diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,11424 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9975429975429977,
+  "eval_steps": 500,
+  "global_step": 1626,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0018427518427518428,
+      "grad_norm": 6.936351874160832,
+      "learning_rate": 6.134969325153375e-08,
+      "loss": 1.1183,
+      "step": 1
+    },
+    {
+      "epoch": 0.0036855036855036856,
+      "grad_norm": 7.158994511409215,
+      "learning_rate": 1.226993865030675e-07,
+      "loss": 1.1817,
+      "step": 2
+    },
+    {
+      "epoch": 0.005528255528255528,
+      "grad_norm": 6.900134700379675,
+      "learning_rate": 1.8404907975460125e-07,
+      "loss": 1.1321,
+      "step": 3
+    },
+    {
+      "epoch": 0.007371007371007371,
+      "grad_norm": 7.166011314686131,
+      "learning_rate": 2.45398773006135e-07,
+      "loss": 1.162,
+      "step": 4
+    },
+    {
+      "epoch": 0.009213759213759214,
+      "grad_norm": 6.926775888552087,
+      "learning_rate": 3.0674846625766876e-07,
+      "loss": 1.1376,
+      "step": 5
+    },
+    {
+      "epoch": 0.011056511056511056,
+      "grad_norm": 6.944284690323478,
+      "learning_rate": 3.680981595092025e-07,
+      "loss": 1.1355,
+      "step": 6
+    },
+    {
+      "epoch": 0.012899262899262898,
+      "grad_norm": 6.899252235817831,
+      "learning_rate": 4.2944785276073624e-07,
+      "loss": 1.1475,
+      "step": 7
+    },
+    {
+      "epoch": 0.014742014742014743,
+      "grad_norm": 6.949372342254332,
+      "learning_rate": 4.9079754601227e-07,
+      "loss": 1.168,
+      "step": 8
+    },
+    {
+      "epoch": 0.016584766584766583,
+      "grad_norm": 6.7926540294370765,
+      "learning_rate": 5.521472392638038e-07,
+      "loss": 1.1229,
+      "step": 9
+    },
+    {
+      "epoch": 0.018427518427518427,
+      "grad_norm": 6.5739409924739585,
+      "learning_rate": 6.134969325153375e-07,
+      "loss": 1.139,
+      "step": 10
+    },
+    {
+      "epoch": 0.02027027027027027,
+      "grad_norm": 6.330974764660368,
+      "learning_rate": 6.748466257668713e-07,
+      "loss": 1.1237,
+      "step": 11
+    },
+    {
+      "epoch": 0.022113022113022112,
+      "grad_norm": 6.04789660942285,
+      "learning_rate": 7.36196319018405e-07,
+      "loss": 1.0987,
+      "step": 12
+    },
+    {
+      "epoch": 0.023955773955773956,
+      "grad_norm": 5.346449489290361,
+      "learning_rate": 7.975460122699387e-07,
+      "loss": 1.0936,
+      "step": 13
+    },
+    {
+      "epoch": 0.025798525798525797,
+      "grad_norm": 4.809089592305406,
+      "learning_rate": 8.588957055214725e-07,
+      "loss": 1.0725,
+      "step": 14
+    },
+    {
+      "epoch": 0.02764127764127764,
+      "grad_norm": 5.070637009206699,
+      "learning_rate": 9.202453987730062e-07,
+      "loss": 1.1232,
+      "step": 15
+    },
+    {
+      "epoch": 0.029484029484029485,
+      "grad_norm": 4.6635708408011896,
+      "learning_rate": 9.8159509202454e-07,
+      "loss": 1.0729,
+      "step": 16
+    },
+    {
+      "epoch": 0.03132678132678133,
+      "grad_norm": 4.455516638294391,
+      "learning_rate": 1.0429447852760736e-06,
+      "loss": 1.0542,
+      "step": 17
+    },
+    {
+      "epoch": 0.033169533169533166,
+      "grad_norm": 3.1158304289507015,
+      "learning_rate": 1.1042944785276075e-06,
+      "loss": 1.0505,
+      "step": 18
+    },
+    {
+      "epoch": 0.03501228501228501,
+      "grad_norm": 3.053075103284106,
+      "learning_rate": 1.165644171779141e-06,
+      "loss": 1.0284,
+      "step": 19
+    },
+    {
+      "epoch": 0.036855036855036855,
+      "grad_norm": 3.020154383597658,
+      "learning_rate": 1.226993865030675e-06,
+      "loss": 1.0471,
+      "step": 20
+    },
+    {
+      "epoch": 0.0386977886977887,
+      "grad_norm": 2.787685107130267,
+      "learning_rate": 1.2883435582822088e-06,
+      "loss": 1.0182,
+      "step": 21
+    },
+    {
+      "epoch": 0.04054054054054054,
+      "grad_norm": 2.697088901105455,
+      "learning_rate": 1.3496932515337425e-06,
+      "loss": 1.0042,
+      "step": 22
+    },
+    {
+      "epoch": 0.04238329238329238,
+      "grad_norm": 2.62049952646934,
+      "learning_rate": 1.411042944785276e-06,
+      "loss": 1.0146,
+      "step": 23
+    },
+    {
+      "epoch": 0.044226044226044224,
+      "grad_norm": 2.1007147795693824,
+      "learning_rate": 1.47239263803681e-06,
+      "loss": 0.9774,
+      "step": 24
+    },
+    {
+      "epoch": 0.04606879606879607,
+      "grad_norm": 2.1071604777232347,
+      "learning_rate": 1.5337423312883437e-06,
+      "loss": 0.9498,
+      "step": 25
+    },
+    {
+      "epoch": 0.04791154791154791,
+      "grad_norm": 2.2415372568761365,
+      "learning_rate": 1.5950920245398775e-06,
+      "loss": 0.9597,
+      "step": 26
+    },
+    {
+      "epoch": 0.049754299754299756,
+      "grad_norm": 2.3232158330556163,
+      "learning_rate": 1.656441717791411e-06,
+      "loss": 0.9531,
+      "step": 27
+    },
+    {
+      "epoch": 0.051597051597051594,
+      "grad_norm": 2.119756267506338,
+      "learning_rate": 1.717791411042945e-06,
+      "loss": 0.9279,
+      "step": 28
+    },
+    {
+      "epoch": 0.05343980343980344,
+      "grad_norm": 1.9256340483319656,
+      "learning_rate": 1.7791411042944787e-06,
+      "loss": 0.944,
+      "step": 29
+    },
+    {
+      "epoch": 0.05528255528255528,
+      "grad_norm": 1.6753643993462066,
+      "learning_rate": 1.8404907975460124e-06,
+      "loss": 0.9362,
+      "step": 30
+    },
+    {
+      "epoch": 0.057125307125307126,
+      "grad_norm": 1.6020227177314084,
+      "learning_rate": 1.9018404907975464e-06,
+      "loss": 0.919,
+      "step": 31
+    },
+    {
+      "epoch": 0.05896805896805897,
+      "grad_norm": 1.3395344004479615,
+      "learning_rate": 1.96319018404908e-06,
+      "loss": 0.877,
+      "step": 32
+    },
+    {
+      "epoch": 0.060810810810810814,
+      "grad_norm": 1.579271323390205,
+      "learning_rate": 2.0245398773006137e-06,
+      "loss": 0.8759,
+      "step": 33
+    },
+    {
+      "epoch": 0.06265356265356266,
+      "grad_norm": 1.625656135789201,
+      "learning_rate": 2.085889570552147e-06,
+      "loss": 0.8433,
+      "step": 34
+    },
+    {
+      "epoch": 0.0644963144963145,
+      "grad_norm": 1.6532364497724479,
+      "learning_rate": 2.147239263803681e-06,
+      "loss": 0.865,
+      "step": 35
+    },
+    {
+      "epoch": 0.06633906633906633,
+      "grad_norm": 1.3392932466475616,
+      "learning_rate": 2.208588957055215e-06,
+      "loss": 0.8652,
+      "step": 36
+    },
+    {
+      "epoch": 0.06818181818181818,
+      "grad_norm": 1.2107215815762167,
+      "learning_rate": 2.2699386503067486e-06,
+      "loss": 0.8471,
+      "step": 37
+    },
+    {
+      "epoch": 0.07002457002457002,
+      "grad_norm": 1.031549014054998,
+      "learning_rate": 2.331288343558282e-06,
+      "loss": 0.7976,
+      "step": 38
+    },
+    {
+      "epoch": 0.07186732186732187,
+      "grad_norm": 1.191899584969997,
+      "learning_rate": 2.392638036809816e-06,
+      "loss": 0.842,
+      "step": 39
+    },
+    {
+      "epoch": 0.07371007371007371,
+      "grad_norm": 1.217214930314069,
+      "learning_rate": 2.45398773006135e-06,
+      "loss": 0.8479,
+      "step": 40
+    },
+    {
+      "epoch": 0.07555282555282555,
+      "grad_norm": 1.1802438597204925,
+      "learning_rate": 2.5153374233128836e-06,
+      "loss": 0.8437,
+      "step": 41
+    },
+    {
+      "epoch": 0.0773955773955774,
+      "grad_norm": 0.9909732109045103,
+      "learning_rate": 2.5766871165644175e-06,
+      "loss": 0.8306,
+      "step": 42
+    },
+    {
+      "epoch": 0.07923832923832924,
+      "grad_norm": 0.9742267400035776,
+      "learning_rate": 2.638036809815951e-06,
+      "loss": 0.8456,
+      "step": 43
+    },
+    {
+      "epoch": 0.08108108108108109,
+      "grad_norm": 0.8587565028364036,
+      "learning_rate": 2.699386503067485e-06,
+      "loss": 0.8004,
+      "step": 44
+    },
+    {
+      "epoch": 0.08292383292383293,
+      "grad_norm": 0.9054071564204477,
+      "learning_rate": 2.760736196319019e-06,
+      "loss": 0.807,
+      "step": 45
+    },
+    {
+      "epoch": 0.08476658476658476,
+      "grad_norm": 0.9538402644325439,
+      "learning_rate": 2.822085889570552e-06,
+      "loss": 0.7893,
+      "step": 46
+    },
+    {
+      "epoch": 0.0866093366093366,
+      "grad_norm": 0.843274932244754,
+      "learning_rate": 2.883435582822086e-06,
+      "loss": 0.8217,
+      "step": 47
+    },
+    {
+      "epoch": 0.08845208845208845,
+      "grad_norm": 0.8189952753848745,
+      "learning_rate": 2.94478527607362e-06,
+      "loss": 0.8143,
+      "step": 48
+    },
+    {
+      "epoch": 0.09029484029484029,
+      "grad_norm": 1.0140362990166192,
+      "learning_rate": 3.0061349693251535e-06,
+      "loss": 0.8021,
+      "step": 49
+    },
+    {
+      "epoch": 0.09213759213759214,
+      "grad_norm": 0.8837423011690458,
+      "learning_rate": 3.0674846625766875e-06,
+      "loss": 0.813,
+      "step": 50
+    },
+    {
+      "epoch": 0.09398034398034398,
+      "grad_norm": 0.8761214471080836,
+      "learning_rate": 3.1288343558282214e-06,
+      "loss": 0.8153,
+      "step": 51
+    },
+    {
+      "epoch": 0.09582309582309582,
+      "grad_norm": 0.9510690176844926,
+      "learning_rate": 3.190184049079755e-06,
+      "loss": 0.7842,
+      "step": 52
+    },
+    {
+      "epoch": 0.09766584766584767,
+      "grad_norm": 0.8155017545341612,
+      "learning_rate": 3.251533742331289e-06,
+      "loss": 0.7771,
+      "step": 53
+    },
+    {
+      "epoch": 0.09950859950859951,
+      "grad_norm": 0.8275968863391634,
+      "learning_rate": 3.312883435582822e-06,
+      "loss": 0.7733,
+      "step": 54
+    },
+    {
+      "epoch": 0.10135135135135136,
+      "grad_norm": 0.8410873610920073,
+      "learning_rate": 3.374233128834356e-06,
+      "loss": 0.7698,
+      "step": 55
+    },
+    {
+      "epoch": 0.10319410319410319,
+      "grad_norm": 0.892526823952353,
+      "learning_rate": 3.43558282208589e-06,
+      "loss": 0.7745,
+      "step": 56
+    },
+    {
+      "epoch": 0.10503685503685503,
+      "grad_norm": 0.8353139757677477,
+      "learning_rate": 3.4969325153374234e-06,
+      "loss": 0.78,
+      "step": 57
+    },
+    {
+      "epoch": 0.10687960687960688,
+      "grad_norm": 0.7838112728518961,
+      "learning_rate": 3.5582822085889574e-06,
+      "loss": 0.7678,
+      "step": 58
+    },
+    {
+      "epoch": 0.10872235872235872,
+      "grad_norm": 0.8047758930431282,
+      "learning_rate": 3.6196319018404913e-06,
+      "loss": 0.7669,
+      "step": 59
+    },
+    {
+      "epoch": 0.11056511056511056,
+      "grad_norm": 0.7985358361684849,
+      "learning_rate": 3.680981595092025e-06,
+      "loss": 0.757,
+      "step": 60
+    },
+    {
+      "epoch": 0.11240786240786241,
+      "grad_norm": 0.7691475271563908,
+      "learning_rate": 3.742331288343559e-06,
+      "loss": 0.7568,
+      "step": 61
+    },
+    {
+      "epoch": 0.11425061425061425,
+      "grad_norm": 0.8751013576486238,
+      "learning_rate": 3.8036809815950928e-06,
+      "loss": 0.7704,
+      "step": 62
+    },
+    {
+      "epoch": 0.1160933660933661,
+      "grad_norm": 0.7342280099332974,
+      "learning_rate": 3.865030674846626e-06,
+      "loss": 0.7764,
+      "step": 63
+    },
+    {
+      "epoch": 0.11793611793611794,
+      "grad_norm": 0.7268967755716395,
+      "learning_rate": 3.92638036809816e-06,
+      "loss": 0.772,
+      "step": 64
+    },
+    {
+      "epoch": 0.11977886977886978,
+      "grad_norm": 0.7302479363998343,
+      "learning_rate": 3.987730061349693e-06,
+      "loss": 0.7458,
+      "step": 65
+    },
+    {
+      "epoch": 0.12162162162162163,
+      "grad_norm": 0.9177156240056192,
+      "learning_rate": 4.049079754601227e-06,
+      "loss": 0.7619,
+      "step": 66
+    },
+    {
+      "epoch": 0.12346437346437346,
+      "grad_norm": 0.7441158010886504,
+      "learning_rate": 4.110429447852761e-06,
+      "loss": 0.7393,
+      "step": 67
+    },
+    {
+      "epoch": 0.12530712530712532,
+      "grad_norm": 0.7679007501482531,
+      "learning_rate": 4.171779141104294e-06,
+      "loss": 0.7707,
+      "step": 68
+    },
+    {
+      "epoch": 0.12714987714987716,
+      "grad_norm": 0.7081815423294785,
+      "learning_rate": 4.233128834355829e-06,
+      "loss": 0.7387,
+      "step": 69
+    },
+    {
+      "epoch": 0.128992628992629,
+      "grad_norm": 0.7405989060425029,
+      "learning_rate": 4.294478527607362e-06,
+      "loss": 0.7546,
+      "step": 70
+    },
+    {
+      "epoch": 0.13083538083538085,
+      "grad_norm": 0.8003646530920945,
+      "learning_rate": 4.355828220858896e-06,
+      "loss": 0.7501,
+      "step": 71
+    },
+    {
+      "epoch": 0.13267813267813267,
+      "grad_norm": 0.7379190837251934,
+      "learning_rate": 4.41717791411043e-06,
+      "loss": 0.7532,
+      "step": 72
+    },
+    {
+      "epoch": 0.1345208845208845,
+      "grad_norm": 0.7238652578013907,
+      "learning_rate": 4.478527607361964e-06,
+      "loss": 0.7466,
+      "step": 73
+    },
+    {
+      "epoch": 0.13636363636363635,
+      "grad_norm": 0.7338969066596017,
+      "learning_rate": 4.539877300613497e-06,
+      "loss": 0.7427,
+      "step": 74
+    },
+    {
+      "epoch": 0.1382063882063882,
+      "grad_norm": 0.7546105668450892,
+      "learning_rate": 4.601226993865031e-06,
+      "loss": 0.7344,
+      "step": 75
+    },
+    {
+      "epoch": 0.14004914004914004,
+      "grad_norm": 0.7071131047892283,
+      "learning_rate": 4.662576687116564e-06,
+      "loss": 0.7784,
+      "step": 76
+    },
+    {
+      "epoch": 0.14189189189189189,
+      "grad_norm": 0.7729673418348249,
+      "learning_rate": 4.723926380368099e-06,
+      "loss": 0.7366,
+      "step": 77
+    },
+    {
+      "epoch": 0.14373464373464373,
+      "grad_norm": 0.7484567741319061,
+      "learning_rate": 4.785276073619632e-06,
+      "loss": 0.7586,
+      "step": 78
+    },
+    {
+      "epoch": 0.14557739557739557,
+      "grad_norm": 0.803054357317792,
+      "learning_rate": 4.846625766871166e-06,
+      "loss": 0.7534,
+      "step": 79
+    },
+    {
+      "epoch": 0.14742014742014742,
+      "grad_norm": 0.7266203197907423,
+      "learning_rate": 4.9079754601227e-06,
+      "loss": 0.7237,
+      "step": 80
+    },
+    {
+      "epoch": 0.14926289926289926,
+      "grad_norm": 0.9496426061914155,
+      "learning_rate": 4.969325153374234e-06,
+      "loss": 0.7614,
+      "step": 81
+    },
+    {
+      "epoch": 0.1511056511056511,
+      "grad_norm": 0.784906575776571,
+      "learning_rate": 5.030674846625767e-06,
+      "loss": 0.7154,
+      "step": 82
+    },
+    {
+      "epoch": 0.15294840294840295,
+      "grad_norm": 0.9272906254175235,
+      "learning_rate": 5.092024539877301e-06,
+      "loss": 0.7551,
+      "step": 83
+    },
+    {
+      "epoch": 0.1547911547911548,
+      "grad_norm": 0.7556504360818758,
+      "learning_rate": 5.153374233128835e-06,
+      "loss": 0.7178,
+      "step": 84
+    },
+    {
+      "epoch": 0.15663390663390664,
+      "grad_norm": 0.9541915899209443,
+      "learning_rate": 5.214723926380369e-06,
+      "loss": 0.7494,
+      "step": 85
+    },
+    {
+      "epoch": 0.15847665847665848,
+      "grad_norm": 0.8301496563130755,
+      "learning_rate": 5.276073619631902e-06,
+      "loss": 0.7038,
+      "step": 86
+    },
+    {
+      "epoch": 0.16031941031941033,
+      "grad_norm": 0.7865792257326523,
+      "learning_rate": 5.3374233128834365e-06,
+      "loss": 0.7289,
+      "step": 87
+    },
+    {
+      "epoch": 0.16216216216216217,
+      "grad_norm": 0.8116971595208069,
+      "learning_rate": 5.39877300613497e-06,
+      "loss": 0.7264,
+      "step": 88
+    },
+    {
+      "epoch": 0.16400491400491402,
+      "grad_norm": 0.7798872449183105,
+      "learning_rate": 5.460122699386503e-06,
+      "loss": 0.7279,
+      "step": 89
+    },
+    {
+      "epoch": 0.16584766584766586,
+      "grad_norm": 0.8031674497121164,
+      "learning_rate": 5.521472392638038e-06,
+      "loss": 0.7285,
+      "step": 90
+    },
+    {
+      "epoch": 0.1676904176904177,
+      "grad_norm": 0.7559509146269632,
+      "learning_rate": 5.582822085889571e-06,
+      "loss": 0.7112,
+      "step": 91
+    },
+    {
+      "epoch": 0.16953316953316952,
+      "grad_norm": 0.7702594848014152,
+      "learning_rate": 5.644171779141104e-06,
+      "loss": 0.7162,
+      "step": 92
+    },
+    {
+      "epoch": 0.17137592137592136,
+      "grad_norm": 0.7975909360641498,
+      "learning_rate": 5.7055214723926385e-06,
+      "loss": 0.7247,
+      "step": 93
+    },
+    {
+      "epoch": 0.1732186732186732,
+      "grad_norm": 0.784639177136631,
+      "learning_rate": 5.766871165644172e-06,
+      "loss": 0.7305,
+      "step": 94
+    },
+    {
+      "epoch": 0.17506142506142505,
+      "grad_norm": 0.884945536860071,
+      "learning_rate": 5.828220858895706e-06,
+      "loss": 0.7144,
+      "step": 95
+    },
+    {
+      "epoch": 0.1769041769041769,
+      "grad_norm": 0.754761466435824,
+      "learning_rate": 5.88957055214724e-06,
+      "loss": 0.7228,
+      "step": 96
+    },
+    {
+      "epoch": 0.17874692874692874,
+      "grad_norm": 0.7290558135483822,
+      "learning_rate": 5.9509202453987735e-06,
+      "loss": 0.7052,
+      "step": 97
+    },
+    {
+      "epoch": 0.18058968058968058,
+      "grad_norm": 0.72135328970469,
+      "learning_rate": 6.012269938650307e-06,
+      "loss": 0.715,
+      "step": 98
+    },
+    {
+      "epoch": 0.18243243243243243,
+      "grad_norm": 0.7950999021050339,
+      "learning_rate": 6.073619631901841e-06,
+      "loss": 0.7235,
+      "step": 99
+    },
+    {
+      "epoch": 0.18427518427518427,
+      "grad_norm": 0.7324129407499743,
+      "learning_rate": 6.134969325153375e-06,
+      "loss": 0.7027,
+      "step": 100
+    },
+    {
+      "epoch": 0.18611793611793612,
+      "grad_norm": 0.8806142140253985,
+      "learning_rate": 6.1963190184049085e-06,
+      "loss": 0.7267,
+      "step": 101
+    },
+    {
+      "epoch": 0.18796068796068796,
+      "grad_norm": 0.9502993375896591,
+      "learning_rate": 6.257668711656443e-06,
+      "loss": 0.7149,
+      "step": 102
+    },
+    {
+      "epoch": 0.1898034398034398,
+      "grad_norm": 0.7503856185197219,
+      "learning_rate": 6.319018404907976e-06,
+      "loss": 0.7318,
+      "step": 103
+    },
+    {
+      "epoch": 0.19164619164619165,
+      "grad_norm": 0.9076999377387144,
+      "learning_rate": 6.38036809815951e-06,
+      "loss": 0.7195,
+      "step": 104
+    },
+    {
+      "epoch": 0.1934889434889435,
+      "grad_norm": 0.9446737115827708,
+      "learning_rate": 6.441717791411043e-06,
+      "loss": 0.7137,
+      "step": 105
+    },
+    {
+      "epoch": 0.19533169533169534,
+      "grad_norm": 0.909559855001219,
+      "learning_rate": 6.503067484662578e-06,
+      "loss": 0.7075,
+      "step": 106
+    },
+    {
+      "epoch": 0.19717444717444718,
+      "grad_norm": 0.9062899775577271,
+      "learning_rate": 6.5644171779141105e-06,
+      "loss": 0.7165,
+      "step": 107
+    },
+    {
+      "epoch": 0.19901719901719903,
+      "grad_norm": 0.8806111586438345,
+      "learning_rate": 6.625766871165644e-06,
+      "loss": 0.7049,
+      "step": 108
+    },
+    {
+      "epoch": 0.20085995085995087,
+      "grad_norm": 0.7662902566868905,
+      "learning_rate": 6.687116564417178e-06,
+      "loss": 0.7072,
+      "step": 109
+    },
+    {
+      "epoch": 0.20270270270270271,
+      "grad_norm": 0.9649792059251925,
+      "learning_rate": 6.748466257668712e-06,
+      "loss": 0.7225,
+      "step": 110
+    },
+    {
+      "epoch": 0.20454545454545456,
+      "grad_norm": 0.7646524557190553,
+      "learning_rate": 6.8098159509202454e-06,
+      "loss": 0.7003,
+      "step": 111
+    },
+    {
+      "epoch": 0.20638820638820637,
+      "grad_norm": 0.9373676279598895,
+      "learning_rate": 6.87116564417178e-06,
+      "loss": 0.712,
+      "step": 112
+    },
+    {
+      "epoch": 0.20823095823095822,
+      "grad_norm": 0.8419987408693567,
+      "learning_rate": 6.932515337423313e-06,
+      "loss": 0.7375,
+      "step": 113
+    },
+    {
+      "epoch": 0.21007371007371006,
+      "grad_norm": 0.7797124959230111,
+      "learning_rate": 6.993865030674847e-06,
+      "loss": 0.7163,
+      "step": 114
+    },
+    {
+      "epoch": 0.2119164619164619,
+      "grad_norm": 0.7709139895272769,
+      "learning_rate": 7.055214723926381e-06,
+      "loss": 0.7173,
+      "step": 115
+    },
+    {
+      "epoch": 0.21375921375921375,
+      "grad_norm": 0.8059618694360401,
+      "learning_rate": 7.116564417177915e-06,
+      "loss": 0.7023,
+      "step": 116
+    },
+    {
+      "epoch": 0.2156019656019656,
+      "grad_norm": 0.7570626280193589,
+      "learning_rate": 7.177914110429448e-06,
+      "loss": 0.7028,
+      "step": 117
+    },
+    {
+      "epoch": 0.21744471744471744,
+      "grad_norm": 0.7490932699028602,
+      "learning_rate": 7.239263803680983e-06,
+      "loss": 0.7057,
+      "step": 118
+    },
+    {
+      "epoch": 0.21928746928746928,
+      "grad_norm": 0.8446232885006794,
+      "learning_rate": 7.300613496932516e-06,
+      "loss": 0.7284,
+      "step": 119
+    },
+    {
+      "epoch": 0.22113022113022113,
+      "grad_norm": 0.7466226311563231,
+      "learning_rate": 7.36196319018405e-06,
+      "loss": 0.7364,
+      "step": 120
+    },
+    {
+      "epoch": 0.22297297297297297,
+      "grad_norm": 0.8721976608869918,
+      "learning_rate": 7.423312883435584e-06,
+      "loss": 0.6977,
+      "step": 121
+    },
+    {
+      "epoch": 0.22481572481572482,
+      "grad_norm": 0.853043109083207,
+      "learning_rate": 7.484662576687118e-06,
+      "loss": 0.6932,
+      "step": 122
+    },
+    {
+      "epoch": 0.22665847665847666,
+      "grad_norm": 0.8479339244570194,
+      "learning_rate": 7.54601226993865e-06,
+      "loss": 0.6977,
+      "step": 123
+    },
+    {
+      "epoch": 0.2285012285012285,
+      "grad_norm": 0.8591983344203613,
+      "learning_rate": 7.6073619631901856e-06,
+      "loss": 0.7153,
+      "step": 124
+    },
+    {
+      "epoch": 0.23034398034398035,
+      "grad_norm": 0.7889778618496568,
+      "learning_rate": 7.668711656441718e-06,
+      "loss": 0.6908,
+      "step": 125
+    },
+    {
+      "epoch": 0.2321867321867322,
+      "grad_norm": 0.7831917399731678,
+      "learning_rate": 7.730061349693252e-06,
+      "loss": 0.7094,
+      "step": 126
+    },
+    {
+      "epoch": 0.23402948402948404,
+      "grad_norm": 1.0249289702537945,
+      "learning_rate": 7.791411042944787e-06,
+      "loss": 0.6988,
+      "step": 127
+    },
+    {
+      "epoch": 0.23587223587223588,
+      "grad_norm": 0.8272383804835828,
+      "learning_rate": 7.85276073619632e-06,
+      "loss": 0.7022,
+      "step": 128
+    },
+    {
+      "epoch": 0.23771498771498772,
+      "grad_norm": 0.8587810595583729,
+      "learning_rate": 7.914110429447854e-06,
+      "loss": 0.6831,
+      "step": 129
+    },
+    {
+      "epoch": 0.23955773955773957,
+      "grad_norm": 0.7620630413219557,
+      "learning_rate": 7.975460122699386e-06,
+      "loss": 0.7037,
+      "step": 130
+    },
+    {
+      "epoch": 0.2414004914004914,
+      "grad_norm": 0.9493896991427978,
+      "learning_rate": 8.036809815950921e-06,
+      "loss": 0.7043,
+      "step": 131
+    },
+    {
+      "epoch": 0.24324324324324326,
+      "grad_norm": 0.7916228013507662,
+      "learning_rate": 8.098159509202455e-06,
+      "loss": 0.7153,
+      "step": 132
+    },
+    {
+      "epoch": 0.24508599508599507,
+      "grad_norm": 0.8866470449252055,
+      "learning_rate": 8.159509202453988e-06,
+      "loss": 0.7292,
+      "step": 133
+    },
+    {
+      "epoch": 0.24692874692874692,
+      "grad_norm": 0.9327786133269829,
+      "learning_rate": 8.220858895705522e-06,
+      "loss": 0.7146,
+      "step": 134
+    },
+    {
+      "epoch": 0.24877149877149876,
+      "grad_norm": 0.875350022903229,
+      "learning_rate": 8.282208588957055e-06,
+      "loss": 0.7068,
+      "step": 135
+    },
+    {
+      "epoch": 0.25061425061425063,
+      "grad_norm": 0.9009363768901936,
+      "learning_rate": 8.343558282208589e-06,
+      "loss": 0.7171,
+      "step": 136
+    },
+    {
+      "epoch": 0.2524570024570025,
+      "grad_norm": 0.8452389477887238,
+      "learning_rate": 8.404907975460124e-06,
+      "loss": 0.7016,
+      "step": 137
+    },
+    {
+      "epoch": 0.2542997542997543,
+      "grad_norm": 0.8367773386273315,
+      "learning_rate": 8.466257668711658e-06,
+      "loss": 0.7277,
+      "step": 138
+    },
+    {
+      "epoch": 0.25614250614250617,
+      "grad_norm": 0.8628318927047455,
+      "learning_rate": 8.527607361963191e-06,
+      "loss": 0.6973,
+      "step": 139
+    },
+    {
+      "epoch": 0.257985257985258,
+      "grad_norm": 0.8735915275325096,
+      "learning_rate": 8.588957055214725e-06,
+      "loss": 0.7208,
+      "step": 140
+    },
+    {
+      "epoch": 0.25982800982800985,
+      "grad_norm": 0.7342838583533036,
+      "learning_rate": 8.650306748466258e-06,
+      "loss": 0.6852,
+      "step": 141
+    },
+    {
+      "epoch": 0.2616707616707617,
+      "grad_norm": 0.7950971811445855,
+      "learning_rate": 8.711656441717792e-06,
+      "loss": 0.7129,
+      "step": 142
+    },
+    {
+      "epoch": 0.2635135135135135,
+      "grad_norm": 0.9142591711511253,
+      "learning_rate": 8.773006134969327e-06,
+      "loss": 0.7167,
+      "step": 143
+    },
+    {
+      "epoch": 0.26535626535626533,
+      "grad_norm": 0.8843276384942427,
+      "learning_rate": 8.83435582822086e-06,
+      "loss": 0.6905,
+      "step": 144
+    },
+    {
+      "epoch": 0.2671990171990172,
+      "grad_norm": 0.8008880420989046,
+      "learning_rate": 8.895705521472394e-06,
+      "loss": 0.6956,
+      "step": 145
+    },
+    {
+      "epoch": 0.269041769041769,
+      "grad_norm": 0.8374970617652128,
+      "learning_rate": 8.957055214723927e-06,
+      "loss": 0.7186,
+      "step": 146
+    },
+    {
+      "epoch": 0.27088452088452086,
+      "grad_norm": 0.8756519567269015,
+      "learning_rate": 9.018404907975461e-06,
+      "loss": 0.6908,
+      "step": 147
+    },
+    {
+      "epoch": 0.2727272727272727,
+      "grad_norm": 0.8188371555501296,
+      "learning_rate": 9.079754601226994e-06,
+      "loss": 0.6842,
+      "step": 148
+    },
+    {
+      "epoch": 0.27457002457002455,
+      "grad_norm": 0.8409342268339723,
+      "learning_rate": 9.14110429447853e-06,
+      "loss": 0.6945,
+      "step": 149
+    },
+    {
+      "epoch": 0.2764127764127764,
+      "grad_norm": 0.8161906327499587,
+      "learning_rate": 9.202453987730062e-06,
+      "loss": 0.6999,
+      "step": 150
+    },
+    {
+      "epoch": 0.27825552825552824,
+      "grad_norm": 1.0366966479465183,
+      "learning_rate": 9.263803680981595e-06,
+      "loss": 0.6939,
+      "step": 151
+    },
+    {
+      "epoch": 0.2800982800982801,
+      "grad_norm": 0.8629477777087785,
+      "learning_rate": 9.325153374233129e-06,
+      "loss": 0.7038,
+      "step": 152
+    },
+    {
+      "epoch": 0.2819410319410319,
+      "grad_norm": 0.7752539673381745,
+      "learning_rate": 9.386503067484664e-06,
+      "loss": 0.7,
+      "step": 153
+    },
+    {
+      "epoch": 0.28378378378378377,
+      "grad_norm": 0.875425080541561,
+      "learning_rate": 9.447852760736197e-06,
+      "loss": 0.72,
+      "step": 154
+    },
+    {
+      "epoch": 0.2856265356265356,
+      "grad_norm": 0.8471817925250265,
+      "learning_rate": 9.509202453987731e-06,
+      "loss": 0.7116,
+      "step": 155
+    },
+    {
+      "epoch": 0.28746928746928746,
+      "grad_norm": 0.8860227273018948,
+      "learning_rate": 9.570552147239264e-06,
+      "loss": 0.6947,
+      "step": 156
+    },
+    {
+      "epoch": 0.2893120393120393,
+      "grad_norm": 0.7699390372791538,
+      "learning_rate": 9.631901840490798e-06,
+      "loss": 0.6924,
+      "step": 157
+    },
+    {
+      "epoch": 0.29115479115479115,
+      "grad_norm": 3.849172436532743,
+      "learning_rate": 9.693251533742331e-06,
+      "loss": 0.6925,
+      "step": 158
+    },
+    {
+      "epoch": 0.292997542997543,
+      "grad_norm": 1.0186542146847384,
+      "learning_rate": 9.754601226993867e-06,
+      "loss": 0.6907,
+      "step": 159
+    },
+    {
+      "epoch": 0.29484029484029484,
+      "grad_norm": 0.7873011703540498,
+      "learning_rate": 9.8159509202454e-06,
+      "loss": 0.6916,
+      "step": 160
+    },
+    {
+      "epoch": 0.2966830466830467,
+      "grad_norm": 0.9233826649734354,
+      "learning_rate": 9.877300613496934e-06,
+      "loss": 0.6955,
+      "step": 161
+    },
+    {
+      "epoch": 0.2985257985257985,
+      "grad_norm": 0.8822552121165639,
+      "learning_rate": 9.938650306748467e-06,
+      "loss": 0.7115,
+      "step": 162
+    },
+    {
+      "epoch": 0.30036855036855037,
+      "grad_norm": 0.9040260973373175,
+      "learning_rate": 1e-05,
+      "loss": 0.6888,
+      "step": 163
+    },
+    {
+      "epoch": 0.3022113022113022,
+      "grad_norm": 0.793792788840252,
+      "learning_rate": 9.999988472080506e-06,
+      "loss": 0.6957,
+      "step": 164
+    },
+    {
+      "epoch": 0.30405405405405406,
+      "grad_norm": 0.8262935005040135,
+      "learning_rate": 9.999953888375178e-06,
+      "loss": 0.6911,
+      "step": 165
+    },
+    {
+      "epoch": 0.3058968058968059,
+      "grad_norm": 0.8203135391383373,
+      "learning_rate": 9.999896249043488e-06,
+      "loss": 0.6974,
+      "step": 166
+    },
+    {
+      "epoch": 0.30773955773955775,
+      "grad_norm": 0.8502083403422361,
+      "learning_rate": 9.999815554351224e-06,
+      "loss": 0.6922,
+      "step": 167
+    },
+    {
+      "epoch": 0.3095823095823096,
+      "grad_norm": 0.8534936750009088,
+      "learning_rate": 9.999711804670478e-06,
+      "loss": 0.6864,
+      "step": 168
+    },
+    {
+      "epoch": 0.31142506142506143,
+      "grad_norm": 0.8243825154971994,
+      "learning_rate": 9.999585000479658e-06,
+      "loss": 0.7022,
+      "step": 169
+    },
+    {
+      "epoch": 0.3132678132678133,
+      "grad_norm": 1.0429334662677958,
+      "learning_rate": 9.999435142363484e-06,
+      "loss": 0.7017,
+      "step": 170
+    },
+    {
+      "epoch": 0.3151105651105651,
+      "grad_norm": 0.8440449415802802,
+      "learning_rate": 9.999262231012971e-06,
+      "loss": 0.703,
+      "step": 171
+    },
+    {
+      "epoch": 0.31695331695331697,
+      "grad_norm": 1.1226970455929344,
+      "learning_rate": 9.999066267225447e-06,
+      "loss": 0.6864,
+      "step": 172
+    },
+    {
+      "epoch": 0.3187960687960688,
+      "grad_norm": 0.9101767374053522,
+      "learning_rate": 9.998847251904529e-06,
+      "loss": 0.6753,
+      "step": 173
+    },
+    {
+      "epoch": 0.32063882063882065,
+      "grad_norm": 0.879648705311873,
+      "learning_rate": 9.998605186060138e-06,
+      "loss": 0.6927,
+      "step": 174
+    },
+    {
+      "epoch": 0.3224815724815725,
+      "grad_norm": 0.9046635948429502,
+      "learning_rate": 9.998340070808478e-06,
+      "loss": 0.6756,
+      "step": 175
+    },
+    {
+      "epoch": 0.32432432432432434,
+      "grad_norm": 0.8369134698042224,
+      "learning_rate": 9.99805190737204e-06,
+      "loss": 0.6882,
+      "step": 176
+    },
+    {
+      "epoch": 0.3261670761670762,
+      "grad_norm": 0.8606891861462971,
+      "learning_rate": 9.997740697079595e-06,
+      "loss": 0.7141,
+      "step": 177
+    },
+    {
+      "epoch": 0.32800982800982803,
+      "grad_norm": 0.8770213273225577,
+      "learning_rate": 9.997406441366182e-06,
+      "loss": 0.6893,
+      "step": 178
+    },
+    {
+      "epoch": 0.3298525798525799,
+      "grad_norm": 0.9739780203593683,
+      "learning_rate": 9.997049141773117e-06,
+      "loss": 0.7074,
+      "step": 179
+    },
+    {
+      "epoch": 0.3316953316953317,
+      "grad_norm": 0.870251343397272,
+      "learning_rate": 9.996668799947962e-06,
+      "loss": 0.6984,
+      "step": 180
+    },
+    {
+      "epoch": 0.33353808353808356,
+      "grad_norm": 1.0959486047020575,
+      "learning_rate": 9.99626541764454e-06,
+      "loss": 0.6829,
+      "step": 181
+    },
+    {
+      "epoch": 0.3353808353808354,
+      "grad_norm": 1.017879732390155,
+      "learning_rate": 9.995838996722916e-06,
+      "loss": 0.6871,
+      "step": 182
+    },
+    {
+      "epoch": 0.33722358722358725,
+      "grad_norm": 1.080307689390482,
+      "learning_rate": 9.995389539149386e-06,
+      "loss": 0.7008,
+      "step": 183
+    },
+    {
+      "epoch": 0.33906633906633904,
+      "grad_norm": 0.909541548051601,
+      "learning_rate": 9.994917046996472e-06,
+      "loss": 0.6949,
+      "step": 184
+    },
+    {
+      "epoch": 0.3409090909090909,
+      "grad_norm": 0.9055598105549953,
+      "learning_rate": 9.99442152244292e-06,
+      "loss": 0.6921,
+      "step": 185
+    },
+    {
+      "epoch": 0.34275184275184273,
+      "grad_norm": 0.9429480376258302,
+      "learning_rate": 9.993902967773674e-06,
+      "loss": 0.6991,
+      "step": 186
+    },
+    {
+      "epoch": 0.34459459459459457,
+      "grad_norm": 0.9275521833452651,
+      "learning_rate": 9.993361385379876e-06,
+      "loss": 0.6891,
+      "step": 187
+    },
+    {
+      "epoch": 0.3464373464373464,
+      "grad_norm": 0.8179435683922109,
+      "learning_rate": 9.992796777758855e-06,
+      "loss": 0.6727,
+      "step": 188
+    },
+    {
+      "epoch": 0.34828009828009826,
+      "grad_norm": 0.930802587368205,
+      "learning_rate": 9.992209147514108e-06,
+      "loss": 0.6786,
+      "step": 189
+    },
+    {
+      "epoch": 0.3501228501228501,
+      "grad_norm": 0.9097350073276268,
+      "learning_rate": 9.991598497355304e-06,
+      "loss": 0.6946,
+      "step": 190
+    },
+    {
+      "epoch": 0.35196560196560195,
+      "grad_norm": 1.0436179097442717,
+      "learning_rate": 9.990964830098246e-06,
+      "loss": 0.7073,
+      "step": 191
+    },
+    {
+      "epoch": 0.3538083538083538,
+      "grad_norm": 0.8375488498707497,
+      "learning_rate": 9.990308148664882e-06,
+      "loss": 0.6816,
+      "step": 192
+    },
+    {
+      "epoch": 0.35565110565110564,
+      "grad_norm": 0.8640261531619103,
+      "learning_rate": 9.989628456083283e-06,
+      "loss": 0.6727,
+      "step": 193
+    },
+    {
+      "epoch": 0.3574938574938575,
+      "grad_norm": 0.8863760340917516,
+      "learning_rate": 9.988925755487622e-06,
+      "loss": 0.6955,
+      "step": 194
+    },
+    {
+      "epoch": 0.3593366093366093,
+      "grad_norm": 0.8477605938532607,
+      "learning_rate": 9.98820005011817e-06,
+      "loss": 0.6852,
+      "step": 195
+    },
+    {
+      "epoch": 0.36117936117936117,
+      "grad_norm": 0.8737677060264961,
+      "learning_rate": 9.98745134332128e-06,
+      "loss": 0.6845,
+      "step": 196
+    },
+    {
+      "epoch": 0.363022113022113,
+      "grad_norm": 1.0020323627576249,
+      "learning_rate": 9.98667963854936e-06,
+      "loss": 0.6896,
+      "step": 197
+    },
+    {
+      "epoch": 0.36486486486486486,
+      "grad_norm": 0.9165561226081099,
+      "learning_rate": 9.985884939360873e-06,
+      "loss": 0.6928,
+      "step": 198
+    },
+    {
+      "epoch": 0.3667076167076167,
+      "grad_norm": 1.0386725004317794,
+      "learning_rate": 9.985067249420308e-06,
+      "loss": 0.6914,
+      "step": 199
+    },
+    {
+      "epoch": 0.36855036855036855,
+      "grad_norm": 1.1004451999100757,
+      "learning_rate": 9.984226572498173e-06,
+      "loss": 0.6921,
+      "step": 200
+    },
+    {
+      "epoch": 0.3703931203931204,
+      "grad_norm": 0.8697793253794458,
+      "learning_rate": 9.983362912470967e-06,
+      "loss": 0.6837,
+      "step": 201
+    },
+    {
+      "epoch": 0.37223587223587223,
+      "grad_norm": 0.9429896462898235,
+      "learning_rate": 9.982476273321175e-06,
