246753244,
+      "grad_norm": 0.6402063965797424,
+      "learning_rate": 2.9028923076923076e-05,
+      "loss": 0.0871,
+      "step": 5659
+    },
+    {
+      "epoch": 36.753246753246756,
+      "grad_norm": 0.6479465365409851,
+      "learning_rate": 2.9028615384615387e-05,
+      "loss": 0.0684,
+      "step": 5660
+    },
+    {
+      "epoch": 36.75974025974026,
+      "grad_norm": 0.43468889594078064,
+      "learning_rate": 2.9028307692307696e-05,
+      "loss": 0.0542,
+      "step": 5661
+    },
+    {
+      "epoch": 36.76623376623377,
+      "grad_norm": 0.41597333550453186,
+      "learning_rate": 2.9028e-05,
+      "loss": 0.0447,
+      "step": 5662
+    },
+    {
+      "epoch": 36.77272727272727,
+      "grad_norm": 0.3900293707847595,
+      "learning_rate": 2.902769230769231e-05,
+      "loss": 0.0328,
+      "step": 5663
+    },
+    {
+      "epoch": 36.77922077922078,
+      "grad_norm": 0.35357940196990967,
+      "learning_rate": 2.9027384615384617e-05,
+      "loss": 0.0254,
+      "step": 5664
+    },
+    {
+      "epoch": 36.785714285714285,
+      "grad_norm": 0.46574369072914124,
+      "learning_rate": 2.9027076923076922e-05,
+      "loss": 0.026,
+      "step": 5665
+    },
+    {
+      "epoch": 36.79220779220779,
+      "grad_norm": 0.6147619485855103,
+      "learning_rate": 2.902676923076923e-05,
+      "loss": 0.0242,
+      "step": 5666
+    },
+    {
+      "epoch": 36.798701298701296,
+      "grad_norm": 0.32326582074165344,
+      "learning_rate": 2.9026461538461542e-05,
+      "loss": 0.0141,
+      "step": 5667
+    },
+    {
+      "epoch": 36.8051948051948,
+      "grad_norm": 0.42057231068611145,
+      "learning_rate": 2.9026153846153847e-05,
+      "loss": 0.0205,
+      "step": 5668
+    },
+    {
+      "epoch": 36.811688311688314,
+      "grad_norm": 0.33616119623184204,
+      "learning_rate": 2.9025846153846155e-05,
+      "loss": 0.016,
+      "step": 5669
+    },
+    {
+      "epoch": 36.81818181818182,
+      "grad_norm": 0.46913179755210876,
+      "learning_rate": 2.9025538461538464e-05,
+      "loss": 0.0241,
+      "step": 5670
+    },
+    {
+      "epoch": 36.824675324675326,
+      "grad_norm": 0.43245717883110046,
+      "learning_rate": 2.902523076923077e-05,
+      "loss": 0.019,
+      "step": 5671
+    },
+    {
+      "epoch": 36.83116883116883,
+      "grad_norm": 0.30958181619644165,
+      "learning_rate": 2.9024923076923077e-05,
+      "loss": 0.0105,
+      "step": 5672
+    },
+    {
+      "epoch": 36.83766233766234,
+      "grad_norm": 0.5329145193099976,
+      "learning_rate": 2.9024615384615385e-05,
+      "loss": 0.0145,
+      "step": 5673
+    },
+    {
+      "epoch": 36.84415584415584,
+      "grad_norm": 0.5764246582984924,
+      "learning_rate": 2.9024307692307694e-05,
+      "loss": 0.0239,
+      "step": 5674
+    },
+    {
+      "epoch": 36.85064935064935,
+      "grad_norm": 0.43031397461891174,
+      "learning_rate": 2.9024000000000002e-05,
+      "loss": 0.0109,
+      "step": 5675
+    },
+    {
+      "epoch": 36.857142857142854,
+      "grad_norm": 0.3727981448173523,
+      "learning_rate": 2.902369230769231e-05,
+      "loss": 0.0124,
+      "step": 5676
+    },
+    {
+      "epoch": 36.86363636363637,
+      "grad_norm": 0.4733588695526123,
+      "learning_rate": 2.9023384615384615e-05,
+      "loss": 0.0181,
+      "step": 5677
+    },
+    {
+      "epoch": 36.87012987012987,
+      "grad_norm": 0.2856893241405487,
+      "learning_rate": 2.9023076923076924e-05,
+      "loss": 0.0085,
+      "step": 5678
+    },
+    {
+      "epoch": 36.87662337662338,
+      "grad_norm": 0.5714840888977051,
+      "learning_rate": 2.9022769230769232e-05,
+      "loss": 0.0158,
+      "step": 5679
+    },
+    {
+      "epoch": 36.883116883116884,
+      "grad_norm": 0.22427834570407867,
+      "learning_rate": 2.9022461538461537e-05,
+      "loss": 0.005,
+      "step": 5680
+    },
+    {
+      "epoch": 36.88961038961039,
+      "grad_norm": 0.3562869429588318,
+      "learning_rate": 2.902215384615385e-05,
+      "loss": 0.0059,
+      "step": 5681
+    },
+    {
+      "epoch": 36.896103896103895,
+      "grad_norm": 0.6766841411590576,
+      "learning_rate": 2.9021846153846157e-05,
+      "loss": 0.0111,
+      "step": 5682
+    },
+    {
+      "epoch": 36.9025974025974,
+      "grad_norm": 1.1050662994384766,
+      "learning_rate": 2.9021538461538462e-05,
+      "loss": 0.039,
+      "step": 5683
+    },
+    {
+      "epoch": 36.90909090909091,
+      "grad_norm": 0.3368038833141327,
+      "learning_rate": 2.902123076923077e-05,
+      "loss": 0.0071,
+      "step": 5684
+    },
+    {
+      "epoch": 36.91558441558441,
+      "grad_norm": 0.31755903363227844,
+      "learning_rate": 2.902092307692308e-05,
+      "loss": 0.0048,
+      "step": 5685
+    },
+    {
+      "epoch": 36.922077922077925,
+      "grad_norm": 0.7562558650970459,
+      "learning_rate": 2.9020615384615383e-05,
+      "loss": 0.0708,
+      "step": 5686
+    },
+    {
+      "epoch": 36.92857142857143,
+      "grad_norm": 0.35310858488082886,
+      "learning_rate": 2.902030769230769e-05,
+      "loss": 0.0162,
+      "step": 5687
+    },
+    {
+      "epoch": 36.935064935064936,
+      "grad_norm": 0.4592399299144745,
+      "learning_rate": 2.9020000000000003e-05,
+      "loss": 0.006,
+      "step": 5688
+    },
+    {
+      "epoch": 36.94155844155844,
+      "grad_norm": 0.07205262780189514,
+      "learning_rate": 2.9019692307692308e-05,
+      "loss": 0.0012,
+      "step": 5689
+    },
+    {
+      "epoch": 36.94805194805195,
+      "grad_norm": 0.21239978075027466,
+      "learning_rate": 2.9019384615384617e-05,
+      "loss": 0.0031,
+      "step": 5690
+    },
+    {
+      "epoch": 36.95454545454545,
+      "grad_norm": 0.6470236778259277,
+      "learning_rate": 2.9019076923076925e-05,
+      "loss": 0.0073,
+      "step": 5691
+    },
+    {
+      "epoch": 36.96103896103896,
+      "grad_norm": 0.5057578086853027,
+      "learning_rate": 2.901876923076923e-05,
+      "loss": 0.0057,
+      "step": 5692
+    },
+    {
+      "epoch": 36.967532467532465,
+      "grad_norm": 1.3637326955795288,
+      "learning_rate": 2.9018461538461538e-05,
+      "loss": 0.0118,
+      "step": 5693
+    },
+    {
+      "epoch": 36.97402597402598,
+      "grad_norm": 0.8590590357780457,
+      "learning_rate": 2.9018153846153847e-05,
+      "loss": 0.0245,
+      "step": 5694
+    },
+    {
+      "epoch": 36.98051948051948,
+      "grad_norm": 0.49676382541656494,
+      "learning_rate": 2.9017846153846155e-05,
+      "loss": 0.0071,
+      "step": 5695
+    },
+    {
+      "epoch": 36.98701298701299,
+      "grad_norm": 0.8142293691635132,
+      "learning_rate": 2.9017538461538463e-05,
+      "loss": 0.0197,
+      "step": 5696
+    },
+    {
+      "epoch": 36.993506493506494,
+      "grad_norm": 0.39492130279541016,
+      "learning_rate": 2.901723076923077e-05,
+      "loss": 0.0233,
+      "step": 5697
+    },
+    {
+      "epoch": 37.0,
+      "grad_norm": 0.42088377475738525,
+      "learning_rate": 2.9016923076923076e-05,
+      "loss": 0.0082,
+      "step": 5698
+    },
+    {
+      "epoch": 37.006493506493506,
+      "grad_norm": 0.5682960152626038,
+      "learning_rate": 2.9016615384615385e-05,
+      "loss": 0.068,
+      "step": 5699
+    },
+    {
+      "epoch": 37.01298701298701,
+      "grad_norm": 0.4266335368156433,
+      "learning_rate": 2.9016307692307693e-05,
+      "loss": 0.0336,
+      "step": 5700
+    },
+    {
+      "epoch": 37.01948051948052,
+      "grad_norm": 0.45381370186805725,
+      "learning_rate": 2.9015999999999998e-05,
+      "loss": 0.0341,
+      "step": 5701
+    },
+    {
+      "epoch": 37.02597402597402,
+      "grad_norm": 0.3741210699081421,
+      "learning_rate": 2.901569230769231e-05,
+      "loss": 0.0339,
+      "step": 5702
+    },
+    {
+      "epoch": 37.032467532467535,
+      "grad_norm": 0.3010711073875427,
+      "learning_rate": 2.9015384615384618e-05,
+      "loss": 0.0234,
+      "step": 5703
+    },
+    {
+      "epoch": 37.03896103896104,
+      "grad_norm": 0.33665797114372253,
+      "learning_rate": 2.9015076923076923e-05,
+      "loss": 0.0206,
+      "step": 5704
+    },
+    {
+      "epoch": 37.04545454545455,
+      "grad_norm": 0.2654063105583191,
+      "learning_rate": 2.901476923076923e-05,
+      "loss": 0.0144,
+      "step": 5705
+    },
+    {
+      "epoch": 37.05194805194805,
+      "grad_norm": 0.3056391477584839,
+      "learning_rate": 2.901446153846154e-05,
+      "loss": 0.0139,
+      "step": 5706
+    },
+    {
+      "epoch": 37.05844155844156,
+      "grad_norm": 0.25488588213920593,
+      "learning_rate": 2.9014153846153845e-05,
+      "loss": 0.0097,
+      "step": 5707
+    },
+    {
+      "epoch": 37.064935064935064,
+      "grad_norm": 0.25274279713630676,
+      "learning_rate": 2.9013846153846153e-05,
+      "loss": 0.0102,
+      "step": 5708
+    },
+    {
+      "epoch": 37.07142857142857,
+      "grad_norm": 0.47891965508461,
+      "learning_rate": 2.9013538461538465e-05,
+      "loss": 0.012,
+      "step": 5709
+    },
+    {
+      "epoch": 37.077922077922075,
+      "grad_norm": 0.4401334822177887,
+      "learning_rate": 2.901323076923077e-05,
+      "loss": 0.0136,
+      "step": 5710
+    },
+    {
+      "epoch": 37.08441558441559,
+      "grad_norm": 0.2747070789337158,
+      "learning_rate": 2.9012923076923078e-05,
+      "loss": 0.0069,
+      "step": 5711
+    },
+    {
+      "epoch": 37.09090909090909,
+      "grad_norm": 0.3973609209060669,
+      "learning_rate": 2.9012615384615386e-05,
+      "loss": 0.0087,
+      "step": 5712
+    },
+    {
+      "epoch": 37.0974025974026,
+      "grad_norm": 0.3059113323688507,
+      "learning_rate": 2.901230769230769e-05,
+      "loss": 0.0101,
+      "step": 5713
+    },
+    {
+      "epoch": 37.103896103896105,
+      "grad_norm": 0.263389527797699,
+      "learning_rate": 2.9012e-05,
+      "loss": 0.007,
+      "step": 5714
+    },
+    {
+      "epoch": 37.11038961038961,
+      "grad_norm": 0.24985991418361664,
+      "learning_rate": 2.9011692307692308e-05,
+      "loss": 0.0081,
+      "step": 5715
+    },
+    {
+      "epoch": 37.116883116883116,
+      "grad_norm": 0.6577467918395996,
+      "learning_rate": 2.9011384615384616e-05,
+      "loss": 0.01,
+      "step": 5716
+    },
+    {
+      "epoch": 37.12337662337662,
+      "grad_norm": 0.6368363499641418,
+      "learning_rate": 2.9011076923076924e-05,
+      "loss": 0.0062,
+      "step": 5717
+    },
+    {
+      "epoch": 37.12987012987013,
+      "grad_norm": 0.2599269151687622,
+      "learning_rate": 2.9010769230769233e-05,
+      "loss": 0.0045,
+      "step": 5718
+    },
+    {
+      "epoch": 37.13636363636363,
+      "grad_norm": 1.2361116409301758,
+      "learning_rate": 2.9010461538461538e-05,
+      "loss": 0.0121,
+      "step": 5719
+    },
+    {
+      "epoch": 37.142857142857146,
+      "grad_norm": 0.44706419110298157,
+      "learning_rate": 2.9010153846153846e-05,
+      "loss": 0.0052,
+      "step": 5720
+    },
+    {
+      "epoch": 37.14935064935065,
+      "grad_norm": 0.28249335289001465,
+      "learning_rate": 2.9009846153846154e-05,
+      "loss": 0.0042,
+      "step": 5721
+    },
+    {
+      "epoch": 37.15584415584416,
+      "grad_norm": 1.5646660327911377,
+      "learning_rate": 2.900953846153846e-05,
+      "loss": 0.0205,
+      "step": 5722
+    },
+    {
+      "epoch": 37.16233766233766,
+      "grad_norm": 0.5014433264732361,
+      "learning_rate": 2.900923076923077e-05,
+      "loss": 0.0103,
+      "step": 5723
+    },
+    {
+      "epoch": 37.16883116883117,
+      "grad_norm": 0.6213587522506714,
+      "learning_rate": 2.900892307692308e-05,
+      "loss": 0.0103,
+      "step": 5724
+    },
+    {
+      "epoch": 37.175324675324674,
+      "grad_norm": 1.0360206365585327,
+      "learning_rate": 2.9008615384615384e-05,
+      "loss": 0.0268,
+      "step": 5725
+    },
+    {
+      "epoch": 37.18181818181818,
+      "grad_norm": 0.5155927538871765,
+      "learning_rate": 2.9008307692307692e-05,
+      "loss": 0.0406,
+      "step": 5726
+    },
+    {
+      "epoch": 37.188311688311686,
+      "grad_norm": 0.4905499517917633,
+      "learning_rate": 2.9008e-05,
+      "loss": 0.0203,
+      "step": 5727
+    },
+    {
+      "epoch": 37.1948051948052,
+      "grad_norm": 0.4875847399234772,
+      "learning_rate": 2.9007692307692306e-05,
+      "loss": 0.0081,
+      "step": 5728
+    },
+    {
+      "epoch": 37.201298701298704,
+      "grad_norm": 0.6300015449523926,
+      "learning_rate": 2.9007384615384614e-05,
+      "loss": 0.009,
+      "step": 5729
+    },
+    {
+      "epoch": 37.20779220779221,
+      "grad_norm": 1.02634596824646,
+      "learning_rate": 2.9007076923076926e-05,
+      "loss": 0.0157,
+      "step": 5730
+    },
+    {
+      "epoch": 37.214285714285715,
+      "grad_norm": 0.350734144449234,
+      "learning_rate": 2.900676923076923e-05,
+      "loss": 0.0023,
+      "step": 5731
+    },
+    {
+      "epoch": 37.22077922077922,
+      "grad_norm": 0.9716202020645142,
+      "learning_rate": 2.900646153846154e-05,
+      "loss": 0.0041,
+      "step": 5732
+    },
+    {
+      "epoch": 37.22727272727273,
+      "grad_norm": 1.7610501050949097,
+      "learning_rate": 2.9006153846153847e-05,
+      "loss": 0.0101,
+      "step": 5733
+    },
+    {
+      "epoch": 37.23376623376623,
+      "grad_norm": 1.2568671703338623,
+      "learning_rate": 2.9005846153846152e-05,
+      "loss": 0.0426,
+      "step": 5734
+    },
+    {
+      "epoch": 37.24025974025974,
+      "grad_norm": 0.8857156038284302,
+      "learning_rate": 2.900553846153846e-05,
+      "loss": 0.0056,
+      "step": 5735
+    },
+    {
+      "epoch": 37.246753246753244,
+      "grad_norm": 0.18274565041065216,
+      "learning_rate": 2.900523076923077e-05,
+      "loss": 0.0026,
+      "step": 5736
+    },
+    {
+      "epoch": 37.253246753246756,
+      "grad_norm": 0.590124785900116,
+      "learning_rate": 2.9004923076923077e-05,
+      "loss": 0.0624,
+      "step": 5737
+    },
+    {
+      "epoch": 37.25974025974026,
+      "grad_norm": 0.4975883960723877,
+      "learning_rate": 2.9004615384615386e-05,
+      "loss": 0.0502,
+      "step": 5738
+    },
+    {
+      "epoch": 37.26623376623377,
+      "grad_norm": 0.43369370698928833,
+      "learning_rate": 2.9004307692307694e-05,
+      "loss": 0.0497,
+      "step": 5739
+    },
+    {
+      "epoch": 37.27272727272727,
+      "grad_norm": 0.3413119912147522,
+      "learning_rate": 2.9004000000000002e-05,
+      "loss": 0.0339,
+      "step": 5740
+    },
+    {
+      "epoch": 37.27922077922078,
+      "grad_norm": 0.32237881422042847,
+      "learning_rate": 2.9003692307692307e-05,
+      "loss": 0.0245,
+      "step": 5741
+    },
+    {
+      "epoch": 37.285714285714285,
+      "grad_norm": 0.38169053196907043,
+      "learning_rate": 2.9003384615384615e-05,
+      "loss": 0.0196,
+      "step": 5742
+    },
+    {
+      "epoch": 37.29220779220779,
+      "grad_norm": 0.3442230224609375,
+      "learning_rate": 2.9003076923076924e-05,
+      "loss": 0.0194,
+      "step": 5743
+    },
+    {
+      "epoch": 37.298701298701296,
+      "grad_norm": 0.2957189679145813,
+      "learning_rate": 2.9002769230769232e-05,
+      "loss": 0.015,
+      "step": 5744
+    },
+    {
+      "epoch": 37.3051948051948,
+      "grad_norm": 0.40390369296073914,