+      "loss": 0.6894,
+      "step": 202
+    },
+    {
+      "epoch": 0.3740786240786241,
+      "grad_norm": 1.0091718669651177,
+      "learning_rate": 9.981566659137238e-06,
+      "loss": 0.6858,
+      "step": 203
+    },
+    {
+      "epoch": 0.3759213759213759,
+      "grad_norm": 0.8273464015973249,
+      "learning_rate": 9.980634074113538e-06,
+      "loss": 0.6874,
+      "step": 204
+    },
+    {
+      "epoch": 0.37776412776412777,
+      "grad_norm": 0.9431112035462864,
+      "learning_rate": 9.979678522550382e-06,
+      "loss": 0.6803,
+      "step": 205
+    },
+    {
+      "epoch": 0.3796068796068796,
+      "grad_norm": 0.8905211285917688,
+      "learning_rate": 9.97870000885398e-06,
+      "loss": 0.6846,
+      "step": 206
+    },
+    {
+      "epoch": 0.38144963144963145,
+      "grad_norm": 0.8060629404214893,
+      "learning_rate": 9.97769853753642e-06,
+      "loss": 0.6964,
+      "step": 207
+    },
+    {
+      "epoch": 0.3832923832923833,
+      "grad_norm": 0.8931670383206097,
+      "learning_rate": 9.976674113215655e-06,
+      "loss": 0.7044,
+      "step": 208
+    },
+    {
+      "epoch": 0.38513513513513514,
+      "grad_norm": 0.8289987846758576,
+      "learning_rate": 9.975626740615478e-06,
+      "loss": 0.6794,
+      "step": 209
+    },
+    {
+      "epoch": 0.386977886977887,
+      "grad_norm": 0.800163925134946,
+      "learning_rate": 9.974556424565503e-06,
+      "loss": 0.6823,
+      "step": 210
+    },
+    {
+      "epoch": 0.38882063882063883,
+      "grad_norm": 0.9236436427198064,
+      "learning_rate": 9.973463170001131e-06,
+      "loss": 0.6675,
+      "step": 211
+    },
+    {
+      "epoch": 0.3906633906633907,
+      "grad_norm": 0.9597616294654956,
+      "learning_rate": 9.972346981963546e-06,
+      "loss": 0.6976,
+      "step": 212
+    },
+    {
+      "epoch": 0.3925061425061425,
+      "grad_norm": 0.9132503443161095,
+      "learning_rate": 9.971207865599679e-06,
+      "loss": 0.7022,
+      "step": 213
+    },
+    {
+      "epoch": 0.39434889434889436,
+      "grad_norm": 1.0941626527227897,
+      "learning_rate": 9.970045826162182e-06,
+      "loss": 0.6796,
+      "step": 214
+    },
+    {
+      "epoch": 0.3961916461916462,
+      "grad_norm": 0.8323222207324584,
+      "learning_rate": 9.96886086900942e-06,
+      "loss": 0.6953,
+      "step": 215
+    },
+    {
+      "epoch": 0.39803439803439805,
+      "grad_norm": 1.0343206669945002,
+      "learning_rate": 9.967652999605424e-06,
+      "loss": 0.684,
+      "step": 216
+    },
+    {
+      "epoch": 0.3998771498771499,
+      "grad_norm": 0.8676624899905876,
+      "learning_rate": 9.966422223519887e-06,
+      "loss": 0.668,
+      "step": 217
+    },
+    {
+      "epoch": 0.40171990171990174,
+      "grad_norm": 0.8955953744674456,
+      "learning_rate": 9.965168546428122e-06,
+      "loss": 0.7064,
+      "step": 218
+    },
+    {
+      "epoch": 0.4035626535626536,
+      "grad_norm": 1.0349054019089599,
+      "learning_rate": 9.963891974111042e-06,
+      "loss": 0.6667,
+      "step": 219
+    },
+    {
+      "epoch": 0.40540540540540543,
+      "grad_norm": 0.7600100070574172,
+      "learning_rate": 9.96259251245514e-06,
+      "loss": 0.6982,
+      "step": 220
+    },
+    {
+      "epoch": 0.40724815724815727,
+      "grad_norm": 0.8502309113207338,
+      "learning_rate": 9.961270167452449e-06,
+      "loss": 0.6851,
+      "step": 221
+    },
+    {
+      "epoch": 0.4090909090909091,
+      "grad_norm": 0.8703951843840896,
+      "learning_rate": 9.959924945200525e-06,
+      "loss": 0.6746,
+      "step": 222
+    },
+    {
+      "epoch": 0.41093366093366096,
+      "grad_norm": 0.8374723364851869,
+      "learning_rate": 9.958556851902414e-06,
+      "loss": 0.7011,
+      "step": 223
+    },
+    {
+      "epoch": 0.41277641277641275,
+      "grad_norm": 0.8899266507186723,
+      "learning_rate": 9.957165893866623e-06,
+      "loss": 0.6874,
+      "step": 224
+    },
+    {
+      "epoch": 0.4146191646191646,
+      "grad_norm": 0.8582092730981327,
+      "learning_rate": 9.955752077507093e-06,
+      "loss": 0.6801,
+      "step": 225
+    },
+    {
+      "epoch": 0.41646191646191644,
+      "grad_norm": 0.9180971339113595,
+      "learning_rate": 9.95431540934317e-06,
+      "loss": 0.6744,
+      "step": 226
+    },
+    {
+      "epoch": 0.4183046683046683,
+      "grad_norm": 0.929807962059908,
+      "learning_rate": 9.952855895999567e-06,
+      "loss": 0.6741,
+      "step": 227
+    },
+    {
+      "epoch": 0.4201474201474201,
+      "grad_norm": 0.7102226264148732,
+      "learning_rate": 9.951373544206352e-06,
+      "loss": 0.6723,
+      "step": 228
+    },
+    {
+      "epoch": 0.42199017199017197,
+      "grad_norm": 0.8676830395725106,
+      "learning_rate": 9.949868360798893e-06,
+      "loss": 0.6984,
+      "step": 229
+    },
+    {
+      "epoch": 0.4238329238329238,
+      "grad_norm": 0.8210850318301138,
+      "learning_rate": 9.948340352717845e-06,
+      "loss": 0.6692,
+      "step": 230
+    },
+    {
+      "epoch": 0.42567567567567566,
+      "grad_norm": 0.7699325526782979,
+      "learning_rate": 9.946789527009108e-06,
+      "loss": 0.6676,
+      "step": 231
+    },
+    {
+      "epoch": 0.4275184275184275,
+      "grad_norm": 0.7962651398970799,
+      "learning_rate": 9.9452158908238e-06,
+      "loss": 0.6796,
+      "step": 232
+    },
+    {
+      "epoch": 0.42936117936117935,
+      "grad_norm": 0.8849100613189843,
+      "learning_rate": 9.943619451418225e-06,
+      "loss": 0.6715,
+      "step": 233
+    },
+    {
+      "epoch": 0.4312039312039312,
+      "grad_norm": 0.7521331373464345,
+      "learning_rate": 9.942000216153829e-06,
+      "loss": 0.6805,
+      "step": 234
+    },
+    {
+      "epoch": 0.43304668304668303,
+      "grad_norm": 0.7984843503768578,
+      "learning_rate": 9.940358192497178e-06,
+      "loss": 0.6684,
+      "step": 235
+    },
+    {
+      "epoch": 0.4348894348894349,
+      "grad_norm": 0.7735650229714421,
+      "learning_rate": 9.93869338801992e-06,
+      "loss": 0.6993,
+      "step": 236
+    },
+    {
+      "epoch": 0.4367321867321867,
+      "grad_norm": 0.8818501863491192,
+      "learning_rate": 9.937005810398747e-06,
+      "loss": 0.6653,
+      "step": 237
+    },
+    {
+      "epoch": 0.43857493857493857,
+      "grad_norm": 0.7517036208199124,
+      "learning_rate": 9.935295467415363e-06,
+      "loss": 0.6621,
+      "step": 238
+    },
+    {
+      "epoch": 0.4404176904176904,
+      "grad_norm": 0.7835671359856426,
+      "learning_rate": 9.933562366956445e-06,
+      "loss": 0.6957,
+      "step": 239
+    },
+    {
+      "epoch": 0.44226044226044225,
+      "grad_norm": 0.7617909165993084,
+      "learning_rate": 9.931806517013612e-06,
+      "loss": 0.6475,
+      "step": 240
+    },
+    {
+      "epoch": 0.4441031941031941,
+      "grad_norm": 0.8811091735405586,
+      "learning_rate": 9.930027925683384e-06,
+      "loss": 0.6871,
+      "step": 241
+    },
+    {
+      "epoch": 0.44594594594594594,
+      "grad_norm": 0.7411597097074916,
+      "learning_rate": 9.928226601167139e-06,
+      "loss": 0.6573,
+      "step": 242
+    },
+    {
+      "epoch": 0.4477886977886978,
+      "grad_norm": 0.8435716739939152,
+      "learning_rate": 9.92640255177109e-06,
+      "loss": 0.6979,
+      "step": 243
+    },
+    {
+      "epoch": 0.44963144963144963,
+      "grad_norm": 0.8370871584022952,
+      "learning_rate": 9.924555785906235e-06,
+      "loss": 0.689,
+      "step": 244
+    },
+    {
+      "epoch": 0.4514742014742015,
+      "grad_norm": 0.8053068638858747,
+      "learning_rate": 9.922686312088323e-06,
+      "loss": 0.6909,
+      "step": 245
+    },
+    {
+      "epoch": 0.4533169533169533,
+      "grad_norm": 0.845959808062115,
+      "learning_rate": 9.920794138937807e-06,
+      "loss": 0.6838,
+      "step": 246
+    },
+    {
+      "epoch": 0.45515970515970516,
+      "grad_norm": 0.835233034245389,
+      "learning_rate": 9.918879275179819e-06,
+      "loss": 0.6533,
+      "step": 247
+    },
+    {
+      "epoch": 0.457002457002457,
+      "grad_norm": 0.8243213839911792,
+      "learning_rate": 9.916941729644112e-06,
+      "loss": 0.6493,
+      "step": 248
+    },
+    {
+      "epoch": 0.45884520884520885,
+      "grad_norm": 0.9094168419301908,
+      "learning_rate": 9.914981511265039e-06,
+      "loss": 0.6784,
+      "step": 249
+    },
+    {
+      "epoch": 0.4606879606879607,
+      "grad_norm": 0.7833684013624378,
+      "learning_rate": 9.912998629081495e-06,
+      "loss": 0.6936,
+      "step": 250
+    },
+    {
+      "epoch": 0.46253071253071254,
+      "grad_norm": 0.9924653990659058,
+      "learning_rate": 9.910993092236878e-06,
+      "loss": 0.6646,
+      "step": 251
+    },
+    {
+      "epoch": 0.4643734643734644,
+      "grad_norm": 0.8712628359075767,
+      "learning_rate": 9.90896490997906e-06,
+      "loss": 0.6886,
+      "step": 252
+    },
+    {
+      "epoch": 0.46621621621621623,
+      "grad_norm": 0.8114558039313291,
+      "learning_rate": 9.906914091660327e-06,
+      "loss": 0.6668,
+      "step": 253
+    },
+    {
+      "epoch": 0.4680589680589681,
+      "grad_norm": 0.8842547128955308,
+      "learning_rate": 9.904840646737346e-06,
+      "loss": 0.6711,
+      "step": 254
+    },
+    {
+      "epoch": 0.4699017199017199,
+      "grad_norm": 0.7594876657071846,
+      "learning_rate": 9.902744584771123e-06,
+      "loss": 0.6782,
+      "step": 255
+    },
+    {
+      "epoch": 0.47174447174447176,
+      "grad_norm": 0.8980126784830716,
+      "learning_rate": 9.900625915426948e-06,
+      "loss": 0.6653,
+      "step": 256
+    },
+    {
+      "epoch": 0.4735872235872236,
+      "grad_norm": 0.8267075761550458,
+      "learning_rate": 9.898484648474362e-06,
+      "loss": 0.7199,
+      "step": 257
+    },
+    {
+      "epoch": 0.47542997542997545,
+      "grad_norm": 0.820870222680857,
+      "learning_rate": 9.896320793787106e-06,
+      "loss": 0.6567,
+      "step": 258
+    },
+    {
+      "epoch": 0.4772727272727273,
+      "grad_norm": 0.9623184616318754,
+      "learning_rate": 9.894134361343077e-06,
+      "loss": 0.6842,
+      "step": 259
+    },
+    {
+      "epoch": 0.47911547911547914,
+      "grad_norm": 1.0138944941279977,
+      "learning_rate": 9.891925361224284e-06,
+      "loss": 0.6869,
+      "step": 260
+    },
+    {
+      "epoch": 0.480958230958231,
+      "grad_norm": 0.7957942146949784,
+      "learning_rate": 9.889693803616793e-06,
+      "loss": 0.6803,
+      "step": 261
+    },
+    {
+      "epoch": 0.4828009828009828,
+      "grad_norm": 0.9646582451232161,
+      "learning_rate": 9.887439698810694e-06,
+      "loss": 0.6581,
+      "step": 262
+    },
+    {
+      "epoch": 0.48464373464373467,
+      "grad_norm": 1.1916684804544473,
+      "learning_rate": 9.88516305720004e-06,
+      "loss": 0.6612,
+      "step": 263
+    },
+    {
+      "epoch": 0.4864864864864865,
+      "grad_norm": 0.8180153291023301,
+      "learning_rate": 9.88286388928281e-06,
+      "loss": 0.6686,
+      "step": 264
+    },
+    {
+      "epoch": 0.4883292383292383,
+      "grad_norm": 1.0805512929701664,
+      "learning_rate": 9.880542205660853e-06,
+      "loss": 0.6798,
+      "step": 265
+    },
+    {
+      "epoch": 0.49017199017199015,
+      "grad_norm": 0.8077071991318745,
+      "learning_rate": 9.878198017039839e-06,
+      "loss": 0.6581,
+      "step": 266
+    },
+    {
+      "epoch": 0.492014742014742,
+      "grad_norm": 0.963722740159231,
+      "learning_rate": 9.875831334229217e-06,
+      "loss": 0.6726,
+      "step": 267
+    },
+    {
+      "epoch": 0.49385749385749383,
+      "grad_norm": 0.9779030133764357,
+      "learning_rate": 9.873442168142158e-06,
+      "loss": 0.6875,
+      "step": 268
+    },
+    {
+      "epoch": 0.4957002457002457,
+      "grad_norm": 0.8225455364911165,
+      "learning_rate": 9.87103052979551e-06,
+      "loss": 0.6664,
+      "step": 269
+    },
+    {
+      "epoch": 0.4975429975429975,
+      "grad_norm": 0.9819822842261648,
+      "learning_rate": 9.868596430309739e-06,
+      "loss": 0.6929,
+      "step": 270
+    },
+    {
+      "epoch": 0.49938574938574937,
+      "grad_norm": 0.8336124137642973,
+      "learning_rate": 9.866139880908887e-06,
+      "loss": 0.6701,
+      "step": 271
+    },
+    {
+      "epoch": 0.5012285012285013,
+      "grad_norm": 0.9759554674220386,
+      "learning_rate": 9.863660892920514e-06,
+      "loss": 0.6745,
+      "step": 272
+    },
+    {
+      "epoch": 0.5030712530712531,
+      "grad_norm": 0.8834618135641567,
+      "learning_rate": 9.861159477775653e-06,
+      "loss": 0.6492,
+      "step": 273
+    },
+    {
+      "epoch": 0.504914004914005,
+      "grad_norm": 1.078396409779586,
+      "learning_rate": 9.858635647008747e-06,
+      "loss": 0.6765,
+      "step": 274
+    },
+    {
+      "epoch": 0.5067567567567568,
+      "grad_norm": 1.0907272401525143,
+      "learning_rate": 9.856089412257605e-06,
+      "loss": 0.689,
+      "step": 275
+    },
+    {
+      "epoch": 0.5085995085995086,
+      "grad_norm": 0.8758238916788025,
+      "learning_rate": 9.85352078526334e-06,
+      "loss": 0.6541,
+      "step": 276
+    },
+    {
+      "epoch": 0.5104422604422605,
+      "grad_norm": 1.0557240417677856,
+      "learning_rate": 9.850929777870324e-06,
+      "loss": 0.6772,
+      "step": 277
+    },
+    {
+      "epoch": 0.5122850122850123,
+      "grad_norm": 0.8426465191321125,
+      "learning_rate": 9.848316402026125e-06,
+      "loss": 0.6848,
+      "step": 278
+    },
+    {
+      "epoch": 0.5141277641277642,
+      "grad_norm": 1.0257365851876739,
+      "learning_rate": 9.845680669781459e-06,
+      "loss": 0.6768,
+      "step": 279
+    },
+    {
+      "epoch": 0.515970515970516,
+      "grad_norm": 0.8047506165723123,
+      "learning_rate": 9.843022593290129e-06,
+      "loss": 0.6636,
+      "step": 280
+    },
+    {
+      "epoch": 0.5178132678132679,
+      "grad_norm": 0.8009707162554297,
+      "learning_rate": 9.840342184808973e-06,
+      "loss": 0.6863,
+      "step": 281
+    },
+    {
+      "epoch": 0.5196560196560197,
+      "grad_norm": 0.9787753659666715,
+      "learning_rate": 9.837639456697802e-06,
+      "loss": 0.6647,
+      "step": 282
+    },
+    {
+      "epoch": 0.5214987714987716,
+      "grad_norm": 0.8610772272821721,
+      "learning_rate": 9.83491442141935e-06,
+      "loss": 0.6781,
+      "step": 283
+    },
+    {
+      "epoch": 0.5233415233415234,
+      "grad_norm": 0.8289276292926917,
+      "learning_rate": 9.832167091539215e-06,
+      "loss": 0.6851,
+      "step": 284
+    },
+    {
+      "epoch": 0.5251842751842751,
+      "grad_norm": 1.0470178148821083,
+      "learning_rate": 9.829397479725791e-06,
+      "loss": 0.6576,
+      "step": 285
+    },
+    {
+      "epoch": 0.527027027027027,
+      "grad_norm": 0.956025821170101,
+      "learning_rate": 9.826605598750223e-06,
+      "loss": 0.6809,
+      "step": 286
+    },
+    {
+      "epoch": 0.5288697788697788,
+      "grad_norm": 0.8103480199952018,
+      "learning_rate": 9.823791461486348e-06,
+      "loss": 0.6591,
+      "step": 287
+    },
+    {
+      "epoch": 0.5307125307125307,
+      "grad_norm": 0.9889979716627547,
+      "learning_rate": 9.82095508091062e-06,
+      "loss": 0.6651,
+      "step": 288
+    },
+    {
+      "epoch": 0.5325552825552825,
+      "grad_norm": 0.8188375016880358,
+      "learning_rate": 9.818096470102067e-06,
+      "loss": 0.6803,
+      "step": 289
+    },
+    {
+      "epoch": 0.5343980343980343,
+      "grad_norm": 0.8530340115544307,
+      "learning_rate": 9.815215642242224e-06,
+      "loss": 0.6798,
+      "step": 290
+    },
+    {
+      "epoch": 0.5362407862407862,
+      "grad_norm": 0.7736885768773254,
+      "learning_rate": 9.812312610615073e-06,
+      "loss": 0.6731,
+      "step": 291
+    },
+    {
+      "epoch": 0.538083538083538,
+      "grad_norm": 0.980220740942662,
+      "learning_rate": 9.809387388606977e-06,
+      "loss": 0.6744,
+      "step": 292
+    },
+    {
+      "epoch": 0.5399262899262899,
+      "grad_norm": 0.9496227476783082,
+      "learning_rate": 9.806439989706626e-06,
+      "loss": 0.7036,
+      "step": 293
+    },
+    {
+      "epoch": 0.5417690417690417,
+      "grad_norm": 0.8880138947725991,
+      "learning_rate": 9.803470427504972e-06,
+      "loss": 0.6825,
+      "step": 294
+    },
+    {
+      "epoch": 0.5436117936117936,
+      "grad_norm": 0.9684159975862886,
+      "learning_rate": 9.800478715695165e-06,
+      "loss": 0.6508,
+      "step": 295
+    },
+    {
+      "epoch": 0.5454545454545454,
+      "grad_norm": 0.8777937622924454,
+      "learning_rate": 9.797464868072489e-06,
+      "loss": 0.6708,
+      "step": 296
+    },
+    {
+      "epoch": 0.5472972972972973,
+      "grad_norm": 0.8429093761893578,
+      "learning_rate": 9.794428898534301e-06,
+      "loss": 0.6705,
+      "step": 297
+    },
+    {
+      "epoch": 0.5491400491400491,
+      "grad_norm": 0.7464187656371628,
+      "learning_rate": 9.791370821079967e-06,
+      "loss": 0.681,
+      "step": 298
+    },
+    {
+      "epoch": 0.550982800982801,
+      "grad_norm": 0.9107955219588116,
+      "learning_rate": 9.788290649810794e-06,
+      "loss": 0.6603,
+      "step": 299
+    },
+    {
+      "epoch": 0.5528255528255528,
+      "grad_norm": 0.861979565380758,
+      "learning_rate": 9.78518839892997e-06,
+      "loss": 0.6559,
+      "step": 300
+    },
+    {
+      "epoch": 0.5546683046683046,
+      "grad_norm": 0.8962986110250479,
+      "learning_rate": 9.782064082742494e-06,
+      "loss": 0.6684,
+      "step": 301
+    },
+    {
+      "epoch": 0.5565110565110565,
+      "grad_norm": 0.865908738869927,
+      "learning_rate": 9.77891771565511e-06,
+      "loss": 0.6947,
+      "step": 302
+    },
+    {
+      "epoch": 0.5583538083538083,
+      "grad_norm": 0.8644713987435614,
+      "learning_rate": 9.775749312176249e-06,
+      "loss": 0.6825,
+      "step": 303
+    },
+    {
+      "epoch": 0.5601965601965602,
+      "grad_norm": 0.8658400850970434,
+      "learning_rate": 9.772558886915946e-06,
+      "loss": 0.6716,
+      "step": 304
+    },
+    {
+      "epoch": 0.562039312039312,
+      "grad_norm": 0.8548103485450707,
+      "learning_rate": 9.76934645458579e-06,
+      "loss": 0.6794,
+      "step": 305
+    },
+    {
+      "epoch": 0.5638820638820639,
+      "grad_norm": 0.8654621047479792,
+      "learning_rate": 9.766112029998847e-06,
+      "loss": 0.6647,
+      "step": 306
+    },
+    {
+      "epoch": 0.5657248157248157,
+      "grad_norm": 0.9471044571012447,
+      "learning_rate": 9.76285562806959e-06,
+      "loss": 0.6611,
+      "step": 307
+    },
+    {
+      "epoch": 0.5675675675675675,
+      "grad_norm": 0.8222653795648578,
+      "learning_rate": 9.759577263813833e-06,
+      "loss": 0.6575,
+      "step": 308
+    },
+    {
+      "epoch": 0.5694103194103194,
+      "grad_norm": 0.991512448211715,
+      "learning_rate": 9.756276952348666e-06,
+      "loss": 0.6606,
+      "step": 309
+    },
+    {
+      "epoch": 0.5712530712530712,
+      "grad_norm": 0.8726170823578269,
+      "learning_rate": 9.752954708892379e-06,
+      "loss": 0.668,
+      "step": 310
+    },
+    {
+      "epoch": 0.5730958230958231,
+      "grad_norm": 0.8034244776639897,
+      "learning_rate": 9.749610548764393e-06,
+      "loss": 0.6668,
+      "step": 311
+    },
+    {
+      "epoch": 0.5749385749385749,
+      "grad_norm": 0.9297516969222446,
+      "learning_rate": 9.74624448738519e-06,
+      "loss": 0.6934,
+      "step": 312
+    },
+    {
+      "epoch": 0.5767813267813268,
+      "grad_norm": 0.9174156290556367,
+      "learning_rate": 9.742856540276246e-06,
+      "loss": 0.6905,
+      "step": 313
+    },
+    {
+      "epoch": 0.5786240786240786,
+      "grad_norm": 0.8669181590348406,
+      "learning_rate": 9.739446723059953e-06,
+      "loss": 0.6569,
+      "step": 314
+    },
+    {
+      "epoch": 0.5804668304668305,
+      "grad_norm": 0.9926815966925522,
+      "learning_rate": 9.736015051459551e-06,
+      "loss": 0.6626,
+      "step": 315
+    },
+    {
+      "epoch": 0.5823095823095823,
+      "grad_norm": 0.8636763778697563,
+      "learning_rate": 9.732561541299053e-06,
+      "loss": 0.6581,
+      "step": 316
+    },
+    {
+      "epoch": 0.5841523341523341,
+      "grad_norm": 0.9064863131071679,
+      "learning_rate": 9.729086208503174e-06,
+      "loss": 0.6656,
+      "step": 317
+    },
+    {
+      "epoch": 0.585995085995086,
+      "grad_norm": 0.930453295225485,
+      "learning_rate": 9.725589069097257e-06,
+      "loss": 0.6686,
+      "step": 318
+    },
+    {
+      "epoch": 0.5878378378378378,
+      "grad_norm": 0.8002110790872421,
+      "learning_rate": 9.722070139207198e-06,
+      "loss": 0.661,
+      "step": 319
+    },
+    {
+      "epoch": 0.5896805896805897,
+      "grad_norm": 0.8206142495981427,
+      "learning_rate": 9.718529435059372e-06,
+      "loss": 0.6577,
+      "step": 320
+    },
+    {
+      "epoch": 0.5915233415233415,
+      "grad_norm": 0.9149459595267929,
+      "learning_rate": 9.714966972980563e-06,
+      "loss": 0.6797,
+      "step": 321
+    },
+    {
+      "epoch": 0.5933660933660934,
+      "grad_norm": 0.7716356354445693,
+      "learning_rate": 9.711382769397879e-06,
+      "loss": 0.6874,
+      "step": 322
+    },
+    {
+      "epoch": 0.5952088452088452,
+      "grad_norm": 0.7966049754065327,
+      "learning_rate": 9.707776840838685e-06,
+      "loss": 0.6465,
+      "step": 323
+    },
+    {
+      "epoch": 0.597051597051597,
+      "grad_norm": 0.8104242640073799,
+      "learning_rate": 9.704149203930522e-06,
+      "loss": 0.6565,
+      "step": 324
+    },
+    {
+      "epoch": 0.5988943488943489,
+      "grad_norm": 0.7192971487379867,
+      "learning_rate": 9.700499875401035e-06,
+      "loss": 0.6705,
+      "step": 325
+    },
+    {
+      "epoch": 0.6007371007371007,
+      "grad_norm": 0.8128722231030346,
+      "learning_rate": 9.696828872077885e-06,
+      "loss": 0.6758,
+      "step": 326
+    },
+    {
+      "epoch": 0.6025798525798526,
+      "grad_norm": 0.774395906070038,
+      "learning_rate": 9.69313621088869e-06,
+      "loss": 0.6772,
+      "step": 327
+    },
+    {
+      "epoch": 0.6044226044226044,
+      "grad_norm": 0.8384106910142824,
+      "learning_rate": 9.689421908860928e-06,
+      "loss": 0.6815,
+      "step": 328
+    },
+    {
+      "epoch": 0.6062653562653563,
+      "grad_norm": 0.7370917293163881,
+      "learning_rate": 9.685685983121867e-06,
+      "loss": 0.6698,
+      "step": 329
+    },
+    {
+      "epoch": 0.6081081081081081,
+      "grad_norm": 0.9415738306119148,
+      "learning_rate": 9.681928450898492e-06,
+      "loss": 0.6544,
+      "step": 330
+    },
+    {
+      "epoch": 0.60995085995086,
+      "grad_norm": 0.8842520203269579,
+      "learning_rate": 9.67814932951741e-06,
+      "loss": 0.6865,
+      "step": 331
+    },
+    {
+      "epoch": 0.6117936117936118,
+      "grad_norm": 0.8375158261232246,
+      "learning_rate": 9.674348636404784e-06,
+      "loss": 0.6541,
+      "step": 332
+    },
+    {
+      "epoch": 0.6136363636363636,
+      "grad_norm": 0.9231259757408194,
+      "learning_rate": 9.670526389086252e-06,
+      "loss": 0.6804,
+      "step": 333
+    },
+    {
+      "epoch": 0.6154791154791155,
+      "grad_norm": 0.9013434907960934,
+      "learning_rate": 9.666682605186834e-06,
+      "loss": 0.6755,
+      "step": 334
+    },
+    {
+      "epoch": 0.6173218673218673,
+      "grad_norm": 0.7591389086417178,
+      "learning_rate": 9.662817302430864e-06,
+      "loss": 0.6595,
+      "step": 335
+    },
+    {
+      "epoch": 0.6191646191646192,
+      "grad_norm": 0.9819771127294318,
+      "learning_rate": 9.658930498641901e-06,
+      "loss": 0.6845,
+      "step": 336
+    },
+    {
+      "epoch": 0.621007371007371,
+      "grad_norm": 0.8518106805206948,
+      "learning_rate": 9.655022211742651e-06,
+      "loss": 0.6556,
+      "step": 337
+    },
+    {
+      "epoch": 0.6228501228501229,
+      "grad_norm": 0.8762545547629078,
+      "learning_rate": 9.651092459754879e-06,
+      "loss": 0.6738,
+      "step": 338
+    },
+    {
+      "epoch": 0.6246928746928747,
+      "grad_norm": 0.8974369439844926,
+      "learning_rate": 9.64714126079933e-06,
+      "loss": 0.6496,
+      "step": 339
+    },
+    {
+      "epoch": 0.6265356265356266,
+      "grad_norm": 0.8492213714812119,
+      "learning_rate": 9.643168633095647e-06,
+      "loss": 0.662,
+      "step": 340
+    },
+    {
+      "epoch": 0.6283783783783784,
+      "grad_norm": 1.2430502901021347,
+      "learning_rate": 9.639174594962283e-06,
+      "loss": 0.6603,
+      "step": 341
+    },
+    {
+      "epoch": 0.6302211302211302,
+      "grad_norm": 0.8748209205075634,
+      "learning_rate": 9.635159164816416e-06,
+      "loss": 0.6632,
+      "step": 342
+    },
+    {
+      "epoch": 0.6320638820638821,
+      "grad_norm": 0.7988282830757295,
+      "learning_rate": 9.631122361173873e-06,
+      "loss": 0.6425,
+      "step": 343
+    },
+    {
+      "epoch": 0.6339066339066339,
+      "grad_norm": 0.8881687986099336,
+      "learning_rate": 9.627064202649027e-06,
+      "loss": 0.653,
+      "step": 344
+    },
+    {
+      "epoch": 0.6357493857493858,
+      "grad_norm": 0.7438494192019424,
+      "learning_rate": 9.622984707954732e-06,
+      "loss": 0.6635,
+      "step": 345
+    },
+    {
+      "epoch": 0.6375921375921376,
+      "grad_norm": 0.8212918326766123,
+      "learning_rate": 9.61888389590222e-06,
+      "loss": 0.6667,
+      "step": 346
+    },
+    {
+      "epoch": 0.6394348894348895,
+      "grad_norm": 0.9107714177102624,
+      "learning_rate": 9.614761785401026e-06,
+      "loss": 0.651,
+      "step": 347
+    },
+    {
+      "epoch": 0.6412776412776413,
+      "grad_norm": 0.7967349906305417,
+      "learning_rate": 9.610618395458892e-06,
+      "loss": 0.6518,
+      "step": 348
+    },
+    {
+      "epoch": 0.6431203931203932,
+      "grad_norm": 0.8287196930571482,
+      "learning_rate": 9.606453745181682e-06,
+      "loss": 0.6891,
+      "step": 349
+    },
+    {
+      "epoch": 0.644963144963145,
+      "grad_norm": 0.8009586618997105,
+      "learning_rate": 9.602267853773301e-06,
+      "loss": 0.6556,
+      "step": 350
+    },
+    {
+      "epoch": 0.6468058968058968,
+      "grad_norm": 0.8392689349596499,
+      "learning_rate": 9.598060740535596e-06,
+      "loss": 0.6619,
+      "step": 351
+    },
+    {
+      "epoch": 0.6486486486486487,
+      "grad_norm": 0.7422576341954112,
+      "learning_rate": 9.593832424868271e-06,
+      "loss": 0.6732,
+      "step": 352
+    },
+    {
+      "epoch": 0.6504914004914005,
+      "grad_norm": 0.8325282008479906,
+      "learning_rate": 9.589582926268798e-06,
+      "loss": 0.6701,
+      "step": 353
+    },
+    {
+      "epoch": 0.6523341523341524,
+      "grad_norm": 0.8895029072593259,
+      "learning_rate": 9.585312264332329e-06,
+      "loss": 0.6742,
+      "step": 354
+    },
+    {
+      "epoch": 0.6541769041769042,
+      "grad_norm": 0.7719426587608895,
+      "learning_rate": 9.581020458751604e-06,
+      "loss": 0.6483,
+      "step": 355
+    },
+    {
+      "epoch": 0.6560196560196561,
+      "grad_norm": 0.8182381229107879,
+      "learning_rate": 9.576707529316857e-06,
+      "loss": 0.6983,
+      "step": 356
+    },
+    {
+      "epoch": 0.6578624078624079,
+      "grad_norm": 0.7334995547553077,
+      "learning_rate": 9.572373495915732e-06,
+      "loss": 0.661,
+      "step": 357
+    },
+    {
+      "epoch": 0.6597051597051597,
+      "grad_norm": 0.7280198804154006,
+      "learning_rate": 9.568018378533181e-06,
+      "loss": 0.6619,
+      "step": 358
+    },
+    {
+      "epoch": 0.6615479115479116,
+      "grad_norm": 1.0666065317779767,
+      "learning_rate": 9.563642197251382e-06,
+      "loss": 0.6493,
+      "step": 359
+    },
+    {
+      "epoch": 0.6633906633906634,
+      "grad_norm": 0.8302814282517763,
+      "learning_rate": 9.559244972249643e-06,
+      "loss": 0.6759,
+      "step": 360
+    },
+    {
+      "epoch": 0.6652334152334153,
+      "grad_norm": 0.8508722596741045,
+      "learning_rate": 9.554826723804304e-06,
+      "loss": 0.6457,
+      "step": 361
+    },
+    {
+      "epoch": 0.6670761670761671,
+      "grad_norm": 1.0010795332269642,
+      "learning_rate": 9.550387472288651e-06,
+      "loss": 0.687,
+      "step": 362
+    },
+    {
+      "epoch": 0.668918918918919,
+      "grad_norm": 0.7456205455604845,
+      "learning_rate": 9.545927238172817e-06,
+      "loss": 0.6597,
+      "step": 363
+    },
+    {
+      "epoch": 0.6707616707616708,
+      "grad_norm": 1.0779738607323948,
+      "learning_rate": 9.541446042023692e-06,
+      "loss": 0.6675,
+      "step": 364
+    },
+    {
+      "epoch": 0.6726044226044227,
+      "grad_norm": 0.8044418663766405,
+      "learning_rate": 9.53694390450482e-06,
+      "loss": 0.6621,
+      "step": 365
+    },
+    {
+      "epoch": 0.6744471744471745,
+      "grad_norm": 0.7517455763908821,
+      "learning_rate": 9.532420846376316e-06,
+      "loss": 0.6873,
+      "step": 366
+    },
+    {
+      "epoch": 0.6762899262899262,
+      "grad_norm": 0.8192277230548196,
+      "learning_rate": 9.527876888494758e-06,
+      "loss": 0.6495,
+      "step": 367
+    },
+    {
+      "epoch": 0.6781326781326781,
+      "grad_norm": 0.7477462475596808,
+      "learning_rate": 9.523312051813097e-06,
+      "loss": 0.6736,
+      "step": 368
+    },
+    {
+      "epoch": 0.6799754299754299,
+      "grad_norm": 0.8249486265661817,
+      "learning_rate": 9.518726357380565e-06,
+      "loss": 0.6617,
+      "step": 369
+    },
+    {
+      "epoch": 0.6818181818181818,
+      "grad_norm": 0.7453668467448247,
+      "learning_rate": 9.514119826342564e-06,
+      "loss": 0.6693,
+      "step": 370
+    },
+    {
+      "epoch": 0.6836609336609336,
+      "grad_norm": 0.7495032228335399,
+      "learning_rate": 9.509492479940585e-06,
+      "loss": 0.6635,
+      "step": 371
+    },
+    {
+      "epoch": 0.6855036855036855,
+      "grad_norm": 0.7749240770655454,
+      "learning_rate": 9.504844339512096e-06,
+      "loss": 0.668,
+      "step": 372
+    },
+    {
+      "epoch": 0.6873464373464373,
+      "grad_norm": 0.7821956468584554,
+      "learning_rate": 9.500175426490455e-06,
+      "loss": 0.664,
+      "step": 373
+    },
+    {
+      "epoch": 0.6891891891891891,
+      "grad_norm": 0.8164847737457118,
+      "learning_rate": 9.495485762404801e-06,
+      "loss": 0.6899,
+      "step": 374
+    },
+    {
+      "epoch": 0.691031941031941,
+      "grad_norm": 0.7997564433278598,
+      "learning_rate": 9.490775368879966e-06,
+      "loss": 0.6444,
+      "step": 375
+    },
+    {
+      "epoch": 0.6928746928746928,
+      "grad_norm": 0.7488792100084561,
+      "learning_rate": 9.486044267636359e-06,
+      "loss": 0.665,
+      "step": 376
+    },
+    {
+      "epoch": 0.6947174447174447,
+      "grad_norm": 0.8057375743505253,
+      "learning_rate": 9.481292480489885e-06,
+      "loss": 0.668,
+      "step": 377
+    },
+    {
+      "epoch": 0.6965601965601965,
+      "grad_norm": 0.7800885000422563,
+      "learning_rate": 9.476520029351834e-06,
+      "loss": 0.6455,
+      "step": 378
+    },
+    {
+      "epoch": 0.6984029484029484,
+      "grad_norm": 0.8646959515644008,
+      "learning_rate": 9.471726936228775e-06,
+      "loss": 0.6688,
+      "step": 379
+    },
+    {
+      "epoch": 0.7002457002457002,
+      "grad_norm": 0.7945688751619612,
+      "learning_rate": 9.466913223222467e-06,
+      "loss": 0.6536,
+      "step": 380
+    },
+    {
+      "epoch": 0.702088452088452,
+      "grad_norm": 0.896746267568911,
+      "learning_rate": 9.462078912529748e-06,
+      "loss": 0.6537,
+      "step": 381
+    },
+    {
+      "epoch": 0.7039312039312039,
+      "grad_norm": 0.7872440156352458,
+      "learning_rate": 9.457224026442435e-06,
+      "loss": 0.6383,
+      "step": 382
+    },
+    {
+      "epoch": 0.7057739557739557,
+      "grad_norm": 0.9124664828293124,
+      "learning_rate": 9.452348587347224e-06,
+      "loss": 0.6814,
+      "step": 383
+    },
+    {
+      "epoch": 0.7076167076167076,
+      "grad_norm": 0.817149332626832,
+      "learning_rate": 9.447452617725581e-06,
+      "loss": 0.6677,
+      "step": 384
+    },
+    {
+      "epoch": 0.7094594594594594,
+      "grad_norm": 0.7773593959059183,
+      "learning_rate": 9.442536140153644e-06,
+      "loss": 0.6628,
+      "step": 385
+    },
+    {
+      "epoch": 0.7113022113022113,
+      "grad_norm": 0.7899094907576322,
+      "learning_rate": 9.437599177302115e-06,
+      "loss": 0.6646,
+      "step": 386
+    },
+    {
+      "epoch": 0.7131449631449631,
+      "grad_norm": 0.8235411101841071,
+      "learning_rate": 9.432641751936162e-06,
+      "loss": 0.6526,
+      "step": 387
+    },
+    {
+      "epoch": 0.714987714987715,
+      "grad_norm": 0.8706869629475792,
+      "learning_rate": 9.427663886915302e-06,
+      "loss": 0.6491,
+      "step": 388
+    },
+    {
+      "epoch": 0.7168304668304668,
+      "grad_norm": 0.7347769186670484,
+      "learning_rate": 9.422665605193304e-06,
+      "loss": 0.6393,
+      "step": 389
+    },
+    {
+      "epoch": 0.7186732186732187,
+      "grad_norm": 0.7827345974475972,
+      "learning_rate": 9.417646929818089e-06,
+      "loss": 0.6662,
+      "step": 390
+    },
+    {
+      "epoch": 0.7205159705159705,
+      "grad_norm": 0.7608286179054605,
+      "learning_rate": 9.412607883931608e-06,
+      "loss": 0.6302,
+      "step": 391
+    },
+    {
+      "epoch": 0.7223587223587223,
+      "grad_norm": 0.7545015929831567,
+      "learning_rate": 9.407548490769747e-06,
+      "loss": 0.6384,
+      "step": 392
+    },
+    {
+      "epoch": 0.7242014742014742,
+      "grad_norm": 0.7672085924812692,
+      "learning_rate": 9.402468773662217e-06,
+      "loss": 0.6789,
+      "step": 393
+    },
+    {
+      "epoch": 0.726044226044226,
+      "grad_norm": 0.7591772174209351,
+      "learning_rate": 9.397368756032445e-06,
+      "loss": 0.6547,
+      "step": 394
+    },
+    {
+      "epoch": 0.7278869778869779,
+      "grad_norm": 0.8524442794608429,
+      "learning_rate": 9.392248461397474e-06,
+      "loss": 0.6615,
+      "step": 395
+    },
+    {
+      "epoch": 0.7297297297297297,
+      "grad_norm": 0.8384850853426337,
+      "learning_rate": 9.387107913367832e-06,
+      "loss": 0.6703,
+      "step": 396
+    },
+    {
+      "epoch": 0.7315724815724816,
+      "grad_norm": 0.8378170096292153,
+      "learning_rate": 9.381947135647457e-06,
+      "loss": 0.6606,
+      "step": 397
+    },
+    {
+      "epoch": 0.7334152334152334,
+      "grad_norm": 0.8384336389649186,
+      "learning_rate": 9.376766152033556e-06,
+      "loss": 0.6784,
+      "step": 398
+    },
+    {
+      "epoch": 0.7352579852579852,
+      "grad_norm": 0.7632782067676832,
+      "learning_rate": 9.371564986416517e-06,
+      "loss": 0.6659,
+      "step": 399
+    },
+    {
+      "epoch": 0.7371007371007371,
+      "grad_norm": 0.7222719560434115,
+      "learning_rate": 9.366343662779784e-06,
+      "loss": 0.6825,
+      "step": 400
+    },
+    {
+      "epoch": 0.7389434889434889,
+      "grad_norm": 0.80585707807462,
+      "learning_rate": 9.361102205199762e-06,
+      "loss": 0.6666,
+      "step": 401
+    },
+    {
+      "epoch": 0.7407862407862408,
+      "grad_norm": 0.8233997208248243,
+      "learning_rate": 9.355840637845683e-06,
+      "loss": 0.6597,
+      "step": 402
+    },
+    {
+      "epoch": 0.7426289926289926,
+      "grad_norm": 0.8422301038097365,
+      "learning_rate": 9.350558984979525e-06,
+      "loss": 0.6776,
+      "step": 403
+    },
+    {
+      "epoch": 0.7444717444717445,
+      "grad_norm": 0.7320999818720708,
+      "learning_rate": 9.345257270955873e-06,
+      "loss": 0.6415,
+      "step": 404
+    },
+    {
+      "epoch": 0.7463144963144963,
+      "grad_norm": 0.9240567974232431,
+      "learning_rate": 9.339935520221816e-06,
+      "loss": 0.6637,
+      "step": 405
+    },
+    {
+      "epoch": 0.7481572481572482,
+      "grad_norm": 0.7390426476384717,
+      "learning_rate": 9.334593757316845e-06,
+      "loss": 0.6527,
+      "step": 406
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.819281702560366,
+      "learning_rate": 9.329232006872722e-06,
+      "loss": 0.6633,
+      "step": 407
+    },
+    {
+      "epoch": 0.7518427518427518,
+      "grad_norm": 0.7970606284875631,
+      "learning_rate": 9.32385029361338e-06,
+      "loss": 0.6751,
+      "step": 408
+    },
+    {
+      "epoch": 0.7536855036855037,
+      "grad_norm": 0.8583350280165865,
+      "learning_rate": 9.3184486423548e-06,
+      "loss": 0.6447,
+      "step": 409
+    },
+    {
+      "epoch": 0.7555282555282555,
+      "grad_norm": 0.7619416643124071,
+      "learning_rate": 9.313027078004903e-06,
+      "loss": 0.6604,
+      "step": 410
+    },
+    {
+      "epoch": 0.7573710073710074,
+      "grad_norm": 0.7762383107391397,
+      "learning_rate": 9.307585625563433e-06,
+      "loss": 0.6796,
+      "step": 411
+    },
+    {
+      "epoch": 0.7592137592137592,
+      "grad_norm": 0.8566669202156035,
+      "learning_rate": 9.302124310121841e-06,
+      "loss": 0.6595,
+      "step": 412
+    },
+    {
+      "epoch": 0.7610565110565111,
+      "grad_norm": 0.7200892712212054,
+      "learning_rate": 9.296643156863166e-06,
+      "loss": 0.67,
+      "step": 413
+    },
+    {
+      "epoch": 0.7628992628992629,
+      "grad_norm": 0.777908130355558,
+      "learning_rate": 9.291142191061927e-06,
+      "loss": 0.6735,
+      "step": 414
+    },
+    {
+      "epoch": 0.7647420147420148,
+      "grad_norm": 0.7745314280032366,
+      "learning_rate": 9.285621438083997e-06,
+      "loss": 0.6521,
+      "step": 415
+    },
+    {
+      "epoch": 0.7665847665847666,
+      "grad_norm": 0.8148954376617112,
+      "learning_rate": 9.280080923386501e-06,
+      "loss": 0.6598,
+      "step": 416
+    },
+    {
+      "epoch": 0.7684275184275184,
+      "grad_norm": 0.9644633774641425,
+      "learning_rate": 9.274520672517677e-06,
+      "loss": 0.6695,
+      "step": 417
+    },
+    {
+      "epoch": 0.7702702702702703,
+      "grad_norm": 1.1800194434068902,
+      "learning_rate": 9.268940711116776e-06,
+      "loss": 0.6466,
+      "step": 418
+    },
+    {
+      "epoch": 0.7721130221130221,
+      "grad_norm": 0.8670124448519689,
+      "learning_rate": 9.263341064913937e-06,
+      "loss": 0.6547,
+      "step": 419
+    },
+    {
+      "epoch": 0.773955773955774,
+      "grad_norm": 0.9141859302036267,
+      "learning_rate": 9.25772175973007e-06,
+      "loss": 0.6256,
+      "step": 420
+    },
+    {
+      "epoch": 0.7757985257985258,
+      "grad_norm": 1.2319090563380988,
+      "learning_rate": 9.252082821476728e-06,
+      "loss": 0.6555,
+      "step": 421
+    },
+    {
+      "epoch": 0.7776412776412777,
+      "grad_norm": 0.8271104498963754,
+      "learning_rate": 9.246424276156008e-06,
+      "loss": 0.6532,
+      "step": 422
+    },
+    {
+      "epoch": 0.7794840294840295,
+      "grad_norm": 1.006108211030373,
+      "learning_rate": 9.240746149860408e-06,
+      "loss": 0.6533,
+      "step": 423
+    },
+    {
+      "epoch": 0.7813267813267813,
+      "grad_norm": 0.814808420878039,
+      "learning_rate": 9.235048468772722e-06,
+      "loss": 0.6518,
+      "step": 424
+    },
+    {
+      "epoch": 0.7831695331695332,
+      "grad_norm": 0.8447351813096302,
+      "learning_rate": 9.229331259165915e-06,
+      "loss": 0.6649,
+      "step": 425
+    },
+    {
+      "epoch": 0.785012285012285,
+      "grad_norm": 0.9316139301451462,
+      "learning_rate": 9.223594547402999e-06,
+      "loss": 0.6555,
+      "step": 426
+    },
+    {
+      "epoch": 0.7868550368550369,
+      "grad_norm": 0.9200122348054227,
+      "learning_rate": 9.217838359936914e-06,
+      "loss": 0.6493,
+      "step": 427
+    },
+    {
+      "epoch": 0.7886977886977887,
+      "grad_norm": 0.8450442934680126,
+      "learning_rate": 9.212062723310406e-06,
+      "loss": 0.685,
+      "step": 428
+    },
+    {
+      "epoch": 0.7905405405405406,
+      "grad_norm": 0.922063047721046,
+      "learning_rate": 9.206267664155906e-06,
+      "loss": 0.6629,
+      "step": 429
+    },
+    {
+      "epoch": 0.7923832923832924,
+      "grad_norm": 0.809214859370997,
+      "learning_rate": 9.200453209195404e-06,
+      "loss": 0.6575,
+      "step": 430
+    },
+    {
+      "epoch": 0.7942260442260443,
+      "grad_norm": 0.9282374972647772,
+      "learning_rate": 9.194619385240327e-06,
+      "loss": 0.6512,
+      "step": 431
+    },
+    {
+      "epoch": 0.7960687960687961,
+      "grad_norm": 1.0144249500003397,
+      "learning_rate": 9.188766219191415e-06,
+      "loss": 0.6551,
+      "step": 432
+    },
+    {
+      "epoch": 0.797911547911548,
+      "grad_norm": 0.735594042862335,
+      "learning_rate": 9.1828937380386e-06,
+      "loss": 0.643,
+      "step": 433
+    },
+    {
+      "epoch": 0.7997542997542998,
+      "grad_norm": 1.0853481916508259,
+      "learning_rate": 9.177001968860878e-06,
+      "loss": 0.6608,
+      "step": 434
+    },
+    {
+      "epoch": 0.8015970515970516,
+      "grad_norm": 0.8027639136627172,
+      "learning_rate": 9.171090938826184e-06,
+      "loss": 0.6332,
+      "step": 435
+    },
+    {
+      "epoch": 0.8034398034398035,
+      "grad_norm": 0.717800309806213,
+      "learning_rate": 9.165160675191272e-06,
+      "loss": 0.6549,
+      "step": 436
+    },
+    {
+      "epoch": 0.8052825552825553,
+      "grad_norm": 0.7498819425939834,
+      "learning_rate": 9.159211205301581e-06,
+      "loss": 0.6577,
+      "step": 437
+    },
+    {
+      "epoch": 0.8071253071253072,
+      "grad_norm": 0.8879854775902885,
+      "learning_rate": 9.153242556591115e-06,
+      "loss": 0.6566,
+      "step": 438
+    },
+    {
+      "epoch": 0.808968058968059,
+      "grad_norm": 0.7494925526236218,
+      "learning_rate": 9.147254756582316e-06,
+      "loss": 0.6761,
+      "step": 439
+    },
+    {
+      "epoch": 0.8108108108108109,
+      "grad_norm": 0.846750301070852,
+      "learning_rate": 9.141247832885932e-06,
+      "loss": 0.6636,
+      "step": 440
+    },
+    {
+      "epoch": 0.8126535626535627,
+      "grad_norm": 0.8057263401550839,
+      "learning_rate": 9.135221813200899e-06,
+      "loss": 0.625,
+      "step": 441
+    },
+    {
+      "epoch": 0.8144963144963145,
+      "grad_norm": 0.7163441674081074,
+      "learning_rate": 9.129176725314201e-06,
+      "loss": 0.635,
+      "step": 442
+    },
+    {
+      "epoch": 0.8163390663390664,
+      "grad_norm": 0.7894050184360264,
+      "learning_rate": 9.123112597100759e-06,
+      "loss": 0.6603,
+      "step": 443
+    },
+    {
+      "epoch": 0.8181818181818182,
+      "grad_norm": 0.8090134243598819,
+      "learning_rate": 9.11702945652328e-06,
+      "loss": 0.6401,
+      "step": 444
+    },
+    {
+      "epoch": 0.8200245700245701,
+      "grad_norm": 0.8577292073870282,
+      "learning_rate": 9.110927331632147e-06,
+      "loss": 0.6502,
+      "step": 445
+    },
+    {
+      "epoch": 0.8218673218673219,
+      "grad_norm": 0.8505268724831258,
+      "learning_rate": 9.104806250565283e-06,
+      "loss": 0.6619,
+      "step": 446
+    },
+    {
+      "epoch": 0.8237100737100738,
+      "grad_norm": 0.7923990421015751,
+      "learning_rate": 9.09866624154802e-06,
+      "loss": 0.6578,
+      "step": 447
+    },
+    {
+      "epoch": 0.8255528255528255,
+      "grad_norm": 0.680774605820965,
+      "learning_rate": 9.092507332892968e-06,
+      "loss": 0.6561,
+      "step": 448
+    },
+    {
+      "epoch": 0.8273955773955773,
+      "grad_norm": 0.8632751994197385,
+      "learning_rate": 9.08632955299989e-06,
+      "loss": 0.6307,
+      "step": 449
+    },
+    {
+      "epoch": 0.8292383292383292,
+      "grad_norm": 0.8362771149068344,
+      "learning_rate": 9.080132930355567e-06,
+      "loss": 0.6733,
+      "step": 450
+    },
+    {
+      "epoch": 0.831081081081081,
+      "grad_norm": 0.7510900302639285,
+      "learning_rate": 9.073917493533663e-06,
+      "loss": 0.656,
+      "step": 451
+    },
+    {
+      "epoch": 0.8329238329238329,
+      "grad_norm": 0.7706027790979336,
+      "learning_rate": 9.067683271194601e-06,
+      "loss": 0.6643,
+      "step": 452
+    },
+    {
+      "epoch": 0.8347665847665847,
+      "grad_norm": 0.7781686573265065,
+      "learning_rate": 9.061430292085427e-06,
+      "loss": 0.6569,
+      "step": 453
+    },
+    {
+      "epoch": 0.8366093366093366,
+      "grad_norm": 0.7427003018068699,
+      "learning_rate": 9.055158585039678e-06,
+      "loss": 0.6521,
+      "step": 454
+    },
+    {
+      "epoch": 0.8384520884520884,
+      "grad_norm": 0.7397457137012208,
+      "learning_rate": 9.048868178977243e-06,
+      "loss": 0.6552,
+      "step": 455
+    },
+    {
+      "epoch": 0.8402948402948403,
+      "grad_norm": 0.8581620237032738,
+      "learning_rate": 9.042559102904245e-06,
+      "loss": 0.6566,
+      "step": 456
+    },
+    {
+      "epoch": 0.8421375921375921,
+      "grad_norm": 0.7712938882011053,
+      "learning_rate": 9.03623138591289e-06,
+      "loss": 0.6468,
+      "step": 457
+    },
+    {
+      "epoch": 0.8439803439803439,
+      "grad_norm": 0.8890510433500383,
+      "learning_rate": 9.029885057181343e-06,
+      "loss": 0.6446,
+      "step": 458
+    },
+    {
+      "epoch": 0.8458230958230958,
+      "grad_norm": 0.7290466358267097,
+      "learning_rate": 9.02352014597359e-06,
+      "loss": 0.6454,
+      "step": 459
+    },
+    {
+      "epoch": 0.8476658476658476,
+      "grad_norm": 0.7146649190580427,
+      "learning_rate": 9.017136681639307e-06,
+      "loss": 0.6718,
+      "step": 460
+    },
+    {
+      "epoch": 0.8495085995085995,
+      "grad_norm": 0.7845525140383119,
+      "learning_rate": 9.010734693613716e-06,
+      "loss": 0.6686,
+      "step": 461
+    },
+    {
+      "epoch": 0.8513513513513513,
+      "grad_norm": 0.7058298908407231,
+      "learning_rate": 9.004314211417461e-06,
+      "loss": 0.6579,
+      "step": 462
+    },
+    {
+      "epoch": 0.8531941031941032,
+      "grad_norm": 0.9275282637380985,
+      "learning_rate": 8.997875264656461e-06,
+      "loss": 0.6593,
+      "step": 463
+    },
+    {
+      "epoch": 0.855036855036855,
+      "grad_norm": 0.7510201338976162,
+      "learning_rate": 8.99141788302178e-06,
+      "loss": 0.6511,
+      "step": 464
+    },
+    {
+      "epoch": 0.8568796068796068,
+      "grad_norm": 0.7924253380281218,
+      "learning_rate": 8.984942096289488e-06,
+      "loss": 0.6569,
+      "step": 465
+    },
+    {
+      "epoch": 0.8587223587223587,
+      "grad_norm": 0.7950614614013062,
+      "learning_rate": 8.978447934320526e-06,
+      "loss": 0.6625,
+      "step": 466
+    },
+    {
+      "epoch": 0.8605651105651105,
+      "grad_norm": 0.783065472883908,
+      "learning_rate": 8.971935427060563e-06,
+      "loss": 0.6532,
+      "step": 467
+    },
+    {
+      "epoch": 0.8624078624078624,
+      "grad_norm": 0.8010941553307661,
+      "learning_rate": 8.965404604539864e-06,
+      "loss": 0.6477,
+      "step": 468
+    },
+    {
+      "epoch": 0.8642506142506142,
+      "grad_norm": 0.7828731416204852,
+      "learning_rate": 8.958855496873146e-06,
+      "loss": 0.6372,
+      "step": 469
+    },
+    {
+      "epoch": 0.8660933660933661,
+      "grad_norm": 0.7597269182760727,
+      "learning_rate": 8.952288134259443e-06,
+      "loss": 0.6448,
+      "step": 470
+    },
+    {
+      "epoch": 0.8679361179361179,
+      "grad_norm": 0.741605049875348,
+      "learning_rate": 8.94570254698197e-06,
+      "loss": 0.6694,
+      "step": 471
+    },
+    {
+      "epoch": 0.8697788697788698,
+      "grad_norm": 0.7181841404443761,
+      "learning_rate": 8.93909876540797e-06,
+      "loss": 0.6609,
+      "step": 472
+    },
+    {
+      "epoch": 0.8716216216216216,
+      "grad_norm": 0.7663473253105653,
+      "learning_rate": 8.932476819988589e-06,
+      "loss": 0.6378,
+      "step": 473
+    },
+    {
+      "epoch": 0.8734643734643734,
+      "grad_norm": 0.8173455908928104,
+      "learning_rate": 8.925836741258729e-06,
+      "loss": 0.6714,
+      "step": 474
+    },
+    {
+      "epoch": 0.8753071253071253,
+      "grad_norm": 0.7583877042704429,
+      "learning_rate": 8.91917855983691e-06,
+      "loss": 0.6726,
+      "step": 475
+    },
+    {
+      "epoch": 0.8771498771498771,
+      "grad_norm": 0.7480765146803664,
+      "learning_rate": 8.912502306425121e-06,
+      "loss": 0.6565,
+      "step": 476
+    },
+    {
+      "epoch": 0.878992628992629,
+      "grad_norm": 0.6859083447971515,
+      "learning_rate": 8.905808011808685e-06,
+      "loss": 0.653,
+      "step": 477
+    },
+    {
+      "epoch": 0.8808353808353808,
+      "grad_norm": 0.8181927675513071,
+      "learning_rate": 8.899095706856122e-06,
+      "loss": 0.6571,
+      "step": 478
+    },
+    {
+      "epoch": 0.8826781326781327,
+      "grad_norm": 0.8229656917976655,
+      "learning_rate": 8.892365422518995e-06,
+      "loss": 0.6598,
+      "step": 479
+    },
+    {
+      "epoch": 0.8845208845208845,
+      "grad_norm": 0.79944667846446,
+      "learning_rate": 8.885617189831772e-06,
+      "loss": 0.6543,
+      "step": 480
+    },
+    {
+      "epoch": 0.8863636363636364,
+      "grad_norm": 0.8000910062918836,
+      "learning_rate": 8.87885103991169e-06,
+      "loss": 0.6604,
+      "step": 481
+    },
+    {
+      "epoch": 0.8882063882063882,
+      "grad_norm": 0.7747380372554368,
+      "learning_rate": 8.872067003958597e-06,
+      "loss": 0.6478,
+      "step": 482
+    },
+    {
+      "epoch": 0.89004914004914,
+      "grad_norm": 0.7711140620131817,
+      "learning_rate": 8.865265113254826e-06,
+      "loss": 0.661,
+      "step": 483
+    },
+    {
+      "epoch": 0.8918918918918919,
+      "grad_norm": 0.8030474255072744,
+      "learning_rate": 8.858445399165033e-06,
+      "loss": 0.6624,
+      "step": 484
+    },
+    {
+      "epoch": 0.8937346437346437,
+      "grad_norm": 0.7330461039637997,
+      "learning_rate": 8.851607893136065e-06,
+      "loss": 0.6701,
+      "step": 485
+    },
+    {
+      "epoch": 0.8955773955773956,
+      "grad_norm": 0.8268965599742298,
+      "learning_rate": 8.84475262669681e-06,
+      "loss": 0.6441,
+      "step": 486
+    },
+    {
+      "epoch": 0.8974201474201474,
+      "grad_norm": 0.7084822010000799,
+      "learning_rate": 8.837879631458053e-06,
+      "loss": 0.667,
+      "step": 487
+    },
+    {
+      "epoch": 0.8992628992628993,
+      "grad_norm": 0.7323963645211595,
+      "learning_rate": 8.830988939112326e-06,
+      "loss": 0.651,
+      "step": 488
+    },
+    {
+      "epoch": 0.9011056511056511,
+      "grad_norm": 0.8349210423913278,
+      "learning_rate": 8.824080581433769e-06,
+      "loss": 0.6627,
+      "step": 489
+    },
+    {
+      "epoch": 0.902948402948403,
+      "grad_norm": 0.7443897083602373,
+      "learning_rate": 8.817154590277976e-06,
+      "loss": 0.6657,
+      "step": 490
+    },
+    {
+      "epoch": 0.9047911547911548,
+      "grad_norm": 0.8338611023908313,
+      "learning_rate": 8.810210997581859e-06,
+      "loss": 0.6442,
+      "step": 491
+    },
+    {
+      "epoch": 0.9066339066339066,
+      "grad_norm": 0.7771582831889335,
+      "learning_rate": 8.803249835363486e-06,
+      "loss": 0.6684,
+      "step": 492
+    },
+    {
+      "epoch": 0.9084766584766585,
+      "grad_norm": 0.7785727523914899,
+      "learning_rate": 8.796271135721944e-06,
+      "loss": 0.6421,
+      "step": 493
+    },
+    {
+      "epoch": 0.9103194103194103,
+      "grad_norm": 0.842337491051386,
+      "learning_rate": 8.789274930837189e-06,
+      "loss": 0.6568,
+      "step": 494
+    },
+    {
+      "epoch": 0.9121621621621622,
+      "grad_norm": 0.6974887020665224,
+      "learning_rate": 8.782261252969895e-06,
+      "loss": 0.6571,
+      "step": 495
+    },
+    {
+      "epoch": 0.914004914004914,
+      "grad_norm": 0.7375887565646241,
+      "learning_rate": 8.775230134461307e-06,
+      "loss": 0.6597,
+      "step": 496
+    },
+    {
+      "epoch": 0.9158476658476659,
+      "grad_norm": 0.7426988288332769,
+      "learning_rate": 8.768181607733095e-06,
+      "loss": 0.6505,
+      "step": 497
+    },
+    {
+      "epoch": 0.9176904176904177,
+      "grad_norm": 0.709332532258604,
+      "learning_rate": 8.761115705287195e-06,
+      "loss": 0.6515,
+      "step": 498
+    },
+    {
+      "epoch": 0.9195331695331695,
+      "grad_norm": 0.8015532638152897,
+      "learning_rate": 8.754032459705672e-06,
+      "loss": 0.6786,
+      "step": 499
+    },
+    {
+      "epoch": 0.9213759213759214,
+      "grad_norm": 0.7507728174041544,
+      "learning_rate": 8.746931903650558e-06,
+      "loss": 0.6584,
+      "step": 500
+    },
+    {
+      "epoch": 0.9232186732186732,
+      "grad_norm": 0.7938224543426052,
+      "learning_rate": 8.739814069863708e-06,
+      "loss": 0.6615,
+      "step": 501
+    },
+    {
+      "epoch": 0.9250614250614251,
+      "grad_norm": 0.7337604678335348,
+      "learning_rate": 8.732678991166647e-06,
+      "loss": 0.6346,
+      "step": 502
+    },
+    {
+      "epoch": 0.9269041769041769,
+      "grad_norm": 0.8491117907059232,
+      "learning_rate": 8.725526700460426e-06,
+      "loss": 0.6518,
+      "step": 503
+    },
+    {
+      "epoch": 0.9287469287469288,
+      "grad_norm": 0.7423777292730689,
+      "learning_rate": 8.71835723072545e-06,
+      "loss": 0.6524,
+      "step": 504
+    },
+    {
+      "epoch": 0.9305896805896806,
+      "grad_norm": 0.9214804967762069,
+      "learning_rate": 8.71117061502135e-06,
+      "loss": 0.6361,
+      "step": 505
+    },
+    {
+      "epoch": 0.9324324324324325,
+      "grad_norm": 0.7368071143800767,
+      "learning_rate": 8.703966886486819e-06,
+      "loss": 0.6392,
+      "step": 506
+    },
+    {
+      "epoch": 0.9342751842751843,
+      "grad_norm": 0.7160710562931542,
+      "learning_rate": 8.696746078339455e-06,
+      "loss": 0.649,
+      "step": 507
+    },
+    {
+      "epoch": 0.9361179361179361,
+      "grad_norm": 0.7467328531981149,
+      "learning_rate": 8.68950822387562e-06,
+      "loss": 0.6537,
+      "step": 508
+    },
+    {
+      "epoch": 0.937960687960688,
+      "grad_norm": 0.7520100918172796,
+      "learning_rate": 8.68225335647027e-06,
+      "loss": 0.6513,
+      "step": 509
+    },
+    {
+      "epoch": 0.9398034398034398,
+      "grad_norm": 0.8004449386543357,
+      "learning_rate": 8.674981509576819e-06,
+      "loss": 0.6476,
+      "step": 510
+    },
+    {
+      "epoch": 0.9416461916461917,
+      "grad_norm": 0.8327269032657233,
+      "learning_rate": 8.667692716726974e-06,
+      "loss": 0.6472,
+      "step": 511
+    },
+    {
+      "epoch": 0.9434889434889435,
+      "grad_norm": 0.8839591618619658,
+      "learning_rate": 8.66038701153058e-06,
+      "loss": 0.6166,
+      "step": 512
+    },
+    {
+      "epoch": 0.9453316953316954,
+      "grad_norm": 0.9379994954764558,
+      "learning_rate": 8.65306442767547e-06,
+      "loss": 0.651,
+      "step": 513
+    },
+    {
+      "epoch": 0.9471744471744472,
+      "grad_norm": 0.7487293293271072,
+      "learning_rate": 8.64572499892731e-06,
+      "loss": 0.66,
+      "step": 514
+    },
+    {
+      "epoch": 0.949017199017199,
+      "grad_norm": 0.7961382561422509,
+      "learning_rate": 8.638368759129433e-06,
+      "loss": 0.664,
+      "step": 515
+    },
+    {
+      "epoch": 0.9508599508599509,
+      "grad_norm": 0.8874779795271696,
+      "learning_rate": 8.630995742202695e-06,
+      "loss": 0.6452,
+      "step": 516
+    },
+    {
+      "epoch": 0.9527027027027027,
+      "grad_norm": 0.6841041864618989,
+      "learning_rate": 8.623605982145318e-06,
+      "loss": 0.6478,
+      "step": 517
+    },
+    {
+      "epoch": 0.9545454545454546,
+      "grad_norm": 0.804797096615849,
+      "learning_rate": 8.616199513032723e-06,
+      "loss": 0.6461,
+      "step": 518
+    },
+    {
+      "epoch": 0.9563882063882064,
+      "grad_norm": 0.880775319209045,
+      "learning_rate": 8.608776369017382e-06,
+      "loss": 0.6621,
+      "step": 519
+    },
+    {
+      "epoch": 0.9582309582309583,
+      "grad_norm": 0.7542987011863731,
+      "learning_rate": 8.601336584328659e-06,
+      "loss": 0.6347,
+      "step": 520
+    },
+    {
+      "epoch": 0.9600737100737101,
+      "grad_norm": 0.7100147589007487,
+      "learning_rate": 8.593880193272649e-06,
+      "loss": 0.6603,
+      "step": 521
+    },
+    {
+      "epoch": 0.961916461916462,
+      "grad_norm": 0.7087889597381314,
+      "learning_rate": 8.58640723023202e-06,
+      "loss": 0.6355,
+      "step": 522
+    },
+    {
+      "epoch": 0.9637592137592138,
+      "grad_norm": 0.9370070620497674,
+      "learning_rate": 8.578917729665863e-06,
+      "loss": 0.6425,
+      "step": 523
+    },
+    {
+      "epoch": 0.9656019656019657,
+      "grad_norm": 0.7000173833609282,
+      "learning_rate": 8.571411726109518e-06,
+      "loss": 0.6356,
+      "step": 524
+    },
+    {
+      "epoch": 0.9674447174447175,
+      "grad_norm": 0.7553763494598327,
+      "learning_rate": 8.563889254174429e-06,
+      "loss": 0.6354,
+      "step": 525
+    },
+    {
+      "epoch": 0.9692874692874693,
+      "grad_norm": 0.8900262312840056,
+      "learning_rate": 8.556350348547978e-06,
+      "loss": 0.6633,
+      "step": 526
+    },
+    {
+      "epoch": 0.9711302211302212,
+      "grad_norm": 0.8539803894597575,
+      "learning_rate": 8.548795043993316e-06,
+      "loss": 0.6459,
+      "step": 527
+    },
+    {
+      "epoch": 0.972972972972973,
+      "grad_norm": 0.7006505594925245,
+      "learning_rate": 8.54122337534923e-06,
+      "loss": 0.6647,
+      "step": 528
+    },
+    {
+      "epoch": 0.9748157248157249,
+      "grad_norm": 1.0891473249129742,
+      "learning_rate": 8.533635377529949e-06,
+      "loss": 0.6507,
+      "step": 529
+    },
+    {
+      "epoch": 0.9766584766584766,
+      "grad_norm": 0.8693595934981059,
+      "learning_rate": 8.526031085525004e-06,
+      "loss": 0.6456,
+      "step": 530
+    },
+    {
+      "epoch": 0.9785012285012284,
+      "grad_norm": 0.7798305556359063,
+      "learning_rate": 8.518410534399063e-06,
+      "loss": 0.6534,
+      "step": 531
+    },
+    {
+      "epoch": 0.9803439803439803,
+      "grad_norm": 0.8628050704444215,
+      "learning_rate": 8.510773759291768e-06,
+      "loss": 0.6453,
+      "step": 532
+    },
+    {
+      "epoch": 0.9821867321867321,
+      "grad_norm": 0.857636379598045,
+      "learning_rate": 8.503120795417568e-06,
+      "loss": 0.6535,
+      "step": 533
+    },
+    {
+      "epoch": 0.984029484029484,
+      "grad_norm": 0.736047660257665,
+      "learning_rate": 8.495451678065563e-06,
+      "loss": 0.6471,
+      "step": 534
+    },
+    {
+      "epoch": 0.9858722358722358,
+      "grad_norm": 0.7485859862167331,
+      "learning_rate": 8.487766442599339e-06,
+      "loss": 0.65,
+      "step": 535
+    },
+    {
+      "epoch": 0.9877149877149877,
+      "grad_norm": 0.7723804121593948,
+      "learning_rate": 8.48006512445681e-06,
+      "loss": 0.6483,
+      "step": 536
+    },
+    {
+      "epoch": 0.9895577395577395,
+      "grad_norm": 0.8296493605761454,
+      "learning_rate": 8.472347759150044e-06,
+      "loss": 0.6646,
+      "step": 537
+    },
+    {
+      "epoch": 0.9914004914004914,
+      "grad_norm": 0.736053638866414,
+      "learning_rate": 8.464614382265107e-06,
+      "loss": 0.6319,
+      "step": 538
+    },
+    {
+      "epoch": 0.9932432432432432,
+      "grad_norm": 0.8864717443449409,
+      "learning_rate": 8.4568650294619e-06,
+      "loss": 0.6573,
+      "step": 539
+    },
+    {
+      "epoch": 0.995085995085995,
+      "grad_norm": 0.7181405986000386,
+      "learning_rate": 8.449099736473986e-06,
+      "loss": 0.6486,
+      "step": 540
+    },
+    {
+      "epoch": 0.9969287469287469,
+      "grad_norm": 0.7832348667472111,
+      "learning_rate": 8.441318539108433e-06,
+      "loss": 0.6589,
+      "step": 541
+    },
+    {
+      "epoch": 0.9987714987714987,
+      "grad_norm": 0.6713344019791813,
+      "learning_rate": 8.433521473245653e-06,
+      "loss": 0.6254,
+      "step": 542
+    },
+    {
+      "epoch": 1.0012285012285012,
+      "grad_norm": 0.8332197235450152,
+      "learning_rate": 8.425708574839221e-06,
+      "loss": 0.6191,
+      "step": 543
+    },
+    {
+      "epoch": 1.003071253071253,
+      "grad_norm": 0.7628170877202599,
+      "learning_rate": 8.417879879915724e-06,
+      "loss": 0.5938,
+      "step": 544
+    },
+    {
+      "epoch": 1.0049140049140048,
+      "grad_norm": 0.6944051274625848,
+      "learning_rate": 8.410035424574587e-06,
+      "loss": 0.5697,
+      "step": 545
+    },
+    {
+      "epoch": 1.0067567567567568,
+      "grad_norm": 0.7689817834192046,
+      "learning_rate": 8.40217524498791e-06,
+      "loss": 0.5782,
+      "step": 546
+    },
+    {
+      "epoch": 1.0085995085995085,
+      "grad_norm": 0.7087929766881554,
+      "learning_rate": 8.394299377400301e-06,
+      "loss": 0.5914,
+      "step": 547
+    },
+    {
+      "epoch": 1.0104422604422605,
+      "grad_norm": 0.7195201848180063,
+      "learning_rate": 8.386407858128707e-06,
+      "loss": 0.5784,
+      "step": 548
+    },
+    {
+      "epoch": 1.0122850122850122,
+      "grad_norm": 0.8324291624053168,
+      "learning_rate": 8.378500723562243e-06,
+      "loss": 0.6016,
+      "step": 549
+    },
+    {
+      "epoch": 1.0141277641277642,
+      "grad_norm": 0.7083654556797554,
+      "learning_rate": 8.370578010162043e-06,
+      "loss": 0.5914,
+      "step": 550
+    },
+    {
+      "epoch": 1.015970515970516,
+      "grad_norm": 0.7334666729181748,
+      "learning_rate": 8.362639754461057e-06,
+      "loss": 0.5856,
+      "step": 551
+    },
+    {
+      "epoch": 1.0178132678132679,
+      "grad_norm": 0.7686093423350041,
+      "learning_rate": 8.354685993063923e-06,
+      "loss": 0.6008,
+      "step": 552
+    },
+    {
+      "epoch": 1.0196560196560196,
+      "grad_norm": 0.7142115222443369,
+      "learning_rate": 8.346716762646763e-06,
+      "loss": 0.5852,
+      "step": 553
+    },
+    {
+      "epoch": 1.0214987714987716,
+      "grad_norm": 0.6924490195107053,
+      "learning_rate": 8.338732099957038e-06,
+      "loss": 0.5902,
+      "step": 554
+    },
+    {
+      "epoch": 1.0233415233415233,
+      "grad_norm": 0.759799739950896,
+      "learning_rate": 8.330732041813367e-06,
+      "loss": 0.5758,
+      "step": 555
+    },
+    {
+      "epoch": 1.0251842751842752,
+      "grad_norm": 0.7187939376898641,
+      "learning_rate": 8.322716625105363e-06,
+      "loss": 0.5779,
+      "step": 556
+    },
+    {
+      "epoch": 1.027027027027027,
+      "grad_norm": 0.6992682343903865,
+      "learning_rate": 8.314685886793456e-06,
+      "loss": 0.6029,
+      "step": 557
+    },
+    {
+      "epoch": 1.028869778869779,
+      "grad_norm": 0.7216544170224443,
+      "learning_rate": 8.306639863908725e-06,
+      "loss": 0.5838,
+      "step": 558
+    },
+    {
+      "epoch": 1.0307125307125307,
+      "grad_norm": 0.6628258282790795,
+      "learning_rate": 8.298578593552737e-06,
+      "loss": 0.591,
+      "step": 559
+    },
+    {
+      "epoch": 1.0325552825552826,
+      "grad_norm": 0.7497972716296265,
+      "learning_rate": 8.290502112897357e-06,
+      "loss": 0.6024,
+      "step": 560
+    },
+    {
+      "epoch": 1.0343980343980343,
+      "grad_norm": 0.6544190372682622,
+      "learning_rate": 8.282410459184597e-06,
+      "loss": 0.578,
+      "step": 561
+    },
+    {
+      "epoch": 1.0362407862407863,
+      "grad_norm": 0.709210608913888,
+      "learning_rate": 8.274303669726427e-06,
+      "loss": 0.5801,
+      "step": 562
+    },
+    {
+      "epoch": 1.038083538083538,
+      "grad_norm": 0.7185181447778201,
+      "learning_rate": 8.266181781904613e-06,
+      "loss": 0.5901,
+      "step": 563
+    },
+    {
+      "epoch": 1.03992628992629,
+      "grad_norm": 0.7265522245225428,
+      "learning_rate": 8.258044833170545e-06,
+      "loss": 0.583,
+      "step": 564
+    },
+    {
+      "epoch": 1.0417690417690417,
+      "grad_norm": 0.7027293993104479,
+      "learning_rate": 8.24989286104506e-06,
+      "loss": 0.604,
+      "step": 565
+    },
+    {
+      "epoch": 1.0436117936117937,
+      "grad_norm": 0.7234476110066482,
+      "learning_rate": 8.241725903118264e-06,
+      "loss": 0.5909,
+      "step": 566
+    },
+    {
+      "epoch": 1.0454545454545454,
+      "grad_norm": 0.6631099917331491,
+      "learning_rate": 8.233543997049376e-06,
+      "loss": 0.592,
+      "step": 567
+    },
+    {
+      "epoch": 1.0472972972972974,
+      "grad_norm": 0.7591111622797487,
+      "learning_rate": 8.225347180566534e-06,
+      "loss": 0.5912,
+      "step": 568
+    },
+    {
+      "epoch": 1.049140049140049,
+      "grad_norm": 0.6529047723798723,
+      "learning_rate": 8.217135491466636e-06,
+      "loss": 0.6124,
+      "step": 569
+    },
+    {
+      "epoch": 1.050982800982801,
+      "grad_norm": 0.6662262695985915,
+      "learning_rate": 8.208908967615159e-06,
+      "loss": 0.5634,
+      "step": 570
+    },
+    {
+      "epoch": 1.0528255528255528,
+      "grad_norm": 0.6492633749107398,
+      "learning_rate": 8.200667646945983e-06,
+      "loss": 0.5729,
+      "step": 571
+    },
+    {
+      "epoch": 1.0546683046683047,
+      "grad_norm": 0.7083584828833062,
+      "learning_rate": 8.192411567461222e-06,
+      "loss": 0.5733,
+      "step": 572
+    },
+    {
+      "epoch": 1.0565110565110565,
+      "grad_norm": 0.7461693823998224,
+      "learning_rate": 8.184140767231044e-06,
+      "loss": 0.5862,
+      "step": 573
+    },
+    {
+      "epoch": 1.0583538083538084,
+      "grad_norm": 0.7307661994991586,
+      "learning_rate": 8.175855284393495e-06,
+      "loss": 0.6048,
+      "step": 574
+    },
+    {
+      "epoch": 1.0601965601965602,
+      "grad_norm": 0.7449249015506589,
+      "learning_rate": 8.167555157154327e-06,
+      "loss": 0.5769,
+      "step": 575
+    },
+    {
+      "epoch": 1.0620393120393121,
+      "grad_norm": 0.676130892422225,
+      "learning_rate": 8.15924042378682e-06,
+      "loss": 0.5935,
+      "step": 576
+    },
+    {
+      "epoch": 1.0638820638820639,
+      "grad_norm": 0.6643833719870827,
+      "learning_rate": 8.150911122631606e-06,
+      "loss": 0.5944,
+      "step": 577
+    },
+    {
+      "epoch": 1.0657248157248158,
+      "grad_norm": 0.6843588500891945,
+      "learning_rate": 8.142567292096488e-06,
+      "loss": 0.5895,
+      "step": 578
+    },
+    {
+      "epoch": 1.0675675675675675,
+      "grad_norm": 0.6268234218030757,
+      "learning_rate": 8.13420897065627e-06,
+      "loss": 0.5953,
+      "step": 579
+    },
+    {
+      "epoch": 1.0694103194103195,
+      "grad_norm": 0.7223266784582031,
+      "learning_rate": 8.125836196852577e-06,
+      "loss": 0.5755,
+      "step": 580
+    },
+    {
+      "epoch": 1.0712530712530712,
+      "grad_norm": 0.6819487348864856,
+      "learning_rate": 8.117449009293668e-06,
+      "loss": 0.5627,
+      "step": 581
+    },
+    {
+      "epoch": 1.0730958230958232,
+      "grad_norm": 0.7362356961372206,
+      "learning_rate": 8.109047446654276e-06,
+      "loss": 0.5928,
+      "step": 582
+    },
+    {
+      "epoch": 1.074938574938575,
+      "grad_norm": 0.7386543254473633,
+      "learning_rate": 8.100631547675417e-06,
+      "loss": 0.5892,
+      "step": 583
+    },
+    {
+      "epoch": 1.0767813267813269,
+      "grad_norm": 0.8293500729055248,
+      "learning_rate": 8.092201351164213e-06,
+      "loss": 0.5894,
+      "step": 584
+    },
+    {
+      "epoch": 1.0786240786240786,
+      "grad_norm": 0.7921767319098849,
+      "learning_rate": 8.083756895993712e-06,
+      "loss": 0.5988,
+      "step": 585
+    },
+    {
+      "epoch": 1.0804668304668306,
+      "grad_norm": 0.7126407964873656,
+      "learning_rate": 8.075298221102714e-06,
+      "loss": 0.5996,
+      "step": 586
+    },
+    {
+      "epoch": 1.0823095823095823,
+      "grad_norm": 0.7643066739205533,
+      "learning_rate": 8.066825365495591e-06,
+      "loss": 0.5925,
+      "step": 587
+    },
+    {
+      "epoch": 1.0841523341523343,
+      "grad_norm": 0.6971456985666573,
+      "learning_rate": 8.058338368242103e-06,
+      "loss": 0.6013,
+      "step": 588
+    },
+    {
+      "epoch": 1.085995085995086,
+      "grad_norm": 0.8770399736805485,
+      "learning_rate": 8.049837268477213e-06,
+      "loss": 0.5796,
+      "step": 589
+    },
+    {
+      "epoch": 1.087837837837838,
+      "grad_norm": 0.8883143731161111,
+      "learning_rate": 8.041322105400923e-06,
+      "loss": 0.5772,
+      "step": 590
+    },
+    {
+      "epoch": 1.0896805896805897,
+      "grad_norm": 0.7412146185642823,
+      "learning_rate": 8.032792918278076e-06,
+      "loss": 0.5714,
+      "step": 591
+    },
+    {
+      "epoch": 1.0915233415233416,
+      "grad_norm": 0.8047716040124842,
+      "learning_rate": 8.024249746438189e-06,
+      "loss": 0.6049,
+      "step": 592
+    },
+    {
+      "epoch": 1.0933660933660934,
+      "grad_norm": 0.8689359917209114,
+      "learning_rate": 8.015692629275256e-06,
+      "loss": 0.6043,
+      "step": 593
+    },
+    {
+      "epoch": 1.095208845208845,
+      "grad_norm": 0.7906889660346444,
+      "learning_rate": 8.007121606247583e-06,
+      "loss": 0.594,
+      "step": 594
+    },
+    {
+      "epoch": 1.097051597051597,
+      "grad_norm": 0.7942556830556607,
+      "learning_rate": 7.998536716877593e-06,
+      "loss": 0.5786,
+      "step": 595
+    },
+    {
+      "epoch": 1.098894348894349,
+      "grad_norm": 0.7572958377237894,
+      "learning_rate": 7.989938000751655e-06,
+      "loss": 0.6026,
+      "step": 596
+    },
+    {
+      "epoch": 1.1007371007371007,
+      "grad_norm": 0.7574838978404311,
+      "learning_rate": 7.981325497519892e-06,
+      "loss": 0.5742,
+      "step": 597
+    },
+    {
+      "epoch": 1.1025798525798525,
+      "grad_norm": 0.7349225675213524,
+      "learning_rate": 7.972699246895996e-06,
+      "loss": 0.5753,
+      "step": 598
+    },
+    {
+      "epoch": 1.1044226044226044,
+      "grad_norm": 0.6932252161043515,
+      "learning_rate": 7.964059288657061e-06,
+      "loss": 0.5758,
+      "step": 599
+    },
+    {
+      "epoch": 1.1062653562653564,
+      "grad_norm": 0.7430101864982959,
+      "learning_rate": 7.955405662643384e-06,
+      "loss": 0.5715,
+      "step": 600
+    },
+    {
+      "epoch": 1.1081081081081081,
+      "grad_norm": 0.6973294734350624,
+      "learning_rate": 7.946738408758283e-06,
+      "loss": 0.5922,
+      "step": 601
+    },
+    {
+      "epoch": 1.1099508599508598,
+      "grad_norm": 0.8373457641145695,
+      "learning_rate": 7.938057566967926e-06,
+      "loss": 0.578,
+      "step": 602
+    },
+    {
+      "epoch": 1.1117936117936118,
+      "grad_norm": 0.7929239116156964,
+      "learning_rate": 7.929363177301124e-06,
+      "loss": 0.5658,
+      "step": 603
+    },
+    {
+      "epoch": 1.1136363636363635,
+      "grad_norm": 0.6162869073280085,
+      "learning_rate": 7.920655279849173e-06,
+      "loss": 0.5771,
+      "step": 604
+    },
+    {
+      "epoch": 1.1154791154791155,
+      "grad_norm": 0.897451119367903,
+      "learning_rate": 7.911933914765645e-06,
+      "loss": 0.6083,
+      "step": 605
+    },
+    {
+      "epoch": 1.1173218673218672,
+      "grad_norm": 0.8364241754499491,
+      "learning_rate": 7.90319912226622e-06,
+      "loss": 0.5933,
+      "step": 606
+    },
+    {
+      "epoch": 1.1191646191646192,
+      "grad_norm": 0.6757258923259274,
+      "learning_rate": 7.894450942628491e-06,
+      "loss": 0.5885,
+      "step": 607
+    },
+    {
+      "epoch": 1.121007371007371,
+      "grad_norm": 0.9017186887468682,
+      "learning_rate": 7.885689416191785e-06,
+      "loss": 0.5824,
+      "step": 608
+    },
+    {
+      "epoch": 1.1228501228501229,