+      "learning_rate": 2.900246153846154e-05,
+      "loss": 0.0183,
+      "step": 5745
+    },
+    {
+      "epoch": 37.311688311688314,
+      "grad_norm": 0.5869516134262085,
+      "learning_rate": 2.900215384615385e-05,
+      "loss": 0.0383,
+      "step": 5746
+    },
+    {
+      "epoch": 37.31818181818182,
+      "grad_norm": 0.5072347521781921,
+      "learning_rate": 2.9001846153846154e-05,
+      "loss": 0.0131,
+      "step": 5747
+    },
+    {
+      "epoch": 37.324675324675326,
+      "grad_norm": 0.357571005821228,
+      "learning_rate": 2.9001538461538462e-05,
+      "loss": 0.0095,
+      "step": 5748
+    },
+    {
+      "epoch": 37.33116883116883,
+      "grad_norm": 0.430806964635849,
+      "learning_rate": 2.900123076923077e-05,
+      "loss": 0.0153,
+      "step": 5749
+    },
+    {
+      "epoch": 37.33766233766234,
+      "grad_norm": 0.35799238085746765,
+      "learning_rate": 2.900092307692308e-05,
+      "loss": 0.0105,
+      "step": 5750
+    },
+    {
+      "epoch": 37.34415584415584,
+      "grad_norm": 0.3252030313014984,
+      "learning_rate": 2.9000615384615387e-05,
+      "loss": 0.0064,
+      "step": 5751
+    },
+    {
+      "epoch": 37.35064935064935,
+      "grad_norm": 0.6061626076698303,
+      "learning_rate": 2.9000307692307695e-05,
+      "loss": 0.009,
+      "step": 5752
+    },
+    {
+      "epoch": 37.357142857142854,
+      "grad_norm": 0.3573189675807953,
+      "learning_rate": 2.9e-05,
+      "loss": 0.0077,
+      "step": 5753
+    },
+    {
+      "epoch": 37.36363636363637,
+      "grad_norm": 1.481142282485962,
+      "learning_rate": 2.899969230769231e-05,
+      "loss": 0.0168,
+      "step": 5754
+    },
+    {
+      "epoch": 37.37012987012987,
+      "grad_norm": 0.536189079284668,
+      "learning_rate": 2.8999384615384617e-05,
+      "loss": 0.007,
+      "step": 5755
+    },
+    {
+      "epoch": 37.37662337662338,
+      "grad_norm": 1.1793863773345947,
+      "learning_rate": 2.8999076923076922e-05,
+      "loss": 0.0278,
+      "step": 5756
+    },
+    {
+      "epoch": 37.383116883116884,
+      "grad_norm": 1.4678716659545898,
+      "learning_rate": 2.8998769230769233e-05,
+      "loss": 0.0197,
+      "step": 5757
+    },
+    {
+      "epoch": 37.38961038961039,
+      "grad_norm": 0.29317334294319153,
+      "learning_rate": 2.8998461538461542e-05,
+      "loss": 0.0067,
+      "step": 5758
+    },
+    {
+      "epoch": 37.396103896103895,
+      "grad_norm": 1.279661774635315,
+      "learning_rate": 2.8998153846153847e-05,
+      "loss": 0.0254,
+      "step": 5759
+    },
+    {
+      "epoch": 37.4025974025974,
+      "grad_norm": 0.7066517472267151,
+      "learning_rate": 2.8997846153846155e-05,
+      "loss": 0.0226,
+      "step": 5760
+    },
+    {
+      "epoch": 37.40909090909091,
+      "grad_norm": 0.8013560175895691,
+      "learning_rate": 2.8997538461538463e-05,
+      "loss": 0.02,
+      "step": 5761
+    },
+    {
+      "epoch": 37.41558441558441,
+      "grad_norm": 0.4228562116622925,
+      "learning_rate": 2.8997230769230768e-05,
+      "loss": 0.0058,
+      "step": 5762
+    },
+    {
+      "epoch": 37.422077922077925,
+      "grad_norm": 0.484175443649292,
+      "learning_rate": 2.8996923076923077e-05,
+      "loss": 0.0057,
+      "step": 5763
+    },
+    {
+      "epoch": 37.42857142857143,
+      "grad_norm": 0.30538302659988403,
+      "learning_rate": 2.899661538461539e-05,
+      "loss": 0.0056,
+      "step": 5764
+    },
+    {
+      "epoch": 37.435064935064936,
+      "grad_norm": 0.9660945534706116,
+      "learning_rate": 2.8996307692307693e-05,
+      "loss": 0.0221,
+      "step": 5765
+    },
+    {
+      "epoch": 37.44155844155844,
+      "grad_norm": 1.1904863119125366,
+      "learning_rate": 2.8996e-05,
+      "loss": 0.0125,
+      "step": 5766
+    },
+    {
+      "epoch": 37.44805194805195,
+      "grad_norm": 0.4259716272354126,
+      "learning_rate": 2.899569230769231e-05,
+      "loss": 0.0049,
+      "step": 5767
+    },
+    {
+      "epoch": 37.45454545454545,
+      "grad_norm": 0.23206937313079834,
+      "learning_rate": 2.8995384615384615e-05,
+      "loss": 0.0027,
+      "step": 5768
+    },
+    {
+      "epoch": 37.46103896103896,
+      "grad_norm": 0.4441584646701813,
+      "learning_rate": 2.8995076923076923e-05,
+      "loss": 0.004,
+      "step": 5769
+    },
+    {
+      "epoch": 37.467532467532465,
+      "grad_norm": 0.5857951641082764,
+      "learning_rate": 2.899476923076923e-05,
+      "loss": 0.0243,
+      "step": 5770
+    },
+    {
+      "epoch": 37.47402597402598,
+      "grad_norm": 0.31035560369491577,
+      "learning_rate": 2.899446153846154e-05,
+      "loss": 0.0033,
+      "step": 5771
+    },
+    {
+      "epoch": 37.48051948051948,
+      "grad_norm": 0.9182817339897156,
+      "learning_rate": 2.8994153846153848e-05,
+      "loss": 0.0065,
+      "step": 5772
+    },
+    {
+      "epoch": 37.48701298701299,
+      "grad_norm": 2.1770718097686768,
+      "learning_rate": 2.8993846153846156e-05,
+      "loss": 0.0168,
+      "step": 5773
+    },
+    {
+      "epoch": 37.493506493506494,
+      "grad_norm": 2.2369048595428467,
+      "learning_rate": 2.899353846153846e-05,
+      "loss": 0.0343,
+      "step": 5774
+    },
+    {
+      "epoch": 37.5,
+      "grad_norm": 0.6820853352546692,
+      "learning_rate": 2.899323076923077e-05,
+      "loss": 0.0905,
+      "step": 5775
+    },
+    {
+      "epoch": 37.506493506493506,
+      "grad_norm": 0.47322067618370056,
+      "learning_rate": 2.8992923076923078e-05,
+      "loss": 0.0595,
+      "step": 5776
+    },
+    {
+      "epoch": 37.51298701298701,
+      "grad_norm": 0.4924643039703369,
+      "learning_rate": 2.8992615384615383e-05,
+      "loss": 0.0374,
+      "step": 5777
+    },
+    {
+      "epoch": 37.51948051948052,
+      "grad_norm": 0.39143550395965576,
+      "learning_rate": 2.8992307692307695e-05,
+      "loss": 0.0302,
+      "step": 5778
+    },
+    {
+      "epoch": 37.52597402597402,
+      "grad_norm": 0.6751034259796143,
+      "learning_rate": 2.8992000000000003e-05,
+      "loss": 0.0334,
+      "step": 5779
+    },
+    {
+      "epoch": 37.532467532467535,
+      "grad_norm": 0.35889044404029846,
+      "learning_rate": 2.8991692307692308e-05,
+      "loss": 0.0224,
+      "step": 5780
+    },
+    {
+      "epoch": 37.53896103896104,
+      "grad_norm": 0.5877482891082764,
+      "learning_rate": 2.8991384615384616e-05,
+      "loss": 0.0326,
+      "step": 5781
+    },
+    {
+      "epoch": 37.54545454545455,
+      "grad_norm": 0.6370552182197571,
+      "learning_rate": 2.8991076923076925e-05,
+      "loss": 0.0331,
+      "step": 5782
+    },
+    {
+      "epoch": 37.55194805194805,
+      "grad_norm": 0.5590326189994812,
+      "learning_rate": 2.899076923076923e-05,
+      "loss": 0.0231,
+      "step": 5783
+    },
+    {
+      "epoch": 37.55844155844156,
+      "grad_norm": 0.4057650864124298,
+      "learning_rate": 2.8990461538461538e-05,
+      "loss": 0.0218,
+      "step": 5784
+    },
+    {
+      "epoch": 37.564935064935064,
+      "grad_norm": 0.39709484577178955,
+      "learning_rate": 2.899015384615385e-05,
+      "loss": 0.016,
+      "step": 5785
+    },
+    {
+      "epoch": 37.57142857142857,
+      "grad_norm": 0.4472017288208008,
+      "learning_rate": 2.8989846153846154e-05,
+      "loss": 0.0198,
+      "step": 5786
+    },
+    {
+      "epoch": 37.577922077922075,
+      "grad_norm": 0.46619942784309387,
+      "learning_rate": 2.8989538461538463e-05,
+      "loss": 0.0143,
+      "step": 5787
+    },
+    {
+      "epoch": 37.58441558441559,
+      "grad_norm": 0.37539732456207275,
+      "learning_rate": 2.898923076923077e-05,
+      "loss": 0.0151,