+      "grad_norm": 1.0069352939418628,
+      "learning_rate": 7.876914583356965e-06,
+      "loss": 0.5852,
+      "step": 609
+    },
+    {
+      "epoch": 1.1246928746928746,
+      "grad_norm": 0.7733653303759642,
+      "learning_rate": 7.868126484586261e-06,
+      "loss": 0.5853,
+      "step": 610
+    },
+    {
+      "epoch": 1.1265356265356266,
+      "grad_norm": 0.8054916816934943,
+      "learning_rate": 7.859325160403073e-06,
+      "loss": 0.558,
+      "step": 611
+    },
+    {
+      "epoch": 1.1283783783783783,
+      "grad_norm": 0.9347093594958867,
+      "learning_rate": 7.850510651391778e-06,
+      "loss": 0.5972,
+      "step": 612
+    },
+    {
+      "epoch": 1.1302211302211302,
+      "grad_norm": 0.689980595344348,
+      "learning_rate": 7.841682998197561e-06,
+      "loss": 0.5817,
+      "step": 613
+    },
+    {
+      "epoch": 1.132063882063882,
+      "grad_norm": 0.8006544074796992,
+      "learning_rate": 7.832842241526212e-06,
+      "loss": 0.5925,
+      "step": 614
+    },
+    {
+      "epoch": 1.133906633906634,
+      "grad_norm": 0.6938530997331321,
+      "learning_rate": 7.823988422143942e-06,
+      "loss": 0.5737,
+      "step": 615
+    },
+    {
+      "epoch": 1.1357493857493857,
+      "grad_norm": 0.7372999061357947,
+      "learning_rate": 7.815121580877197e-06,
+      "loss": 0.5907,
+      "step": 616
+    },
+    {
+      "epoch": 1.1375921375921376,
+      "grad_norm": 0.8171425296804689,
+      "learning_rate": 7.806241758612471e-06,
+      "loss": 0.5743,
+      "step": 617
+    },
+    {
+      "epoch": 1.1394348894348894,
+      "grad_norm": 0.7290280666023506,
+      "learning_rate": 7.797348996296116e-06,
+      "loss": 0.5906,
+      "step": 618
+    },
+    {
+      "epoch": 1.1412776412776413,
+      "grad_norm": 0.7842030226269,
+      "learning_rate": 7.788443334934148e-06,
+      "loss": 0.5934,
+      "step": 619
+    },
+    {
+      "epoch": 1.143120393120393,
+      "grad_norm": 0.710439520490786,
+      "learning_rate": 7.779524815592068e-06,
+      "loss": 0.591,
+      "step": 620
+    },
+    {
+      "epoch": 1.144963144963145,
+      "grad_norm": 0.7337512345879669,
+      "learning_rate": 7.770593479394664e-06,
+      "loss": 0.5861,
+      "step": 621
+    },
+    {
+      "epoch": 1.1468058968058967,
+      "grad_norm": 0.819390009792322,
+      "learning_rate": 7.761649367525828e-06,
+      "loss": 0.5811,
+      "step": 622
+    },
+    {
+      "epoch": 1.1486486486486487,
+      "grad_norm": 0.6752381838949487,
+      "learning_rate": 7.752692521228357e-06,
+      "loss": 0.5883,
+      "step": 623
+    },
+    {
+      "epoch": 1.1504914004914004,
+      "grad_norm": 0.8273900001913623,
+      "learning_rate": 7.743722981803777e-06,
+      "loss": 0.5938,
+      "step": 624
+    },
+    {
+      "epoch": 1.1523341523341524,
+      "grad_norm": 0.7398209032825053,
+      "learning_rate": 7.734740790612137e-06,
+      "loss": 0.6121,
+      "step": 625
+    },
+    {
+      "epoch": 1.154176904176904,
+      "grad_norm": 0.7082944856041072,
+      "learning_rate": 7.725745989071825e-06,
+      "loss": 0.5988,
+      "step": 626
+    },
+    {
+      "epoch": 1.156019656019656,
+      "grad_norm": 0.7163721766401956,
+      "learning_rate": 7.716738618659382e-06,
+      "loss": 0.568,
+      "step": 627
+    },
+    {
+      "epoch": 1.1578624078624078,
+      "grad_norm": 0.7325130300125747,
+      "learning_rate": 7.707718720909308e-06,
+      "loss": 0.5679,
+      "step": 628
+    },
+    {
+      "epoch": 1.1597051597051597,
+      "grad_norm": 0.6479429762213357,
+      "learning_rate": 7.698686337413862e-06,
+      "loss": 0.5987,
+      "step": 629
+    },
+    {
+      "epoch": 1.1615479115479115,
+      "grad_norm": 0.6704328240426215,
+      "learning_rate": 7.689641509822879e-06,
+      "loss": 0.6136,
+      "step": 630
+    },
+    {
+      "epoch": 1.1633906633906634,
+      "grad_norm": 0.7557653479667956,
+      "learning_rate": 7.680584279843579e-06,
+      "loss": 0.5668,
+      "step": 631
+    },
+    {
+      "epoch": 1.1652334152334152,
+      "grad_norm": 0.7340272801504989,
+      "learning_rate": 7.671514689240366e-06,
+      "loss": 0.5821,
+      "step": 632
+    },
+    {
+      "epoch": 1.1670761670761671,
+      "grad_norm": 0.7281414615788497,
+      "learning_rate": 7.662432779834648e-06,
+      "loss": 0.6002,
+      "step": 633
+    },
+    {
+      "epoch": 1.1689189189189189,
+      "grad_norm": 0.8081042371855466,
+      "learning_rate": 7.653338593504632e-06,
+      "loss": 0.5945,
+      "step": 634
+    },
+    {
+      "epoch": 1.1707616707616708,
+      "grad_norm": 0.859314062758383,
+      "learning_rate": 7.644232172185133e-06,
+      "loss": 0.5823,
+      "step": 635
+    },
+    {
+      "epoch": 1.1726044226044225,
+      "grad_norm": 0.6789895374195197,
+      "learning_rate": 7.635113557867395e-06,
+      "loss": 0.5938,
+      "step": 636
+    },
+    {
+      "epoch": 1.1744471744471745,
+      "grad_norm": 0.8768994803839908,
+      "learning_rate": 7.625982792598874e-06,
+      "loss": 0.5817,
+      "step": 637
+    },
+    {
+      "epoch": 1.1762899262899262,
+      "grad_norm": 0.809744151641018,
+      "learning_rate": 7.616839918483061e-06,
+      "loss": 0.5875,
+      "step": 638
+    },
+    {
+      "epoch": 1.1781326781326782,
+      "grad_norm": 0.7374621318812015,
+      "learning_rate": 7.607684977679284e-06,
+      "loss": 0.5894,
+      "step": 639
+    },
+    {
+      "epoch": 1.17997542997543,
+      "grad_norm": 0.7619009223430059,
+      "learning_rate": 7.598518012402509e-06,
+      "loss": 0.6166,
+      "step": 640
+    },
+    {
+      "epoch": 1.1818181818181819,
+      "grad_norm": 0.7450642306551846,
+      "learning_rate": 7.5893390649231555e-06,
+      "loss": 0.5838,
+      "step": 641
+    },
+    {
+      "epoch": 1.1836609336609336,
+      "grad_norm": 0.7309501204730096,
+      "learning_rate": 7.580148177566886e-06,
+      "loss": 0.6018,
+      "step": 642
+    },
+    {
+      "epoch": 1.1855036855036856,
+      "grad_norm": 0.762104150936085,
+      "learning_rate": 7.5709453927144275e-06,
+      "loss": 0.6108,
+      "step": 643
+    },
+    {
+      "epoch": 1.1873464373464373,
+      "grad_norm": 0.7422010891344527,
+      "learning_rate": 7.561730752801364e-06,
+      "loss": 0.5831,
+      "step": 644
+    },
+    {
+      "epoch": 1.1891891891891893,
+      "grad_norm": 0.6896570290139655,
+      "learning_rate": 7.552504300317944e-06,
+      "loss": 0.5897,
+      "step": 645
+    },
+    {
+      "epoch": 1.191031941031941,
+      "grad_norm": 0.7186121582430534,
+      "learning_rate": 7.543266077808893e-06,
+      "loss": 0.5938,
+      "step": 646
+    },
+    {
+      "epoch": 1.192874692874693,
+      "grad_norm": 0.7665082858666719,
+      "learning_rate": 7.5340161278732e-06,
+      "loss": 0.5817,
+      "step": 647
+    },
+    {
+      "epoch": 1.1947174447174447,
+      "grad_norm": 0.7087566747464374,
+      "learning_rate": 7.524754493163939e-06,
+      "loss": 0.5852,
+      "step": 648
+    },
+    {
+      "epoch": 1.1965601965601966,
+      "grad_norm": 0.6498300070966749,
+      "learning_rate": 7.515481216388063e-06,
+      "loss": 0.5921,
+      "step": 649
+    },
+    {
+      "epoch": 1.1984029484029484,
+      "grad_norm": 0.6096417054527123,
+      "learning_rate": 7.506196340306204e-06,
+      "loss": 0.595,
+      "step": 650
+    },
+    {
+      "epoch": 1.2002457002457003,
+      "grad_norm": 0.707629935690763,
+      "learning_rate": 7.496899907732485e-06,
+      "loss": 0.5822,
+      "step": 651
+    },
+    {
+      "epoch": 1.202088452088452,
+      "grad_norm": 0.7445237063671669,
+      "learning_rate": 7.487591961534319e-06,
+      "loss": 0.5907,
+      "step": 652
+    },
+    {
+      "epoch": 1.203931203931204,
+      "grad_norm": 0.6136199061216261,
+      "learning_rate": 7.478272544632204e-06,
+      "loss": 0.5981,
+      "step": 653
+    },
+    {
+      "epoch": 1.2057739557739557,
+      "grad_norm": 0.7525220190785968,
+      "learning_rate": 7.468941699999535e-06,
+      "loss": 0.5729,
+      "step": 654
+    },
+    {
+      "epoch": 1.2076167076167077,
+      "grad_norm": 0.6754488330040109,
+      "learning_rate": 7.4595994706624065e-06,
+      "loss": 0.606,
+      "step": 655
+    },
+    {
+      "epoch": 1.2094594594594594,
+      "grad_norm": 0.7164041932079737,
+      "learning_rate": 7.450245899699401e-06,
+      "loss": 0.5811,
+      "step": 656
+    },
+    {
+      "epoch": 1.2113022113022114,
+      "grad_norm": 0.7083998124810403,
+      "learning_rate": 7.440881030241407e-06,
+      "loss": 0.6075,
+      "step": 657
+    },
+    {
+      "epoch": 1.2131449631449631,
+      "grad_norm": 0.771782398674858,
+      "learning_rate": 7.431504905471407e-06,
+      "loss": 0.5796,
+      "step": 658
+    },
+    {
+      "epoch": 1.214987714987715,
+      "grad_norm": 0.7298859220304117,
+      "learning_rate": 7.422117568624288e-06,
+      "loss": 0.5636,
+      "step": 659
+    },
+    {
+      "epoch": 1.2168304668304668,
+      "grad_norm": 0.8498778532851193,
+      "learning_rate": 7.412719062986632e-06,
+      "loss": 0.572,
+      "step": 660
+    },
+    {
+      "epoch": 1.2186732186732188,
+      "grad_norm": 0.6990543959610374,
+      "learning_rate": 7.403309431896528e-06,
+      "loss": 0.5881,
+      "step": 661
+    },
+    {
+      "epoch": 1.2205159705159705,
+      "grad_norm": 0.8529587651681191,
+      "learning_rate": 7.393888718743362e-06,
+      "loss": 0.5633,
+      "step": 662
+    },
+    {
+      "epoch": 1.2223587223587224,
+      "grad_norm": 0.6953247955555371,
+      "learning_rate": 7.384456966967624e-06,
+      "loss": 0.589,
+      "step": 663
+    },
+    {
+      "epoch": 1.2242014742014742,
+      "grad_norm": 0.7162481842235201,
+      "learning_rate": 7.375014220060705e-06,
+      "loss": 0.5737,
+      "step": 664
+    },
+    {
+      "epoch": 1.2260442260442261,
+      "grad_norm": 0.7726067957285083,
+      "learning_rate": 7.365560521564696e-06,
+      "loss": 0.6048,
+      "step": 665
+    },
+    {
+      "epoch": 1.2278869778869779,
+      "grad_norm": 0.6606102482189198,
+      "learning_rate": 7.3560959150721844e-06,
+      "loss": 0.592,
+      "step": 666
+    },
+    {
+      "epoch": 1.2297297297297298,
+      "grad_norm": 0.7684828715827158,
+      "learning_rate": 7.3466204442260605e-06,
+      "loss": 0.5862,
+      "step": 667
+    },
+    {
+      "epoch": 1.2315724815724816,
+      "grad_norm": 0.6708427152730285,
+      "learning_rate": 7.337134152719312e-06,
+      "loss": 0.5734,
+      "step": 668
+    },
+    {
+      "epoch": 1.2334152334152333,
+      "grad_norm": 0.7116323357176465,
+      "learning_rate": 7.327637084294818e-06,
+      "loss": 0.5881,
+      "step": 669
+    },
+    {
+      "epoch": 1.2352579852579852,
+      "grad_norm": 0.657461692482056,
+      "learning_rate": 7.318129282745152e-06,
+      "loss": 0.5758,
+      "step": 670
+    },
+    {
+      "epoch": 1.2371007371007372,
+      "grad_norm": 0.7282900652944116,
+      "learning_rate": 7.30861079191239e-06,
+      "loss": 0.5822,
+      "step": 671
+    },
+    {
+      "epoch": 1.238943488943489,
+      "grad_norm": 0.6778664827662574,
+      "learning_rate": 7.299081655687885e-06,
+      "loss": 0.5637,
+      "step": 672
+    },
+    {
+      "epoch": 1.2407862407862407,
+      "grad_norm": 0.6642424034365546,
+      "learning_rate": 7.2895419180120855e-06,
+      "loss": 0.6017,
+      "step": 673
+    },
+    {
+      "epoch": 1.2426289926289926,
+      "grad_norm": 0.8295412980014947,
+      "learning_rate": 7.279991622874319e-06,
+      "loss": 0.579,
+      "step": 674
+    },
+    {
+      "epoch": 1.2444717444717446,
+      "grad_norm": 0.7373158276797465,
+      "learning_rate": 7.2704308143126035e-06,
+      "loss": 0.5587,
+      "step": 675
+    },
+    {
+      "epoch": 1.2463144963144963,
+      "grad_norm": 0.6578441359489082,
+      "learning_rate": 7.260859536413429e-06,
+      "loss": 0.5849,
+      "step": 676
+    },
+    {
+      "epoch": 1.248157248157248,
+      "grad_norm": 0.7205641663176956,
+      "learning_rate": 7.251277833311565e-06,
+      "loss": 0.5774,
+      "step": 677
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.682369474009883,
+      "learning_rate": 7.241685749189851e-06,
+      "loss": 0.5812,
+      "step": 678
+    },
+    {
+      "epoch": 1.251842751842752,
+      "grad_norm": 0.7869564605836236,
+      "learning_rate": 7.2320833282789985e-06,
+      "loss": 0.5745,
+      "step": 679
+    },
+    {
+      "epoch": 1.2536855036855037,
+      "grad_norm": 0.7191245488199602,
+      "learning_rate": 7.22247061485738e-06,
+      "loss": 0.5876,
+      "step": 680
+    },
+    {
+      "epoch": 1.2555282555282554,
+      "grad_norm": 0.756694930937436,
+      "learning_rate": 7.212847653250828e-06,
+      "loss": 0.5845,
+      "step": 681
+    },
+    {
+      "epoch": 1.2573710073710074,
+      "grad_norm": 0.7026190506066463,
+      "learning_rate": 7.203214487832437e-06,
+      "loss": 0.5954,
+      "step": 682
+    },
+    {
+      "epoch": 1.2592137592137593,
+      "grad_norm": 0.7314954897246114,
+      "learning_rate": 7.193571163022348e-06,
+      "loss": 0.5879,
+      "step": 683
+    },
+    {
+      "epoch": 1.261056511056511,
+      "grad_norm": 0.7667065179071673,
+      "learning_rate": 7.18391772328755e-06,
+      "loss": 0.5844,
+      "step": 684
+    },
+    {
+      "epoch": 1.2628992628992628,
+      "grad_norm": 0.7698895195700798,
+      "learning_rate": 7.174254213141671e-06,
+      "loss": 0.5964,
+      "step": 685
+    },
+    {
+      "epoch": 1.2647420147420148,
+      "grad_norm": 0.7162170267715975,
+      "learning_rate": 7.164580677144781e-06,
+      "loss": 0.6089,
+      "step": 686
+    },
+    {
+      "epoch": 1.2665847665847667,
+      "grad_norm": 0.7908655756956552,
+      "learning_rate": 7.154897159903178e-06,
+      "loss": 0.5997,
+      "step": 687
+    },
+    {
+      "epoch": 1.2684275184275184,
+      "grad_norm": 0.759378654935679,
+      "learning_rate": 7.145203706069183e-06,
+      "loss": 0.5818,
+      "step": 688
+    },
+    {
+      "epoch": 1.2702702702702702,
+      "grad_norm": 0.7142032188493702,
+      "learning_rate": 7.135500360340937e-06,
+      "loss": 0.5726,
+      "step": 689
+    },
+    {
+      "epoch": 1.2721130221130221,
+      "grad_norm": 0.8624013521616145,
+      "learning_rate": 7.125787167462197e-06,
+      "loss": 0.5916,
+      "step": 690
+    },
+    {
+      "epoch": 1.273955773955774,
+      "grad_norm": 0.7076547720902582,
+      "learning_rate": 7.1160641722221255e-06,
+      "loss": 0.5609,
+      "step": 691
+    },
+    {
+      "epoch": 1.2757985257985258,
+      "grad_norm": 0.7202411254415638,
+      "learning_rate": 7.106331419455085e-06,
+      "loss": 0.591,
+      "step": 692
+    },
+    {
+      "epoch": 1.2776412776412776,
+      "grad_norm": 0.9058948059262203,
+      "learning_rate": 7.09658895404043e-06,
+      "loss": 0.5691,
+      "step": 693
+    },
+    {
+      "epoch": 1.2794840294840295,
+      "grad_norm": 0.7478171360763246,
+      "learning_rate": 7.086836820902305e-06,
+      "loss": 0.5923,
+      "step": 694
+    },
+    {
+      "epoch": 1.2813267813267815,
+      "grad_norm": 0.7606748930966204,
+      "learning_rate": 7.0770750650094335e-06,
+      "loss": 0.5746,
+      "step": 695
+    },
+    {
+      "epoch": 1.2831695331695332,
+      "grad_norm": 0.8167318259982754,
+      "learning_rate": 7.067303731374907e-06,
+      "loss": 0.5882,
+      "step": 696
+    },
+    {
+      "epoch": 1.285012285012285,
+      "grad_norm": 0.6974533093489234,
+      "learning_rate": 7.057522865055985e-06,
+      "loss": 0.5885,
+      "step": 697
+    },
+    {
+      "epoch": 1.2868550368550369,
+      "grad_norm": 0.7372445398582096,
+      "learning_rate": 7.047732511153885e-06,
+      "loss": 0.5891,
+      "step": 698
+    },
+    {
+      "epoch": 1.2886977886977886,
+      "grad_norm": 0.7577221629788058,
+      "learning_rate": 7.037932714813572e-06,
+      "loss": 0.5739,
+      "step": 699
+    },
+    {
+      "epoch": 1.2905405405405406,
+      "grad_norm": 0.7515259577033213,
+      "learning_rate": 7.028123521223548e-06,
+      "loss": 0.5939,
+      "step": 700
+    },
+    {
+      "epoch": 1.2923832923832923,
+      "grad_norm": 0.6962113297042494,
+      "learning_rate": 7.018304975615653e-06,
+      "loss": 0.5767,
+      "step": 701
+    },
+    {
+      "epoch": 1.2942260442260443,
+      "grad_norm": 0.6831604283229421,
+      "learning_rate": 7.008477123264849e-06,
+      "loss": 0.5785,
+      "step": 702
+    },
+    {
+      "epoch": 1.296068796068796,
+      "grad_norm": 0.7492555846668243,
+      "learning_rate": 6.998640009489009e-06,
+      "loss": 0.6156,
+      "step": 703
+    },
+    {
+      "epoch": 1.297911547911548,
+      "grad_norm": 0.6411727075977792,
+      "learning_rate": 6.98879367964872e-06,
+      "loss": 0.5759,
+      "step": 704
+    },
+    {
+      "epoch": 1.2997542997542997,
+      "grad_norm": 0.8679241340318938,
+      "learning_rate": 6.9789381791470575e-06,
+      "loss": 0.6188,
+      "step": 705
+    },
+    {
+      "epoch": 1.3015970515970516,
+      "grad_norm": 0.6265658540339792,
+      "learning_rate": 6.969073553429388e-06,
+      "loss": 0.574,
+      "step": 706
+    },
+    {
+      "epoch": 1.3034398034398034,
+      "grad_norm": 0.7788300277331245,
+      "learning_rate": 6.95919984798316e-06,
+      "loss": 0.5957,
+      "step": 707
+    },
+    {
+      "epoch": 1.3052825552825553,
+      "grad_norm": 0.7510584559977606,
+      "learning_rate": 6.949317108337681e-06,
+      "loss": 0.5839,
+      "step": 708
+    },
+    {
+      "epoch": 1.307125307125307,
+      "grad_norm": 0.729674723074692,
+      "learning_rate": 6.939425380063924e-06,
+      "loss": 0.5896,
+      "step": 709
+    },
+    {
+      "epoch": 1.308968058968059,
+      "grad_norm": 0.7955153607137859,
+      "learning_rate": 6.92952470877431e-06,
+      "loss": 0.5826,
+      "step": 710
+    },
+    {
+      "epoch": 1.3108108108108107,
+      "grad_norm": 0.684842631889432,
+      "learning_rate": 6.919615140122492e-06,
+      "loss": 0.5851,
+      "step": 711
+    },
+    {
+      "epoch": 1.3126535626535627,
+      "grad_norm": 0.7738436512462774,
+      "learning_rate": 6.909696719803156e-06,
+      "loss": 0.5767,
+      "step": 712
+    },
+    {
+      "epoch": 1.3144963144963144,
+      "grad_norm": 0.6404541329771748,
+      "learning_rate": 6.8997694935518e-06,
+      "loss": 0.5844,
+      "step": 713
+    },
+    {
+      "epoch": 1.3163390663390664,
+      "grad_norm": 0.7606980179005055,
+      "learning_rate": 6.889833507144534e-06,
+      "loss": 0.5859,
+      "step": 714
+    },
+    {
+      "epoch": 1.3181818181818181,
+      "grad_norm": 0.6899351180489574,
+      "learning_rate": 6.879888806397855e-06,
+      "loss": 0.5995,
+      "step": 715
+    },
+    {
+      "epoch": 1.32002457002457,
+      "grad_norm": 0.6344575142665548,
+      "learning_rate": 6.869935437168449e-06,
+      "loss": 0.584,
+      "step": 716
+    },
+    {
+      "epoch": 1.3218673218673218,
+      "grad_norm": 0.7040037933261688,
+      "learning_rate": 6.85997344535297e-06,
+      "loss": 0.5929,
+      "step": 717
+    },
+    {
+      "epoch": 1.3237100737100738,
+      "grad_norm": 0.6951059936475811,
+      "learning_rate": 6.850002876887836e-06,
+      "loss": 0.5608,
+      "step": 718
+    },
+    {
+      "epoch": 1.3255528255528255,
+      "grad_norm": 0.7169749207461239,
+      "learning_rate": 6.840023777749008e-06,
+      "loss": 0.5942,
+      "step": 719
+    },
+    {
+      "epoch": 1.3273955773955775,
+      "grad_norm": 0.644583241280748,
+      "learning_rate": 6.83003619395179e-06,
+      "loss": 0.5931,
+      "step": 720
+    },
+    {
+      "epoch": 1.3292383292383292,
+      "grad_norm": 0.6451801287732833,
+      "learning_rate": 6.8200401715506045e-06,
+      "loss": 0.5894,
+      "step": 721
+    },
+    {
+      "epoch": 1.3310810810810811,
+      "grad_norm": 0.6859015630335887,
+      "learning_rate": 6.81003575663879e-06,
+      "loss": 0.5649,
+      "step": 722
+    },
+    {
+      "epoch": 1.3329238329238329,
+      "grad_norm": 0.6801293171698467,
+      "learning_rate": 6.800022995348381e-06,
+      "loss": 0.5843,
+      "step": 723
+    },
+    {
+      "epoch": 1.3347665847665848,
+      "grad_norm": 0.6766891059391945,
+      "learning_rate": 6.7900019338499005e-06,
+      "loss": 0.5577,
+      "step": 724
+    },
+    {
+      "epoch": 1.3366093366093366,
+      "grad_norm": 0.5958237951571397,
+      "learning_rate": 6.779972618352143e-06,
+      "loss": 0.5871,
+      "step": 725
+    },
+    {
+      "epoch": 1.3384520884520885,
+      "grad_norm": 0.6711827535540938,
+      "learning_rate": 6.7699350951019685e-06,
+      "loss": 0.5746,
+      "step": 726
+    },
+    {
+      "epoch": 1.3402948402948403,
+      "grad_norm": 0.7118957126980248,
+      "learning_rate": 6.759889410384078e-06,
+      "loss": 0.608,
+      "step": 727
+    },
+    {
+      "epoch": 1.3421375921375922,
+      "grad_norm": 0.6990802422223146,
+      "learning_rate": 6.74983561052081e-06,
+      "loss": 0.5914,
+      "step": 728
+    },
+    {
+      "epoch": 1.343980343980344,
+      "grad_norm": 0.6906329293293347,
+      "learning_rate": 6.739773741871924e-06,
+      "loss": 0.5971,
+      "step": 729
+    },
+    {
+      "epoch": 1.345823095823096,
+      "grad_norm": 0.672320738875141,
+      "learning_rate": 6.729703850834381e-06,
+      "loss": 0.5789,
+      "step": 730
+    },
+    {
+      "epoch": 1.3476658476658476,
+      "grad_norm": 0.7124716264156697,
+      "learning_rate": 6.719625983842144e-06,
+      "loss": 0.578,
+      "step": 731
+    },
+    {
+      "epoch": 1.3495085995085996,
+      "grad_norm": 0.6486728828288537,
+      "learning_rate": 6.7095401873659446e-06,
+      "loss": 0.566,
+      "step": 732
+    },
+    {
+      "epoch": 1.3513513513513513,
+      "grad_norm": 0.7087663086859939,
+      "learning_rate": 6.699446507913083e-06,
+      "loss": 0.5889,
+      "step": 733
+    },
+    {
+      "epoch": 1.3531941031941033,
+      "grad_norm": 0.6756856448868066,
+      "learning_rate": 6.689344992027213e-06,
+      "loss": 0.5917,
+      "step": 734
+    },
+    {
+      "epoch": 1.355036855036855,
+      "grad_norm": 0.6292880111415169,
+      "learning_rate": 6.6792356862881144e-06,
+      "loss": 0.6071,
+      "step": 735
+    },
+    {
+      "epoch": 1.356879606879607,
+      "grad_norm": 0.7511619756093283,
+      "learning_rate": 6.669118637311491e-06,
+      "loss": 0.5652,
+      "step": 736
+    },
+    {
+      "epoch": 1.3587223587223587,
+      "grad_norm": 0.7461040416296377,
+      "learning_rate": 6.65899389174876e-06,
+      "loss": 0.5871,
+      "step": 737
+    },
+    {
+      "epoch": 1.3605651105651106,
+      "grad_norm": 0.8618483377640704,
+      "learning_rate": 6.648861496286818e-06,
+      "loss": 0.5975,
+      "step": 738
+    },
+    {
+      "epoch": 1.3624078624078624,
+      "grad_norm": 0.6772109776980443,
+      "learning_rate": 6.63872149764784e-06,
+      "loss": 0.5676,
+      "step": 739
+    },
+    {
+      "epoch": 1.3642506142506141,
+      "grad_norm": 0.7582253047943485,
+      "learning_rate": 6.628573942589062e-06,
+      "loss": 0.5836,
+      "step": 740
+    },
+    {
+      "epoch": 1.366093366093366,
+      "grad_norm": 0.7001023519740849,
+      "learning_rate": 6.618418877902563e-06,
+      "loss": 0.5755,
+      "step": 741
+    },
+    {
+      "epoch": 1.367936117936118,
+      "grad_norm": 0.7720183245041078,
+      "learning_rate": 6.608256350415052e-06,
+      "loss": 0.5762,
+      "step": 742
+    },
+    {
+      "epoch": 1.3697788697788698,
+      "grad_norm": 0.6854391608921228,
+      "learning_rate": 6.598086406987647e-06,
+      "loss": 0.5866,
+      "step": 743
+    },
+    {
+      "epoch": 1.3716216216216215,
+      "grad_norm": 0.7759907535771228,
+      "learning_rate": 6.587909094515663e-06,
+      "loss": 0.5848,
+      "step": 744
+    },
+    {
+      "epoch": 1.3734643734643734,
+      "grad_norm": 0.6648421640807161,
+      "learning_rate": 6.577724459928396e-06,
+      "loss": 0.5762,
+      "step": 745
+    },
+    {
+      "epoch": 1.3753071253071254,
+      "grad_norm": 0.7499067451731978,
+      "learning_rate": 6.567532550188908e-06,
+      "loss": 0.5843,
+      "step": 746
+    },
+    {
+      "epoch": 1.3771498771498771,
+      "grad_norm": 0.6787942225277,
+      "learning_rate": 6.5573334122938e-06,
+      "loss": 0.5807,
+      "step": 747
+    },
+    {
+      "epoch": 1.3789926289926289,
+      "grad_norm": 0.6842755174707729,
+      "learning_rate": 6.547127093273009e-06,
+      "loss": 0.5699,
+      "step": 748
+    },
+    {
+      "epoch": 1.3808353808353808,
+      "grad_norm": 0.6202134616428023,
+      "learning_rate": 6.536913640189588e-06,
+      "loss": 0.5792,
+      "step": 749
+    },
+    {
+      "epoch": 1.3826781326781328,
+      "grad_norm": 0.6542547821911993,
+      "learning_rate": 6.526693100139481e-06,
+      "loss": 0.5923,
+      "step": 750
+    },
+    {
+      "epoch": 1.3845208845208845,
+      "grad_norm": 0.6400183581506397,
+      "learning_rate": 6.5164655202513135e-06,
+      "loss": 0.5923,
+      "step": 751
+    },
+    {
+      "epoch": 1.3863636363636362,
+      "grad_norm": 0.6864775428675023,
+      "learning_rate": 6.5062309476861714e-06,
+      "loss": 0.5982,
+      "step": 752
+    },
+    {
+      "epoch": 1.3882063882063882,
+      "grad_norm": 0.6815538958226465,
+      "learning_rate": 6.495989429637388e-06,
+      "loss": 0.582,
+      "step": 753
+    },
+    {
+      "epoch": 1.3900491400491402,
+      "grad_norm": 0.6369454996129288,
+      "learning_rate": 6.485741013330321e-06,
+      "loss": 0.5825,
+      "step": 754
+    },
+    {
+      "epoch": 1.3918918918918919,
+      "grad_norm": 0.6322598020673428,
+      "learning_rate": 6.475485746022136e-06,
+      "loss": 0.5668,
+      "step": 755
+    },
+    {
+      "epoch": 1.3937346437346436,
+      "grad_norm": 0.7029350343158693,
+      "learning_rate": 6.465223675001593e-06,
+      "loss": 0.5944,
+      "step": 756
+    },
+    {
+      "epoch": 1.3955773955773956,
+      "grad_norm": 0.7088574774017515,
+      "learning_rate": 6.454954847588824e-06,
+      "loss": 0.5751,
+      "step": 757
+    },
+    {
+      "epoch": 1.3974201474201475,
+      "grad_norm": 0.6680915404265156,
+      "learning_rate": 6.444679311135112e-06,
+      "loss": 0.5844,
+      "step": 758
+    },
+    {
+      "epoch": 1.3992628992628993,
+      "grad_norm": 0.7095108532836971,
+      "learning_rate": 6.434397113022683e-06,
+      "loss": 0.5902,
+      "step": 759
+    },
+    {
+      "epoch": 1.401105651105651,
+      "grad_norm": 0.714642320553394,
+      "learning_rate": 6.424108300664479e-06,
+      "loss": 0.586,
+      "step": 760
+    },
+    {
+      "epoch": 1.402948402948403,
+      "grad_norm": 0.6920271232939866,
+      "learning_rate": 6.4138129215039356e-06,
+      "loss": 0.572,
+      "step": 761
+    },
+    {
+      "epoch": 1.404791154791155,
+      "grad_norm": 0.6950425391357339,
+      "learning_rate": 6.403511023014778e-06,
+      "loss": 0.5785,
+      "step": 762
+    },
+    {
+      "epoch": 1.4066339066339066,
+      "grad_norm": 0.6909676822018072,
+      "learning_rate": 6.393202652700785e-06,
+      "loss": 0.5831,
+      "step": 763
+    },
+    {
+      "epoch": 1.4084766584766584,
+      "grad_norm": 0.7580948312242047,
+      "learning_rate": 6.382887858095585e-06,
+      "loss": 0.5672,
+      "step": 764
+    },
+    {
+      "epoch": 1.4103194103194103,
+      "grad_norm": 0.6632369481109529,
+      "learning_rate": 6.372566686762427e-06,
+      "loss": 0.5883,
+      "step": 765
+    },
+    {
+      "epoch": 1.4121621621621623,
+      "grad_norm": 0.7210928746751901,
+      "learning_rate": 6.3622391862939606e-06,
+      "loss": 0.5655,
+      "step": 766
+    },
+    {
+      "epoch": 1.414004914004914,
+      "grad_norm": 0.709800113717765,
+      "learning_rate": 6.351905404312026e-06,
+      "loss": 0.5942,
+      "step": 767
+    },
+    {
+      "epoch": 1.4158476658476657,
+      "grad_norm": 0.6919402408558638,
+      "learning_rate": 6.341565388467425e-06,
+      "loss": 0.586,
+      "step": 768
+    },
+    {
+      "epoch": 1.4176904176904177,
+      "grad_norm": 0.6904703182446977,
+      "learning_rate": 6.3312191864397045e-06,
+      "loss": 0.5952,
+      "step": 769
+    },
+    {
+      "epoch": 1.4195331695331697,
+      "grad_norm": 0.684741561088499,
+      "learning_rate": 6.320866845936942e-06,
+      "loss": 0.5909,
+      "step": 770
+    },
+    {
+      "epoch": 1.4213759213759214,
+      "grad_norm": 0.7122396738129608,
+      "learning_rate": 6.310508414695511e-06,
+      "loss": 0.5964,
+      "step": 771
+    },
+    {
+      "epoch": 1.4232186732186731,
+      "grad_norm": 0.7932775122176792,
+      "learning_rate": 6.300143940479881e-06,
+      "loss": 0.5599,
+      "step": 772
+    },
+    {
+      "epoch": 1.425061425061425,
+      "grad_norm": 0.6712170997768473,
+      "learning_rate": 6.289773471082381e-06,
+      "loss": 0.5815,
+      "step": 773
+    },
+    {
+      "epoch": 1.426904176904177,
+      "grad_norm": 0.7946225600459027,
+      "learning_rate": 6.279397054322983e-06,
+      "loss": 0.5817,
+      "step": 774
+    },
+    {
+      "epoch": 1.4287469287469288,
+      "grad_norm": 0.8909755581923532,
+      "learning_rate": 6.269014738049088e-06,
+      "loss": 0.5848,
+      "step": 775
+    },
+    {
+      "epoch": 1.4305896805896805,
+      "grad_norm": 0.7318436991678973,
+      "learning_rate": 6.2586265701352976e-06,
+      "loss": 0.5862,
+      "step": 776
+    },
+    {
+      "epoch": 1.4324324324324325,
+      "grad_norm": 0.8293893955559066,
+      "learning_rate": 6.248232598483196e-06,
+      "loss": 0.5845,
+      "step": 777
+    },
+    {
+      "epoch": 1.4342751842751844,
+      "grad_norm": 0.976129444996388,
+      "learning_rate": 6.2378328710211345e-06,
+      "loss": 0.5837,
+      "step": 778
+    },
+    {
+      "epoch": 1.4361179361179361,
+      "grad_norm": 0.7385990873516664,
+      "learning_rate": 6.227427435703997e-06,
+      "loss": 0.5887,
+      "step": 779
+    },
+    {
+      "epoch": 1.4379606879606879,
+      "grad_norm": 0.8264368704045605,
+      "learning_rate": 6.217016340512993e-06,
+      "loss": 0.5909,
+      "step": 780
+    },
+    {
+      "epoch": 1.4398034398034398,
+      "grad_norm": 0.8966607196489723,
+      "learning_rate": 6.2065996334554325e-06,
+      "loss": 0.5854,
+      "step": 781
+    },
+    {
+      "epoch": 1.4416461916461916,
+      "grad_norm": 0.8715559949223112,
+      "learning_rate": 6.196177362564495e-06,
+      "loss": 0.5981,
+      "step": 782
+    },
+    {
+      "epoch": 1.4434889434889435,
+      "grad_norm": 0.7224758520159814,
+      "learning_rate": 6.185749575899023e-06,
+      "loss": 0.5706,
+      "step": 783
+    },
+    {
+      "epoch": 1.4453316953316953,
+      "grad_norm": 1.2425969424060408,
+      "learning_rate": 6.175316321543291e-06,
+      "loss": 0.5998,
+      "step": 784
+    },
+    {
+      "epoch": 1.4471744471744472,
+      "grad_norm": 1.0783787125603415,
+      "learning_rate": 6.164877647606784e-06,
+      "loss": 0.5615,
+      "step": 785
+    },
+    {
+      "epoch": 1.449017199017199,
+      "grad_norm": 0.7086876646634345,
+      "learning_rate": 6.154433602223979e-06,
+      "loss": 0.5802,
+      "step": 786
+    },
+    {
+      "epoch": 1.450859950859951,
+      "grad_norm": 0.8734024567896126,
+      "learning_rate": 6.143984233554125e-06,
+      "loss": 0.5838,
+      "step": 787
+    },
+    {
+      "epoch": 1.4527027027027026,
+      "grad_norm": 0.830451535389402,
+      "learning_rate": 6.13352958978101e-06,
+      "loss": 0.579,
+      "step": 788
+    },
+    {
+      "epoch": 1.4545454545454546,
+      "grad_norm": 0.7350857239028357,
+      "learning_rate": 6.123069719112754e-06,
+      "loss": 0.5609,
+      "step": 789
+    },
+    {
+      "epoch": 1.4563882063882063,
+      "grad_norm": 0.7693220616269851,
+      "learning_rate": 6.112604669781572e-06,
+      "loss": 0.5952,
+      "step": 790
+    },
+    {
+      "epoch": 1.4582309582309583,
+      "grad_norm": 0.9285086160309332,
+      "learning_rate": 6.102134490043566e-06,
+      "loss": 0.5535,
+      "step": 791
+    },
+    {
+      "epoch": 1.46007371007371,
+      "grad_norm": 0.6624485518707897,
+      "learning_rate": 6.091659228178492e-06,
+      "loss": 0.5737,
+      "step": 792
+    },
+    {
+      "epoch": 1.461916461916462,
+      "grad_norm": 0.8066502427277187,
+      "learning_rate": 6.0811789324895365e-06,
+      "loss": 0.5871,
+      "step": 793
+    },
+    {
+      "epoch": 1.4637592137592137,
+      "grad_norm": 0.70832317660221,
+      "learning_rate": 6.070693651303105e-06,
+      "loss": 0.5846,
+      "step": 794
+    },
+    {
+      "epoch": 1.4656019656019657,
+      "grad_norm": 0.6732795988815229,
+      "learning_rate": 6.060203432968587e-06,
+      "loss": 0.5855,
+      "step": 795
+    },
+    {
+      "epoch": 1.4674447174447174,
+      "grad_norm": 0.6905870509835268,
+      "learning_rate": 6.049708325858139e-06,
+      "loss": 0.5789,
+      "step": 796
+    },
+    {
+      "epoch": 1.4692874692874693,
+      "grad_norm": 0.7529264184619736,
+      "learning_rate": 6.039208378366462e-06,
+      "loss": 0.5858,
+      "step": 797
+    },
+    {
+      "epoch": 1.471130221130221,
+      "grad_norm": 0.6703277466652326,
+      "learning_rate": 6.028703638910574e-06,
+      "loss": 0.5909,
+      "step": 798
+    },
+    {
+      "epoch": 1.472972972972973,
+      "grad_norm": 0.7469768331120241,
+      "learning_rate": 6.018194155929594e-06,
+      "loss": 0.5807,
+      "step": 799
+    },
+    {
+      "epoch": 1.4748157248157248,
+      "grad_norm": 0.7592030533290773,
+      "learning_rate": 6.0076799778845105e-06,
+      "loss": 0.5871,
+      "step": 800
+    },
+    {
+      "epoch": 1.4766584766584767,
+      "grad_norm": 0.6708929773564863,
+      "learning_rate": 5.997161153257963e-06,
+      "loss": 0.6023,
+      "step": 801
+    },
+    {
+      "epoch": 1.4785012285012284,
+      "grad_norm": 0.7209080458763151,
+      "learning_rate": 5.986637730554014e-06,
+      "loss": 0.5966,
+      "step": 802
+    },
+    {
+      "epoch": 1.4803439803439804,
+      "grad_norm": 0.717575132158687,
+      "learning_rate": 5.976109758297934e-06,
+      "loss": 0.5842,
+      "step": 803
+    },
+    {
+      "epoch": 1.4821867321867321,
+      "grad_norm": 0.6342098795349664,
+      "learning_rate": 5.96557728503597e-06,
+      "loss": 0.5805,
+      "step": 804
+    },
+    {
+      "epoch": 1.484029484029484,
+      "grad_norm": 0.8076408317819502,
+      "learning_rate": 5.955040359335123e-06,
+      "loss": 0.5693,
+      "step": 805
+    },
+    {
+      "epoch": 1.4858722358722358,
+      "grad_norm": 0.6833830509145711,
+      "learning_rate": 5.944499029782924e-06,
+      "loss": 0.56,
+      "step": 806
+    },
+    {
+      "epoch": 1.4877149877149878,
+      "grad_norm": 0.7228725600540951,
+      "learning_rate": 5.933953344987215e-06,
+      "loss": 0.5803,
+      "step": 807
+    },
+    {
+      "epoch": 1.4895577395577395,
+      "grad_norm": 0.8524275811738037,
+      "learning_rate": 5.923403353575916e-06,
+      "loss": 0.5788,
+      "step": 808
+    },
+    {
+      "epoch": 1.4914004914004915,
+      "grad_norm": 0.6887209625606867,
+      "learning_rate": 5.91284910419681e-06,
+      "loss": 0.5755,
+      "step": 809
+    },
+    {
+      "epoch": 1.4932432432432432,
+      "grad_norm": 0.6640601591239226,
+      "learning_rate": 5.902290645517308e-06,
+      "loss": 0.5673,
+      "step": 810
+    },
+    {
+      "epoch": 1.4950859950859952,
+      "grad_norm": 0.7678119416122671,
+      "learning_rate": 5.891728026224238e-06,
+      "loss": 0.5821,
+      "step": 811
+    },
+    {
+      "epoch": 1.496928746928747,
+      "grad_norm": 0.6164640514943673,
+      "learning_rate": 5.88116129502361e-06,
+      "loss": 0.591,
+      "step": 812
+    },
+    {
+      "epoch": 1.4987714987714988,
+      "grad_norm": 0.7317848166836791,
+      "learning_rate": 5.870590500640392e-06,
+      "loss": 0.5825,
+      "step": 813
+    },
+    {
+      "epoch": 1.5006142506142506,
+      "grad_norm": 0.6611720486910339,
+      "learning_rate": 5.860015691818292e-06,
+      "loss": 0.5575,
+      "step": 814
+    },
+    {
+      "epoch": 1.5024570024570023,
+      "grad_norm": 0.7924868715028064,
+      "learning_rate": 5.84943691731953e-06,
+      "loss": 0.6036,
+      "step": 815
+    },
+    {
+      "epoch": 1.5042997542997543,
+      "grad_norm": 0.6560447727716534,
+      "learning_rate": 5.838854225924607e-06,
+      "loss": 0.5708,
+      "step": 816
+    },
+    {
+      "epoch": 1.5061425061425062,
+      "grad_norm": 0.6786202715711611,
+      "learning_rate": 5.82826766643209e-06,
+      "loss": 0.5793,
+      "step": 817
+    },
+    {
+      "epoch": 1.507985257985258,
+      "grad_norm": 0.6665658840065142,
+      "learning_rate": 5.817677287658382e-06,
+      "loss": 0.5838,
+      "step": 818
+    },
+    {
+      "epoch": 1.5098280098280097,
+      "grad_norm": 0.6855138549349598,
+      "learning_rate": 5.807083138437497e-06,
+      "loss": 0.5726,
+      "step": 819
+    },
+    {
+      "epoch": 1.5116707616707616,
+      "grad_norm": 0.7981302622883958,
+      "learning_rate": 5.796485267620834e-06,
+      "loss": 0.5701,
+      "step": 820
+    },
+    {
+      "epoch": 1.5135135135135136,
+      "grad_norm": 0.6425742078712001,
+      "learning_rate": 5.78588372407695e-06,
+      "loss": 0.6104,
+      "step": 821
+    },
+    {
+      "epoch": 1.5153562653562653,
+      "grad_norm": 0.706245985754697,
+      "learning_rate": 5.7752785566913484e-06,
+      "loss": 0.585,
+      "step": 822
+    },
+    {
+      "epoch": 1.517199017199017,
+      "grad_norm": 0.7505765911219308,
+      "learning_rate": 5.764669814366231e-06,
+      "loss": 0.5861,
+      "step": 823
+    },
+    {
+      "epoch": 1.519041769041769,
+      "grad_norm": 0.6243965189744589,
+      "learning_rate": 5.754057546020289e-06,
+      "loss": 0.5946,
+      "step": 824
+    },
+    {
+      "epoch": 1.520884520884521,
+      "grad_norm": 0.7319415910420785,
+      "learning_rate": 5.743441800588472e-06,
+      "loss": 0.5646,
+      "step": 825
+    },
+    {
+      "epoch": 1.5227272727272727,
+      "grad_norm": 0.6693155357160733,
+      "learning_rate": 5.732822627021765e-06,
+      "loss": 0.6035,
+      "step": 826
+    },
+    {
+      "epoch": 1.5245700245700244,
+      "grad_norm": 0.7069871224647403,
+      "learning_rate": 5.722200074286959e-06,
+      "loss": 0.5711,
+      "step": 827
+    },
+    {
+      "epoch": 1.5264127764127764,
+      "grad_norm": 0.6272699366845061,
+      "learning_rate": 5.711574191366427e-06,
+      "loss": 0.5803,
+      "step": 828
+    },
+    {
+      "epoch": 1.5282555282555284,
+      "grad_norm": 0.6793133850639186,
+      "learning_rate": 5.700945027257896e-06,
+      "loss": 0.5778,
+      "step": 829
+    },
+    {
+      "epoch": 1.53009828009828,
+      "grad_norm": 0.6651555339869941,
+      "learning_rate": 5.690312630974229e-06,
+      "loss": 0.5831,
+      "step": 830
+    },
+    {
+      "epoch": 1.5319410319410318,
+      "grad_norm": 0.6368846727367478,
+      "learning_rate": 5.6796770515431886e-06,
+      "loss": 0.5893,
+      "step": 831
+    },
+    {
+      "epoch": 1.5337837837837838,
+      "grad_norm": 0.6350329440117563,
+      "learning_rate": 5.6690383380072136e-06,
+      "loss": 0.587,
+      "step": 832
+    },
+    {
+      "epoch": 1.5356265356265357,
+      "grad_norm": 0.6412376662341064,
+      "learning_rate": 5.658396539423197e-06,
+      "loss": 0.5605,
+      "step": 833
+    },
+    {
+      "epoch": 1.5374692874692875,
+      "grad_norm": 0.636369662332138,
+      "learning_rate": 5.647751704862263e-06,
+      "loss": 0.556,
+      "step": 834
+    },
+    {
+      "epoch": 1.5393120393120392,
+      "grad_norm": 0.7186072521354805,
+      "learning_rate": 5.637103883409525e-06,
+      "loss": 0.588,
+      "step": 835
+    },
+    {
+      "epoch": 1.5411547911547911,
+      "grad_norm": 0.6365846282742865,
+      "learning_rate": 5.626453124163876e-06,
+      "loss": 0.6009,
+      "step": 836
+    },
+    {
+      "epoch": 1.542997542997543,
+      "grad_norm": 0.6196504713747927,
+      "learning_rate": 5.615799476237752e-06,
+      "loss": 0.5883,
+      "step": 837
+    },
+    {
+      "epoch": 1.5448402948402948,
+      "grad_norm": 0.7404649709741298,
+      "learning_rate": 5.605142988756917e-06,
+      "loss": 0.5804,
+      "step": 838
+    },
+    {
+      "epoch": 1.5466830466830466,
+      "grad_norm": 0.6170048737075667,
+      "learning_rate": 5.594483710860217e-06,
+      "loss": 0.5946,
+      "step": 839
+    },
+    {
+      "epoch": 1.5485257985257985,
+      "grad_norm": 0.651209039766793,
+      "learning_rate": 5.583821691699373e-06,
+      "loss": 0.5996,
+      "step": 840
+    },
+    {
+      "epoch": 1.5503685503685505,
+      "grad_norm": 0.6675614524873646,
+      "learning_rate": 5.573156980438743e-06,
+      "loss": 0.5973,
+      "step": 841
+    },
+    {
+      "epoch": 1.5522113022113022,
+      "grad_norm": 0.6595460960460664,
+      "learning_rate": 5.562489626255104e-06,
+      "loss": 0.5724,
+      "step": 842
+    },
+    {
+      "epoch": 1.554054054054054,
+      "grad_norm": 0.6628234955679191,
+      "learning_rate": 5.5518196783374126e-06,
+      "loss": 0.5854,
+      "step": 843
+    },
+    {
+      "epoch": 1.555896805896806,
+      "grad_norm": 0.6629686530324079,
+      "learning_rate": 5.541147185886591e-06,
+      "loss": 0.5953,
+      "step": 844
+    },
+    {
+      "epoch": 1.5577395577395579,
+      "grad_norm": 0.6241562841465624,
+      "learning_rate": 5.530472198115291e-06,
+      "loss": 0.572,
+      "step": 845
+    },
+    {
+      "epoch": 1.5595823095823096,
+      "grad_norm": 0.6643560690544034,
+      "learning_rate": 5.519794764247673e-06,
+      "loss": 0.5938,
+      "step": 846
+    },
+    {
+      "epoch": 1.5614250614250613,
+      "grad_norm": 0.681768626365906,
+      "learning_rate": 5.509114933519179e-06,
+      "loss": 0.5928,
+      "step": 847
+    },
+    {
+      "epoch": 1.5632678132678133,
+      "grad_norm": 0.6216547936607765,
+      "learning_rate": 5.498432755176295e-06,
+      "loss": 0.5926,
+      "step": 848
+    },
+    {
+      "epoch": 1.5651105651105652,
+      "grad_norm": 0.6534708138171155,
+      "learning_rate": 5.487748278476342e-06,
+      "loss": 0.5897,
+      "step": 849
+    },
+    {
+      "epoch": 1.566953316953317,
+      "grad_norm": 0.6399046253519763,
+      "learning_rate": 5.477061552687234e-06,
+      "loss": 0.5874,
+      "step": 850
+    },
+    {
+      "epoch": 1.5687960687960687,
+      "grad_norm": 0.6402819154027296,
+      "learning_rate": 5.466372627087256e-06,
+      "loss": 0.59,
+      "step": 851
+    },
+    {
+      "epoch": 1.5706388206388207,
+      "grad_norm": 0.6515559492607708,
+      "learning_rate": 5.455681550964839e-06,
+      "loss": 0.5839,
+      "step": 852
+    },
+    {
+      "epoch": 1.5724815724815726,
+      "grad_norm": 0.7209106441986273,
+      "learning_rate": 5.4449883736183286e-06,
+      "loss": 0.5709,
+      "step": 853
+    },
+    {
+      "epoch": 1.5743243243243243,
+      "grad_norm": 0.6362720942009615,
+      "learning_rate": 5.4342931443557576e-06,
+      "loss": 0.5871,
+      "step": 854
+    },
+    {
+      "epoch": 1.576167076167076,
+      "grad_norm": 0.6326057201947428,
+      "learning_rate": 5.423595912494626e-06,
+      "loss": 0.5776,
+      "step": 855
+    },
+    {
+      "epoch": 1.578009828009828,
+      "grad_norm": 0.6484962177927928,
+      "learning_rate": 5.412896727361663e-06,
+      "loss": 0.5887,
+      "step": 856
+    },
+    {
+      "epoch": 1.57985257985258,
+      "grad_norm": 0.6446672965334984,
+      "learning_rate": 5.402195638292605e-06,
+      "loss": 0.5816,
+      "step": 857
+    },
+    {
+      "epoch": 1.5816953316953317,
+      "grad_norm": 0.7081991495637252,
+      "learning_rate": 5.391492694631975e-06,
+      "loss": 0.5875,
+      "step": 858
+    },
+    {
+      "epoch": 1.5835380835380835,
+      "grad_norm": 0.6132397841354326,
+      "learning_rate": 5.380787945732835e-06,
+      "loss": 0.585,
+      "step": 859
+    },
+    {
+      "epoch": 1.5853808353808354,
+      "grad_norm": 0.6139618909567908,
+      "learning_rate": 5.370081440956582e-06,
+      "loss": 0.566,
+      "step": 860
+    },
+    {
+      "epoch": 1.5872235872235874,
+      "grad_norm": 0.6561112855140604,
+      "learning_rate": 5.359373229672708e-06,
+      "loss": 0.5811,
+      "step": 861
+    },
+    {
+      "epoch": 1.589066339066339,
+      "grad_norm": 0.6032733826245348,
+      "learning_rate": 5.348663361258568e-06,
+      "loss": 0.5946,
+      "step": 862
+    },
+    {
+      "epoch": 1.5909090909090908,
+      "grad_norm": 0.638352661722549,
+      "learning_rate": 5.337951885099167e-06,
+      "loss": 0.5731,
+      "step": 863
+    },
+    {
+      "epoch": 1.5927518427518428,
+      "grad_norm": 0.7132015345485251,
+      "learning_rate": 5.327238850586912e-06,
+      "loss": 0.5758,
+      "step": 864
+    },
+    {
+      "epoch": 1.5945945945945947,
+      "grad_norm": 0.7634231115664125,
+      "learning_rate": 5.316524307121409e-06,
+      "loss": 0.5675,
+      "step": 865
+    },
+    {
+      "epoch": 1.5964373464373465,
+      "grad_norm": 0.6346796439103365,
+      "learning_rate": 5.3058083041092145e-06,
+      "loss": 0.5739,
+      "step": 866
+    },
+    {
+      "epoch": 1.5982800982800982,
+      "grad_norm": 0.6712870224541247,
+      "learning_rate": 5.2950908909636144e-06,
+      "loss": 0.5788,
+      "step": 867
+    },
+    {
+      "epoch": 1.6001228501228502,
+      "grad_norm": 0.6434198861436975,
+      "learning_rate": 5.2843721171044e-06,
+      "loss": 0.5518,
+      "step": 868
+    },
+    {
+      "epoch": 1.6019656019656021,
+      "grad_norm": 0.6599169839940391,
+      "learning_rate": 5.273652031957639e-06,
+      "loss": 0.5964,
+      "step": 869
+    },
+    {
+      "epoch": 1.6038083538083538,
+      "grad_norm": 0.6577728377163973,
+      "learning_rate": 5.262930684955439e-06,
+      "loss": 0.5731,
+      "step": 870
+    },
+    {
+      "epoch": 1.6056511056511056,
+      "grad_norm": 0.6319469409991316,
+      "learning_rate": 5.252208125535731e-06,
+      "loss": 0.5863,
+      "step": 871
+    },
+    {
+      "epoch": 1.6074938574938575,
+      "grad_norm": 0.6115300856649373,
+      "learning_rate": 5.241484403142036e-06,
+      "loss": 0.582,
+      "step": 872
+    },
+    {
+      "epoch": 1.6093366093366095,
+      "grad_norm": 0.6322024113402958,
+      "learning_rate": 5.2307595672232384e-06,
+      "loss": 0.6077,
+      "step": 873
+    },
+    {
+      "epoch": 1.6111793611793612,
+      "grad_norm": 0.6427328448424448,
+      "learning_rate": 5.220033667233356e-06,
+      "loss": 0.5898,
+      "step": 874
+    },
+    {
+      "epoch": 1.613022113022113,
+      "grad_norm": 0.6406167958716241,
+      "learning_rate": 5.209306752631312e-06,
+      "loss": 0.5845,
+      "step": 875
+    },
+    {
+      "epoch": 1.614864864864865,
+      "grad_norm": 0.6073567795352578,
+      "learning_rate": 5.198578872880709e-06,
+      "loss": 0.6023,
+      "step": 876
+    },
+    {
+      "epoch": 1.6167076167076169,
+      "grad_norm": 0.6679097013776044,
+      "learning_rate": 5.187850077449604e-06,
+      "loss": 0.5857,
+      "step": 877
+    },
+    {
+      "epoch": 1.6185503685503686,
+      "grad_norm": 0.6580449701819567,
+      "learning_rate": 5.177120415810271e-06,
+      "loss": 0.5723,
+      "step": 878
+    },
+    {
+      "epoch": 1.6203931203931203,
+      "grad_norm": 0.604628847679587,
+      "learning_rate": 5.1663899374389795e-06,
+      "loss": 0.5525,
+      "step": 879
+    },
+    {
+      "epoch": 1.6222358722358723,
+      "grad_norm": 0.6580850695721256,
+      "learning_rate": 5.155658691815766e-06,
+      "loss": 0.58,
+      "step": 880
+    },
+    {
+      "epoch": 1.6240786240786242,
+      "grad_norm": 0.632808742935975,
+      "learning_rate": 5.144926728424205e-06,
+      "loss": 0.5691,
+      "step": 881
+    },
+    {
+      "epoch": 1.6259213759213758,
+      "grad_norm": 0.6172241861463336,
+      "learning_rate": 5.134194096751182e-06,
+      "loss": 0.5709,
+      "step": 882
+    },
+    {
+      "epoch": 1.6277641277641277,
+      "grad_norm": 0.6620437417613627,
+      "learning_rate": 5.123460846286661e-06,
+      "loss": 0.5788,
+      "step": 883
+    },
+    {
+      "epoch": 1.6296068796068797,
+      "grad_norm": 0.6334339291233978,
+      "learning_rate": 5.112727026523461e-06,
+      "loss": 0.6023,
+      "step": 884
+    },
+    {
+      "epoch": 1.6314496314496314,
+      "grad_norm": 0.6676232909457676,
+      "learning_rate": 5.101992686957028e-06,
+      "loss": 0.5943,
+      "step": 885
+    },
+    {
+      "epoch": 1.6332923832923831,
+      "grad_norm": 0.6435412800870824,
+      "learning_rate": 5.091257877085201e-06,
+      "loss": 0.5941,
+      "step": 886
+    },
+    {
+      "epoch": 1.635135135135135,
+      "grad_norm": 0.639059435404824,
+      "learning_rate": 5.080522646407987e-06,
+      "loss": 0.5984,
+      "step": 887
+    },
+    {
+      "epoch": 1.636977886977887,
+      "grad_norm": 0.5601444728586226,
+      "learning_rate": 5.069787044427341e-06,
+      "loss": 0.5725,
+      "step": 888
+    },
+    {
+      "epoch": 1.6388206388206388,
+      "grad_norm": 0.6009043035660707,
+      "learning_rate": 5.059051120646924e-06,
+      "loss": 0.5875,
+      "step": 889
+    },
+    {
+      "epoch": 1.6406633906633905,
+      "grad_norm": 0.7538845068329936,
+      "learning_rate": 5.0483149245718785e-06,
+      "loss": 0.565,
+      "step": 890
+    },
+    {
+      "epoch": 1.6425061425061425,
+      "grad_norm": 0.6354307360070197,
+      "learning_rate": 5.03757850570861e-06,
+      "loss": 0.6021,
+      "step": 891
+    },
+    {
+      "epoch": 1.6443488943488944,
+      "grad_norm": 0.7525918863757648,
+      "learning_rate": 5.0268419135645455e-06,
+      "loss": 0.5896,
+      "step": 892
+    },
+    {
+      "epoch": 1.6461916461916462,
+      "grad_norm": 0.6417137011586802,
+      "learning_rate": 5.016105197647914e-06,
+      "loss": 0.5805,
+      "step": 893
+    },
+    {
+      "epoch": 1.6480343980343979,
+      "grad_norm": 0.6196631650149972,
+      "learning_rate": 5.005368407467515e-06,
+      "loss": 0.588,
+      "step": 894
+    },
+    {
+      "epoch": 1.6498771498771498,
+      "grad_norm": 0.6258590391492487,
+      "learning_rate": 4.994631592532487e-06,
+      "loss": 0.5665,
+      "step": 895
+    },
+    {
+      "epoch": 1.6517199017199018,
+      "grad_norm": 0.7050483169043805,
+      "learning_rate": 4.9838948023520864e-06,
+      "loss": 0.5824,
+      "step": 896
+    },
+    {
+      "epoch": 1.6535626535626535,
+      "grad_norm": 0.6445632616299034,
+      "learning_rate": 4.9731580864354544e-06,
+      "loss": 0.5832,
+      "step": 897
+    },
+    {
+      "epoch": 1.6554054054054053,
+      "grad_norm": 0.6973263362820544,
+      "learning_rate": 4.9624214942913916e-06,
+      "loss": 0.5747,
+      "step": 898
+    },
+    {
+      "epoch": 1.6572481572481572,
+      "grad_norm": 0.7101943374199985,
+      "learning_rate": 4.951685075428122e-06,
+      "loss": 0.575,
+      "step": 899
+    },
+    {
+      "epoch": 1.6590909090909092,
+      "grad_norm": 0.6168420160869215,
+      "learning_rate": 4.940948879353078e-06,
+      "loss": 0.5774,
+      "step": 900
+    },
+    {
+      "epoch": 1.660933660933661,
+      "grad_norm": 0.6954309517281461,
+      "learning_rate": 4.9302129555726615e-06,
+      "loss": 0.5991,
+      "step": 901
+    },
+    {
+      "epoch": 1.6627764127764126,
+      "grad_norm": 0.6896652109831068,
+      "learning_rate": 4.919477353592015e-06,
+      "loss": 0.5884,
+      "step": 902
+    },
+    {
+      "epoch": 1.6646191646191646,
+      "grad_norm": 0.6908958319862214,
+      "learning_rate": 4.908742122914802e-06,
+      "loss": 0.5783,
+      "step": 903
+    },
+    {
+      "epoch": 1.6664619164619165,
+      "grad_norm": 0.6647676052477414,
+      "learning_rate": 4.898007313042975e-06,
+      "loss": 0.579,
+      "step": 904
+    },
+    {
+      "epoch": 1.6683046683046683,
+      "grad_norm": 0.6324913946477013,
+      "learning_rate": 4.88727297347654e-06,
+      "loss": 0.5735,
+      "step": 905
+    },
+    {
+      "epoch": 1.67014742014742,
+      "grad_norm": 0.6589492050603363,
+      "learning_rate": 4.87653915371334e-06,
+      "loss": 0.5779,
+      "step": 906
+    },
+    {
+      "epoch": 1.671990171990172,
+      "grad_norm": 0.6856802576174922,
+      "learning_rate": 4.8658059032488185e-06,
+      "loss": 0.5861,
+      "step": 907
+    },
+    {
+      "epoch": 1.673832923832924,
+      "grad_norm": 0.6282062511860548,
+      "learning_rate": 4.8550732715757965e-06,
+      "loss": 0.5688,
+      "step": 908
+    },
+    {
+      "epoch": 1.6756756756756757,
+      "grad_norm": 0.6493190620724124,
+      "learning_rate": 4.8443413081842354e-06,
+      "loss": 0.5723,
+      "step": 909
+    },
+    {
+      "epoch": 1.6775184275184274,
+      "grad_norm": 0.648392481511256,
+      "learning_rate": 4.833610062561022e-06,
+      "loss": 0.5897,
+      "step": 910
+    },
+    {
+      "epoch": 1.6793611793611793,
+      "grad_norm": 0.6249606050033639,
+      "learning_rate": 4.822879584189732e-06,
+      "loss": 0.5928,
+      "step": 911
+    },
+    {
+      "epoch": 1.6812039312039313,
+      "grad_norm": 0.5948680704599261,
+      "learning_rate": 4.8121499225503974e-06,
+      "loss": 0.5865,
+      "step": 912
+    },
+    {
+      "epoch": 1.683046683046683,
+      "grad_norm": 0.6164085434666802,
+      "learning_rate": 4.8014211271192914e-06,
+      "loss": 0.577,
+      "step": 913
+    },
+    {
+      "epoch": 1.6848894348894348,
+      "grad_norm": 0.615346959218037,
+      "learning_rate": 4.79069324736869e-06,
+      "loss": 0.573,
+      "step": 914
+    },
+    {
+      "epoch": 1.6867321867321867,
+      "grad_norm": 0.6277706682502251,
+      "learning_rate": 4.7799663327666455e-06,
+      "loss": 0.5693,
+      "step": 915
+    },
+    {
+      "epoch": 1.6885749385749387,
+      "grad_norm": 0.7191099854543174,
+      "learning_rate": 4.7692404327767615e-06,
+      "loss": 0.5899,
+      "step": 916
+    },
+    {
+      "epoch": 1.6904176904176904,
+      "grad_norm": 0.6504535341336625,
+      "learning_rate": 4.758515596857965e-06,
+      "loss": 0.5821,
+      "step": 917
+    },
+    {
+      "epoch": 1.6922604422604421,
+      "grad_norm": 0.6720071139445757,
+      "learning_rate": 4.747791874464269e-06,
+      "loss": 0.5871,
+      "step": 918
+    },
+    {
+      "epoch": 1.694103194103194,
+      "grad_norm": 0.6586019434588468,
+      "learning_rate": 4.737069315044562e-06,
+      "loss": 0.5533,
+      "step": 919
+    },
+    {
+      "epoch": 1.695945945945946,
+      "grad_norm": 0.691746349526742,
+      "learning_rate": 4.726347968042364e-06,
+      "loss": 0.5834,
+      "step": 920
+    },
+    {
+      "epoch": 1.6977886977886978,
+      "grad_norm": 0.6996428974156904,
+      "learning_rate": 4.7156278828956005e-06,
+      "loss": 0.5732,
+      "step": 921
+    },
+    {
+      "epoch": 1.6996314496314495,
+      "grad_norm": 0.6947874591361619,
+      "learning_rate": 4.704909109036387e-06,
+      "loss": 0.5778,
+      "step": 922
+    },
+    {
+      "epoch": 1.7014742014742015,
+      "grad_norm": 0.6535886886763463,
+      "learning_rate": 4.694191695890788e-06,
+      "loss": 0.5779,
+      "step": 923
+    },
+    {
+      "epoch": 1.7033169533169534,
+      "grad_norm": 0.637759430064651,
+      "learning_rate": 4.683475692878592e-06,
+      "loss": 0.5693,
+      "step": 924
+    },
+    {
+      "epoch": 1.7051597051597052,
+      "grad_norm": 0.6238486355200112,
+      "learning_rate": 4.672761149413089e-06,
+      "loss": 0.5679,
+      "step": 925
+    },
+    {
+      "epoch": 1.707002457002457,
+      "grad_norm": 0.6458127670354362,
+      "learning_rate": 4.662048114900837e-06,
+      "loss": 0.5929,
+      "step": 926
+    },
+    {
+      "epoch": 1.7088452088452089,
+      "grad_norm": 0.6937885665721389,
+      "learning_rate": 4.651336638741433e-06,
+      "loss": 0.6042,
+      "step": 927
+    },
+    {
+      "epoch": 1.7106879606879608,
+      "grad_norm": 0.6523655420638866,
+      "learning_rate": 4.640626770327293e-06,
+      "loss": 0.5678,
+      "step": 928
+    },
+    {
+      "epoch": 1.7125307125307125,
+      "grad_norm": 0.6154623517494733,
+      "learning_rate": 4.6299185590434185e-06,
+      "loss": 0.5965,
+      "step": 929
+    },
+    {
+      "epoch": 1.7143734643734643,
+      "grad_norm": 0.6177063085735546,
+      "learning_rate": 4.6192120542671665e-06,
+      "loss": 0.5822,
+      "step": 930
+    },
+    {
+      "epoch": 1.7162162162162162,
+      "grad_norm": 0.6835915316644277,
+      "learning_rate": 4.608507305368027e-06,
+      "loss": 0.5758,
+      "step": 931
+    },
+    {
+      "epoch": 1.7180589680589682,
+      "grad_norm": 0.6706276839434657,
+      "learning_rate": 4.597804361707395e-06,
+      "loss": 0.5727,
+      "step": 932
+    },
+    {
+      "epoch": 1.71990171990172,
+      "grad_norm": 0.6692802182464253,
+      "learning_rate": 4.587103272638339e-06,
+      "loss": 0.5823,
+      "step": 933
+    },
+    {
+      "epoch": 1.7217444717444716,
+      "grad_norm": 0.6749292554803209,
+      "learning_rate": 4.576404087505375e-06,
+      "loss": 0.5491,
+      "step": 934
+    },
+    {
+      "epoch": 1.7235872235872236,
+      "grad_norm": 0.6821108347658321,
+      "learning_rate": 4.565706855644242e-06,
+      "loss": 0.5825,
+      "step": 935
+    },
+    {
+      "epoch": 1.7254299754299756,
+      "grad_norm": 0.6841121381839927,
+      "learning_rate": 4.555011626381674e-06,
+      "loss": 0.5591,
+      "step": 936
+    },
+    {
+      "epoch": 1.7272727272727273,
+      "grad_norm": 0.6327359519179588,
+      "learning_rate": 4.544318449035163e-06,
+      "loss": 0.5721,
+      "step": 937
+    },
+    {
+      "epoch": 1.729115479115479,
+      "grad_norm": 0.6712975316191102,
+      "learning_rate": 4.533627372912744e-06,
+      "loss": 0.5645,
+      "step": 938
+    },
+    {
+      "epoch": 1.730958230958231,
+      "grad_norm": 0.753286423073289,
+      "learning_rate": 4.522938447312768e-06,
+      "loss": 0.5656,
+      "step": 939
+    },
+    {
+      "epoch": 1.732800982800983,
+      "grad_norm": 0.5779706923995587,
+      "learning_rate": 4.512251721523659e-06,
+      "loss": 0.585,
+      "step": 940
+    },
+    {
+      "epoch": 1.7346437346437347,
+      "grad_norm": 0.7396713658623506,
+      "learning_rate": 4.501567244823706e-06,
+      "loss": 0.6036,
+      "step": 941
+    },
+    {
+      "epoch": 1.7364864864864864,
+      "grad_norm": 0.7149955656451358,
+      "learning_rate": 4.4908850664808245e-06,
+      "loss": 0.5749,
+      "step": 942
+    },
+    {
+      "epoch": 1.7383292383292384,
+      "grad_norm": 0.619646302655871,
+      "learning_rate": 4.480205235752328e-06,
+      "loss": 0.564,
+      "step": 943
+    },
+    {
+      "epoch": 1.7401719901719903,
+      "grad_norm": 0.6725119379510272,
+      "learning_rate": 4.46952780188471e-06,
+      "loss": 0.5723,
+      "step": 944
+    },
+    {
+      "epoch": 1.742014742014742,
+      "grad_norm": 0.7998557395120705,
+      "learning_rate": 4.458852814113412e-06,
+      "loss": 0.591,
+      "step": 945
+    },
+    {
+      "epoch": 1.7438574938574938,
+      "grad_norm": 0.6586825498306963,
+      "learning_rate": 4.448180321662588e-06,
+      "loss": 0.5782,
+      "step": 946
+    },
+    {
+      "epoch": 1.7457002457002457,
+      "grad_norm": 0.608242985075989,
+      "learning_rate": 4.437510373744897e-06,
+      "loss": 0.5747,
+      "step": 947
+    },
+    {
+      "epoch": 1.7475429975429977,
+      "grad_norm": 0.6959368270233507,
+      "learning_rate": 4.426843019561259e-06,
+      "loss": 0.5866,
+      "step": 948
+    },
+    {
+      "epoch": 1.7493857493857494,
+      "grad_norm": 0.7206742268586664,
+      "learning_rate": 4.41617830830063e-06,
+      "loss": 0.6008,
+      "step": 949
+    },
+    {
+      "epoch": 1.7512285012285012,
+      "grad_norm": 0.7138493055522486,
+      "learning_rate": 4.405516289139785e-06,
+      "loss": 0.583,
+      "step": 950
+    },
+    {
+      "epoch": 1.753071253071253,
+      "grad_norm": 0.6522113216573334,
+      "learning_rate": 4.394857011243084e-06,
+      "loss": 0.5558,
+      "step": 951
+    },
+    {
+      "epoch": 1.754914004914005,
+      "grad_norm": 0.7030766067538453,
+      "learning_rate": 4.384200523762249e-06,
+      "loss": 0.5983,
+      "step": 952
+    },
+    {
+      "epoch": 1.7567567567567568,
+      "grad_norm": 0.7597665310348779,
+      "learning_rate": 4.373546875836125e-06,
+      "loss": 0.5911,
+      "step": 953
+    },
+    {
+      "epoch": 1.7585995085995085,
+      "grad_norm": 0.6823676544085036,
+      "learning_rate": 4.362896116590475e-06,
+      "loss": 0.5676,
+      "step": 954
+    },
+    {
+      "epoch": 1.7604422604422605,
+      "grad_norm": 0.7614725967043108,
+      "learning_rate": 4.352248295137739e-06,
+      "loss": 0.5787,
+      "step": 955
+    },
+    {
+      "epoch": 1.7622850122850124,
+      "grad_norm": 0.6487005077957451,
+      "learning_rate": 4.3416034605768035e-06,
+      "loss": 0.5774,
+      "step": 956
+    },
+    {
+      "epoch": 1.7641277641277642,
+      "grad_norm": 0.6413607169967719,
+      "learning_rate": 4.330961661992788e-06,
+      "loss": 0.5794,
+      "step": 957
+    },
+    {
+      "epoch": 1.765970515970516,
+      "grad_norm": 0.742701172203094,
+      "learning_rate": 4.320322948456814e-06,
+      "loss": 0.5934,
+      "step": 958
+    },
+    {
+      "epoch": 1.7678132678132679,
+      "grad_norm": 0.7040796796036775,
+      "learning_rate": 4.309687369025772e-06,
+      "loss": 0.5835,
+      "step": 959
+    },
+    {
+      "epoch": 1.7696560196560198,
+      "grad_norm": 0.6148692890971733,
+      "learning_rate": 4.299054972742104e-06,
+      "loss": 0.5821,
+      "step": 960
+    },
+    {
+      "epoch": 1.7714987714987716,
+      "grad_norm": 0.7271240213649599,
+      "learning_rate": 4.2884258086335755e-06,
+      "loss": 0.5748,
+      "step": 961
+    },
+    {
+      "epoch": 1.7733415233415233,
+      "grad_norm": 0.6188873193642994,
+      "learning_rate": 4.277799925713043e-06,
+      "loss": 0.587,
+      "step": 962
+    },
+    {
+      "epoch": 1.7751842751842752,
+      "grad_norm": 0.6289926818329122,
+      "learning_rate": 4.267177372978236e-06,
+      "loss": 0.5825,
+      "step": 963
+    },
+    {
+      "epoch": 1.777027027027027,
+      "grad_norm": 0.6879078023917552,
+      "learning_rate": 4.25655819941153e-06,
+      "loss": 0.5803,
+      "step": 964
+    },
+    {
+      "epoch": 1.7788697788697787,
+      "grad_norm": 0.6350383266108354,
+      "learning_rate": 4.245942453979713e-06,
+      "loss": 0.5835,
+      "step": 965
+    },
+    {
+      "epoch": 1.7807125307125307,
+      "grad_norm": 0.6320773688252834,
+      "learning_rate": 4.23533018563377e-06,
+      "loss": 0.5669,
+      "step": 966
+    },
+    {
+      "epoch": 1.7825552825552826,
+      "grad_norm": 0.639042617895869,
+      "learning_rate": 4.224721443308654e-06,
+      "loss": 0.5832,
+      "step": 967
+    },
+    {
+      "epoch": 1.7843980343980343,
+      "grad_norm": 0.6494245445409529,
+      "learning_rate": 4.214116275923051e-06,
+      "loss": 0.5641,
+      "step": 968
+    },
+    {
+      "epoch": 1.786240786240786,
+      "grad_norm": 0.5996452021208784,
+      "learning_rate": 4.203514732379168e-06,
+      "loss": 0.5878,
+      "step": 969
+    },
+    {
+      "epoch": 1.788083538083538,
+      "grad_norm": 0.6517450455988659,
+      "learning_rate": 4.192916861562503e-06,
+      "loss": 0.5867,
+      "step": 970
+    },
+    {
+      "epoch": 1.78992628992629,
+      "grad_norm": 0.673683684240893,
+      "learning_rate": 4.1823227123416185e-06,
+      "loss": 0.5682,
+      "step": 971
+    },
+    {
+      "epoch": 1.7917690417690417,
+      "grad_norm": 0.6039660481793007,
+      "learning_rate": 4.171732333567911e-06,
+      "loss": 0.5885,
+      "step": 972
+    },
+    {
+      "epoch": 1.7936117936117935,
+      "grad_norm": 0.6132385316255978,
+      "learning_rate": 4.161145774075393e-06,
+      "loss": 0.5695,
+      "step": 973
+    },
+    {
+      "epoch": 1.7954545454545454,
+      "grad_norm": 0.6376244576455573,
+      "learning_rate": 4.150563082680472e-06,
+      "loss": 0.5705,
+      "step": 974
+    },
+    {
+      "epoch": 1.7972972972972974,
+      "grad_norm": 0.6034351746190463,
+      "learning_rate": 4.1399843081817085e-06,
+      "loss": 0.5951,
+      "step": 975
+    },
+    {
+      "epoch": 1.799140049140049,
+      "grad_norm": 0.6598525945886587,
+      "learning_rate": 4.129409499359609e-06,
+      "loss": 0.5746,
+      "step": 976
+    },
+    {
+      "epoch": 1.8009828009828008,
+      "grad_norm": 0.6528338903815515,
+      "learning_rate": 4.118838704976392e-06,
+      "loss": 0.5871,
+      "step": 977
+    },
+    {
+      "epoch": 1.8028255528255528,
+      "grad_norm": 0.585033662554439,
+      "learning_rate": 4.108271973775763e-06,
+      "loss": 0.5723,
+      "step": 978
+    },
+    {
+      "epoch": 1.8046683046683047,
+      "grad_norm": 0.6905237041164953,
+      "learning_rate": 4.097709354482693e-06,
+      "loss": 0.5998,
+      "step": 979
+    },
+    {
+      "epoch": 1.8065110565110565,
+      "grad_norm": 0.6304277435264687,
+      "learning_rate": 4.087150895803192e-06,
+      "loss": 0.5774,
+      "step": 980
+    },
+    {
+      "epoch": 1.8083538083538082,
+      "grad_norm": 0.606968520915071,
+      "learning_rate": 4.076596646424085e-06,
+      "loss": 0.5749,
+      "step": 981
+    },
+    {
+      "epoch": 1.8101965601965602,
+      "grad_norm": 0.6470333868857457,
+      "learning_rate": 4.066046655012786e-06,
+      "loss": 0.5922,
+      "step": 982
+    },
+    {
+      "epoch": 1.8120393120393121,
+      "grad_norm": 0.7093025908628318,
+      "learning_rate": 4.055500970217078e-06,
+      "loss": 0.5658,
+      "step": 983
+    },
+    {
+      "epoch": 1.8138820638820639,
+      "grad_norm": 0.6465664330474143,
+      "learning_rate": 4.044959640664878e-06,
+      "loss": 0.5814,
+      "step": 984
+    },
+    {
+      "epoch": 1.8157248157248156,
+      "grad_norm": 0.6486023445844586,
+      "learning_rate": 4.03442271496403e-06,
+      "loss": 0.5712,
+      "step": 985
+    },
+    {
+      "epoch": 1.8175675675675675,
+      "grad_norm": 0.6886682958936369,
+      "learning_rate": 4.023890241702068e-06,
+      "loss": 0.5682,
+      "step": 986
+    },
+    {
+      "epoch": 1.8194103194103195,
+      "grad_norm": 0.6489324622673824,
+      "learning_rate": 4.013362269445988e-06,
+      "loss": 0.588,
+      "step": 987
+    },
+    {
+      "epoch": 1.8212530712530712,
+      "grad_norm": 0.6551696811009383,
+      "learning_rate": 4.002838846742039e-06,
+      "loss": 0.6008,
+      "step": 988
+    },
+    {
+      "epoch": 1.823095823095823,
+      "grad_norm": 0.5920992981703637,
+      "learning_rate": 3.992320022115492e-06,
+      "loss": 0.6019,
+      "step": 989
+    },
+    {
+      "epoch": 1.824938574938575,
+      "grad_norm": 0.6739457232432299,
+      "learning_rate": 3.981805844070407e-06,
+      "loss": 0.5851,
+      "step": 990
+    },
+    {
+      "epoch": 1.8267813267813269,
+      "grad_norm": 0.6166145252403737,
+      "learning_rate": 3.971296361089427e-06,
+      "loss": 0.574,
+      "step": 991
+    },
+    {
+      "epoch": 1.8286240786240786,
+      "grad_norm": 0.666538828692749,
+      "learning_rate": 3.960791621633539e-06,
+      "loss": 0.5832,
+      "step": 992
+    },
+    {
+      "epoch": 1.8304668304668303,
+      "grad_norm": 0.6525292615614627,
+      "learning_rate": 3.950291674141863e-06,
+      "loss": 0.5672,
+      "step": 993
+    },
+    {
+      "epoch": 1.8323095823095823,
+      "grad_norm": 0.6265224172638881,
+      "learning_rate": 3.939796567031414e-06,
+      "loss": 0.5969,
+      "step": 994
+    },
+    {
+      "epoch": 1.8341523341523343,
+      "grad_norm": 0.6380901046403779,
+      "learning_rate": 3.929306348696896e-06,
+      "loss": 0.5768,
+      "step": 995
+    },
+    {
+      "epoch": 1.835995085995086,
+      "grad_norm": 0.6663488883341968,
+      "learning_rate": 3.918821067510464e-06,
+      "loss": 0.5792,
+      "step": 996
+    },
+    {
+      "epoch": 1.8378378378378377,
+      "grad_norm": 0.6659220499927639,
+      "learning_rate": 3.908340771821509e-06,
+      "loss": 0.6074,
+      "step": 997
+    },
+    {
+      "epoch": 1.8396805896805897,
+      "grad_norm": 0.587493916526858,
+      "learning_rate": 3.897865509956434e-06,
+      "loss": 0.5875,
+      "step": 998
+    },
+    {
+      "epoch": 1.8415233415233416,
+      "grad_norm": 0.6400034309470728,
+      "learning_rate": 3.887395330218429e-06,
+      "loss": 0.5715,
+      "step": 999
+    },
+    {
+      "epoch": 1.8433660933660934,
+      "grad_norm": 0.6322046714174706,
+      "learning_rate": 3.876930280887247e-06,
+      "loss": 0.5832,
+      "step": 1000
+    },
+    {
+      "epoch": 1.845208845208845,
+      "grad_norm": 0.6518368784902464,
+      "learning_rate": 3.86647041021899e-06,
+      "loss": 0.5842,
+      "step": 1001
+    },
+    {
+      "epoch": 1.847051597051597,
+      "grad_norm": 0.6435618971934893,
+      "learning_rate": 3.856015766445877e-06,
+      "loss": 0.5886,
+      "step": 1002
+    },
+    {
+      "epoch": 1.848894348894349,
+      "grad_norm": 0.6058950547118837,
+      "learning_rate": 3.845566397776022e-06,
+      "loss": 0.5762,
+      "step": 1003
+    },
+    {
+      "epoch": 1.8507371007371007,
+      "grad_norm": 0.5841738644674215,
+      "learning_rate": 3.835122352393216e-06,
+      "loss": 0.5859,
+      "step": 1004
+    },
+    {
+      "epoch": 1.8525798525798525,
+      "grad_norm": 0.6643484114036269,
+      "learning_rate": 3.824683678456711e-06,
+      "loss": 0.5822,
+      "step": 1005
+    },
+    {
+      "epoch": 1.8544226044226044,
+      "grad_norm": 0.668023683363932,
+      "learning_rate": 3.814250424100978e-06,
+      "loss": 0.5775,
+      "step": 1006
+    },
+    {
+      "epoch": 1.8562653562653564,
+      "grad_norm": 0.6330987654173601,
+      "learning_rate": 3.8038226374355064e-06,
+      "loss": 0.5799,
+      "step": 1007
+    },
+    {
+      "epoch": 1.8581081081081081,
+      "grad_norm": 0.6057876029783777,
+      "learning_rate": 3.79340036654457e-06,
+      "loss": 0.6,
+      "step": 1008
+    },
+    {
+      "epoch": 1.8599508599508598,
+      "grad_norm": 0.5652712441111964,
+      "learning_rate": 3.7829836594870077e-06,
+      "loss": 0.5611,
+      "step": 1009
+    },
+    {
+      "epoch": 1.8617936117936118,
+      "grad_norm": 0.6664602576659144,
+      "learning_rate": 3.7725725642960047e-06,
+      "loss": 0.5698,
+      "step": 1010
+    },
+    {
+      "epoch": 1.8636363636363638,
+      "grad_norm": 0.6626430399664703,
+      "learning_rate": 3.7621671289788685e-06,
+      "loss": 0.6133,
+      "step": 1011
+    },
+    {
+      "epoch": 1.8654791154791155,
+      "grad_norm": 0.6357161657532953,
+      "learning_rate": 3.751767401516805e-06,
+      "loss": 0.5674,
+      "step": 1012
+    },
+    {
+      "epoch": 1.8673218673218672,
+      "grad_norm": 0.6922867224618765,
+      "learning_rate": 3.741373429864704e-06,
+      "loss": 0.5888,
+      "step": 1013
+    },
+    {
+      "epoch": 1.8691646191646192,
+      "grad_norm": 0.5961305129547646,
+      "learning_rate": 3.7309852619509136e-06,
+      "loss": 0.5845,
+      "step": 1014
+    },
+    {
+      "epoch": 1.8710073710073711,
+      "grad_norm": 0.6366340755853362,
+      "learning_rate": 3.7206029456770186e-06,