+      "step": 5788
+    },
+    {
+      "epoch": 37.59090909090909,
+      "grad_norm": 0.5927746295928955,
+      "learning_rate": 2.8988923076923076e-05,
+      "loss": 0.0139,
+      "step": 5789
+    },
+    {
+      "epoch": 37.5974025974026,
+      "grad_norm": 0.7920203804969788,
+      "learning_rate": 2.8988615384615384e-05,
+      "loss": 0.0391,
+      "step": 5790
+    },
+    {
+      "epoch": 37.603896103896105,
+      "grad_norm": 0.5341692566871643,
+      "learning_rate": 2.8988307692307693e-05,
+      "loss": 0.0177,
+      "step": 5791
+    },
+    {
+      "epoch": 37.61038961038961,
+      "grad_norm": 0.21405696868896484,
+      "learning_rate": 2.8988e-05,
+      "loss": 0.0067,
+      "step": 5792
+    },
+    {
+      "epoch": 37.616883116883116,
+      "grad_norm": 0.36822637915611267,
+      "learning_rate": 2.898769230769231e-05,
+      "loss": 0.0077,
+      "step": 5793
+    },
+    {
+      "epoch": 37.62337662337662,
+      "grad_norm": 0.31057003140449524,
+      "learning_rate": 2.8987384615384618e-05,
+      "loss": 0.0079,
+      "step": 5794
+    },
+    {
+      "epoch": 37.62987012987013,
+      "grad_norm": 0.8950670957565308,
+      "learning_rate": 2.8987076923076923e-05,
+      "loss": 0.0106,
+      "step": 5795
+    },
+    {
+      "epoch": 37.63636363636363,
+      "grad_norm": 0.32633891701698303,
+      "learning_rate": 2.898676923076923e-05,
+      "loss": 0.0066,
+      "step": 5796
+    },
+    {
+      "epoch": 37.642857142857146,
+      "grad_norm": 0.6017307043075562,
+      "learning_rate": 2.898646153846154e-05,
+      "loss": 0.0125,
+      "step": 5797
+    },
+    {
+      "epoch": 37.64935064935065,
+      "grad_norm": 0.5148427486419678,
+      "learning_rate": 2.8986153846153844e-05,
+      "loss": 0.0135,
+      "step": 5798
+    },
+    {
+      "epoch": 37.65584415584416,
+      "grad_norm": 0.5627039670944214,
+      "learning_rate": 2.8985846153846156e-05,
+      "loss": 0.0081,
+      "step": 5799
+    },
+    {
+      "epoch": 37.66233766233766,
+      "grad_norm": 0.7895825505256653,
+      "learning_rate": 2.8985538461538464e-05,
+      "loss": 0.0159,
+      "step": 5800
+    },
+    {
+      "epoch": 37.66883116883117,
+      "grad_norm": 2.810176372528076,
+      "learning_rate": 2.898523076923077e-05,
+      "loss": 0.0147,
+      "step": 5801
+    },
+    {
+      "epoch": 37.675324675324674,
+      "grad_norm": 1.2674649953842163,
+      "learning_rate": 2.8984923076923077e-05,
+      "loss": 0.031,
+      "step": 5802
+    },
+    {
+      "epoch": 37.68181818181818,
+      "grad_norm": 0.717071533203125,
+      "learning_rate": 2.8984615384615386e-05,
+      "loss": 0.0096,
+      "step": 5803
+    },
+    {
+      "epoch": 37.688311688311686,
+      "grad_norm": 0.44485998153686523,
+      "learning_rate": 2.898430769230769e-05,
+      "loss": 0.0099,
+      "step": 5804
+    },
+    {
+      "epoch": 37.6948051948052,
+      "grad_norm": 0.44072046875953674,
+      "learning_rate": 2.8984e-05,
+      "loss": 0.0084,
+      "step": 5805
+    },
+    {
+      "epoch": 37.701298701298704,
+      "grad_norm": 1.7065715789794922,
+      "learning_rate": 2.898369230769231e-05,
+      "loss": 0.0554,
+      "step": 5806
+    },
+    {
+      "epoch": 37.70779220779221,
+      "grad_norm": 5.627781867980957,
+      "learning_rate": 2.8983384615384616e-05,
+      "loss": 0.0125,
+      "step": 5807
+    },
+    {
+      "epoch": 37.714285714285715,
+      "grad_norm": 1.938300371170044,
+      "learning_rate": 2.8983076923076924e-05,
+      "loss": 0.017,
+      "step": 5808
+    },
+    {
+      "epoch": 37.72077922077922,
+      "grad_norm": 0.5042718052864075,
+      "learning_rate": 2.8982769230769232e-05,
+      "loss": 0.0036,
+      "step": 5809
+    },
+    {
+      "epoch": 37.72727272727273,
+      "grad_norm": 0.9170762896537781,
+      "learning_rate": 2.8982461538461537e-05,
+      "loss": 0.0091,
+      "step": 5810
+    },
+    {
+      "epoch": 37.73376623376623,
+      "grad_norm": 1.4180288314819336,
+      "learning_rate": 2.8982153846153846e-05,
+      "loss": 0.013,
+      "step": 5811
+    },
+    {
+      "epoch": 37.74025974025974,
+      "grad_norm": 1.590042233467102,
+      "learning_rate": 2.8981846153846154e-05,
+      "loss": 0.0321,
+      "step": 5812
+    },
+    {
+      "epoch": 37.746753246753244,
+      "grad_norm": 0.664558470249176,
+      "learning_rate": 2.8981538461538462e-05,
+      "loss": 0.0704,
+      "step": 5813
+    },
+    {
+      "epoch": 37.753246753246756,
+      "grad_norm": 0.5193811058998108,
+      "learning_rate": 2.898123076923077e-05,
+      "loss": 0.0571,
+      "step": 5814
+    },
+    {
+      "epoch": 37.75974025974026,
+      "grad_norm": 0.39887556433677673,
+      "learning_rate": 2.898092307692308e-05,
+      "loss": 0.044,
+      "step": 5815
+    },
+    {
+      "epoch": 37.76623376623377,
+      "grad_norm": 0.4859084188938141,
+      "learning_rate": 2.8980615384615384e-05,
+      "loss": 0.0472,
+      "step": 5816
+    },
+    {
+      "epoch": 37.77272727272727,
+      "grad_norm": 0.4902481734752655,
+      "learning_rate": 2.8980307692307692e-05,
+      "loss": 0.0341,
+      "step": 5817
+    },
+    {
+      "epoch": 37.77922077922078,
+      "grad_norm": 0.4954281151294708,
+      "learning_rate": 2.898e-05,
+      "loss": 0.0294,
+      "step": 5818
+    },
+    {
+      "epoch": 37.785714285714285,
+      "grad_norm": 0.4322144091129303,
+      "learning_rate": 2.897969230769231e-05,
+      "loss": 0.0255,
+      "step": 5819
+    },
+    {
+      "epoch": 37.79220779220779,
+      "grad_norm": 0.42813196778297424,
+      "learning_rate": 2.8979384615384617e-05,
+      "loss": 0.0258,
+      "step": 5820
+    },
+    {
+      "epoch": 37.798701298701296,
+      "grad_norm": 0.3918134272098541,
+      "learning_rate": 2.8979076923076925e-05,
+      "loss": 0.0158,
+      "step": 5821
+    },
+    {
+      "epoch": 37.8051948051948,
+      "grad_norm": 0.7231109142303467,
+      "learning_rate": 2.8978769230769234e-05,
+      "loss": 0.0252,
+      "step": 5822
+    },
+    {
+      "epoch": 37.811688311688314,
+      "grad_norm": 0.40218809247016907,
+      "learning_rate": 2.897846153846154e-05,
+      "loss": 0.012,
+      "step": 5823
+    },
+    {
+      "epoch": 37.81818181818182,
+      "grad_norm": 0.5218591094017029,
+      "learning_rate": 2.8978153846153847e-05,
+      "loss": 0.0114,
+      "step": 5824
+    },
+    {
+      "epoch": 37.824675324675326,
+      "grad_norm": 0.9542838335037231,
+      "learning_rate": 2.8977846153846155e-05,
+      "loss": 0.0213,
+      "step": 5825
+    },
+    {
+      "epoch": 37.83116883116883,
+      "grad_norm": 0.29812315106391907,
+      "learning_rate": 2.897753846153846e-05,
+      "loss": 0.0156,
+      "step": 5826
+    },
+    {
+      "epoch": 37.83766233766234,
+      "grad_norm": 0.5161244869232178,
+      "learning_rate": 2.8977230769230772e-05,
+      "loss": 0.0194,
+      "step": 5827
+    },
+    {
+      "epoch": 37.84415584415584,
+      "grad_norm": 0.4323521852493286,
+      "learning_rate": 2.897692307692308e-05,
+      "loss": 0.0173,
+      "step": 5828
+    },
+    {
+      "epoch": 37.85064935064935,
+      "grad_norm": 0.27389127016067505,
+      "learning_rate": 2.8976615384615385e-05,
+      "loss": 0.0072,
+      "step": 5829
+    },
+    {
+      "epoch": 37.857142857142854,
+      "grad_norm": 1.3271141052246094,
+      "learning_rate": 2.8976307692307693e-05,
+      "loss": 0.0233,
+      "step": 5830
+    },
+    {
+      "epoch": 37.86363636363637,
+      "grad_norm": 1.1395535469055176,
+      "learning_rate": 2.8976000000000002e-05,
+      "loss": 0.0099,
+      "step": 5831
+    },
+    {