+      "loss": 0.5741,
+      "step": 1015
+    },
+    {
+      "epoch": 1.8728501228501229,
+      "grad_norm": 0.7083678594133035,
+      "learning_rate": 3.71022652891762e-06,
+      "loss": 0.585,
+      "step": 1016
+    },
+    {
+      "epoch": 1.8746928746928746,
+      "grad_norm": 0.6860796899614187,
+      "learning_rate": 3.6998560595201188e-06,
+      "loss": 0.5816,
+      "step": 1017
+    },
+    {
+      "epoch": 1.8765356265356266,
+      "grad_norm": 0.663212153598199,
+      "learning_rate": 3.689491585304491e-06,
+      "loss": 0.563,
+      "step": 1018
+    },
+    {
+      "epoch": 1.8783783783783785,
+      "grad_norm": 0.7532046080076186,
+      "learning_rate": 3.6791331540630597e-06,
+      "loss": 0.5791,
+      "step": 1019
+    },
+    {
+      "epoch": 1.8802211302211302,
+      "grad_norm": 0.6732993229830843,
+      "learning_rate": 3.6687808135602955e-06,
+      "loss": 0.576,
+      "step": 1020
+    },
+    {
+      "epoch": 1.882063882063882,
+      "grad_norm": 0.6399565546331056,
+      "learning_rate": 3.658434611532578e-06,
+      "loss": 0.5819,
+      "step": 1021
+    },
+    {
+      "epoch": 1.883906633906634,
+      "grad_norm": 0.6966308570541498,
+      "learning_rate": 3.6480945956879765e-06,
+      "loss": 0.5844,
+      "step": 1022
+    },
+    {
+      "epoch": 1.8857493857493859,
+      "grad_norm": 0.846705959707348,
+      "learning_rate": 3.637760813706041e-06,
+      "loss": 0.5593,
+      "step": 1023
+    },
+    {
+      "epoch": 1.8875921375921376,
+      "grad_norm": 0.5670725277467041,
+      "learning_rate": 3.627433313237576e-06,
+      "loss": 0.5592,
+      "step": 1024
+    },
+    {
+      "epoch": 1.8894348894348894,
+      "grad_norm": 0.6517843289498574,
+      "learning_rate": 3.617112141904416e-06,
+      "loss": 0.5722,
+      "step": 1025
+    },
+    {
+      "epoch": 1.8912776412776413,
+      "grad_norm": 0.6473904560914079,
+      "learning_rate": 3.606797347299216e-06,
+      "loss": 0.5624,
+      "step": 1026
+    },
+    {
+      "epoch": 1.8931203931203933,
+      "grad_norm": 0.7060636367207498,
+      "learning_rate": 3.596488976985224e-06,
+      "loss": 0.578,
+      "step": 1027
+    },
+    {
+      "epoch": 1.894963144963145,
+      "grad_norm": 0.7349575293099392,
+      "learning_rate": 3.5861870784960657e-06,
+      "loss": 0.5972,
+      "step": 1028
+    },
+    {
+      "epoch": 1.8968058968058967,
+      "grad_norm": 0.6761476900938991,
+      "learning_rate": 3.575891699335523e-06,
+      "loss": 0.5886,
+      "step": 1029
+    },
+    {
+      "epoch": 1.8986486486486487,
+      "grad_norm": 0.7451566010924302,
+      "learning_rate": 3.565602886977318e-06,
+      "loss": 0.5648,
+      "step": 1030
+    },
+    {
+      "epoch": 1.9004914004914006,
+      "grad_norm": 0.6268625984763108,
+      "learning_rate": 3.555320688864889e-06,
+      "loss": 0.5862,
+      "step": 1031
+    },
+    {
+      "epoch": 1.9023341523341524,
+      "grad_norm": 0.5961906369242848,
+      "learning_rate": 3.545045152411178e-06,
+      "loss": 0.5725,
+      "step": 1032
+    },
+    {
+      "epoch": 1.904176904176904,
+      "grad_norm": 0.6677657078981969,
+      "learning_rate": 3.5347763249984097e-06,
+      "loss": 0.5736,
+      "step": 1033
+    },
+    {
+      "epoch": 1.906019656019656,
+      "grad_norm": 0.6737269389859477,
+      "learning_rate": 3.5245142539778655e-06,
+      "loss": 0.5781,
+      "step": 1034
+    },
+    {
+      "epoch": 1.907862407862408,
+      "grad_norm": 0.5860649715714522,
+      "learning_rate": 3.5142589866696804e-06,
+      "loss": 0.5841,
+      "step": 1035
+    },
+    {
+      "epoch": 1.9097051597051597,
+      "grad_norm": 0.6642789141320249,
+      "learning_rate": 3.504010570362612e-06,
+      "loss": 0.5889,
+      "step": 1036
+    },
+    {
+      "epoch": 1.9115479115479115,
+      "grad_norm": 0.6494473057614752,
+      "learning_rate": 3.4937690523138302e-06,
+      "loss": 0.5871,
+      "step": 1037
+    },
+    {
+      "epoch": 1.9133906633906634,
+      "grad_norm": 0.6366805677164155,
+      "learning_rate": 3.483534479748688e-06,
+      "loss": 0.5787,
+      "step": 1038
+    },
+    {
+      "epoch": 1.9152334152334154,
+      "grad_norm": 0.7079322527154163,
+      "learning_rate": 3.473306899860519e-06,
+      "loss": 0.5818,
+      "step": 1039
+    },
+    {
+      "epoch": 1.9170761670761671,
+      "grad_norm": 0.7178576897787424,
+      "learning_rate": 3.4630863598104137e-06,
+      "loss": 0.5724,
+      "step": 1040
+    },
+    {
+      "epoch": 1.9189189189189189,
+      "grad_norm": 0.7038536235599043,
+      "learning_rate": 3.4528729067269916e-06,
+      "loss": 0.6017,
+      "step": 1041
+    },
+    {
+      "epoch": 1.9207616707616708,
+      "grad_norm": 0.6696400688704139,
+      "learning_rate": 3.442666587706203e-06,
+      "loss": 0.5743,
+      "step": 1042
+    },
+    {
+      "epoch": 1.9226044226044228,
+      "grad_norm": 0.7239245124165259,
+      "learning_rate": 3.4324674498110956e-06,
+      "loss": 0.5637,
+      "step": 1043
+    },
+    {
+      "epoch": 1.9244471744471745,
+      "grad_norm": 0.7241759165284539,
+      "learning_rate": 3.4222755400716044e-06,
+      "loss": 0.5638,
+      "step": 1044
+    },
+    {
+      "epoch": 1.9262899262899262,
+      "grad_norm": 0.6229074950628511,
+      "learning_rate": 3.4120909054843375e-06,
+      "loss": 0.5789,
+      "step": 1045
+    },
+    {
+      "epoch": 1.9281326781326782,
+      "grad_norm": 0.6526435845819397,
+      "learning_rate": 3.401913593012355e-06,
+      "loss": 0.5713,
+      "step": 1046
+    },
+    {
+      "epoch": 1.92997542997543,
+      "grad_norm": 0.6427890048379266,
+      "learning_rate": 3.3917436495849486e-06,
+      "loss": 0.5955,
+      "step": 1047
+    },
+    {
+      "epoch": 1.9318181818181817,
+      "grad_norm": 0.5648314597262253,
+      "learning_rate": 3.381581122097437e-06,
+      "loss": 0.5937,
+      "step": 1048
+    },
+    {
+      "epoch": 1.9336609336609336,
+      "grad_norm": 0.6541148129821951,
+      "learning_rate": 3.37142605741094e-06,
+      "loss": 0.5766,
+      "step": 1049
+    },
+    {
+      "epoch": 1.9355036855036856,
+      "grad_norm": 0.6413144105139179,
+      "learning_rate": 3.361278502352161e-06,
+      "loss": 0.5663,
+      "step": 1050
+    },
+    {
+      "epoch": 1.9373464373464373,
+      "grad_norm": 0.563339075640476,
+      "learning_rate": 3.3511385037131827e-06,
+      "loss": 0.6137,
+      "step": 1051
+    },
+    {
+      "epoch": 1.939189189189189,
+      "grad_norm": 0.6382975969811641,
+      "learning_rate": 3.3410061082512422e-06,
+      "loss": 0.5496,
+      "step": 1052
+    },
+    {
+      "epoch": 1.941031941031941,
+      "grad_norm": 0.6216090237034129,
+      "learning_rate": 3.3308813626885096e-06,
+      "loss": 0.5769,
+      "step": 1053
+    },
+    {
+      "epoch": 1.942874692874693,
+      "grad_norm": 0.5947402240225198,
+      "learning_rate": 3.3207643137118872e-06,
+      "loss": 0.5907,
+      "step": 1054
+    },
+    {
+      "epoch": 1.9447174447174447,
+      "grad_norm": 0.6490497895108405,
+      "learning_rate": 3.31065500797279e-06,
+      "loss": 0.5839,
+      "step": 1055
+    },
+    {
+      "epoch": 1.9465601965601964,
+      "grad_norm": 0.6394483770659186,
+      "learning_rate": 3.3005534920869175e-06,
+      "loss": 0.5866,
+      "step": 1056
+    },
+    {
+      "epoch": 1.9484029484029484,
+      "grad_norm": 0.6380535758531071,
+      "learning_rate": 3.290459812634057e-06,
+      "loss": 0.5819,
+      "step": 1057
+    },
+    {
+      "epoch": 1.9502457002457003,
+      "grad_norm": 0.5795032496796598,
+      "learning_rate": 3.2803740161578566e-06,
+      "loss": 0.558,
+      "step": 1058
+    },
+    {
+      "epoch": 1.952088452088452,
+      "grad_norm": 0.6064447035483613,
+      "learning_rate": 3.2702961491656197e-06,
+      "loss": 0.5692,
+      "step": 1059
+    },
+    {
+      "epoch": 1.9539312039312038,
+      "grad_norm": 0.6402796876447354,
+      "learning_rate": 3.2602262581280785e-06,
+      "loss": 0.5844,
+      "step": 1060
+    },
+    {
+      "epoch": 1.9557739557739557,
+      "grad_norm": 0.6532772354485824,
+      "learning_rate": 3.2501643894791912e-06,
+      "loss": 0.5936,
+      "step": 1061
+    },
+    {
+      "epoch": 1.9576167076167077,
+      "grad_norm": 0.5614942215516173,
+      "learning_rate": 3.2401105896159245e-06,
+      "loss": 0.5887,
+      "step": 1062
+    },
+    {
+      "epoch": 1.9594594594594594,
+      "grad_norm": 0.6614002535334561,
+      "learning_rate": 3.2300649048980336e-06,
+      "loss": 0.5946,
+      "step": 1063
+    },
+    {
+      "epoch": 1.9613022113022112,
+      "grad_norm": 0.572629590127392,
+      "learning_rate": 3.220027381647857e-06,
+      "loss": 0.5701,
+      "step": 1064
+    },
+    {
+      "epoch": 1.9631449631449631,
+      "grad_norm": 0.5869817829266226,
+      "learning_rate": 3.2099980661501016e-06,
+      "loss": 0.5773,
+      "step": 1065
+    },
+    {
+      "epoch": 1.964987714987715,
+      "grad_norm": 0.6424212263141368,
+      "learning_rate": 3.1999770046516198e-06,
+      "loss": 0.5593,
+      "step": 1066
+    },
+    {
+      "epoch": 1.9668304668304668,
+      "grad_norm": 0.5533199913209818,
+      "learning_rate": 3.1899642433612104e-06,
+      "loss": 0.5701,
+      "step": 1067
+    },
+    {
+      "epoch": 1.9686732186732185,
+      "grad_norm": 0.5613174626784465,
+      "learning_rate": 3.179959828449397e-06,
+      "loss": 0.5734,
+      "step": 1068
+    },
+    {
+      "epoch": 1.9705159705159705,
+      "grad_norm": 0.5945259848995165,
+      "learning_rate": 3.1699638060482118e-06,
+      "loss": 0.5551,
+      "step": 1069
+    },
+    {
+      "epoch": 1.9723587223587224,
+      "grad_norm": 0.5904424729639253,
+      "learning_rate": 3.159976222250992e-06,
+      "loss": 0.588,
+      "step": 1070
+    },
+    {
+      "epoch": 1.9742014742014742,
+      "grad_norm": 0.5704180073020404,
+      "learning_rate": 3.1499971231121672e-06,
+      "loss": 0.585,
+      "step": 1071
+    },
+    {
+      "epoch": 1.976044226044226,
+      "grad_norm": 0.5909995400011172,
+      "learning_rate": 3.140026554647032e-06,
+      "loss": 0.5695,
+      "step": 1072
+    },
+    {
+      "epoch": 1.9778869778869779,
+      "grad_norm": 0.6084838511727857,
+      "learning_rate": 3.130064562831553e-06,
+      "loss": 0.5656,
+      "step": 1073
+    },
+    {
+      "epoch": 1.9797297297297298,
+      "grad_norm": 0.5649766511573239,
+      "learning_rate": 3.1201111936021467e-06,
+      "loss": 0.5841,
+      "step": 1074
+    },
+    {
+      "epoch": 1.9815724815724816,
+      "grad_norm": 0.5468371302080948,
+      "learning_rate": 3.110166492855468e-06,
+      "loss": 0.602,
+      "step": 1075
+    },
+    {
+      "epoch": 1.9834152334152333,
+      "grad_norm": 0.5908994659248187,
+      "learning_rate": 3.1002305064482006e-06,
+      "loss": 0.5799,
+      "step": 1076
+    },
+    {
+      "epoch": 1.9852579852579852,
+      "grad_norm": 0.6305539980943611,
+      "learning_rate": 3.090303280196847e-06,
+      "loss": 0.5883,
+      "step": 1077
+    },
+    {
+      "epoch": 1.9871007371007372,
+      "grad_norm": 0.5911801242800868,
+      "learning_rate": 3.08038485987751e-06,
+      "loss": 0.5942,
+      "step": 1078
+    },
+    {
+      "epoch": 1.988943488943489,
+      "grad_norm": 0.6102377318976688,
+      "learning_rate": 3.070475291225692e-06,
+      "loss": 0.5641,
+      "step": 1079
+    },
+    {
+      "epoch": 1.9907862407862407,
+      "grad_norm": 0.6042391334701448,
+      "learning_rate": 3.0605746199360755e-06,
+      "loss": 0.5845,
+      "step": 1080
+    },
+    {
+      "epoch": 1.9926289926289926,
+      "grad_norm": 0.5668871098154529,
+      "learning_rate": 3.0506828916623198e-06,
+      "loss": 0.6101,
+      "step": 1081
+    },
+    {
+      "epoch": 1.9944717444717446,
+      "grad_norm": 0.6119708164941998,
+      "learning_rate": 3.040800152016841e-06,
+      "loss": 0.5806,
+      "step": 1082
+    },
+    {
+      "epoch": 1.9963144963144963,
+      "grad_norm": 0.5738911863206702,
+      "learning_rate": 3.030926446570611e-06,
+      "loss": 0.5813,
+      "step": 1083
+    },
+    {
+      "epoch": 1.998157248157248,
+      "grad_norm": 0.6245132749588947,
+      "learning_rate": 3.021061820852945e-06,
+      "loss": 0.5703,
+      "step": 1084
+    },
+    {
+      "epoch": 2.000614250614251,
+      "grad_norm": 0.7151103698901529,
+      "learning_rate": 3.0112063203512808e-06,
+      "loss": 0.5428,
+      "step": 1085
+    },
+    {
+      "epoch": 2.0024570024570023,
+      "grad_norm": 0.6212787733051,
+      "learning_rate": 3.001359990510991e-06,
+      "loss": 0.5192,
+      "step": 1086
+    },
+    {
+      "epoch": 2.0042997542997543,
+      "grad_norm": 0.6179758231338596,
+      "learning_rate": 2.991522876735154e-06,
+      "loss": 0.522,
+      "step": 1087
+    },
+    {
+      "epoch": 2.006142506142506,
+      "grad_norm": 0.5865910448898164,
+      "learning_rate": 2.981695024384349e-06,
+      "loss": 0.5141,
+      "step": 1088
+    },
+    {
+      "epoch": 2.007985257985258,
+      "grad_norm": 0.6826462049445012,
+      "learning_rate": 2.9718764787764542e-06,
+      "loss": 0.5044,
+      "step": 1089
+    },
+    {
+      "epoch": 2.0098280098280097,
+      "grad_norm": 0.778138126215274,
+      "learning_rate": 2.9620672851864313e-06,
+      "loss": 0.5269,
+      "step": 1090
+    },
+    {
+      "epoch": 2.0116707616707616,
+      "grad_norm": 0.6589746947336519,
+      "learning_rate": 2.9522674888461168e-06,
+      "loss": 0.521,
+      "step": 1091
+    },
+    {
+      "epoch": 2.0135135135135136,
+      "grad_norm": 0.5812432357258854,
+      "learning_rate": 2.942477134944016e-06,
+      "loss": 0.5514,
+      "step": 1092
+    },
+    {
+      "epoch": 2.0153562653562656,
+      "grad_norm": 0.6678925793667279,
+      "learning_rate": 2.9326962686250954e-06,
+      "loss": 0.5181,
+      "step": 1093
+    },
+    {
+      "epoch": 2.017199017199017,
+      "grad_norm": 0.7909254354989775,
+      "learning_rate": 2.9229249349905686e-06,
+      "loss": 0.5177,
+      "step": 1094
+    },
+    {
+      "epoch": 2.019041769041769,
+      "grad_norm": 0.5969270044596482,
+      "learning_rate": 2.9131631790976954e-06,
+      "loss": 0.5216,
+      "step": 1095
+    },
+    {
+      "epoch": 2.020884520884521,
+      "grad_norm": 0.6667804499588313,
+      "learning_rate": 2.9034110459595716e-06,
+      "loss": 0.4999,
+      "step": 1096
+    },
+    {
+      "epoch": 2.022727272727273,
+      "grad_norm": 0.6094696936069005,
+      "learning_rate": 2.893668580544918e-06,
+      "loss": 0.5257,
+      "step": 1097
+    },
+    {
+      "epoch": 2.0245700245700244,
+      "grad_norm": 0.5966471750919379,
+      "learning_rate": 2.8839358277778758e-06,
+      "loss": 0.5387,
+      "step": 1098
+    },
+    {
+      "epoch": 2.0264127764127764,
+      "grad_norm": 0.6757865344340327,
+      "learning_rate": 2.874212832537805e-06,
+      "loss": 0.4805,
+      "step": 1099
+    },
+    {
+      "epoch": 2.0282555282555284,
+      "grad_norm": 0.6473047043114694,
+      "learning_rate": 2.8644996396590657e-06,
+      "loss": 0.5249,
+      "step": 1100
+    },
+    {
+      "epoch": 2.0300982800982803,
+      "grad_norm": 0.6035938882626676,
+      "learning_rate": 2.8547962939308187e-06,
+      "loss": 0.5405,
+      "step": 1101
+    },
+    {
+      "epoch": 2.031941031941032,
+      "grad_norm": 0.584414907096397,
+      "learning_rate": 2.8451028400968233e-06,
+      "loss": 0.5005,
+      "step": 1102
+    },
+    {
+      "epoch": 2.0337837837837838,
+      "grad_norm": 0.5837151379319605,
+      "learning_rate": 2.8354193228552195e-06,
+      "loss": 0.5485,
+      "step": 1103
+    },
+    {
+      "epoch": 2.0356265356265357,
+      "grad_norm": 0.6279721939237544,
+      "learning_rate": 2.8257457868583283e-06,
+      "loss": 0.5299,
+      "step": 1104
+    },
+    {
+      "epoch": 2.0374692874692877,
+      "grad_norm": 0.5953485787205862,
+      "learning_rate": 2.816082276712452e-06,
+      "loss": 0.5137,
+      "step": 1105
+    },
+    {
+      "epoch": 2.039312039312039,
+      "grad_norm": 0.6063034571501409,
+      "learning_rate": 2.806428836977654e-06,
+      "loss": 0.524,
+      "step": 1106
+    },
+    {
+      "epoch": 2.041154791154791,
+      "grad_norm": 0.6028434423369535,
+      "learning_rate": 2.796785512167563e-06,
+      "loss": 0.5143,
+      "step": 1107
+    },
+    {
+      "epoch": 2.042997542997543,
+      "grad_norm": 0.6153119981116221,
+      "learning_rate": 2.787152346749173e-06,
+      "loss": 0.4966,
+      "step": 1108
+    },
+    {
+      "epoch": 2.044840294840295,
+      "grad_norm": 0.5876485088413107,
+      "learning_rate": 2.7775293851426233e-06,
+      "loss": 0.5106,
+      "step": 1109
+    },
+    {
+      "epoch": 2.0466830466830466,
+      "grad_norm": 0.5671110802346316,
+      "learning_rate": 2.7679166717210027e-06,
+      "loss": 0.5102,
+      "step": 1110
+    },
+    {
+      "epoch": 2.0485257985257985,
+      "grad_norm": 0.5578858905926076,
+      "learning_rate": 2.7583142508101495e-06,
+      "loss": 0.4939,
+      "step": 1111
+    },
+    {
+      "epoch": 2.0503685503685505,
+      "grad_norm": 0.6008371188920137,
+      "learning_rate": 2.748722166688437e-06,
+      "loss": 0.5255,
+      "step": 1112
+    },
+    {
+      "epoch": 2.052211302211302,
+      "grad_norm": 0.5973411746553632,
+      "learning_rate": 2.7391404635865725e-06,
+      "loss": 0.5094,
+      "step": 1113
+    },
+    {
+      "epoch": 2.054054054054054,
+      "grad_norm": 0.5791057174947255,
+      "learning_rate": 2.7295691856873973e-06,
+      "loss": 0.514,
+      "step": 1114
+    },
+    {
+      "epoch": 2.055896805896806,
+      "grad_norm": 0.6388875054001032,
+      "learning_rate": 2.720008377125682e-06,
+      "loss": 0.4979,
+      "step": 1115
+    },
+    {
+      "epoch": 2.057739557739558,
+      "grad_norm": 0.5205414697967802,
+      "learning_rate": 2.710458081987918e-06,
+      "loss": 0.5523,
+      "step": 1116
+    },
+    {
+      "epoch": 2.0595823095823094,
+      "grad_norm": 0.5708294547131824,
+      "learning_rate": 2.7009183443121154e-06,
+      "loss": 0.5218,
+      "step": 1117
+    },
+    {
+      "epoch": 2.0614250614250613,
+      "grad_norm": 0.6554067933040685,
+      "learning_rate": 2.6913892080876116e-06,
+      "loss": 0.521,
+      "step": 1118
+    },
+    {
+      "epoch": 2.0632678132678133,
+      "grad_norm": 0.5472828157533463,
+      "learning_rate": 2.6818707172548486e-06,
+      "loss": 0.5408,
+      "step": 1119
+    },
+    {
+      "epoch": 2.0651105651105652,
+      "grad_norm": 0.5672331816792499,
+      "learning_rate": 2.6723629157051844e-06,
+      "loss": 0.5164,
+      "step": 1120
+    },
+    {
+      "epoch": 2.0669533169533167,
+      "grad_norm": 0.5311493414067971,
+      "learning_rate": 2.66286584728069e-06,
+      "loss": 0.5345,
+      "step": 1121
+    },
+    {
+      "epoch": 2.0687960687960687,
+      "grad_norm": 0.6317598859437235,
+      "learning_rate": 2.6533795557739407e-06,
+      "loss": 0.5151,
+      "step": 1122
+    },
+    {
+      "epoch": 2.0706388206388207,
+      "grad_norm": 0.5618564156104144,
+      "learning_rate": 2.643904084927815e-06,
+      "loss": 0.5142,
+      "step": 1123
+    },
+    {
+      "epoch": 2.0724815724815726,
+      "grad_norm": 0.5884166255646663,
+      "learning_rate": 2.634439478435305e-06,
+      "loss": 0.5065,
+      "step": 1124
+    },
+    {
+      "epoch": 2.074324324324324,
+      "grad_norm": 0.5890430918570801,
+      "learning_rate": 2.624985779939295e-06,
+      "loss": 0.5106,
+      "step": 1125
+    },
+    {
+      "epoch": 2.076167076167076,
+      "grad_norm": 0.605377741625182,
+      "learning_rate": 2.6155430330323756e-06,
+      "loss": 0.5059,
+      "step": 1126
+    },
+    {
+      "epoch": 2.078009828009828,
+      "grad_norm": 0.5231337633268289,
+      "learning_rate": 2.606111281256639e-06,
+      "loss": 0.5217,
+      "step": 1127
+    },
+    {
+      "epoch": 2.07985257985258,
+      "grad_norm": 0.5926145129194215,
+      "learning_rate": 2.596690568103474e-06,
+      "loss": 0.5108,
+      "step": 1128
+    },
+    {
+      "epoch": 2.0816953316953315,
+      "grad_norm": 0.541273027340509,
+      "learning_rate": 2.5872809370133704e-06,
+      "loss": 0.5183,
+      "step": 1129
+    },
+    {
+      "epoch": 2.0835380835380835,
+      "grad_norm": 0.5827360052111786,
+      "learning_rate": 2.5778824313757136e-06,
+      "loss": 0.5183,
+      "step": 1130
+    },
+    {
+      "epoch": 2.0853808353808354,
+      "grad_norm": 0.5997049062092007,
+      "learning_rate": 2.5684950945285937e-06,
+      "loss": 0.5269,
+      "step": 1131
+    },
+    {
+      "epoch": 2.0872235872235874,
+      "grad_norm": 0.5911092906210775,
+      "learning_rate": 2.559118969758595e-06,
+      "loss": 0.5027,
+      "step": 1132
+    },
+    {
+      "epoch": 2.089066339066339,
+      "grad_norm": 0.5923417866871526,
+      "learning_rate": 2.549754100300599e-06,
+      "loss": 0.5264,
+      "step": 1133
+    },
+    {
+      "epoch": 2.090909090909091,
+      "grad_norm": 0.6054347760170429,
+      "learning_rate": 2.5404005293375955e-06,
+      "loss": 0.5195,
+      "step": 1134
+    },
+    {
+      "epoch": 2.092751842751843,
+      "grad_norm": 0.5505120721028511,
+      "learning_rate": 2.5310583000004664e-06,
+      "loss": 0.5115,
+      "step": 1135
+    },
+    {
+      "epoch": 2.0945945945945947,
+      "grad_norm": 0.5338351663034379,
+      "learning_rate": 2.5217274553677975e-06,
+      "loss": 0.5072,
+      "step": 1136
+    },
+    {
+      "epoch": 2.0964373464373462,
+      "grad_norm": 0.6071538425139407,
+      "learning_rate": 2.512408038465683e-06,
+      "loss": 0.5148,
+      "step": 1137
+    },
+    {
+      "epoch": 2.098280098280098,
+      "grad_norm": 0.5329022625896453,
+      "learning_rate": 2.5031000922675164e-06,
+      "loss": 0.5359,
+      "step": 1138
+    },
+    {
+      "epoch": 2.10012285012285,
+      "grad_norm": 0.5666343067662356,
+      "learning_rate": 2.4938036596937966e-06,
+      "loss": 0.5355,
+      "step": 1139
+    },
+    {
+      "epoch": 2.101965601965602,
+      "grad_norm": 0.606439431637795,
+      "learning_rate": 2.4845187836119404e-06,
+      "loss": 0.529,
+      "step": 1140
+    },
+    {
+      "epoch": 2.1038083538083536,
+      "grad_norm": 0.5686581891994442,
+      "learning_rate": 2.475245506836062e-06,
+      "loss": 0.5261,
+      "step": 1141
+    },
+    {
+      "epoch": 2.1056511056511056,
+      "grad_norm": 0.5834399980313787,
+      "learning_rate": 2.4659838721268005e-06,
+      "loss": 0.5187,
+      "step": 1142
+    },
+    {
+      "epoch": 2.1074938574938575,
+      "grad_norm": 0.5662742798434274,
+      "learning_rate": 2.4567339221911086e-06,
+      "loss": 0.523,
+      "step": 1143
+    },
+    {
+      "epoch": 2.1093366093366095,
+      "grad_norm": 0.6135398969568191,
+      "learning_rate": 2.4474956996820564e-06,
+      "loss": 0.521,
+      "step": 1144
+    },
+    {
+      "epoch": 2.111179361179361,
+      "grad_norm": 0.5995910512735231,
+      "learning_rate": 2.438269247198637e-06,
+      "loss": 0.5207,
+      "step": 1145
+    },
+    {
+      "epoch": 2.113022113022113,
+      "grad_norm": 0.5579767219614573,
+      "learning_rate": 2.4290546072855733e-06,
+      "loss": 0.5277,
+      "step": 1146
+    },
+    {
+      "epoch": 2.114864864864865,
+      "grad_norm": 0.5905925464894933,
+      "learning_rate": 2.419851822433114e-06,
+      "loss": 0.508,
+      "step": 1147
+    },
+    {
+      "epoch": 2.116707616707617,
+      "grad_norm": 0.5988224969332959,
+      "learning_rate": 2.410660935076846e-06,
+      "loss": 0.5192,
+      "step": 1148
+    },
+    {
+      "epoch": 2.1185503685503684,
+      "grad_norm": 0.5586228839014108,
+      "learning_rate": 2.4014819875974905e-06,
+      "loss": 0.4933,
+      "step": 1149
+    },
+    {
+      "epoch": 2.1203931203931203,
+      "grad_norm": 0.5852746359918719,
+      "learning_rate": 2.3923150223207176e-06,
+      "loss": 0.5317,
+      "step": 1150
+    },
+    {
+      "epoch": 2.1222358722358723,
+      "grad_norm": 0.6220286028572775,
+      "learning_rate": 2.383160081516941e-06,
+      "loss": 0.5346,
+      "step": 1151
+    },
+    {
+      "epoch": 2.1240786240786242,
+      "grad_norm": 0.5486004249305663,
+      "learning_rate": 2.3740172074011268e-06,
+      "loss": 0.5295,
+      "step": 1152
+    },
+    {
+      "epoch": 2.1259213759213758,
+      "grad_norm": 0.5688636792988703,
+      "learning_rate": 2.364886442132606e-06,
+      "loss": 0.5101,
+      "step": 1153
+    },
+    {
+      "epoch": 2.1277641277641277,
+      "grad_norm": 0.5970115321149821,
+      "learning_rate": 2.3557678278148675e-06,
+      "loss": 0.545,
+      "step": 1154
+    },
+    {
+      "epoch": 2.1296068796068797,
+      "grad_norm": 0.5838605260321369,
+      "learning_rate": 2.3466614064953697e-06,
+      "loss": 0.4949,
+      "step": 1155
+    },
+    {
+      "epoch": 2.1314496314496316,
+      "grad_norm": 0.5418892814790799,
+      "learning_rate": 2.337567220165353e-06,
+      "loss": 0.5346,
+      "step": 1156
+    },
+    {
+      "epoch": 2.133292383292383,
+      "grad_norm": 0.6026519254665637,
+      "learning_rate": 2.328485310759635e-06,
+      "loss": 0.5387,
+      "step": 1157
+    },
+    {
+      "epoch": 2.135135135135135,
+      "grad_norm": 0.5810054522557371,
+      "learning_rate": 2.319415720156422e-06,
+      "loss": 0.5081,
+      "step": 1158
+    },
+    {
+      "epoch": 2.136977886977887,
+      "grad_norm": 0.5847508638469887,
+      "learning_rate": 2.310358490177124e-06,
+      "loss": 0.5197,
+      "step": 1159
+    },
+    {
+      "epoch": 2.138820638820639,
+      "grad_norm": 0.5652781096409624,
+      "learning_rate": 2.3013136625861397e-06,
+      "loss": 0.5284,
+      "step": 1160
+    },
+    {
+      "epoch": 2.1406633906633905,
+      "grad_norm": 0.5921811146313812,
+      "learning_rate": 2.292281279090692e-06,
+      "loss": 0.5374,
+      "step": 1161
+    },
+    {
+      "epoch": 2.1425061425061425,
+      "grad_norm": 0.5793768540870367,
+      "learning_rate": 2.2832613813406197e-06,
+      "loss": 0.4929,
+      "step": 1162
+    },
+    {
+      "epoch": 2.1443488943488944,
+      "grad_norm": 0.557636111375881,
+      "learning_rate": 2.274254010928177e-06,
+      "loss": 0.5367,
+      "step": 1163
+    },
+    {
+      "epoch": 2.1461916461916464,
+      "grad_norm": 0.5675790348857245,
+      "learning_rate": 2.265259209387867e-06,
+      "loss": 0.5289,
+      "step": 1164
+    },
+    {
+      "epoch": 2.148034398034398,
+      "grad_norm": 0.5383489526575537,
+      "learning_rate": 2.256277018196224e-06,
+      "loss": 0.5434,
+      "step": 1165
+    },
+    {
+      "epoch": 2.14987714987715,
+      "grad_norm": 0.5998794263035954,
+      "learning_rate": 2.247307478771643e-06,
+      "loss": 0.5165,
+      "step": 1166
+    },
+    {
+      "epoch": 2.151719901719902,
+      "grad_norm": 0.5629180781775777,
+      "learning_rate": 2.2383506324741742e-06,
+      "loss": 0.5211,
+      "step": 1167
+    },
+    {
+      "epoch": 2.1535626535626538,
+      "grad_norm": 0.5589908083584408,
+      "learning_rate": 2.229406520605336e-06,
+      "loss": 0.522,
+      "step": 1168
+    },
+    {
+      "epoch": 2.1554054054054053,
+      "grad_norm": 0.5754613044906794,
+      "learning_rate": 2.220475184407933e-06,
+      "loss": 0.5161,
+      "step": 1169
+    },
+    {
+      "epoch": 2.157248157248157,
+      "grad_norm": 0.5778337255607942,
+      "learning_rate": 2.211556665065854e-06,
+      "loss": 0.5624,
+      "step": 1170
+    },
+    {
+      "epoch": 2.159090909090909,
+      "grad_norm": 0.5644667253497806,
+      "learning_rate": 2.202651003703885e-06,
+      "loss": 0.53,
+      "step": 1171
+    },
+    {
+      "epoch": 2.160933660933661,
+      "grad_norm": 0.5629686879518514,
+      "learning_rate": 2.193758241387529e-06,
+      "loss": 0.4804,
+      "step": 1172
+    },
+    {
+      "epoch": 2.1627764127764126,
+      "grad_norm": 0.5407277191623603,
+      "learning_rate": 2.184878419122804e-06,
+      "loss": 0.5269,
+      "step": 1173
+    },
+    {
+      "epoch": 2.1646191646191646,
+      "grad_norm": 0.5457802418155687,
+      "learning_rate": 2.176011577856058e-06,
+      "loss": 0.5067,
+      "step": 1174
+    },
+    {
+      "epoch": 2.1664619164619165,
+      "grad_norm": 0.566041905704649,
+      "learning_rate": 2.16715775847379e-06,
+      "loss": 0.521,
+      "step": 1175
+    },
+    {
+      "epoch": 2.1683046683046685,
+      "grad_norm": 0.5897296118964628,
+      "learning_rate": 2.158317001802439e-06,
+      "loss": 0.515,
+      "step": 1176
+    },
+    {
+      "epoch": 2.17014742014742,
+      "grad_norm": 0.5862365091580539,
+      "learning_rate": 2.149489348608221e-06,
+      "loss": 0.5156,
+      "step": 1177
+    },
+    {
+      "epoch": 2.171990171990172,
+      "grad_norm": 0.6351610819535158,
+      "learning_rate": 2.140674839596931e-06,
+      "loss": 0.5102,
+      "step": 1178
+    },
+    {
+      "epoch": 2.173832923832924,
+      "grad_norm": 0.578736760357454,
+      "learning_rate": 2.1318735154137405e-06,
+      "loss": 0.5436,
+      "step": 1179
+    },
+    {
+      "epoch": 2.175675675675676,
+      "grad_norm": 0.5202251509691359,
+      "learning_rate": 2.1230854166430374e-06,
+      "loss": 0.5217,
+      "step": 1180
+    },
+    {
+      "epoch": 2.1775184275184274,
+      "grad_norm": 0.5427089783201401,
+      "learning_rate": 2.114310583808219e-06,
+      "loss": 0.5176,
+      "step": 1181
+    },
+    {
+      "epoch": 2.1793611793611793,
+      "grad_norm": 0.5522980366025072,
+      "learning_rate": 2.10554905737151e-06,
+      "loss": 0.5166,
+      "step": 1182
+    },
+    {
+      "epoch": 2.1812039312039313,
+      "grad_norm": 0.557287180794352,
+      "learning_rate": 2.0968008777337818e-06,
+      "loss": 0.5218,
+      "step": 1183
+    },
+    {
+      "epoch": 2.1830466830466833,
+      "grad_norm": 0.5436878047426557,
+      "learning_rate": 2.088066085234357e-06,
+      "loss": 0.5275,
+      "step": 1184
+    },
+    {
+      "epoch": 2.1848894348894348,
+      "grad_norm": 0.5966695384500458,
+      "learning_rate": 2.0793447201508288e-06,
+      "loss": 0.5357,
+      "step": 1185
+    },
+    {
+      "epoch": 2.1867321867321867,
+      "grad_norm": 0.5627392534462152,
+      "learning_rate": 2.0706368226988772e-06,
+      "loss": 0.5142,
+      "step": 1186
+    },
+    {
+      "epoch": 2.1885749385749387,
+      "grad_norm": 0.5668398578112559,
+      "learning_rate": 2.0619424330320752e-06,
+      "loss": 0.5274,
+      "step": 1187
+    },
+    {
+      "epoch": 2.19041769041769,
+      "grad_norm": 0.6244966921911362,
+      "learning_rate": 2.053261591241717e-06,
+      "loss": 0.4928,
+      "step": 1188
+    },
+    {
+      "epoch": 2.192260442260442,
+      "grad_norm": 0.5798662390807398,
+      "learning_rate": 2.0445943373566178e-06,
+      "loss": 0.5103,
+      "step": 1189
+    },
+    {
+      "epoch": 2.194103194103194,
+      "grad_norm": 0.5279141194447864,
+      "learning_rate": 2.0359407113429385e-06,
+      "loss": 0.5105,
+      "step": 1190
+    },
+    {
+      "epoch": 2.195945945945946,
+      "grad_norm": 0.5599190511338266,
+      "learning_rate": 2.0273007531040044e-06,
+      "loss": 0.5054,
+      "step": 1191
+    },
+    {
+      "epoch": 2.197788697788698,
+      "grad_norm": 0.5874355885919127,
+      "learning_rate": 2.01867450248011e-06,
+      "loss": 0.4857,
+      "step": 1192
+    },
+    {
+      "epoch": 2.1996314496314495,
+      "grad_norm": 0.6276722336229632,
+      "learning_rate": 2.010061999248344e-06,
+      "loss": 0.5085,
+      "step": 1193
+    },
+    {
+      "epoch": 2.2014742014742015,
+      "grad_norm": 0.5982452972822442,
+      "learning_rate": 2.001463283122408e-06,
+      "loss": 0.5084,
+      "step": 1194
+    },
+    {
+      "epoch": 2.2033169533169534,
+      "grad_norm": 0.5546082691796922,
+      "learning_rate": 1.9928783937524184e-06,
+      "loss": 0.5316,
+      "step": 1195
+    },
+    {
+      "epoch": 2.205159705159705,
+      "grad_norm": 0.5308655101194789,
+      "learning_rate": 1.984307370724744e-06,
+      "loss": 0.5411,
+      "step": 1196
+    },
+    {
+      "epoch": 2.207002457002457,
+      "grad_norm": 0.551045489426061,
+      "learning_rate": 1.9757502535618137e-06,
+      "loss": 0.5194,
+      "step": 1197
+    },
+    {
+      "epoch": 2.208845208845209,
+      "grad_norm": 0.5279517381055028,
+      "learning_rate": 1.9672070817219242e-06,
+      "loss": 0.5079,
+      "step": 1198
+    },
+    {
+      "epoch": 2.210687960687961,
+      "grad_norm": 0.5672064814093597,
+      "learning_rate": 1.9586778945990785e-06,
+      "loss": 0.5188,
+      "step": 1199
+    },
+    {
+      "epoch": 2.2125307125307128,
+      "grad_norm": 0.5654634262394465,
+      "learning_rate": 1.9501627315227887e-06,
+      "loss": 0.5379,
+      "step": 1200
+    },
+    {
+      "epoch": 2.2143734643734643,
+      "grad_norm": 0.4931701400642895,
+      "learning_rate": 1.941661631757899e-06,
+      "loss": 0.5447,
+      "step": 1201
+    },
+    {
+      "epoch": 2.2162162162162162,
+      "grad_norm": 0.5385434687107005,
+      "learning_rate": 1.9331746345044095e-06,
+      "loss": 0.5299,
+      "step": 1202
+    },
+    {
+      "epoch": 2.218058968058968,
+      "grad_norm": 0.5443422653037998,
+      "learning_rate": 1.9247017788972873e-06,
+      "loss": 0.526,
+      "step": 1203
+    },
+    {
+      "epoch": 2.2199017199017197,
+      "grad_norm": 0.5321049197303005,
+      "learning_rate": 1.9162431040062897e-06,
+      "loss": 0.5325,
+      "step": 1204
+    },
+    {
+      "epoch": 2.2217444717444716,
+      "grad_norm": 0.519092045860744,
+      "learning_rate": 1.907798648835789e-06,
+      "loss": 0.5087,
+      "step": 1205
+    },
+    {
+      "epoch": 2.2235872235872236,
+      "grad_norm": 0.5878238806520232,
+      "learning_rate": 1.8993684523245842e-06,
+      "loss": 0.5116,
+      "step": 1206
+    },
+    {
+      "epoch": 2.2254299754299756,
+      "grad_norm": 0.5438006753138266,
+      "learning_rate": 1.8909525533457234e-06,
+      "loss": 0.5335,
+      "step": 1207
+    },
+    {
+      "epoch": 2.227272727272727,
+      "grad_norm": 0.5738693326338455,
+      "learning_rate": 1.8825509907063328e-06,
+      "loss": 0.5277,
+      "step": 1208
+    },
+    {
+      "epoch": 2.229115479115479,
+      "grad_norm": 0.5754954261773169,
+      "learning_rate": 1.874163803147424e-06,
+      "loss": 0.5156,
+      "step": 1209
+    },
+    {
+      "epoch": 2.230958230958231,
+      "grad_norm": 0.6017931457565313,
+      "learning_rate": 1.865791029343731e-06,
+      "loss": 0.5175,
+      "step": 1210
+    },
+    {
+      "epoch": 2.232800982800983,
+      "grad_norm": 0.5387144748433348,
+      "learning_rate": 1.857432707903512e-06,
+      "loss": 0.5176,
+      "step": 1211
+    },
+    {
+      "epoch": 2.2346437346437344,
+      "grad_norm": 0.5331214059856532,
+      "learning_rate": 1.8490888773683935e-06,
+      "loss": 0.5243,
+      "step": 1212
+    },
+    {
+      "epoch": 2.2364864864864864,
+      "grad_norm": 0.5591187903738337,
+      "learning_rate": 1.8407595762131814e-06,
+      "loss": 0.5308,
+      "step": 1213
+    },
+    {
+      "epoch": 2.2383292383292384,
+      "grad_norm": 0.5724450703047224,
+      "learning_rate": 1.8324448428456742e-06,
+      "loss": 0.5106,
+      "step": 1214
+    },
+    {
+      "epoch": 2.2401719901719903,
+      "grad_norm": 0.570517620587082,
+      "learning_rate": 1.8241447156065073e-06,
+      "loss": 0.5189,
+      "step": 1215
+    },
+    {
+      "epoch": 2.242014742014742,
+      "grad_norm": 0.5930208686401871,
+      "learning_rate": 1.815859232768959e-06,
+      "loss": 0.5236,
+      "step": 1216
+    },
+    {
+      "epoch": 2.243857493857494,
+      "grad_norm": 0.5990095902012088,
+      "learning_rate": 1.8075884325387788e-06,
+      "loss": 0.5002,
+      "step": 1217
+    },
+    {
+      "epoch": 2.2457002457002457,
+      "grad_norm": 0.5500522789796249,
+      "learning_rate": 1.7993323530540185e-06,
+      "loss": 0.5284,
+      "step": 1218
+    },
+    {
+      "epoch": 2.2475429975429977,
+      "grad_norm": 0.5582256264534219,
+      "learning_rate": 1.7910910323848435e-06,
+      "loss": 0.5111,
+      "step": 1219
+    },
+    {
+      "epoch": 2.249385749385749,
+      "grad_norm": 0.5587994124133144,
+      "learning_rate": 1.7828645085333645e-06,
+      "loss": 0.5366,
+      "step": 1220
+    },
+    {
+      "epoch": 2.251228501228501,
+      "grad_norm": 0.5512930345718272,
+      "learning_rate": 1.7746528194334678e-06,
+      "loss": 0.5073,
+      "step": 1221
+    },
+    {
+      "epoch": 2.253071253071253,
+      "grad_norm": 0.5796829196802177,
+      "learning_rate": 1.7664560029506268e-06,
+      "loss": 0.5086,
+      "step": 1222
+    },
+    {
+      "epoch": 2.254914004914005,
+      "grad_norm": 0.5553930701167301,
+      "learning_rate": 1.758274096881737e-06,
+      "loss": 0.5128,
+      "step": 1223
+    },
+    {
+      "epoch": 2.2567567567567566,
+      "grad_norm": 0.6068263809880168,
+      "learning_rate": 1.7501071389549429e-06,
+      "loss": 0.5352,
+      "step": 1224
+    },
+    {
+      "epoch": 2.2585995085995085,
+      "grad_norm": 0.7880292733925895,
+      "learning_rate": 1.7419551668294566e-06,
+      "loss": 0.5334,
+      "step": 1225
+    },
+    {
+      "epoch": 2.2604422604422605,
+      "grad_norm": 0.5548092694749028,
+      "learning_rate": 1.7338182180953889e-06,
+      "loss": 0.4858,
+      "step": 1226
+    },
+    {
+      "epoch": 2.2622850122850124,
+      "grad_norm": 0.5821867651962561,
+      "learning_rate": 1.7256963302735752e-06,
+      "loss": 0.521,
+      "step": 1227
+    },
+    {
+      "epoch": 2.264127764127764,
+      "grad_norm": 0.5828035568733742,
+      "learning_rate": 1.7175895408154037e-06,
+      "loss": 0.5315,
+      "step": 1228
+    },
+    {
+      "epoch": 2.265970515970516,
+      "grad_norm": 0.5500828125710677,
+      "learning_rate": 1.7094978871026447e-06,
+      "loss": 0.5565,
+      "step": 1229
+    },
+    {
+      "epoch": 2.267813267813268,
+      "grad_norm": 0.5512535674789529,
+      "learning_rate": 1.7014214064472646e-06,
+      "loss": 0.5411,
+      "step": 1230
+    },
+    {
+      "epoch": 2.26965601965602,
+      "grad_norm": 0.5872935486862136,
+      "learning_rate": 1.6933601360912737e-06,
+      "loss": 0.5355,
+      "step": 1231
+    },
+    {
+      "epoch": 2.2714987714987713,
+      "grad_norm": 0.5707371713911397,
+      "learning_rate": 1.6853141132065465e-06,
+      "loss": 0.5059,
+      "step": 1232
+    },
+    {
+      "epoch": 2.2733415233415233,
+      "grad_norm": 0.5494383172009434,
+      "learning_rate": 1.6772833748946377e-06,
+      "loss": 0.5295,
+      "step": 1233
+    },
+    {
+      "epoch": 2.2751842751842752,
+      "grad_norm": 0.5757294030357671,
+      "learning_rate": 1.6692679581866334e-06,
+      "loss": 0.5186,
+      "step": 1234
+    },
+    {
+      "epoch": 2.277027027027027,
+      "grad_norm": 0.5477640055164246,
+      "learning_rate": 1.6612679000429644e-06,
+      "loss": 0.5098,
+      "step": 1235
+    },
+    {
+      "epoch": 2.2788697788697787,
+      "grad_norm": 0.5820195644725414,
+      "learning_rate": 1.6532832373532388e-06,
+      "loss": 0.5345,
+      "step": 1236
+    },
+    {
+      "epoch": 2.2807125307125307,
+      "grad_norm": 0.5637409178779166,
+      "learning_rate": 1.6453140069360795e-06,
+      "loss": 0.5301,
+      "step": 1237
+    },
+    {
+      "epoch": 2.2825552825552826,
+      "grad_norm": 0.5783895058900196,
+      "learning_rate": 1.6373602455389443e-06,
+      "loss": 0.5314,
+      "step": 1238
+    },
+    {
+      "epoch": 2.2843980343980346,
+      "grad_norm": 0.569708547362023,
+      "learning_rate": 1.6294219898379587e-06,
+      "loss": 0.5129,
+      "step": 1239
+    },
+    {
+      "epoch": 2.286240786240786,
+      "grad_norm": 0.5277332143091021,
+      "learning_rate": 1.6214992764377563e-06,
+      "loss": 0.5336,
+      "step": 1240
+    },
+    {
+      "epoch": 2.288083538083538,
+      "grad_norm": 0.5995226466344877,
+      "learning_rate": 1.6135921418712959e-06,
+      "loss": 0.5094,
+      "step": 1241
+    },
+    {
+      "epoch": 2.28992628992629,
+      "grad_norm": 0.5598636158247345,
+      "learning_rate": 1.605700622599699e-06,
+      "loss": 0.5146,
+      "step": 1242
+    },
+    {
+      "epoch": 2.291769041769042,
+      "grad_norm": 0.5734850528507046,
+      "learning_rate": 1.5978247550120906e-06,
+      "loss": 0.5013,
+      "step": 1243
+    },
+    {
+      "epoch": 2.2936117936117935,
+      "grad_norm": 0.5362882637500892,
+      "learning_rate": 1.5899645754254144e-06,
+      "loss": 0.526,
+      "step": 1244
+    },
+    {
+      "epoch": 2.2954545454545454,
+      "grad_norm": 0.5425905797996385,
+      "learning_rate": 1.5821201200842779e-06,
+      "loss": 0.5286,
+      "step": 1245
+    },
+    {
+      "epoch": 2.2972972972972974,
+      "grad_norm": 0.566524472751281,
+      "learning_rate": 1.5742914251607794e-06,
+      "loss": 0.5167,
+      "step": 1246
+    },
+    {
+      "epoch": 2.2991400491400493,
+      "grad_norm": 0.5862545032979197,
+      "learning_rate": 1.566478526754348e-06,
+      "loss": 0.5215,
+      "step": 1247
+    },
+    {
+      "epoch": 2.300982800982801,
+      "grad_norm": 0.5698265223741127,
+      "learning_rate": 1.5586814608915673e-06,
+      "loss": 0.5585,
+      "step": 1248
+    },
+    {
+      "epoch": 2.302825552825553,
+      "grad_norm": 0.5483486344026527,
+      "learning_rate": 1.5509002635260156e-06,
+      "loss": 0.5297,
+      "step": 1249
+    },
+    {
+      "epoch": 2.3046683046683047,
+      "grad_norm": 0.5481540852750904,
+      "learning_rate": 1.5431349705381021e-06,
+      "loss": 0.5186,
+      "step": 1250
+    },
+    {
+      "epoch": 2.3065110565110567,
+      "grad_norm": 0.5318081616997268,
+      "learning_rate": 1.5353856177348942e-06,
+      "loss": 0.5444,
+      "step": 1251
+    },
+    {
+      "epoch": 2.308353808353808,
+      "grad_norm": 0.5586360751815707,
+      "learning_rate": 1.5276522408499567e-06,
+      "loss": 0.5168,
+      "step": 1252
+    },
+    {
+      "epoch": 2.31019656019656,
+      "grad_norm": 0.5479270193522238,
+      "learning_rate": 1.5199348755431915e-06,
+      "loss": 0.5413,
+      "step": 1253
+    },
+    {
+      "epoch": 2.312039312039312,
+      "grad_norm": 0.5219023561514311,
+      "learning_rate": 1.5122335574006624e-06,
+      "loss": 0.5306,
+      "step": 1254
+    },
+    {
+      "epoch": 2.3138820638820636,
+      "grad_norm": 0.5504143709403099,
+      "learning_rate": 1.5045483219344387e-06,
+      "loss": 0.5134,
+      "step": 1255
+    },
+    {
+      "epoch": 2.3157248157248156,
+      "grad_norm": 0.565115369890699,
+      "learning_rate": 1.4968792045824338e-06,
+      "loss": 0.513,
+      "step": 1256
+    },
+    {
+      "epoch": 2.3175675675675675,
+      "grad_norm": 0.5577947797751039,
+      "learning_rate": 1.4892262407082336e-06,
+      "loss": 0.5278,
+      "step": 1257
+    },
+    {
+      "epoch": 2.3194103194103195,
+      "grad_norm": 0.5663300636987936,
+      "learning_rate": 1.4815894656009361e-06,
+      "loss": 0.5165,
+      "step": 1258
+    },
+    {
+      "epoch": 2.3212530712530715,
+      "grad_norm": 0.5656798893326416,
+      "learning_rate": 1.4739689144749963e-06,
+      "loss": 0.4957,
+      "step": 1259
+    },
+    {
+      "epoch": 2.323095823095823,
+      "grad_norm": 0.5766785748493637,
+      "learning_rate": 1.4663646224700534e-06,
+      "loss": 0.5201,
+      "step": 1260
+    },
+    {
+      "epoch": 2.324938574938575,
+      "grad_norm": 0.5855696215187147,
+      "learning_rate": 1.4587766246507723e-06,
+      "loss": 0.5214,
+      "step": 1261
+    },
+    {
+      "epoch": 2.326781326781327,
+      "grad_norm": 0.5684800739549024,
+      "learning_rate": 1.4512049560066837e-06,
+      "loss": 0.528,
+      "step": 1262
+    },
+    {
+      "epoch": 2.3286240786240784,
+      "grad_norm": 0.5418105798232908,
+      "learning_rate": 1.4436496514520253e-06,
+      "loss": 0.5197,
+      "step": 1263
+    },
+    {
+      "epoch": 2.3304668304668303,
+      "grad_norm": 0.5591126951147007,
+      "learning_rate": 1.4361107458255723e-06,
+      "loss": 0.4929,
+      "step": 1264
+    },
+    {
+      "epoch": 2.3323095823095823,
+      "grad_norm": 0.5782620524752954,
+      "learning_rate": 1.4285882738904822e-06,
+      "loss": 0.503,
+      "step": 1265
+    },
+    {
+      "epoch": 2.3341523341523343,
+      "grad_norm": 0.5728256147075521,
+      "learning_rate": 1.421082270334138e-06,
+      "loss": 0.527,
+      "step": 1266
+    },
+    {
+      "epoch": 2.335995085995086,
+      "grad_norm": 0.5611381718732993,
+      "learning_rate": 1.4135927697679807e-06,
+      "loss": 0.5296,
+      "step": 1267
+    },
+    {
+      "epoch": 2.3378378378378377,
+      "grad_norm": 0.561904452108065,
+      "learning_rate": 1.4061198067273519e-06,
+      "loss": 0.5189,
+      "step": 1268
+    },
+    {
+      "epoch": 2.3396805896805897,
+      "grad_norm": 0.5691739423273924,
+      "learning_rate": 1.3986634156713418e-06,
+      "loss": 0.5073,
+      "step": 1269
+    },
+    {
+      "epoch": 2.3415233415233416,
+      "grad_norm": 0.5520137569709835,
+      "learning_rate": 1.3912236309826193e-06,
+      "loss": 0.5041,
+      "step": 1270
+    },
+    {
+      "epoch": 2.343366093366093,
+      "grad_norm": 0.5585445018729781,
+      "learning_rate": 1.3838004869672778e-06,
+      "loss": 0.5268,
+      "step": 1271
+    },
+    {
+      "epoch": 2.345208845208845,
+      "grad_norm": 0.5812491140152132,
+      "learning_rate": 1.3763940178546836e-06,
+      "loss": 0.5199,
+      "step": 1272
+    },
+    {
+      "epoch": 2.347051597051597,
+      "grad_norm": 0.5912827124035164,
+      "learning_rate": 1.3690042577973062e-06,
+      "loss": 0.4975,
+      "step": 1273
+    },
+    {
+      "epoch": 2.348894348894349,
+      "grad_norm": 0.5677115163159278,
+      "learning_rate": 1.361631240870569e-06,
+      "loss": 0.4982,
+      "step": 1274
+    },
+    {
+      "epoch": 2.350737100737101,
+      "grad_norm": 0.5655224688887805,
+      "learning_rate": 1.3542750010726919e-06,
+      "loss": 0.5356,
+      "step": 1275
+    },
+    {
+      "epoch": 2.3525798525798525,
+      "grad_norm": 0.5399817898861393,
+      "learning_rate": 1.3469355723245303e-06,
+      "loss": 0.5382,
+      "step": 1276
+    },
+    {
+      "epoch": 2.3544226044226044,
+      "grad_norm": 0.5656896785876854,
+      "learning_rate": 1.3396129884694197e-06,
+      "loss": 0.5192,
+      "step": 1277
+    },
+    {
+      "epoch": 2.3562653562653564,
+      "grad_norm": 0.5588057994634386,
+      "learning_rate": 1.3323072832730267e-06,
+      "loss": 0.4785,
+      "step": 1278
+    },
+    {
+      "epoch": 2.358108108108108,
+      "grad_norm": 0.5402339190440288,
+      "learning_rate": 1.3250184904231816e-06,
+      "loss": 0.5162,
+      "step": 1279
+    },
+    {
+      "epoch": 2.35995085995086,
+      "grad_norm": 0.5519837052680514,
+      "learning_rate": 1.317746643529732e-06,
+      "loss": 0.5248,
+      "step": 1280
+    },
+    {
+      "epoch": 2.361793611793612,
+      "grad_norm": 0.5428453058670644,
+      "learning_rate": 1.310491776124382e-06,
+      "loss": 0.5007,
+      "step": 1281
+    },
+    {
+      "epoch": 2.3636363636363638,
+      "grad_norm": 0.5279216272745427,
+      "learning_rate": 1.3032539216605456e-06,
+      "loss": 0.5223,
+      "step": 1282
+    },
+    {
+      "epoch": 2.3654791154791157,
+      "grad_norm": 0.5357723146846121,
+      "learning_rate": 1.2960331135131826e-06,
+      "loss": 0.5159,
+      "step": 1283
+    },
+    {
+      "epoch": 2.3673218673218672,
+      "grad_norm": 0.5533772971235306,
+      "learning_rate": 1.2888293849786503e-06,
+      "loss": 0.5208,
+      "step": 1284
+    },
+    {
+      "epoch": 2.369164619164619,
+      "grad_norm": 0.559520755979308,
+      "learning_rate": 1.281642769274552e-06,
+      "loss": 0.5145,
+      "step": 1285
+    },
+    {
+      "epoch": 2.371007371007371,
+      "grad_norm": 0.5082431949273397,
+      "learning_rate": 1.2744732995395776e-06,
+      "loss": 0.5284,
+      "step": 1286
+    },
+    {
+      "epoch": 2.3728501228501226,
+      "grad_norm": 0.5363137069961286,
+      "learning_rate": 1.2673210088333526e-06,
+      "loss": 0.526,
+      "step": 1287
+    },
+    {
+      "epoch": 2.3746928746928746,
+      "grad_norm": 0.5736154478012215,
+      "learning_rate": 1.260185930136294e-06,
+      "loss": 0.5313,
+      "step": 1288
+    },
+    {
+      "epoch": 2.3765356265356266,
+      "grad_norm": 0.5720308264692338,
+      "learning_rate": 1.2530680963494445e-06,
+      "loss": 0.5153,
+      "step": 1289
+    },
+    {
+      "epoch": 2.3783783783783785,
+      "grad_norm": 0.5153042538105618,
+      "learning_rate": 1.245967540294329e-06,
+      "loss": 0.5261,
+      "step": 1290
+    },
+    {
+      "epoch": 2.3802211302211305,
+      "grad_norm": 0.5757622093200251,
+      "learning_rate": 1.2388842947128065e-06,
+      "loss": 0.5243,
+      "step": 1291
+    },
+    {
+      "epoch": 2.382063882063882,
+      "grad_norm": 0.5479849453697248,
+      "learning_rate": 1.2318183922669065e-06,
+      "loss": 0.5323,
+      "step": 1292
+    },
+    {
+      "epoch": 2.383906633906634,
+      "grad_norm": 0.5559310169533165,
+      "learning_rate": 1.2247698655386926e-06,
+      "loss": 0.5034,
+      "step": 1293
+    },
+    {
+      "epoch": 2.385749385749386,
+      "grad_norm": 0.5464405786828462,
+      "learning_rate": 1.2177387470301055e-06,
+      "loss": 0.4988,
+      "step": 1294
+    },
+    {
+      "epoch": 2.3875921375921374,
+      "grad_norm": 0.5407195843950098,
+      "learning_rate": 1.2107250691628125e-06,
+      "loss": 0.512,
+      "step": 1295
+    },
+    {
+      "epoch": 2.3894348894348894,
+      "grad_norm": 0.5203140872927583,
+      "learning_rate": 1.2037288642780575e-06,
+      "loss": 0.5135,
+      "step": 1296
+    },
+    {
+      "epoch": 2.3912776412776413,
+      "grad_norm": 0.5326331995897293,
+      "learning_rate": 1.1967501646365147e-06,
+      "loss": 0.5256,
+      "step": 1297
+    },
+    {
+      "epoch": 2.3931203931203933,
+      "grad_norm": 0.54724392010294,
+      "learning_rate": 1.1897890024181418e-06,
+      "loss": 0.5324,
+      "step": 1298
+    },
+    {
+      "epoch": 2.3949631449631448,
+      "grad_norm": 0.5657402896980751,
+      "learning_rate": 1.1828454097220242e-06,
+      "loss": 0.5228,
+      "step": 1299
+    },
+    {
+      "epoch": 2.3968058968058967,
+      "grad_norm": 0.5412915050460042,
+      "learning_rate": 1.175919418566232e-06,
+      "loss": 0.5237,
+      "step": 1300
+    },
+    {
+      "epoch": 2.3986486486486487,
+      "grad_norm": 0.5464765569949633,
+      "learning_rate": 1.1690110608876753e-06,
+      "loss": 0.5219,
+      "step": 1301
+    },
+    {
+      "epoch": 2.4004914004914006,
+      "grad_norm": 0.5417068325019825,
+      "learning_rate": 1.1621203685419485e-06,
+      "loss": 0.5123,
+      "step": 1302
+    },
+    {
+      "epoch": 2.402334152334152,
+      "grad_norm": 0.5331674145053917,
+      "learning_rate": 1.1552473733031893e-06,
+      "loss": 0.5395,
+      "step": 1303
+    },
+    {
+      "epoch": 2.404176904176904,
+      "grad_norm": 0.5438881566563263,
+      "learning_rate": 1.1483921068639353e-06,
+      "loss": 0.5209,
+      "step": 1304
+    },
+    {
+      "epoch": 2.406019656019656,
+      "grad_norm": 0.5788514914566472,
+      "learning_rate": 1.141554600834968e-06,
+      "loss": 0.518,
+      "step": 1305
+    },
+    {
+      "epoch": 2.407862407862408,
+      "grad_norm": 0.5512756162489538,
+      "learning_rate": 1.1347348867451745e-06,
+      "loss": 0.5327,
+      "step": 1306
+    },
+    {
+      "epoch": 2.4097051597051595,
+      "grad_norm": 0.5512345761941424,
+      "learning_rate": 1.1279329960414047e-06,
+      "loss": 0.5085,
+      "step": 1307
+    },
+    {
+      "epoch": 2.4115479115479115,
+      "grad_norm": 0.5513822201780654,
+      "learning_rate": 1.1211489600883124e-06,
+      "loss": 0.5373,
+      "step": 1308
+    },
+    {
+      "epoch": 2.4133906633906634,
+      "grad_norm": 0.5453125933303177,
+      "learning_rate": 1.114382810168228e-06,
+      "loss": 0.5247,
+      "step": 1309
+    },
+    {
+      "epoch": 2.4152334152334154,
+      "grad_norm": 0.5721672291790646,
+      "learning_rate": 1.1076345774810077e-06,
+      "loss": 0.5089,
+      "step": 1310
+    },
+    {
+      "epoch": 2.417076167076167,
+      "grad_norm": 0.5878860669647571,
+      "learning_rate": 1.1009042931438784e-06,
+      "loss": 0.5242,
+      "step": 1311
+    },
+    {
+      "epoch": 2.418918918918919,
+      "grad_norm": 0.5446595093720938,
+      "learning_rate": 1.0941919881913142e-06,
+      "loss": 0.5259,
+      "step": 1312
+    },
+    {
+      "epoch": 2.420761670761671,
+      "grad_norm": 0.5470781542045658,
+      "learning_rate": 1.0874976935748816e-06,
+      "loss": 0.5242,
+      "step": 1313
+    },
+    {
+      "epoch": 2.4226044226044228,
+      "grad_norm": 0.569178719100136,
+      "learning_rate": 1.0808214401630913e-06,
+      "loss": 0.5234,
+      "step": 1314
+    },
+    {
+      "epoch": 2.4244471744471743,
+      "grad_norm": 0.58082239404479,
+      "learning_rate": 1.0741632587412715e-06,
+      "loss": 0.5017,
+      "step": 1315
+    },
+    {
+      "epoch": 2.4262899262899262,
+      "grad_norm": 0.5522199856056926,
+      "learning_rate": 1.0675231800114116e-06,
+      "loss": 0.527,
+      "step": 1316
+    },
+    {
+      "epoch": 2.428132678132678,
+      "grad_norm": 0.5329048976118389,
+      "learning_rate": 1.0609012345920323e-06,
+      "loss": 0.5303,
+      "step": 1317
+    },
+    {
+      "epoch": 2.42997542997543,
+      "grad_norm": 0.586690659518715,
+      "learning_rate": 1.0542974530180327e-06,
+      "loss": 0.5067,
+      "step": 1318
+    },
+    {
+      "epoch": 2.4318181818181817,
+      "grad_norm": 0.592660205582941,
+      "learning_rate": 1.0477118657405566e-06,
+      "loss": 0.4948,
+      "step": 1319
+    },
+    {
+      "epoch": 2.4336609336609336,
+      "grad_norm": 0.5574194885705247,
+      "learning_rate": 1.0411445031268553e-06,
+      "loss": 0.5056,
+      "step": 1320
+    },
+    {
+      "epoch": 2.4355036855036856,
+      "grad_norm": 0.5540311496198144,
+      "learning_rate": 1.0345953954601372e-06,
+      "loss": 0.5018,
+      "step": 1321
+    },
+    {
+      "epoch": 2.4373464373464375,
+      "grad_norm": 0.5391650566397757,
+      "learning_rate": 1.0280645729394368e-06,
+      "loss": 0.5231,
+      "step": 1322
+    },
+    {
+      "epoch": 2.439189189189189,
+      "grad_norm": 0.536717138337082,
+      "learning_rate": 1.0215520656794741e-06,
+      "loss": 0.5233,
+      "step": 1323
+    },
+    {
+      "epoch": 2.441031941031941,
+      "grad_norm": 0.5413787254548732,
+      "learning_rate": 1.0150579037105123e-06,
+      "loss": 0.5359,
+      "step": 1324
+    },
+    {
+      "epoch": 2.442874692874693,
+      "grad_norm": 0.5487879455956987,
+      "learning_rate": 1.00858211697822e-06,
+      "loss": 0.5206,
+      "step": 1325
+    },
+    {
+      "epoch": 2.444717444717445,
+      "grad_norm": 0.5762486373327458,
+      "learning_rate": 1.0021247353435408e-06,
+      "loss": 0.5208,
+      "step": 1326
+    },
+    {
+      "epoch": 2.4465601965601964,
+      "grad_norm": 0.4949918997453795,
+      "learning_rate": 9.956857885825399e-07,
+      "loss": 0.536,
+      "step": 1327
+    },
+    {
+      "epoch": 2.4484029484029484,
+      "grad_norm": 0.5142678253871459,
+      "learning_rate": 9.892653063862834e-07,
+      "loss": 0.5338,
+      "step": 1328
+    },
+    {
+      "epoch": 2.4502457002457003,
+      "grad_norm": 0.5128867657307047,
+      "learning_rate": 9.82863318360695e-07,
+      "loss": 0.5234,
+      "step": 1329
+    },
+    {
+      "epoch": 2.4520884520884523,
+      "grad_norm": 0.5402083523918946,
+      "learning_rate": 9.764798540264103e-07,
+      "loss": 0.525,
+      "step": 1330
+    },
+    {
+      "epoch": 2.453931203931204,
+      "grad_norm": 0.5386657902359987,
+      "learning_rate": 9.701149428186584e-07,
+      "loss": 0.5395,
+      "step": 1331
+    },
+    {
+      "epoch": 2.4557739557739557,
+      "grad_norm": 0.5299352250692322,
+      "learning_rate": 9.637686140871121e-07,
+      "loss": 0.521,
+      "step": 1332
+    },
+    {
+      "epoch": 2.4576167076167077,
+      "grad_norm": 0.550773298326422,
+      "learning_rate": 9.57440897095756e-07,
+      "loss": 0.5419,
+      "step": 1333
+    },
+    {
+      "epoch": 2.4594594594594597,
+      "grad_norm": 0.5434331748865798,
+      "learning_rate": 9.511318210227577e-07,
+      "loss": 0.537,
+      "step": 1334
+    },
+    {
+      "epoch": 2.461302211302211,
+      "grad_norm": 0.5068608658625061,
+      "learning_rate": 9.448414149603247e-07,
+      "loss": 0.5222,
+      "step": 1335
+    },
+    {
+      "epoch": 2.463144963144963,
+      "grad_norm": 0.5344297216482027,
+      "learning_rate": 9.385697079145734e-07,
+      "loss": 0.5434,
+      "step": 1336
+    },
+    {
+      "epoch": 2.464987714987715,
+      "grad_norm": 0.5511526853470056,
+      "learning_rate": 9.323167288053997e-07,
+      "loss": 0.501,
+      "step": 1337
+    },
+    {
+      "epoch": 2.4668304668304666,
+      "grad_norm": 0.5590501093935379,
+      "learning_rate": 9.26082506466337e-07,
+      "loss": 0.5418,
+      "step": 1338
+    },
+    {
+      "epoch": 2.4686732186732185,
+      "grad_norm": 0.574115809382087,
+      "learning_rate": 9.198670696444339e-07,
+      "loss": 0.5077,
+      "step": 1339
+    },
+    {
+      "epoch": 2.4705159705159705,
+      "grad_norm": 0.5414941713354048,
+      "learning_rate": 9.136704470001101e-07,
+      "loss": 0.5091,
+      "step": 1340
+    },
+    {
+      "epoch": 2.4723587223587224,
+      "grad_norm": 0.5580447780339524,
+      "learning_rate": 9.074926671070322e-07,
+      "loss": 0.529,
+      "step": 1341
+    },
+    {
+      "epoch": 2.4742014742014744,
+      "grad_norm": 0.5670338425529446,
+      "learning_rate": 9.013337584519827e-07,
+      "loss": 0.5218,
+      "step": 1342
+    },
+    {
+      "epoch": 2.476044226044226,
+      "grad_norm": 0.610166930609582,
+      "learning_rate": 8.951937494347185e-07,
+      "loss": 0.5034,
+      "step": 1343
+    },
+    {
+      "epoch": 2.477886977886978,
+      "grad_norm": 0.5573616704381394,
+      "learning_rate": 8.890726683678541e-07,
+      "loss": 0.5168,
+      "step": 1344
+    },
+    {
+      "epoch": 2.47972972972973,
+      "grad_norm": 0.5103635104616873,
+      "learning_rate": 8.829705434767233e-07,
+      "loss": 0.5249,
+      "step": 1345
+    },
+    {
+      "epoch": 2.4815724815724813,
+      "grad_norm": 0.5233394870581846,
+      "learning_rate": 8.768874028992431e-07,
+      "loss": 0.5273,
+      "step": 1346
+    },
+    {
+      "epoch": 2.4834152334152333,
+      "grad_norm": 0.5540971035686267,
+      "learning_rate": 8.708232746857992e-07,
+      "loss": 0.5234,
+      "step": 1347
+    },
+    {
+      "epoch": 2.4852579852579852,
+      "grad_norm": 0.8887508797490826,
+      "learning_rate": 8.647781867991034e-07,
+      "loss": 0.5022,
+      "step": 1348
+    },
+    {
+      "epoch": 2.487100737100737,
+      "grad_norm": 0.5574739697769302,
+      "learning_rate": 8.587521671140686e-07,
+      "loss": 0.5345,
+      "step": 1349
+    },
+    {
+      "epoch": 2.488943488943489,
+      "grad_norm": 0.5538616098120768,
+      "learning_rate": 8.527452434176858e-07,
+      "loss": 0.5022,
+      "step": 1350
+    },
+    {
+      "epoch": 2.4907862407862407,
+      "grad_norm": 0.5500589045521471,
+      "learning_rate": 8.46757443408886e-07,
+      "loss": 0.4959,
+      "step": 1351
+    },
+    {
+      "epoch": 2.4926289926289926,
+      "grad_norm": 0.5437738085202242,
+      "learning_rate": 8.407887946984195e-07,
+      "loss": 0.5238,
+      "step": 1352
+    },
+    {
+      "epoch": 2.4944717444717446,
+      "grad_norm": 0.5484524328665054,
+      "learning_rate": 8.348393248087289e-07,
+      "loss": 0.5098,
+      "step": 1353
+    },
+    {
+      "epoch": 2.496314496314496,
+      "grad_norm": 0.5573211453155351,
+      "learning_rate": 8.28909061173817e-07,
+      "loss": 0.5069,
+      "step": 1354
+    },
+    {
+      "epoch": 2.498157248157248,
+      "grad_norm": 0.5239027420486853,
+      "learning_rate": 8.229980311391233e-07,
+      "loss": 0.5217,
+      "step": 1355
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.5845356833924413,
+      "learning_rate": 8.171062619614017e-07,
+      "loss": 0.5236,
+      "step": 1356
+    },
+    {
+      "epoch": 2.501842751842752,
+      "grad_norm": 0.5293058047454984,
+      "learning_rate": 8.112337808085869e-07,
+      "loss": 0.5009,
+      "step": 1357
+    },
+    {
+      "epoch": 2.503685503685504,
+      "grad_norm": 0.5633449925187162,
+      "learning_rate": 8.053806147596743e-07,
+      "loss": 0.5132,
+      "step": 1358
+    },
+    {
+      "epoch": 2.5055282555282554,
+      "grad_norm": 0.580062441038288,
+      "learning_rate": 7.995467908045973e-07,
+      "loss": 0.5037,
+      "step": 1359
+    },
+    {
+      "epoch": 2.5073710073710074,
+      "grad_norm": 0.5240191814186952,
+      "learning_rate": 7.937323358440935e-07,
+      "loss": 0.5184,
+      "step": 1360
+    },
+    {
+      "epoch": 2.5092137592137593,
+      "grad_norm": 0.5498584217272254,
+      "learning_rate": 7.879372766895948e-07,
+      "loss": 0.5356,
+      "step": 1361
+    },
+    {
+      "epoch": 2.511056511056511,
+      "grad_norm": 0.583111193803761,
+      "learning_rate": 7.821616400630866e-07,
+      "loss": 0.5148,
+      "step": 1362
+    },
+    {
+      "epoch": 2.512899262899263,
+      "grad_norm": 0.5390207790968203,
+      "learning_rate": 7.764054525970011e-07,
+      "loss": 0.5044,
+      "step": 1363
+    },
+    {
+      "epoch": 2.5147420147420148,
+      "grad_norm": 0.5649960244033627,
+      "learning_rate": 7.706687408340862e-07,
+      "loss": 0.5238,
+      "step": 1364
+    },
+    {
+      "epoch": 2.5165847665847667,
+      "grad_norm": 0.5424491725053029,
+      "learning_rate": 7.649515312272781e-07,
+      "loss": 0.5037,
+      "step": 1365
+    },
+    {
+      "epoch": 2.5184275184275187,
+      "grad_norm": 0.544822169122595,
+      "learning_rate": 7.59253850139593e-07,
+      "loss": 0.5092,
+      "step": 1366
+    },
+    {
+      "epoch": 2.52027027027027,
+      "grad_norm": 0.5326164341077295,
+      "learning_rate": 7.535757238439939e-07,
+      "loss": 0.543,
+      "step": 1367
+    },
+    {
+      "epoch": 2.522113022113022,
+      "grad_norm": 0.5495756111099781,
+      "learning_rate": 7.47917178523272e-07,
+      "loss": 0.5099,
+      "step": 1368
+    },
+    {
+      "epoch": 2.523955773955774,
+      "grad_norm": 0.5388891006832571,
+      "learning_rate": 7.422782402699318e-07,
+      "loss": 0.4953,
+      "step": 1369
+    },
+    {
+      "epoch": 2.5257985257985256,
+      "grad_norm": 0.5539029254411925,
+      "learning_rate": 7.366589350860631e-07,
+      "loss": 0.5177,
+      "step": 1370
+    },
+    {
+      "epoch": 2.5276412776412776,
+      "grad_norm": 0.6027554658433644,
+      "learning_rate": 7.310592888832235e-07,
+      "loss": 0.5002,
+      "step": 1371
+    },
+    {
+      "epoch": 2.5294840294840295,
+      "grad_norm": 0.5556385334175498,
+      "learning_rate": 7.254793274823235e-07,
+      "loss": 0.5187,
+      "step": 1372
+    },
+    {
+      "epoch": 2.5313267813267815,
+      "grad_norm": 0.5600475314628117,
+      "learning_rate": 7.199190766135001e-07,
+      "loss": 0.504,
+      "step": 1373
+    },
+    {
+      "epoch": 2.5331695331695334,
+      "grad_norm": 0.48720948908547856,
+      "learning_rate": 7.143785619160026e-07,
+      "loss": 0.53,
+      "step": 1374
+    },
+    {
+      "epoch": 2.535012285012285,
+      "grad_norm": 0.5368768632605015,
+      "learning_rate": 7.088578089380754e-07,
+      "loss": 0.5137,
+      "step": 1375
+    },
+    {
+      "epoch": 2.536855036855037,
+      "grad_norm": 0.5308365558659569,
+      "learning_rate": 7.033568431368359e-07,
+      "loss": 0.5207,
+      "step": 1376
+    },
+    {
+      "epoch": 2.538697788697789,
+      "grad_norm": 0.5581722781828659,
+      "learning_rate": 6.978756898781613e-07,
+      "loss": 0.5103,
+      "step": 1377
+    },
+    {
+      "epoch": 2.5405405405405403,
+      "grad_norm": 0.5211463910398807,
+      "learning_rate": 6.924143744365669e-07,
+      "loss": 0.5041,
+      "step": 1378
+    },
+    {
+      "epoch": 2.5423832923832923,
+      "grad_norm": 0.5509637193653638,
+      "learning_rate": 6.86972921995096e-07,
+      "loss": 0.5311,
+      "step": 1379
+    },
+    {
+      "epoch": 2.5442260442260443,
+      "grad_norm": 0.5465797583718852,
+      "learning_rate": 6.81551357645201e-07,
+      "loss": 0.4967,
+      "step": 1380
+    },
+    {
+      "epoch": 2.546068796068796,
+      "grad_norm": 0.5737104129480678,
+      "learning_rate": 6.761497063866207e-07,
+      "loss": 0.4982,
+      "step": 1381
+    },
+    {
+      "epoch": 2.547911547911548,
+      "grad_norm": 0.5256441761235395,
+      "learning_rate": 6.707679931272787e-07,
+      "loss": 0.5192,
+      "step": 1382
+    },
+    {
+      "epoch": 2.5497542997542997,
+      "grad_norm": 0.5494787678461711,
+      "learning_rate": 6.654062426831571e-07,
+      "loss": 0.5307,
+      "step": 1383
+    },
+    {
+      "epoch": 2.5515970515970516,
+      "grad_norm": 0.5571851335088197,
+      "learning_rate": 6.600644797781847e-07,
+      "loss": 0.5268,
+      "step": 1384
+    },
+    {
+      "epoch": 2.5534398034398036,
+      "grad_norm": 0.532273417634668,
+      "learning_rate": 6.547427290441299e-07,
+      "loss": 0.5005,
+      "step": 1385
+    },
+    {
+      "epoch": 2.555282555282555,
+      "grad_norm": 0.5716047507205246,
+      "learning_rate": 6.494410150204766e-07,
+      "loss": 0.4998,
+      "step": 1386
+    },
+    {
+      "epoch": 2.557125307125307,
+      "grad_norm": 0.5390957427859147,
+      "learning_rate": 6.441593621543169e-07,
+      "loss": 0.5177,
+      "step": 1387
+    },
+    {
+      "epoch": 2.558968058968059,
+      "grad_norm": 0.5450045472605696,
+      "learning_rate": 6.388977948002406e-07,
+      "loss": 0.5173,
+      "step": 1388
+    },
+    {
+      "epoch": 2.560810810810811,
+      "grad_norm": 0.5231994635753375,
+      "learning_rate": 6.336563372202165e-07,
+      "loss": 0.5271,
+      "step": 1389
+    },
+    {
+      "epoch": 2.562653562653563,
+      "grad_norm": 0.5459386206408792,
+      "learning_rate": 6.284350135834838e-07,
+      "loss": 0.5375,
+      "step": 1390
+    },
+    {
+      "epoch": 2.5644963144963144,
+      "grad_norm": 0.515129799125257,
+      "learning_rate": 6.232338479664446e-07,
+      "loss": 0.5158,
+      "step": 1391
+    },
+    {
+      "epoch": 2.5663390663390664,
+      "grad_norm": 0.5132356558561333,
+      "learning_rate": 6.180528643525446e-07,
+      "loss": 0.5304,
+      "step": 1392
+    },
+    {
+      "epoch": 2.5681818181818183,
+      "grad_norm": 0.5264064570717434,
+      "learning_rate": 6.128920866321691e-07,
+      "loss": 0.4978,
+      "step": 1393
+    },
+    {
+      "epoch": 2.57002457002457,
+      "grad_norm": 0.5124626667780037,
+      "learning_rate": 6.077515386025284e-07,
+      "loss": 0.537,
+      "step": 1394
+    },
+    {
+      "epoch": 2.571867321867322,
+      "grad_norm": 0.5074345924898198,
+      "learning_rate": 6.026312439675553e-07,
+      "loss": 0.5248,
+      "step": 1395
+    },
+    {
+      "epoch": 2.5737100737100738,
+      "grad_norm": 0.5828198046918084,
+      "learning_rate": 5.975312263377853e-07,
+      "loss": 0.5074,
+      "step": 1396
+    },
+    {
+      "epoch": 2.5755528255528253,
+      "grad_norm": 0.5412468343075626,
+      "learning_rate": 5.924515092302546e-07,
+      "loss": 0.5219,
+      "step": 1397
+    },
+    {
+      "epoch": 2.5773955773955772,
+      "grad_norm": 0.558125630647803,
+      "learning_rate": 5.873921160683943e-07,
+      "loss": 0.5371,
+      "step": 1398
+    },
+    {
+      "epoch": 2.579238329238329,
+      "grad_norm": 0.51365903926961,
+      "learning_rate": 5.82353070181913e-07,
+      "loss": 0.5475,
+      "step": 1399
+    },
+    {
+      "epoch": 2.581081081081081,
+      "grad_norm": 0.5175830564425453,
+      "learning_rate": 5.773343948066962e-07,
+      "loss": 0.5345,
+      "step": 1400
+    },
+    {
+      "epoch": 2.582923832923833,
+      "grad_norm": 0.5054979986702276,
+      "learning_rate": 5.723361130846999e-07,
+      "loss": 0.5321,
+      "step": 1401
+    },
+    {
+      "epoch": 2.5847665847665846,
+      "grad_norm": 0.5407312621464927,
+      "learning_rate": 5.673582480638395e-07,
+      "loss": 0.5119,
+      "step": 1402
+    },
+    {
+      "epoch": 2.5866093366093366,
+      "grad_norm": 0.5566778420219478,
+      "learning_rate": 5.62400822697885e-07,
+      "loss": 0.4804,
+      "step": 1403
+    },
+    {
+      "epoch": 2.5884520884520885,
+      "grad_norm": 0.514820808386191,
+      "learning_rate": 5.574638598463578e-07,
+      "loss": 0.5244,
+      "step": 1404
+    },
+    {
+      "epoch": 2.59029484029484,
+      "grad_norm": 0.4950517731540446,
+      "learning_rate": 5.525473822744209e-07,
+      "loss": 0.5358,
+      "step": 1405
+    },
+    {
+      "epoch": 2.592137592137592,
+      "grad_norm": 0.5515780693768244,
+      "learning_rate": 5.476514126527771e-07,
+      "loss": 0.5083,
+      "step": 1406
+    },
+    {
+      "epoch": 2.593980343980344,
+      "grad_norm": 0.533468762037984,
+      "learning_rate": 5.427759735575655e-07,
+      "loss": 0.5225,
+      "step": 1407
+    },
+    {
+      "epoch": 2.595823095823096,
+      "grad_norm": 0.48280653630675135,
+      "learning_rate": 5.379210874702534e-07,
+      "loss": 0.536,
+      "step": 1408
+    },
+    {
+      "epoch": 2.597665847665848,
+      "grad_norm": 0.5475201412957846,
+      "learning_rate": 5.330867767775333e-07,
+      "loss": 0.5255,
+      "step": 1409
+    },
+    {
+      "epoch": 2.5995085995085994,
+      "grad_norm": 0.5354509887014567,
+      "learning_rate": 5.282730637712252e-07,
+      "loss": 0.5297,
+      "step": 1410
+    },
+    {
+      "epoch": 2.6013513513513513,
+      "grad_norm": 0.5969112402319641,
+      "learning_rate": 5.234799706481669e-07,
+      "loss": 0.5101,
+      "step": 1411
+    },
+    {
+      "epoch": 2.6031941031941033,
+      "grad_norm": 0.5595657406685925,
+      "learning_rate": 5.187075195101154e-07,
+      "loss": 0.5353,
+      "step": 1412
+    },
+    {
+      "epoch": 2.605036855036855,
+      "grad_norm": 0.560912000906855,
+      "learning_rate": 5.139557323636418e-07,
+      "loss": 0.5105,
+      "step": 1413
+    },
+    {
+      "epoch": 2.6068796068796067,
+      "grad_norm": 0.5289118972315553,
+      "learning_rate": 5.09224631120036e-07,
+      "loss": 0.5172,
+      "step": 1414
+    },
+    {
+      "epoch": 2.6087223587223587,
+      "grad_norm": 0.5099619546963888,
+      "learning_rate": 5.045142375951994e-07,
+      "loss": 0.5384,
+      "step": 1415
+    },
+    {
+      "epoch": 2.6105651105651106,
+      "grad_norm": 0.4737354304863168,
+      "learning_rate": 4.998245735095459e-07,
+      "loss": 0.5185,
+      "step": 1416
+    },
+    {
+      "epoch": 2.6124078624078626,
+      "grad_norm": 0.5188795037095131,
+      "learning_rate": 4.951556604879049e-07,
+      "loss": 0.5357,
+      "step": 1417
+    },
+    {
+      "epoch": 2.614250614250614,
+      "grad_norm": 0.5434297578263628,
+      "learning_rate": 4.905075200594167e-07,
+      "loss": 0.5185,
+      "step": 1418
+    },
+    {
+      "epoch": 2.616093366093366,
+      "grad_norm": 0.5506053455503561,