+      "epoch": 37.87012987012987,
+      "grad_norm": 0.7338045239448547,
+      "learning_rate": 2.8975692307692307e-05,
+      "loss": 0.0156,
+      "step": 5832
+    },
+    {
+      "epoch": 37.87662337662338,
+      "grad_norm": 0.4919627606868744,
+      "learning_rate": 2.8975384615384615e-05,
+      "loss": 0.0075,
+      "step": 5833
+    },
+    {
+      "epoch": 37.883116883116884,
+      "grad_norm": 0.8875814080238342,
+      "learning_rate": 2.8975076923076927e-05,
+      "loss": 0.0189,
+      "step": 5834
+    },
+    {
+      "epoch": 37.88961038961039,
+      "grad_norm": 0.5302218198776245,
+      "learning_rate": 2.897476923076923e-05,
+      "loss": 0.0103,
+      "step": 5835
+    },
+    {
+      "epoch": 37.896103896103895,
+      "grad_norm": 0.46994951367378235,
+      "learning_rate": 2.897446153846154e-05,
+      "loss": 0.0063,
+      "step": 5836
+    },
+    {
+      "epoch": 37.9025974025974,
+      "grad_norm": 0.16305571794509888,
+      "learning_rate": 2.8974153846153848e-05,
+      "loss": 0.0023,
+      "step": 5837
+    },
+    {
+      "epoch": 37.90909090909091,
+      "grad_norm": 0.30576035380363464,
+      "learning_rate": 2.8973846153846153e-05,
+      "loss": 0.0053,
+      "step": 5838
+    },
+    {
+      "epoch": 37.91558441558441,
+      "grad_norm": 0.4666111171245575,
+      "learning_rate": 2.897353846153846e-05,
+      "loss": 0.0143,
+      "step": 5839
+    },
+    {
+      "epoch": 37.922077922077925,
+      "grad_norm": 0.7408274412155151,
+      "learning_rate": 2.897323076923077e-05,
+      "loss": 0.0139,
+      "step": 5840
+    },
+    {
+      "epoch": 37.92857142857143,
+      "grad_norm": 0.7977734804153442,
+      "learning_rate": 2.8972923076923078e-05,
+      "loss": 0.0059,
+      "step": 5841
+    },
+    {
+      "epoch": 37.935064935064936,
+      "grad_norm": 0.6239389181137085,
+      "learning_rate": 2.8972615384615387e-05,
+      "loss": 0.0089,
+      "step": 5842
+    },
+    {
+      "epoch": 37.94155844155844,
+      "grad_norm": 0.18126250803470612,
+      "learning_rate": 2.8972307692307695e-05,
+      "loss": 0.002,
+      "step": 5843
+    },
+    {
+      "epoch": 37.94805194805195,
+      "grad_norm": 1.3525784015655518,
+      "learning_rate": 2.8972e-05,
+      "loss": 0.0071,
+      "step": 5844
+    },
+    {
+      "epoch": 37.95454545454545,
+      "grad_norm": 0.4028265178203583,
+      "learning_rate": 2.8971692307692308e-05,
+      "loss": 0.0032,
+      "step": 5845
+    },
+    {
+      "epoch": 37.96103896103896,
+      "grad_norm": 0.37903255224227905,
+      "learning_rate": 2.8971384615384616e-05,
+      "loss": 0.0084,
+      "step": 5846
+    },
+    {
+      "epoch": 37.967532467532465,
+      "grad_norm": 0.39638563990592957,
+      "learning_rate": 2.897107692307692e-05,
+      "loss": 0.0048,
+      "step": 5847
+    },
+    {
+      "epoch": 37.97402597402598,
+      "grad_norm": 0.5073239803314209,
+      "learning_rate": 2.8970769230769233e-05,
+      "loss": 0.0039,
+      "step": 5848
+    },
+    {
+      "epoch": 37.98051948051948,
+      "grad_norm": 1.5499805212020874,
+      "learning_rate": 2.897046153846154e-05,
+      "loss": 0.0081,
+      "step": 5849
+    },
+    {
+      "epoch": 37.98701298701299,
+      "grad_norm": 8.752338409423828,
+      "learning_rate": 2.8970153846153846e-05,
+      "loss": 0.1089,
+      "step": 5850
+    },
+    {
+      "epoch": 37.993506493506494,
+      "grad_norm": 0.5049633383750916,
+      "learning_rate": 2.8969846153846155e-05,
+      "loss": 0.0226,
+      "step": 5851
+    },
+    {
+      "epoch": 38.0,
+      "grad_norm": 0.4144134521484375,
+      "learning_rate": 2.8969538461538463e-05,
+      "loss": 0.0047,
+      "step": 5852
+    },
+    {
+      "epoch": 38.006493506493506,
+      "grad_norm": 0.8269168138504028,
+      "learning_rate": 2.8969230769230768e-05,
+      "loss": 0.0579,
+      "step": 5853
+    },
+    {
+      "epoch": 38.01298701298701,
+      "grad_norm": 0.5803773403167725,
+      "learning_rate": 2.8968923076923076e-05,
+      "loss": 0.0441,
+      "step": 5854
+    },
+    {
+      "epoch": 38.01948051948052,
+      "grad_norm": 1.5662708282470703,
+      "learning_rate": 2.8968615384615388e-05,
+      "loss": 0.0452,
+      "step": 5855
+    },
+    {
+      "epoch": 38.02597402597402,
+      "grad_norm": 0.508958637714386,
+      "learning_rate": 2.8968307692307693e-05,
+      "loss": 0.0451,
+      "step": 5856
+    },
+    {
+      "epoch": 38.032467532467535,
+      "grad_norm": 0.5083428621292114,
+      "learning_rate": 2.8968e-05,
+      "loss": 0.0309,
+      "step": 5857
+    },
+    {
+      "epoch": 38.03896103896104,
+      "grad_norm": 0.3908602297306061,
+      "learning_rate": 2.896769230769231e-05,
+      "loss": 0.0213,
+      "step": 5858
+    },
+    {
+      "epoch": 38.04545454545455,
+      "grad_norm": 0.3186749815940857,
+      "learning_rate": 2.8967384615384614e-05,
+      "loss": 0.0257,
+      "step": 5859
+    },
+    {
+      "epoch": 38.05194805194805,
+      "grad_norm": 0.33088552951812744,
+      "learning_rate": 2.8967076923076923e-05,
+      "loss": 0.0191,
+      "step": 5860
+    },
+    {
+      "epoch": 38.05844155844156,
+      "grad_norm": 0.3540251553058624,
+      "learning_rate": 2.896676923076923e-05,
+      "loss": 0.014,
+      "step": 5861
+    },
+    {
+      "epoch": 38.064935064935064,
+      "grad_norm": 0.48857712745666504,
+      "learning_rate": 2.896646153846154e-05,
+      "loss": 0.0151,
+      "step": 5862
+    },
+    {
+      "epoch": 38.07142857142857,
+      "grad_norm": 0.4713970124721527,
+      "learning_rate": 2.8966153846153848e-05,
+      "loss": 0.0128,
+      "step": 5863
+    },
+    {
+      "epoch": 38.077922077922075,
+      "grad_norm": 0.31919345259666443,
+      "learning_rate": 2.8965846153846156e-05,
+      "loss": 0.0119,
+      "step": 5864
+    },
+    {
+      "epoch": 38.08441558441559,
+      "grad_norm": 0.7923030257225037,
+      "learning_rate": 2.896553846153846e-05,
+      "loss": 0.0165,
+      "step": 5865
+    },
+    {
+      "epoch": 38.09090909090909,
+      "grad_norm": 0.25103074312210083,
+      "learning_rate": 2.896523076923077e-05,
+      "loss": 0.0076,
+      "step": 5866
+    },
+    {
+      "epoch": 38.0974025974026,
+      "grad_norm": 0.9443399906158447,
+      "learning_rate": 2.8964923076923078e-05,
+      "loss": 0.0208,
+      "step": 5867
+    },
+    {
+      "epoch": 38.103896103896105,
+      "grad_norm": 0.7961001396179199,
+      "learning_rate": 2.8964615384615383e-05,
+      "loss": 0.0303,
+      "step": 5868
+    },
+    {
+      "epoch": 38.11038961038961,
+      "grad_norm": 0.8084138035774231,
+      "learning_rate": 2.8964307692307694e-05,
+      "loss": 0.0133,
+      "step": 5869
+    },
+    {
+      "epoch": 38.116883116883116,
+      "grad_norm": 0.5023177266120911,
+      "learning_rate": 2.8964000000000003e-05,
+      "loss": 0.009,
+      "step": 5870
+    },
+    {
+      "epoch": 38.12337662337662,
+      "grad_norm": 1.263300895690918,
+      "learning_rate": 2.8963692307692307e-05,
+      "loss": 0.01,
+      "step": 5871
+    },
+    {
+      "epoch": 38.12987012987013,
+      "grad_norm": 0.15346093475818634,
+      "learning_rate": 2.8963384615384616e-05,
+      "loss": 0.0029,
+      "step": 5872
+    },
+    {
+      "epoch": 38.13636363636363,
+      "grad_norm": 0.6068188548088074,
+      "learning_rate": 2.8963076923076924e-05,
+      "loss": 0.0121,
+      "step": 5873
+    },
+    {
+      "epoch": 38.142857142857146,
+      "grad_norm": 0.6815487146377563,
+      "learning_rate": 2.896276923076923e-05,
+      "loss": 0.0122,
+      "step": 5874
+    },
+    {
+      "epoch": 38.14935064935065,
+      "grad_norm": 0.6805928349494934,
+      "learning_rate": 2.8962461538461537e-05,
+      "loss": 0.0098,
+      "step": 5875
+    },
+    {
+      "epoch": 38.15584415584416,
+      "grad_norm": 0.5679097175598145,
+      "learning_rate": 2.896215384615385e-05,
+      "loss": 0.0156,
+      "step": 5876
+    },
+    {
+      "epoch": 38.16233766233766,
+      "grad_norm": 0.40975064039230347,
+      "learning_rate": 2.8961846153846154e-05,
+      "loss": 0.0045,
+      "step": 5877
+    },
+    {
+      "epoch": 38.16883116883117,
+      "grad_norm": 0.8027766346931458,
+      "learning_rate": 2.8961538461538462e-05,
+      "loss": 0.0105,
+      "step": 5878
+    },
+    {
+      "epoch": 38.175324675324674,
+      "grad_norm": 0.30105969309806824,
+      "learning_rate": 2.896123076923077e-05,
+      "loss": 0.0049,
+      "step": 5879
+    },
+    {
+      "epoch": 38.18181818181818,
+      "grad_norm": 0.4990985095500946,
+      "learning_rate": 2.8960923076923076e-05,
+      "loss": 0.0321,
+      "step": 5880
+    },
+    {
+      "epoch": 38.188311688311686,
+      "grad_norm": 0.8950565457344055,
+      "learning_rate": 2.8960615384615384e-05,
+      "loss": 0.0137,
+      "step": 5881
+    },
+    {
+      "epoch": 38.1948051948052,
+      "grad_norm": 0.21951325237751007,
+      "learning_rate": 2.8960307692307692e-05,
+      "loss": 0.003,
+      "step": 5882
+    },
+    {
+      "epoch": 38.201298701298704,
+      "grad_norm": 0.8957061171531677,
+      "learning_rate": 2.896e-05,
+      "loss": 0.0346,
+      "step": 5883
+    },
+    {
+      "epoch": 38.20779220779221,
+      "grad_norm": 0.23368746042251587,
+      "learning_rate": 2.895969230769231e-05,
+      "loss": 0.0039,
+      "step": 5884
+    },
+    {
+      "epoch": 38.214285714285715,
+      "grad_norm": 0.3076799213886261,
+      "learning_rate": 2.8959384615384617e-05,
+      "loss": 0.0086,
+      "step": 5885
+    },
+    {
+      "epoch": 38.22077922077922,
+      "grad_norm": 0.6283580660820007,
+      "learning_rate": 2.8959076923076922e-05,
+      "loss": 0.0077,
+      "step": 5886
+    },
+    {
+      "epoch": 38.22727272727273,
+      "grad_norm": 0.48088017106056213,
+      "learning_rate": 2.895876923076923e-05,
+      "loss": 0.0068,
+      "step": 5887
+    },
+    {
+      "epoch": 38.23376623376623,
+      "grad_norm": 1.0112690925598145,
+      "learning_rate": 2.895846153846154e-05,
+      "loss": 0.0069,
+      "step": 5888
+    },
+    {
+      "epoch": 38.24025974025974,
+      "grad_norm": 3.1249289512634277,
+      "learning_rate": 2.8958153846153844e-05,
+      "loss": 0.0456,
+      "step": 5889
+    },
+    {
+      "epoch": 38.246753246753244,
+      "grad_norm": 1.8980683088302612,
+      "learning_rate": 2.8957846153846155e-05,
+      "loss": 0.049,
+      "step": 5890
+    },
+    {
+      "epoch": 38.253246753246756,
+      "grad_norm": 0.6966969966888428,
+      "learning_rate": 2.8957538461538464e-05,
+      "loss": 0.079,
+      "step": 5891
+    },
+    {
+      "epoch": 38.25974025974026,
+      "grad_norm": 0.50810307264328,
+      "learning_rate": 2.895723076923077e-05,
+      "loss": 0.0532,
+      "step": 5892
+    },
+    {
+      "epoch": 38.26623376623377,
+      "grad_norm": 0.6730650067329407,
+      "learning_rate": 2.8956923076923077e-05,
+      "loss": 0.0406,
+      "step": 5893
+    },
+    {
+      "epoch": 38.27272727272727,
+      "grad_norm": 0.8003307580947876,
+      "learning_rate": 2.8956615384615385e-05,
+      "loss": 0.0473,
+      "step": 5894
+    },
+    {
+      "epoch": 38.27922077922078,
+      "grad_norm": 0.36403682827949524,
+      "learning_rate": 2.895630769230769e-05,
+      "loss": 0.0263,
+      "step": 5895
+    },
+    {
+      "epoch": 38.285714285714285,
+      "grad_norm": 0.359247088432312,
+      "learning_rate": 2.8956e-05,
+      "loss": 0.0172,
+      "step": 5896
+    },
+    {
+      "epoch": 38.29220779220779,
+      "grad_norm": 0.26991915702819824,
+      "learning_rate": 2.895569230769231e-05,
+      "loss": 0.0153,
+      "step": 5897
+    },
+    {
+      "epoch": 38.298701298701296,
+      "grad_norm": 0.4487290680408478,
+      "learning_rate": 2.895538461538462e-05,
+      "loss": 0.02,
+      "step": 5898
+    },
+    {
+      "epoch": 38.3051948051948,
+      "grad_norm": 0.41255083680152893,
+      "learning_rate": 2.8955076923076924e-05,
+      "loss": 0.0127,
+      "step": 5899
+    },
+    {
+      "epoch": 38.311688311688314,
+      "grad_norm": 0.4111977815628052,
+      "learning_rate": 2.8954769230769232e-05,
+      "loss": 0.0173,
+      "step": 5900
+    },
+    {
+      "epoch": 38.31818181818182,
+      "grad_norm": 0.32981088757514954,
+      "learning_rate": 2.895446153846154e-05,
+      "loss": 0.0126,
+      "step": 5901
+    },
+    {
+      "epoch": 38.324675324675326,
+      "grad_norm": 0.37807318568229675,
+      "learning_rate": 2.8954153846153845e-05,
+      "loss": 0.0152,
+      "step": 5902
+    },
+    {
+      "epoch": 38.33116883116883,
+      "grad_norm": 0.43366217613220215,
+      "learning_rate": 2.8953846153846153e-05,
+      "loss": 0.0146,
+      "step": 5903
+    },
+    {
+      "epoch": 38.33766233766234,
+      "grad_norm": 0.3417205214500427,
+      "learning_rate": 2.8953538461538465e-05,
+      "loss": 0.0093,
+      "step": 5904
+    },
+    {
+      "epoch": 38.34415584415584,
+      "grad_norm": 0.5234576463699341,
+      "learning_rate": 2.895323076923077e-05,
+      "loss": 0.0083,
+      "step": 5905
+    },
+    {
+      "epoch": 38.35064935064935,
+      "grad_norm": 0.2839077115058899,
+      "learning_rate": 2.895292307692308e-05,
+      "loss": 0.0076,
+      "step": 5906
+    },
+    {
+      "epoch": 38.357142857142854,
+      "grad_norm": 0.35945460200309753,
+      "learning_rate": 2.8952615384615387e-05,
+      "loss": 0.0085,
+      "step": 5907
+    },
+    {
+      "epoch": 38.36363636363637,
+      "grad_norm": 0.5638504028320312,
+      "learning_rate": 2.895230769230769e-05,
+      "loss": 0.0093,
+      "step": 5908
+    },
+    {
+      "epoch": 38.37012987012987,
+      "grad_norm": 1.1708959341049194,
+      "learning_rate": 2.8952e-05,
+      "loss": 0.0206,
+      "step": 5909
+    },
+    {
+      "epoch": 38.37662337662338,
+      "grad_norm": 0.12267431616783142,
+      "learning_rate": 2.8951692307692308e-05,
+      "loss": 0.0016,
+      "step": 5910
+    },
+    {
+      "epoch": 38.383116883116884,
+      "grad_norm": 0.5171460509300232,
+      "learning_rate": 2.8951384615384617e-05,
+      "loss": 0.0113,
+      "step": 5911
+    },
+    {
+      "epoch": 38.38961038961039,
+      "grad_norm": 0.5932318568229675,
+      "learning_rate": 2.8951076923076925e-05,
+      "loss": 0.0059,
+      "step": 5912
+    },
+    {
+      "epoch": 38.396103896103895,
+      "grad_norm": 0.5508298873901367,
+      "learning_rate": 2.8950769230769233e-05,
+      "loss": 0.0046,
+      "step": 5913
+    },
+    {
+      "epoch": 38.4025974025974,
+      "grad_norm": 0.5931794047355652,
+      "learning_rate": 2.8950461538461538e-05,
+      "loss": 0.0061,
+      "step": 5914
+    },
+    {
+      "epoch": 38.40909090909091,
+      "grad_norm": 1.0745666027069092,
+      "learning_rate": 2.8950153846153846e-05,
+      "loss": 0.0184,
+      "step": 5915
+    },
+    {
+      "epoch": 38.41558441558441,
+      "grad_norm": 1.1727707386016846,
+      "learning_rate": 2.8949846153846155e-05,
+      "loss": 0.0113,
+      "step": 5916
+    },
+    {
+      "epoch": 38.422077922077925,
+      "grad_norm": 1.9281693696975708,
+      "learning_rate": 2.894953846153846e-05,
+      "loss": 0.028,
+      "step": 5917
+    },
+    {
+      "epoch": 38.42857142857143,
+      "grad_norm": 0.5420227646827698,
+      "learning_rate": 2.894923076923077e-05,
+      "loss": 0.0056,
+      "step": 5918
+    },
+    {
+      "epoch": 38.435064935064936,
+      "grad_norm": 0.42538145184516907,
+      "learning_rate": 2.894892307692308e-05,
+      "loss": 0.0079,
+      "step": 5919
+    },
+    {
+      "epoch": 38.44155844155844,
+      "grad_norm": 1.2814815044403076,
+      "learning_rate": 2.8948615384615385e-05,
+      "loss": 0.0142,
+      "step": 5920
+    },
+    {
+      "epoch": 38.44805194805195,
+      "grad_norm": 0.6370809078216553,
+      "learning_rate": 2.8948307692307693e-05,
+      "loss": 0.0208,
+      "step": 5921
+    },
+    {
+      "epoch": 38.45454545454545,
+      "grad_norm": 1.0014452934265137,
+      "learning_rate": 2.8948e-05,
+      "loss": 0.0525,
+      "step": 5922
+    },
+    {
+      "epoch": 38.46103896103896,
+      "grad_norm": 0.9764885306358337,
+      "learning_rate": 2.8947692307692306e-05,
+      "loss": 0.0139,
+      "step": 5923
+    },
+    {
+      "epoch": 38.467532467532465,
+      "grad_norm": 1.2797648906707764,
+      "learning_rate": 2.8947384615384618e-05,
+      "loss": 0.0118,
+      "step": 5924
+    },
+    {
+      "epoch": 38.47402597402598,
+      "grad_norm": 0.826373279094696,
+      "learning_rate": 2.8947076923076926e-05,
+      "loss": 0.0089,
+      "step": 5925
+    },
+    {
+      "epoch": 38.48051948051948,
+      "grad_norm": 1.1263399124145508,
+      "learning_rate": 2.894676923076923e-05,
+      "loss": 0.0097,
+      "step": 5926
+    },
+    {
+      "epoch": 38.48701298701299,
+      "grad_norm": 1.4181472063064575,
+      "learning_rate": 2.894646153846154e-05,
+      "loss": 0.0233,
+      "step": 5927
+    },
+    {
+      "epoch": 38.493506493506494,
+      "grad_norm": 0.6018449664115906,
+      "learning_rate": 2.8946153846153848e-05,
+      "loss": 0.0187,
+      "step": 5928
+    },
+    {
+      "epoch": 38.5,
+      "grad_norm": 0.7165797352790833,
+      "learning_rate": 2.8945846153846153e-05,
+      "loss": 0.0856,
+      "step": 5929
+    },
+    {
+      "epoch": 38.506493506493506,
+      "grad_norm": 0.7134907841682434,
+      "learning_rate": 2.894553846153846e-05,
+      "loss": 0.0648,
+      "step": 5930
+    },
+    {
+      "epoch": 38.51298701298701,
+      "grad_norm": 0.4745461642742157,
+      "learning_rate": 2.8945230769230773e-05,
+      "loss": 0.0426,
+      "step": 5931
+    },
+    {
+      "epoch": 38.51948051948052,
+      "grad_norm": 0.40484580397605896,
+      "learning_rate": 2.8944923076923078e-05,
+      "loss": 0.0441,
+      "step": 5932
+    },
+    {
+      "epoch": 38.52597402597402,
+      "grad_norm": 0.3975403308868408,
+      "learning_rate": 2.8944615384615386e-05,
+      "loss": 0.0317,
+      "step": 5933
+    },
+    {
+      "epoch": 38.532467532467535,
+      "grad_norm": 0.4806075692176819,
+      "learning_rate": 2.8944307692307694e-05,
+      "loss": 0.029,
+      "step": 5934
+    },
+    {
+      "epoch": 38.53896103896104,
+      "grad_norm": 0.4230036437511444,
+      "learning_rate": 2.8944e-05,
+      "loss": 0.0254,
+      "step": 5935
+    },
+    {
+      "epoch": 38.54545454545455,
+      "grad_norm": 0.40213966369628906,
+      "learning_rate": 2.8943692307692308e-05,
+      "loss": 0.0261,
+      "step": 5936
+    },
+    {
+      "epoch": 38.55194805194805,
+      "grad_norm": 0.3931240737438202,
+      "learning_rate": 2.8943384615384616e-05,
+      "loss": 0.0212,
+      "step": 5937
+    },
+    {
+      "epoch": 38.55844155844156,
+      "grad_norm": 0.3137282431125641,
+      "learning_rate": 2.8943076923076924e-05,
+      "loss": 0.0155,
+      "step": 5938
+    },
+    {
+      "epoch": 38.564935064935064,
+      "grad_norm": 0.43438827991485596,
+      "learning_rate": 2.8942769230769233e-05,
+      "loss": 0.0154,
+      "step": 5939
+    },
+    {
+      "epoch": 38.57142857142857,
+      "grad_norm": 0.4714788794517517,
+      "learning_rate": 2.894246153846154e-05,
+      "loss": 0.0216,
+      "step": 5940
+    },
+    {
+      "epoch": 38.577922077922075,
+      "grad_norm": 0.4681532084941864,
+      "learning_rate": 2.8942153846153846e-05,
+      "loss": 0.0188,
+      "step": 5941
+    },
+    {
+      "epoch": 38.58441558441559,
+      "grad_norm": 0.4101855754852295,
+      "learning_rate": 2.8941846153846154e-05,
+      "loss": 0.0316,
+      "step": 5942
+    },
+    {
+      "epoch": 38.59090909090909,
+      "grad_norm": 0.8212214112281799,
+      "learning_rate": 2.8941538461538463e-05,
+      "loss": 0.0098,
+      "step": 5943
+    },
+    {
+      "epoch": 38.5974025974026,
+      "grad_norm": 0.3889029622077942,
+      "learning_rate": 2.8941230769230767e-05,
+      "loss": 0.0125,
+      "step": 5944
+    },
+    {
+      "epoch": 38.603896103896105,
+      "grad_norm": 0.31291624903678894,
+      "learning_rate": 2.894092307692308e-05,
+      "loss": 0.0087,
+      "step": 5945
+    },
+    {
+      "epoch": 38.61038961038961,
+      "grad_norm": 0.4163176119327545,
+      "learning_rate": 2.8940615384615387e-05,
+      "loss": 0.0151,
+      "step": 5946
+    },
+    {
+      "epoch": 38.616883116883116,
+      "grad_norm": 0.3109338879585266,
+      "learning_rate": 2.8940307692307692e-05,
+      "loss": 0.005,
+      "step": 5947
+    },
+    {
+      "epoch": 38.62337662337662,
+      "grad_norm": 0.4024008810520172,
+      "learning_rate": 2.894e-05,
+      "loss": 0.0078,
+      "step": 5948
+    },
+    {
+      "epoch": 38.62987012987013,
+      "grad_norm": 0.2557741105556488,
+      "learning_rate": 2.893969230769231e-05,
+      "loss": 0.0047,
+      "step": 5949
+    },
+    {
+      "epoch": 38.63636363636363,
+      "grad_norm": 0.5303022861480713,
+      "learning_rate": 2.8939384615384614e-05,
+      "loss": 0.0155,
+      "step": 5950
+    },
+    {
+      "epoch": 38.642857142857146,
+      "grad_norm": 0.8672530651092529,
+      "learning_rate": 2.8939076923076922e-05,
+      "loss": 0.0201,
+      "step": 5951
+    },
+    {
+      "epoch": 38.64935064935065,
+      "grad_norm": 0.8533628582954407,
+      "learning_rate": 2.8938769230769234e-05,
+      "loss": 0.0081,
+      "step": 5952
+    },
+    {
+      "epoch": 38.65584415584416,
+      "grad_norm": 0.48627012968063354,
+      "learning_rate": 2.893846153846154e-05,
+      "loss": 0.0142,
+      "step": 5953
+    },
+    {
+      "epoch": 38.66233766233766,
+      "grad_norm": 0.8721800446510315,
+      "learning_rate": 2.8938153846153847e-05,
+      "loss": 0.0129,
+      "step": 5954
+    },
+    {
+      "epoch": 38.66883116883117,
+      "grad_norm": 0.5573851466178894,
+      "learning_rate": 2.8937846153846156e-05,
+      "loss": 0.009,
+      "step": 5955
+    },
+    {
+      "epoch": 38.675324675324674,
+      "grad_norm": 0.35611802339553833,
+      "learning_rate": 2.893753846153846e-05,
+      "loss": 0.0108,
+      "step": 5956
+    },
+    {
+      "epoch": 38.68181818181818,
+      "grad_norm": 0.7225804328918457,
+      "learning_rate": 2.893723076923077e-05,
+      "loss": 0.0091,
+      "step": 5957
+    },
+    {
+      "epoch": 38.688311688311686,
+      "grad_norm": 1.0647177696228027,
+      "learning_rate": 2.8936923076923077e-05,
+      "loss": 0.008,
+      "step": 5958
+    },
+    {
+      "epoch": 38.6948051948052,
+      "grad_norm": 0.27271005511283875,
+      "learning_rate": 2.8936615384615385e-05,
+      "loss": 0.0041,
+      "step": 5959
+    },
+    {
+      "epoch": 38.701298701298704,
+      "grad_norm": 0.8564441800117493,
+      "learning_rate": 2.8936307692307694e-05,
+      "loss": 0.0104,
+      "step": 5960
+    },
+    {
+      "epoch": 38.70779220779221,
+      "grad_norm": 0.8845938444137573,
+      "learning_rate": 2.8936000000000002e-05,
+      "loss": 0.0077,
+      "step": 5961
+    },
+    {
+      "epoch": 38.714285714285715,
+      "grad_norm": 0.567851185798645,
+      "learning_rate": 2.8935692307692307e-05,
+      "loss": 0.0068,
+      "step": 5962
+    },
+    {
+      "epoch": 38.72077922077922,
+      "grad_norm": 0.7509989738464355,
+      "learning_rate": 2.8935384615384615e-05,
+      "loss": 0.0043,
+      "step": 5963
+    },
+    {
+      "epoch": 38.72727272727273,
+      "grad_norm": 0.5271767973899841,
+      "learning_rate": 2.8935076923076924e-05,
+      "loss": 0.005,
+      "step": 5964
+    },
+    {
+      "epoch": 38.73376623376623,
+      "grad_norm": 0.5564625263214111,
+      "learning_rate": 2.893476923076923e-05,
+      "loss": 0.0171,
+      "step": 5965
+    },
+    {
+      "epoch": 38.74025974025974,
+      "grad_norm": 0.4576219618320465,
+      "learning_rate": 2.893446153846154e-05,
+      "loss": 0.0045,
+      "step": 5966
+    },
+    {
+      "epoch": 38.746753246753244,
+      "grad_norm": 0.6022686958312988,
+      "learning_rate": 2.893415384615385e-05,
+      "loss": 0.0684,
+      "step": 5967
+    },
+    {
+      "epoch": 38.753246753246756,
+      "grad_norm": 0.532085657119751,
+      "learning_rate": 2.8933846153846154e-05,
+      "loss": 0.0574,
+      "step": 5968
+    },
+    {
+      "epoch": 38.75974025974026,
+      "grad_norm": 0.5232948064804077,
+      "learning_rate": 2.8933538461538462e-05,
+      "loss": 0.0385,
+      "step": 5969
+    },
+    {
+      "epoch": 38.76623376623377,
+      "grad_norm": 0.40600380301475525,
+      "learning_rate": 2.893323076923077e-05,
+      "loss": 0.0252,
+      "step": 5970
+    },
+    {
+      "epoch": 38.77272727272727,
+      "grad_norm": 0.4164242744445801,
+      "learning_rate": 2.8932923076923075e-05,
+      "loss": 0.03,
+      "step": 5971
+    },
+    {
+      "epoch": 38.77922077922078,
+      "grad_norm": 0.3534366488456726,
+      "learning_rate": 2.8932615384615384e-05,
+      "loss": 0.0163,
+      "step": 5972
+    },
+    {
+      "epoch": 38.785714285714285,
+      "grad_norm": 0.32166191935539246,
+      "learning_rate": 2.8932307692307695e-05,
+      "loss": 0.0223,
+      "step": 5973
+    },
+    {
+      "epoch": 38.79220779220779,
+      "grad_norm": 0.39083895087242126,
+      "learning_rate": 2.8932e-05,
+      "loss": 0.0165,
+      "step": 5974
+    },
+    {
+      "epoch": 38.798701298701296,
+      "grad_norm": 0.3537064790725708,
+      "learning_rate": 2.893169230769231e-05,
+      "loss": 0.0325,
+      "step": 5975
+    },
+    {
+      "epoch": 38.8051948051948,
+      "grad_norm": 0.5228418707847595,
+      "learning_rate": 2.8931384615384617e-05,
+      "loss": 0.0201,
+      "step": 5976
+    },
+    {
+      "epoch": 38.811688311688314,
+      "grad_norm": 0.2531750500202179,
+      "learning_rate": 2.8931076923076925e-05,
+      "loss": 0.0098,
+      "step": 5977
+    },
+    {
+      "epoch": 38.81818181818182,
+      "grad_norm": 0.3136075437068939,
+      "learning_rate": 2.893076923076923e-05,
+      "loss": 0.0091,
+      "step": 5978
+    },
+    {
+      "epoch": 38.824675324675326,
+      "grad_norm": 0.31537479162216187,
+      "learning_rate": 2.893046153846154e-05,
+      "loss": 0.0096,
+      "step": 5979
+    },
+    {
+      "epoch": 38.83116883116883,
+      "grad_norm": 0.41998496651649475,
+      "learning_rate": 2.893015384615385e-05,
+      "loss": 0.0138,
+      "step": 5980
+    },
+    {
+      "epoch": 38.83766233766234,
+      "grad_norm": 0.25479060411453247,
+      "learning_rate": 2.8929846153846155e-05,
+      "loss": 0.0073,
+      "step": 5981
+    },
+    {
+      "epoch": 38.84415584415584,
+      "grad_norm": 0.382602721452713,
+      "learning_rate": 2.8929538461538463e-05,
+      "loss": 0.011,
+      "step": 5982
+    },
+    {
+      "epoch": 38.85064935064935,
+      "grad_norm": 0.6440215706825256,
+      "learning_rate": 2.892923076923077e-05,
+      "loss": 0.0103,
+      "step": 5983
+    },
+    {
+      "epoch": 38.857142857142854,
+      "grad_norm": 0.31958869099617004,
+      "learning_rate": 2.8928923076923077e-05,
+      "loss": 0.0103,
+      "step": 5984
+    },
+    {
+      "epoch": 38.86363636363637,
+      "grad_norm": 0.28973206877708435,
+      "learning_rate": 2.8928615384615385e-05,
+      "loss": 0.0086,
+      "step": 5985
+    },
+    {
+      "epoch": 38.87012987012987,
+      "grad_norm": 0.2433740496635437,
+      "learning_rate": 2.8928307692307693e-05,
+      "loss": 0.007,
+      "step": 5986
+    },
+    {
+      "epoch": 38.87662337662338,
+      "grad_norm": 0.3104799687862396,
+      "learning_rate": 2.8928e-05,
+      "loss": 0.0064,
+      "step": 5987
+    },
+    {
+      "epoch": 38.883116883116884,
+      "grad_norm": 0.2099333554506302,
+      "learning_rate": 2.892769230769231e-05,
+      "loss": 0.0038,
+      "step": 5988
+    },
+    {
+      "epoch": 38.88961038961039,
+      "grad_norm": 0.2746211588382721,
+      "learning_rate": 2.8927384615384618e-05,
+      "loss": 0.0037,
+      "step": 5989
+    },
+    {
+      "epoch": 38.896103896103895,
+      "grad_norm": 0.4701051414012909,
+      "learning_rate": 2.8927076923076923e-05,
+      "loss": 0.0192,
+      "step": 5990
+    },
+    {
+      "epoch": 38.9025974025974,
+      "grad_norm": 0.8103092312812805,
+      "learning_rate": 2.892676923076923e-05,
+      "loss": 0.0038,
+      "step": 5991
+    },
+    {
+      "epoch": 38.90909090909091,
+      "grad_norm": 0.2507099509239197,
+      "learning_rate": 2.892646153846154e-05,
+      "loss": 0.0024,
+      "step": 5992
+    },
+    {
+      "epoch": 38.91558441558441,
+      "grad_norm": 0.6929746270179749,
+      "learning_rate": 2.8926153846153845e-05,
+      "loss": 0.0083,
+      "step": 5993
+    },
+    {
+      "epoch": 38.922077922077925,
+      "grad_norm": 1.030951976776123,
+      "learning_rate": 2.8925846153846156e-05,
+      "loss": 0.0186,
+      "step": 5994
+    },
+    {
+      "epoch": 38.92857142857143,
+      "grad_norm": 0.7033496499061584,
+      "learning_rate": 2.8925538461538465e-05,
+      "loss": 0.0207,
+      "step": 5995
+    },
+    {
+      "epoch": 38.935064935064936,
+      "grad_norm": 0.38966307044029236,
+      "learning_rate": 2.892523076923077e-05,
+      "loss": 0.0044,
+      "step": 5996
+    },
+    {
+      "epoch": 38.94155844155844,
+      "grad_norm": 0.5722068548202515,
+      "learning_rate": 2.8924923076923078e-05,
+      "loss": 0.0061,
+      "step": 5997
+    },
+    {
+      "epoch": 38.94805194805195,
+      "grad_norm": 0.504027247428894,
+      "learning_rate": 2.8924615384615386e-05,
+      "loss": 0.0072,
+      "step": 5998
+    },
+    {
+      "epoch": 38.95454545454545,
+      "grad_norm": 1.0058695077896118,
+      "learning_rate": 2.892430769230769e-05,
+      "loss": 0.0085,
+      "step": 5999
+    },
+    {
+      "epoch": 38.96103896103896,
+      "grad_norm": 7.794951438903809,
+      "learning_rate": 2.8924e-05,
+      "loss": 0.0138,
+      "step": 6000
+    },
+    {
+      "epoch": 38.96103896103896,
+      "eval_cer": 0.08826600273070435,
+      "eval_loss": 0.4513417184352875,
+      "eval_runtime": 8.2069,
+      "eval_samples_per_second": 12.307,
+      "eval_steps_per_second": 0.487,
+      "eval_wer": 0.28174603174603174,
+      "step": 6000
     }
   ],
   "logging_steps": 1.0,
@@ -35071,7 +42081,7 @@
         "early_stopping_threshold": 0.0
       },
       "attributes": {
-        "early_stopping_patience_counter": 3
+        "early_stopping_patience_counter": 4
       }
     },
     "TrainerControl": {
@@ -35085,7 +42095,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 6.805472800258521e+19,
+  "total_flos": 8.165773678594713e+19,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null