+      "learning_rate": 4.858801736574364e-07,
+      "loss": 0.5235,
+      "step": 1419
+    },
+    {
+      "epoch": 2.617936117936118,
+      "grad_norm": 0.5899529240228132,
+      "learning_rate": 4.812736426194369e-07,
+      "loss": 0.4931,
+      "step": 1420
+    },
+    {
+      "epoch": 2.6197788697788695,
+      "grad_norm": 0.5449072335635499,
+      "learning_rate": 4.766879481869041e-07,
+      "loss": 0.529,
+      "step": 1421
+    },
+    {
+      "epoch": 2.6216216216216215,
+      "grad_norm": 0.514098260894945,
+      "learning_rate": 4.721231115052438e-07,
+      "loss": 0.5056,
+      "step": 1422
+    },
+    {
+      "epoch": 2.6234643734643734,
+      "grad_norm": 0.5307576223327942,
+      "learning_rate": 4.6757915362368567e-07,
+      "loss": 0.527,
+      "step": 1423
+    },
+    {
+      "epoch": 2.6253071253071254,
+      "grad_norm": 0.548972665933976,
+      "learning_rate": 4.6305609549518127e-07,
+      "loss": 0.5013,
+      "step": 1424
+    },
+    {
+      "epoch": 2.6271498771498774,
+      "grad_norm": 0.5454137970017471,
+      "learning_rate": 4.585539579763093e-07,
+      "loss": 0.4909,
+      "step": 1425
+    },
+    {
+      "epoch": 2.628992628992629,
+      "grad_norm": 0.5231396911931239,
+      "learning_rate": 4.540727618271834e-07,
+      "loss": 0.5106,
+      "step": 1426
+    },
+    {
+      "epoch": 2.630835380835381,
+      "grad_norm": 0.5528030699492519,
+      "learning_rate": 4.4961252771135057e-07,
+      "loss": 0.5171,
+      "step": 1427
+    },
+    {
+      "epoch": 2.6326781326781328,
+      "grad_norm": 0.5140211403725814,
+      "learning_rate": 4.4517327619569784e-07,
+      "loss": 0.5332,
+      "step": 1428
+    },
+    {
+      "epoch": 2.6345208845208843,
+      "grad_norm": 0.5430868455795982,
+      "learning_rate": 4.4075502775035806e-07,
+      "loss": 0.4979,
+      "step": 1429
+    },
+    {
+      "epoch": 2.6363636363636362,
+      "grad_norm": 0.5415444224088569,
+      "learning_rate": 4.363578027486187e-07,
+      "loss": 0.5342,
+      "step": 1430
+    },
+    {
+      "epoch": 2.638206388206388,
+      "grad_norm": 0.4948885450647579,
+      "learning_rate": 4.3198162146682044e-07,
+      "loss": 0.5192,
+      "step": 1431
+    },
+    {
+      "epoch": 2.64004914004914,
+      "grad_norm": 0.5200393102613293,
+      "learning_rate": 4.276265040842692e-07,
+      "loss": 0.5364,
+      "step": 1432
+    },
+    {
+      "epoch": 2.641891891891892,
+      "grad_norm": 0.5502554645612471,
+      "learning_rate": 4.232924706831432e-07,
+      "loss": 0.4963,
+      "step": 1433
+    },
+    {
+      "epoch": 2.6437346437346436,
+      "grad_norm": 0.5235057509484665,
+      "learning_rate": 4.189795412483971e-07,
+      "loss": 0.5013,
+      "step": 1434
+    },
+    {
+      "epoch": 2.6455773955773956,
+      "grad_norm": 0.5595876898969606,
+      "learning_rate": 4.146877356676715e-07,
+      "loss": 0.506,
+      "step": 1435
+    },
+    {
+      "epoch": 2.6474201474201475,
+      "grad_norm": 0.520672552118964,
+      "learning_rate": 4.1041707373120354e-07,
+      "loss": 0.5041,
+      "step": 1436
+    },
+    {
+      "epoch": 2.649262899262899,
+      "grad_norm": 0.5383308363838436,
+      "learning_rate": 4.0616757513173123e-07,
+      "loss": 0.5134,
+      "step": 1437
+    },
+    {
+      "epoch": 2.651105651105651,
+      "grad_norm": 0.5109536834988514,
+      "learning_rate": 4.0193925946440447e-07,
+      "loss": 0.516,
+      "step": 1438
+    },
+    {
+      "epoch": 2.652948402948403,
+      "grad_norm": 0.48915661834971175,
+      "learning_rate": 3.9773214622669974e-07,
+      "loss": 0.5501,
+      "step": 1439
+    },
+    {
+      "epoch": 2.654791154791155,
+      "grad_norm": 0.4923552399908929,
+      "learning_rate": 3.935462548183183e-07,
+      "loss": 0.5351,
+      "step": 1440
+    },
+    {
+      "epoch": 2.656633906633907,
+      "grad_norm": 0.5524809505429511,
+      "learning_rate": 3.8938160454110916e-07,
+      "loss": 0.5095,
+      "step": 1441
+    },
+    {
+      "epoch": 2.6584766584766584,
+      "grad_norm": 0.5207800852590828,
+      "learning_rate": 3.852382145989758e-07,
+      "loss": 0.5214,
+      "step": 1442
+    },
+    {
+      "epoch": 2.6603194103194103,
+      "grad_norm": 0.5281684078757936,
+      "learning_rate": 3.8111610409778065e-07,
+      "loss": 0.5339,
+      "step": 1443
+    },
+    {
+      "epoch": 2.6621621621621623,
+      "grad_norm": 0.5260749887389743,
+      "learning_rate": 3.7701529204526856e-07,
+      "loss": 0.5308,
+      "step": 1444
+    },
+    {
+      "epoch": 2.664004914004914,
+      "grad_norm": 0.5520851520559095,
+      "learning_rate": 3.729357973509734e-07,
+      "loss": 0.5007,
+      "step": 1445
+    },
+    {
+      "epoch": 2.6658476658476657,
+      "grad_norm": 0.5511355007365665,
+      "learning_rate": 3.6887763882612835e-07,
+      "loss": 0.5176,
+      "step": 1446
+    },
+    {
+      "epoch": 2.6676904176904177,
+      "grad_norm": 0.6251980846306995,
+      "learning_rate": 3.6484083518358416e-07,
+      "loss": 0.496,
+      "step": 1447
+    },
+    {
+      "epoch": 2.6695331695331697,
+      "grad_norm": 0.508753644296028,
+      "learning_rate": 3.60825405037718e-07,
+      "loss": 0.5492,
+      "step": 1448
+    },
+    {
+      "epoch": 2.6713759213759216,
+      "grad_norm": 0.5321113248129838,
+      "learning_rate": 3.568313669043538e-07,
+      "loss": 0.5117,
+      "step": 1449
+    },
+    {
+      "epoch": 2.673218673218673,
+      "grad_norm": 0.5029555458760624,
+      "learning_rate": 3.528587392006716e-07,
+      "loss": 0.5196,
+      "step": 1450
+    },
+    {
+      "epoch": 2.675061425061425,
+      "grad_norm": 0.5127911290006885,
+      "learning_rate": 3.4890754024512254e-07,
+      "loss": 0.5224,
+      "step": 1451
+    },
+    {
+      "epoch": 2.676904176904177,
+      "grad_norm": 0.5357389917138566,
+      "learning_rate": 3.449777882573502e-07,
+      "loss": 0.5309,
+      "step": 1452
+    },
+    {
+      "epoch": 2.6787469287469285,
+      "grad_norm": 0.5481288328763365,
+      "learning_rate": 3.4106950135809944e-07,
+      "loss": 0.5226,
+      "step": 1453
+    },
+    {
+      "epoch": 2.6805896805896805,
+      "grad_norm": 0.5472598346839228,
+      "learning_rate": 3.3718269756913615e-07,
+      "loss": 0.5203,
+      "step": 1454
+    },
+    {
+      "epoch": 2.6824324324324325,
+      "grad_norm": 0.509557581087114,
+      "learning_rate": 3.333173948131663e-07,
+      "loss": 0.5281,
+      "step": 1455
+    },
+    {
+      "epoch": 2.6842751842751844,
+      "grad_norm": 0.5343237970201113,
+      "learning_rate": 3.294736109137486e-07,
+      "loss": 0.5249,
+      "step": 1456
+    },
+    {
+      "epoch": 2.6861179361179364,
+      "grad_norm": 0.5214112854682433,
+      "learning_rate": 3.2565136359521575e-07,
+      "loss": 0.5336,
+      "step": 1457
+    },
+    {
+      "epoch": 2.687960687960688,
+      "grad_norm": 0.5272548594472223,
+      "learning_rate": 3.2185067048259245e-07,
+      "loss": 0.5195,
+      "step": 1458
+    },
+    {
+      "epoch": 2.68980343980344,
+      "grad_norm": 0.540051382771758,
+      "learning_rate": 3.1807154910151017e-07,
+      "loss": 0.5073,
+      "step": 1459
+    },
+    {
+      "epoch": 2.691646191646192,
+      "grad_norm": 0.5054216265172371,
+      "learning_rate": 3.143140168781328e-07,
+      "loss": 0.5176,
+      "step": 1460
+    },
+    {
+      "epoch": 2.6934889434889433,
+      "grad_norm": 0.599911466994551,
+      "learning_rate": 3.105780911390738e-07,
+      "loss": 0.4749,
+      "step": 1461
+    },
+    {
+      "epoch": 2.6953316953316953,
+      "grad_norm": 0.5471874541374816,
+      "learning_rate": 3.068637891113108e-07,
+      "loss": 0.511,
+      "step": 1462
+    },
+    {
+      "epoch": 2.697174447174447,
+      "grad_norm": 0.5440649471073515,
+      "learning_rate": 3.03171127922115e-07,
+      "loss": 0.5176,
+      "step": 1463
+    },
+    {
+      "epoch": 2.699017199017199,
+      "grad_norm": 0.6739323123202369,
+      "learning_rate": 2.9950012459896704e-07,
+      "loss": 0.5067,
+      "step": 1464
+    },
+    {
+      "epoch": 2.700859950859951,
+      "grad_norm": 0.5179674692471761,
+      "learning_rate": 2.9585079606947843e-07,
+      "loss": 0.522,
+      "step": 1465
+    },
+    {
+      "epoch": 2.7027027027027026,
+      "grad_norm": 0.5233506940975264,
+      "learning_rate": 2.922231591613162e-07,
+      "loss": 0.5247,
+      "step": 1466
+    },
+    {
+      "epoch": 2.7045454545454546,
+      "grad_norm": 0.5323434132130863,
+      "learning_rate": 2.8861723060212175e-07,
+      "loss": 0.5216,
+      "step": 1467
+    },
+    {
+      "epoch": 2.7063882063882065,
+      "grad_norm": 0.5355687478109868,
+      "learning_rate": 2.850330270194379e-07,
+      "loss": 0.5028,
+      "step": 1468
+    },
+    {
+      "epoch": 2.708230958230958,
+      "grad_norm": 0.5358707016951616,
+      "learning_rate": 2.814705649406285e-07,
+      "loss": 0.5012,
+      "step": 1469
+    },
+    {
+      "epoch": 2.71007371007371,
+      "grad_norm": 0.5318264635548231,
+      "learning_rate": 2.779298607928033e-07,
+      "loss": 0.5047,
+      "step": 1470
+    },
+    {
+      "epoch": 2.711916461916462,
+      "grad_norm": 0.506932030946583,
+      "learning_rate": 2.744109309027443e-07,
+      "loss": 0.5109,
+      "step": 1471
+    },
+    {
+      "epoch": 2.713759213759214,
+      "grad_norm": 0.5361424344061636,
+      "learning_rate": 2.7091379149682683e-07,
+      "loss": 0.5148,
+      "step": 1472
+    },
+    {
+      "epoch": 2.715601965601966,
+      "grad_norm": 0.5060813500688558,
+      "learning_rate": 2.674384587009471e-07,
+      "loss": 0.4952,
+      "step": 1473
+    },
+    {
+      "epoch": 2.7174447174447174,
+      "grad_norm": 0.5398256468790039,
+      "learning_rate": 2.6398494854045055e-07,
+      "loss": 0.5083,
+      "step": 1474
+    },
+    {
+      "epoch": 2.7192874692874693,
+      "grad_norm": 0.5210260140264493,
+      "learning_rate": 2.605532769400476e-07,
+      "loss": 0.5469,
+      "step": 1475
+    },
+    {
+      "epoch": 2.7211302211302213,
+      "grad_norm": 0.5405692081264444,
+      "learning_rate": 2.5714345972375486e-07,
+      "loss": 0.4896,
+      "step": 1476
+    },
+    {
+      "epoch": 2.722972972972973,
+      "grad_norm": 0.5232083879353613,
+      "learning_rate": 2.5375551261481224e-07,
+      "loss": 0.53,
+      "step": 1477
+    },
+    {
+      "epoch": 2.7248157248157248,
+      "grad_norm": 0.5558868238313144,
+      "learning_rate": 2.5038945123560976e-07,
+      "loss": 0.5221,
+      "step": 1478
+    },
+    {
+      "epoch": 2.7266584766584767,
+      "grad_norm": 0.5201686152773398,
+      "learning_rate": 2.470452911076227e-07,
+      "loss": 0.5117,
+      "step": 1479
+    },
+    {
+      "epoch": 2.7285012285012282,
+      "grad_norm": 0.5691422296922358,
+      "learning_rate": 2.4372304765133525e-07,
+      "loss": 0.542,
+      "step": 1480
+    },
+    {
+      "epoch": 2.73034398034398,
+      "grad_norm": 0.5229159658063525,
+      "learning_rate": 2.4042273618616773e-07,
+      "loss": 0.5117,
+      "step": 1481
+    },
+    {
+      "epoch": 2.732186732186732,
+      "grad_norm": 0.5163731483384187,
+      "learning_rate": 2.3714437193041174e-07,
+      "loss": 0.5262,
+      "step": 1482
+    },
+    {
+      "epoch": 2.734029484029484,
+      "grad_norm": 0.5097803007051829,
+      "learning_rate": 2.3388797000115427e-07,
+      "loss": 0.5205,
+      "step": 1483
+    },
+    {
+      "epoch": 2.735872235872236,
+      "grad_norm": 0.5549796452460358,
+      "learning_rate": 2.3065354541421003e-07,
+      "loss": 0.5004,
+      "step": 1484
+    },
+    {
+      "epoch": 2.7377149877149876,
+      "grad_norm": 0.4995612274169777,
+      "learning_rate": 2.2744111308405536e-07,
+      "loss": 0.5386,
+      "step": 1485
+    },
+    {
+      "epoch": 2.7395577395577395,
+      "grad_norm": 0.5345126061688347,
+      "learning_rate": 2.242506878237538e-07,
+      "loss": 0.5059,
+      "step": 1486
+    },
+    {
+      "epoch": 2.7414004914004915,
+      "grad_norm": 0.5280477519760196,
+      "learning_rate": 2.2108228434489121e-07,
+      "loss": 0.5389,
+      "step": 1487
+    },
+    {
+      "epoch": 2.743243243243243,
+      "grad_norm": 0.5328406147358064,
+      "learning_rate": 2.1793591725750851e-07,
+      "loss": 0.5135,
+      "step": 1488
+    },
+    {
+      "epoch": 2.745085995085995,
+      "grad_norm": 0.518738137144819,
+      "learning_rate": 2.1481160107003185e-07,
+      "loss": 0.5305,
+      "step": 1489
+    },
+    {
+      "epoch": 2.746928746928747,
+      "grad_norm": 0.570126363434237,
+      "learning_rate": 2.1170935018920758e-07,
+      "loss": 0.511,
+      "step": 1490
+    },
+    {
+      "epoch": 2.748771498771499,
+      "grad_norm": 0.5428468187439561,
+      "learning_rate": 2.0862917892003508e-07,
+      "loss": 0.5201,
+      "step": 1491
+    },
+    {
+      "epoch": 2.750614250614251,
+      "grad_norm": 0.5641270756122503,
+      "learning_rate": 2.055711014657008e-07,
+      "loss": 0.4936,
+      "step": 1492
+    },
+    {
+      "epoch": 2.7524570024570023,
+      "grad_norm": 0.5729351774826397,
+      "learning_rate": 2.0253513192751374e-07,
+      "loss": 0.5222,
+      "step": 1493
+    },
+    {
+      "epoch": 2.7542997542997543,
+      "grad_norm": 0.5385365255486991,
+      "learning_rate": 1.9952128430483718e-07,
+      "loss": 0.527,
+      "step": 1494
+    },
+    {
+      "epoch": 2.756142506142506,
+      "grad_norm": 0.5534741998423213,
+      "learning_rate": 1.965295724950289e-07,
+      "loss": 0.4979,
+      "step": 1495
+    },
+    {
+      "epoch": 2.7579852579852577,
+      "grad_norm": 0.5227660598052728,
+      "learning_rate": 1.93560010293376e-07,
+      "loss": 0.528,
+      "step": 1496
+    },
+    {
+      "epoch": 2.7598280098280097,
+      "grad_norm": 0.5394256733379909,
+      "learning_rate": 1.9061261139302512e-07,
+      "loss": 0.5302,
+      "step": 1497
+    },
+    {
+      "epoch": 2.7616707616707616,
+      "grad_norm": 0.5330336116320038,
+      "learning_rate": 1.8768738938492903e-07,
+      "loss": 0.498,
+      "step": 1498
+    },
+    {
+      "epoch": 2.7635135135135136,
+      "grad_norm": 0.567931117824737,
+      "learning_rate": 1.8478435775777682e-07,
+      "loss": 0.5152,
+      "step": 1499
+    },
+    {
+      "epoch": 2.7653562653562656,
+      "grad_norm": 0.5290955422326936,
+      "learning_rate": 1.8190352989793325e-07,
+      "loss": 0.5279,
+      "step": 1500
+    },
+    {
+      "epoch": 2.767199017199017,
+      "grad_norm": 0.5335019973371169,
+      "learning_rate": 1.7904491908938116e-07,
+      "loss": 0.4912,
+      "step": 1501
+    },
+    {
+      "epoch": 2.769041769041769,
+      "grad_norm": 0.5342900178741219,
+      "learning_rate": 1.762085385136536e-07,
+      "loss": 0.5027,
+      "step": 1502
+    },
+    {
+      "epoch": 2.770884520884521,
+      "grad_norm": 0.5484789231059105,
+      "learning_rate": 1.733944012497768e-07,
+      "loss": 0.5131,
+      "step": 1503
+    },
+    {
+      "epoch": 2.7727272727272725,
+      "grad_norm": 0.5378265973937162,
+      "learning_rate": 1.706025202742112e-07,
+      "loss": 0.5431,
+      "step": 1504
+    },
+    {
+      "epoch": 2.7745700245700244,
+      "grad_norm": 0.5439890138233454,
+      "learning_rate": 1.6783290846078714e-07,
+      "loss": 0.5318,
+      "step": 1505
+    },
+    {
+      "epoch": 2.7764127764127764,
+      "grad_norm": 0.5399892336186415,
+      "learning_rate": 1.6508557858064988e-07,
+      "loss": 0.5017,
+      "step": 1506
+    },
+    {
+      "epoch": 2.7782555282555284,
+      "grad_norm": 0.5107292400197293,
+      "learning_rate": 1.6236054330219853e-07,
+      "loss": 0.5141,
+      "step": 1507
+    },
+    {
+      "epoch": 2.7800982800982803,
+      "grad_norm": 0.5220629869223243,
+      "learning_rate": 1.596578151910283e-07,
+      "loss": 0.5326,
+      "step": 1508
+    },
+    {
+      "epoch": 2.781941031941032,
+      "grad_norm": 0.5344819254208754,
+      "learning_rate": 1.5697740670987227e-07,
+      "loss": 0.521,
+      "step": 1509
+    },
+    {
+      "epoch": 2.7837837837837838,
+      "grad_norm": 0.49739959819162516,
+      "learning_rate": 1.5431933021854196e-07,
+      "loss": 0.5284,
+      "step": 1510
+    },
+    {
+      "epoch": 2.7856265356265357,
+      "grad_norm": 0.5252793334717664,
+      "learning_rate": 1.5168359797387567e-07,
+      "loss": 0.5319,
+      "step": 1511
+    },
+    {
+      "epoch": 2.7874692874692872,
+      "grad_norm": 0.5162365634544558,
+      "learning_rate": 1.4907022212967803e-07,
+      "loss": 0.5157,
+      "step": 1512
+    },
+    {
+      "epoch": 2.789312039312039,
+      "grad_norm": 0.5435620810032266,
+      "learning_rate": 1.4647921473666115e-07,
+      "loss": 0.5173,
+      "step": 1513
+    },
+    {
+      "epoch": 2.791154791154791,
+      "grad_norm": 0.5437936492257905,
+      "learning_rate": 1.439105877423963e-07,
+      "loss": 0.5073,
+      "step": 1514
+    },
+    {
+      "epoch": 2.792997542997543,
+      "grad_norm": 0.5400119134114907,
+      "learning_rate": 1.4136435299125338e-07,
+      "loss": 0.5236,
+      "step": 1515
+    },
+    {
+      "epoch": 2.794840294840295,
+      "grad_norm": 0.5263997106929648,
+      "learning_rate": 1.388405222243472e-07,
+      "loss": 0.5208,
+      "step": 1516
+    },
+    {
+      "epoch": 2.7966830466830466,
+      "grad_norm": 0.5238943065166844,
+      "learning_rate": 1.3633910707948672e-07,
+      "loss": 0.5221,
+      "step": 1517
+    },
+    {
+      "epoch": 2.7985257985257985,
+      "grad_norm": 0.5209531777061343,
+      "learning_rate": 1.338601190911154e-07,
+      "loss": 0.5319,
+      "step": 1518
+    },
+    {
+      "epoch": 2.8003685503685505,
+      "grad_norm": 0.5348114178801304,
+      "learning_rate": 1.3140356969026324e-07,
+      "loss": 0.5205,
+      "step": 1519
+    },
+    {
+      "epoch": 2.802211302211302,
+      "grad_norm": 0.5149435481155648,
+      "learning_rate": 1.2896947020449192e-07,
+      "loss": 0.5368,
+      "step": 1520
+    },
+    {
+      "epoch": 2.804054054054054,
+      "grad_norm": 0.49589239705557614,
+      "learning_rate": 1.2655783185784253e-07,
+      "loss": 0.5296,
+      "step": 1521
+    },
+    {
+      "epoch": 2.805896805896806,
+      "grad_norm": 0.5527478019400617,
+      "learning_rate": 1.2416866577078358e-07,
+      "loss": 0.4831,
+      "step": 1522
+    },
+    {
+      "epoch": 2.807739557739558,
+      "grad_norm": 0.5258660752526185,
+      "learning_rate": 1.2180198296016188e-07,
+      "loss": 0.505,
+      "step": 1523
+    },
+    {
+      "epoch": 2.80958230958231,
+      "grad_norm": 0.5311800207286351,
+      "learning_rate": 1.194577943391484e-07,
+      "loss": 0.5084,
+      "step": 1524
+    },
+    {
+      "epoch": 2.8114250614250613,
+      "grad_norm": 0.5570052126541479,
+      "learning_rate": 1.1713611071719033e-07,
+      "loss": 0.5004,
+      "step": 1525
+    },
+    {
+      "epoch": 2.8132678132678133,
+      "grad_norm": 0.5084028623206239,
+      "learning_rate": 1.1483694279996016e-07,
+      "loss": 0.5355,
+      "step": 1526
+    },
+    {
+      "epoch": 2.8151105651105652,
+      "grad_norm": 0.5177247225316407,
+      "learning_rate": 1.1256030118930727e-07,
+      "loss": 0.5478,
+      "step": 1527
+    },
+    {
+      "epoch": 2.8169533169533167,
+      "grad_norm": 0.5227991364704099,
+      "learning_rate": 1.1030619638320805e-07,
+      "loss": 0.5289,
+      "step": 1528
+    },
+    {
+      "epoch": 2.8187960687960687,
+      "grad_norm": 0.5010162145293622,
+      "learning_rate": 1.0807463877571755e-07,
+      "loss": 0.5214,
+      "step": 1529
+    },
+    {
+      "epoch": 2.8206388206388207,
+      "grad_norm": 0.5161191509826484,
+      "learning_rate": 1.058656386569229e-07,
+      "loss": 0.5017,
+      "step": 1530
+    },
+    {
+      "epoch": 2.8224815724815726,
+      "grad_norm": 0.5151668467598891,
+      "learning_rate": 1.0367920621289496e-07,
+      "loss": 0.5216,
+      "step": 1531
+    },
+    {
+      "epoch": 2.8243243243243246,
+      "grad_norm": 0.5167518987913083,
+      "learning_rate": 1.015153515256384e-07,
+      "loss": 0.5392,
+      "step": 1532
+    },
+    {
+      "epoch": 2.826167076167076,
+      "grad_norm": 0.5585868037090572,
+      "learning_rate": 9.937408457305286e-08,
+      "loss": 0.4852,
+      "step": 1533
+    },
+    {
+      "epoch": 2.828009828009828,
+      "grad_norm": 0.49455726101376457,
+      "learning_rate": 9.725541522887794e-08,
+      "loss": 0.5167,
+      "step": 1534
+    },
+    {
+      "epoch": 2.82985257985258,
+      "grad_norm": 0.5343007973294925,
+      "learning_rate": 9.51593532626538e-08,
+      "loss": 0.52,
+      "step": 1535
+    },
+    {
+      "epoch": 2.8316953316953315,
+      "grad_norm": 0.5447202014765054,
+      "learning_rate": 9.308590833967457e-08,
+      "loss": 0.5133,
+      "step": 1536
+    },
+    {
+      "epoch": 2.8335380835380835,
+      "grad_norm": 0.48136640766506666,
+      "learning_rate": 9.103509002094169e-08,
+      "loss": 0.5573,
+      "step": 1537
+    },
+    {
+      "epoch": 2.8353808353808354,
+      "grad_norm": 0.5475979901870444,
+      "learning_rate": 8.900690776312282e-08,
+      "loss": 0.5166,
+      "step": 1538
+    },
+    {
+      "epoch": 2.8372235872235874,
+      "grad_norm": 0.5714590601520692,
+      "learning_rate": 8.700137091850747e-08,
+      "loss": 0.5256,
+      "step": 1539
+    },
+    {
+      "epoch": 2.8390663390663393,
+      "grad_norm": 0.5169609156800679,
+      "learning_rate": 8.501848873496199e-08,
+      "loss": 0.5142,
+      "step": 1540
+    },
+    {
+      "epoch": 2.840909090909091,
+      "grad_norm": 0.5221017533602311,
+      "learning_rate": 8.305827035588799e-08,
+      "loss": 0.5282,
+      "step": 1541
+    },
+    {
+      "epoch": 2.842751842751843,
+      "grad_norm": 0.5163216920647394,
+      "learning_rate": 8.11207248201834e-08,
+      "loss": 0.5087,
+      "step": 1542
+    },
+    {
+      "epoch": 2.8445945945945947,
+      "grad_norm": 0.4778265428342873,
+      "learning_rate": 7.920586106219375e-08,
+      "loss": 0.5302,
+      "step": 1543
+    },
+    {
+      "epoch": 2.8464373464373462,
+      "grad_norm": 0.5159474389117635,
+      "learning_rate": 7.731368791167815e-08,
+      "loss": 0.5197,
+      "step": 1544
+    },
+    {
+      "epoch": 2.848280098280098,
+      "grad_norm": 0.5190857284210453,
+      "learning_rate": 7.544421409376446e-08,
+      "loss": 0.5077,
+      "step": 1545
+    },
+    {
+      "epoch": 2.85012285012285,
+      "grad_norm": 0.5201156568028776,
+      "learning_rate": 7.35974482289098e-08,
+      "loss": 0.5314,
+      "step": 1546
+    },
+    {
+      "epoch": 2.851965601965602,
+      "grad_norm": 0.5352676143567097,
+      "learning_rate": 7.177339883286172e-08,
+      "loss": 0.5021,
+      "step": 1547
+    },
+    {
+      "epoch": 2.853808353808354,
+      "grad_norm": 0.5623957580025978,
+      "learning_rate": 6.997207431661768e-08,
+      "loss": 0.5388,
+      "step": 1548
+    },
+    {
+      "epoch": 2.8556511056511056,
+      "grad_norm": 0.5656830610195335,
+      "learning_rate": 6.819348298638839e-08,
+      "loss": 0.5217,
+      "step": 1549
+    },
+    {
+      "epoch": 2.8574938574938575,
+      "grad_norm": 0.5250423884999545,
+      "learning_rate": 6.643763304355566e-08,
+      "loss": 0.5079,
+      "step": 1550
+    },
+    {
+      "epoch": 2.8593366093366095,
+      "grad_norm": 0.523794662101541,
+      "learning_rate": 6.470453258463849e-08,
+      "loss": 0.5399,
+      "step": 1551
+    },
+    {
+      "epoch": 2.861179361179361,
+      "grad_norm": 0.5357681490605517,
+      "learning_rate": 6.299418960125425e-08,
+      "loss": 0.5046,
+      "step": 1552
+    },
+    {
+      "epoch": 2.863022113022113,
+      "grad_norm": 0.5003415131752551,
+      "learning_rate": 6.130661198008148e-08,
+      "loss": 0.5109,
+      "step": 1553
+    },
+    {
+      "epoch": 2.864864864864865,
+      "grad_norm": 0.6102703196761635,
+      "learning_rate": 5.964180750282323e-08,
+      "loss": 0.4931,
+      "step": 1554
+    },
+    {
+      "epoch": 2.866707616707617,
+      "grad_norm": 0.5117995766059695,
+      "learning_rate": 5.799978384617266e-08,
+      "loss": 0.5151,
+      "step": 1555
+    },
+    {
+      "epoch": 2.868550368550369,
+      "grad_norm": 0.5085899565767487,
+      "learning_rate": 5.638054858177644e-08,
+      "loss": 0.5233,
+      "step": 1556
+    },
+    {
+      "epoch": 2.8703931203931203,
+      "grad_norm": 0.5658111747265913,
+      "learning_rate": 5.478410917620025e-08,
+      "loss": 0.5196,
+      "step": 1557
+    },
+    {
+      "epoch": 2.8722358722358723,
+      "grad_norm": 0.5259577754515197,
+      "learning_rate": 5.3210472990893324e-08,
+      "loss": 0.5165,
+      "step": 1558
+    },
+    {
+      "epoch": 2.8740786240786242,
+      "grad_norm": 0.5065693364517446,
+      "learning_rate": 5.165964728215678e-08,
+      "loss": 0.5383,
+      "step": 1559
+    },
+    {
+      "epoch": 2.8759213759213758,
+      "grad_norm": 0.5304753947056791,
+      "learning_rate": 5.013163920110864e-08,
+      "loss": 0.5119,
+      "step": 1560
+    },
+    {
+      "epoch": 2.8777641277641277,
+      "grad_norm": 0.5902237964401794,
+      "learning_rate": 4.8626455793648885e-08,
+      "loss": 0.5276,
+      "step": 1561
+    },
+    {
+      "epoch": 2.8796068796068797,
+      "grad_norm": 0.5241069372099676,
+      "learning_rate": 4.714410400043279e-08,
+      "loss": 0.516,
+      "step": 1562
+    },
+    {
+      "epoch": 2.881449631449631,
+      "grad_norm": 0.5345543291105886,
+      "learning_rate": 4.568459065683206e-08,
+      "loss": 0.5347,
+      "step": 1563
+    },
+    {
+      "epoch": 2.883292383292383,
+      "grad_norm": 0.5112722186633386,
+      "learning_rate": 4.42479224929071e-08,
+      "loss": 0.5123,
+      "step": 1564
+    },
+    {
+      "epoch": 2.885135135135135,
+      "grad_norm": 0.5582137640547699,
+      "learning_rate": 4.283410613337757e-08,
+      "loss": 0.5001,
+      "step": 1565
+    },
+    {
+      "epoch": 2.886977886977887,
+      "grad_norm": 0.5180327071055143,
+      "learning_rate": 4.144314809758632e-08,
+      "loss": 0.5004,
+      "step": 1566
+    },
+    {
+      "epoch": 2.888820638820639,
+      "grad_norm": 0.5500565309937185,
+      "learning_rate": 4.007505479947493e-08,
+      "loss": 0.5171,
+      "step": 1567
+    },
+    {
+      "epoch": 2.8906633906633905,
+      "grad_norm": 0.5257480695683946,
+      "learning_rate": 3.872983254755158e-08,
+      "loss": 0.4842,
+      "step": 1568
+    },
+    {
+      "epoch": 2.8925061425061425,
+      "grad_norm": 0.518257644119193,
+      "learning_rate": 3.7407487544861565e-08,
+      "loss": 0.5241,
+      "step": 1569
+    },
+    {
+      "epoch": 2.8943488943488944,
+      "grad_norm": 0.5293424654092045,
+      "learning_rate": 3.610802588895845e-08,
+      "loss": 0.4821,
+      "step": 1570
+    },
+    {
+      "epoch": 2.896191646191646,
+      "grad_norm": 0.5322583536376432,
+      "learning_rate": 3.483145357187967e-08,
+      "loss": 0.5129,
+      "step": 1571
+    },
+    {
+      "epoch": 2.898034398034398,
+      "grad_norm": 0.5225929079816026,
+      "learning_rate": 3.357777648011373e-08,
+      "loss": 0.532,
+      "step": 1572
+    },
+    {
+      "epoch": 2.89987714987715,
+      "grad_norm": 0.517341134696403,
+      "learning_rate": 3.234700039457528e-08,
+      "loss": 0.5172,
+      "step": 1573
+    },
+    {
+      "epoch": 2.901719901719902,
+      "grad_norm": 0.4979836681119717,
+      "learning_rate": 3.1139130990580633e-08,
+      "loss": 0.5085,
+      "step": 1574
+    },
+    {
+      "epoch": 2.9035626535626538,
+      "grad_norm": 0.5518846914296488,
+      "learning_rate": 2.995417383781785e-08,
+      "loss": 0.5118,
+      "step": 1575
+    },
+    {
+      "epoch": 2.9054054054054053,
+      "grad_norm": 0.5496644654499139,
+      "learning_rate": 2.8792134400322803e-08,
+      "loss": 0.5196,
+      "step": 1576
+    },
+    {
+      "epoch": 2.907248157248157,
+      "grad_norm": 0.5474855701979505,
+      "learning_rate": 2.765301803645426e-08,
+      "loss": 0.5036,
+      "step": 1577
+    },
+    {
+      "epoch": 2.909090909090909,
+      "grad_norm": 0.5089872760743619,
+      "learning_rate": 2.6536829998869972e-08,
+      "loss": 0.5291,
+      "step": 1578
+    },
+    {
+      "epoch": 2.9109336609336607,
+      "grad_norm": 0.5132817868117517,
+      "learning_rate": 2.5443575434498934e-08,
+      "loss": 0.5293,
+      "step": 1579
+    },
+    {
+      "epoch": 2.9127764127764126,
+      "grad_norm": 0.536605997939143,
+      "learning_rate": 2.437325938452195e-08,
+      "loss": 0.5004,
+      "step": 1580
+    },
+    {
+      "epoch": 2.9146191646191646,
+      "grad_norm": 0.5217051541943999,
+      "learning_rate": 2.3325886784346107e-08,
+      "loss": 0.5178,
+      "step": 1581
+    },
+    {
+      "epoch": 2.9164619164619165,
+      "grad_norm": 0.5213989025905075,
+      "learning_rate": 2.230146246358256e-08,
+      "loss": 0.5035,
+      "step": 1582
+    },
+    {
+      "epoch": 2.9183046683046685,
+      "grad_norm": 0.5269062911960729,
+      "learning_rate": 2.129999114602266e-08,
+      "loss": 0.5295,
+      "step": 1583
+    },
+    {
+      "epoch": 2.92014742014742,
+      "grad_norm": 0.5223271241646864,
+      "learning_rate": 2.0321477449619098e-08,
+      "loss": 0.5125,
+      "step": 1584
+    },
+    {
+      "epoch": 2.921990171990172,
+      "grad_norm": 0.49949021621663886,
+      "learning_rate": 1.9365925886463686e-08,
+      "loss": 0.5256,
+      "step": 1585
+    },
+    {
+      "epoch": 2.923832923832924,
+      "grad_norm": 0.5268748809558064,
+      "learning_rate": 1.8433340862763493e-08,
+      "loss": 0.5182,
+      "step": 1586
+    },
+    {
+      "epoch": 2.9256756756756754,
+      "grad_norm": 0.518246467414494,
+      "learning_rate": 1.7523726678825846e-08,
+      "loss": 0.5139,
+      "step": 1587
+    },
+    {
+      "epoch": 2.9275184275184274,
+      "grad_norm": 0.5267644430462356,
+      "learning_rate": 1.6637087529033925e-08,
+      "loss": 0.5218,
+      "step": 1588
+    },
+    {
+      "epoch": 2.9293611793611793,
+      "grad_norm": 0.5280183145470049,
+      "learning_rate": 1.5773427501828976e-08,
+      "loss": 0.5229,
+      "step": 1589
+    },
+    {
+      "epoch": 2.9312039312039313,
+      "grad_norm": 0.5335958126517655,
+      "learning_rate": 1.4932750579693677e-08,
+      "loss": 0.5013,
+      "step": 1590
+    },
+    {
+      "epoch": 2.9330466830466833,
+      "grad_norm": 0.5306814136213432,
+      "learning_rate": 1.411506063912882e-08,
+      "loss": 0.5137,
+      "step": 1591
+    },
+    {
+      "epoch": 2.9348894348894348,
+      "grad_norm": 0.49670791706482154,
+      "learning_rate": 1.3320361450641085e-08,
+      "loss": 0.5432,
+      "step": 1592
+    },
+    {
+      "epoch": 2.9367321867321867,
+      "grad_norm": 0.4989920826680736,
+      "learning_rate": 1.2548656678721404e-08,
+      "loss": 0.5173,
+      "step": 1593
+    },
+    {
+      "epoch": 2.9385749385749387,
+      "grad_norm": 0.5021104577156021,
+      "learning_rate": 1.179994988182942e-08,
+      "loss": 0.5065,
+      "step": 1594
+    },
+    {
+      "epoch": 2.94041769041769,
+      "grad_norm": 0.5190484328481474,
+      "learning_rate": 1.107424451237904e-08,
+      "loss": 0.5092,
+      "step": 1595
+    },
+    {
+      "epoch": 2.942260442260442,
+      "grad_norm": 0.5201870198337205,
+      "learning_rate": 1.0371543916718462e-08,
+      "loss": 0.5158,
+      "step": 1596
+    },
+    {
+      "epoch": 2.944103194103194,
+      "grad_norm": 0.5339930911537096,
+      "learning_rate": 9.691851335117963e-09,
+      "loss": 0.4951,
+      "step": 1597
+    },
+    {
+      "epoch": 2.945945945945946,
+      "grad_norm": 0.5372129665623002,
+      "learning_rate": 9.035169901754902e-09,
+      "loss": 0.5363,
+      "step": 1598
+    },
+    {
+      "epoch": 2.947788697788698,
+      "grad_norm": 0.5294411970009515,
+      "learning_rate": 8.40150264469708e-09,
+      "loss": 0.5381,
+      "step": 1599
+    },
+    {
+      "epoch": 2.9496314496314495,
+      "grad_norm": 0.528662651229835,
+      "learning_rate": 7.790852485891064e-09,
+      "loss": 0.5346,
+      "step": 1600
+    },
+    {
+      "epoch": 2.9514742014742015,
+      "grad_norm": 0.510744248932066,
+      "learning_rate": 7.2032222411461124e-09,
+      "loss": 0.5378,
+      "step": 1601
+    },
+    {
+      "epoch": 2.9533169533169534,
+      "grad_norm": 0.5227799149717537,
+      "learning_rate": 6.638614620124717e-09,
+      "loss": 0.4963,
+      "step": 1602
+    },
+    {
+      "epoch": 2.955159705159705,
+      "grad_norm": 0.5032792782878184,
+      "learning_rate": 6.097032226327626e-09,
+      "loss": 0.5237,
+      "step": 1603
+    },
+    {
+      "epoch": 2.957002457002457,
+      "grad_norm": 0.5391047469852651,
+      "learning_rate": 5.578477557081074e-09,
+      "loss": 0.4974,
+      "step": 1604
+    },
+    {
+      "epoch": 2.958845208845209,
+      "grad_norm": 0.5161973157365932,
+      "learning_rate": 5.082953003528457e-09,
+      "loss": 0.5326,
+      "step": 1605
+    },
+    {
+      "epoch": 2.960687960687961,
+      "grad_norm": 0.563132075275518,
+      "learning_rate": 4.61046085061645e-09,
+      "loss": 0.529,
+      "step": 1606
+    },
+    {
+      "epoch": 2.9625307125307128,
+      "grad_norm": 0.4847676884854679,
+      "learning_rate": 4.161003277085574e-09,
+      "loss": 0.5283,
+      "step": 1607
+    },
+    {
+      "epoch": 2.9643734643734643,
+      "grad_norm": 0.5160519048655445,
+      "learning_rate": 3.7345823554602036e-09,
+      "loss": 0.5207,
+      "step": 1608
+    },
+    {
+      "epoch": 2.9662162162162162,
+      "grad_norm": 0.5289805831213371,
+      "learning_rate": 3.3312000520391295e-09,
+      "loss": 0.5196,
+      "step": 1609
+    },
+    {
+      "epoch": 2.968058968058968,
+      "grad_norm": 0.5038729046555345,
+      "learning_rate": 2.9508582268850116e-09,
+      "loss": 0.5268,
+      "step": 1610
+    },
+    {
+      "epoch": 2.9699017199017197,
+      "grad_norm": 0.542469008488981,
+      "learning_rate": 2.5935586338182717e-09,
+      "loss": 0.5172,
+      "step": 1611
+    },
+    {
+      "epoch": 2.9717444717444716,
+      "grad_norm": 0.5307097575155351,
+      "learning_rate": 2.2593029204076578e-09,
+      "loss": 0.5312,
+      "step": 1612
+    },
+    {
+      "epoch": 2.9735872235872236,
+      "grad_norm": 0.5360023089158333,
+      "learning_rate": 1.948092627961362e-09,
+      "loss": 0.5281,
+      "step": 1613
+    },
+    {
+      "epoch": 2.9754299754299756,
+      "grad_norm": 0.5332878840925898,
+      "learning_rate": 1.6599291915231352e-09,
+      "loss": 0.5098,
+      "step": 1614
+    },
+    {
+      "epoch": 2.9772727272727275,
+      "grad_norm": 0.5340358689685448,
+      "learning_rate": 1.3948139398628492e-09,
+      "loss": 0.508,
+      "step": 1615
+    },
+    {
+      "epoch": 2.979115479115479,
+      "grad_norm": 0.552203625624729,
+      "learning_rate": 1.1527480954715009e-09,
+      "loss": 0.5025,
+      "step": 1616
+    },
+    {
+      "epoch": 2.980958230958231,
+      "grad_norm": 0.5292466518856,
+      "learning_rate": 9.337327745545522e-10,
+      "loss": 0.4898,
+      "step": 1617
+    },
+    {
+      "epoch": 2.982800982800983,
+      "grad_norm": 0.5235238784727385,
+      "learning_rate": 7.377689870291527e-10,
+      "loss": 0.5279,
+      "step": 1618
+    },
+    {
+      "epoch": 2.9846437346437344,
+      "grad_norm": 0.5418699376359378,
+      "learning_rate": 5.648576365169245e-10,
+      "loss": 0.5108,
+      "step": 1619
+    },
+    {
+      "epoch": 2.9864864864864864,
+      "grad_norm": 0.5261328512969875,
+      "learning_rate": 4.1499952034118605e-10,
+      "loss": 0.5017,
+      "step": 1620
+    },
+    {
+      "epoch": 2.9883292383292384,
+      "grad_norm": 0.6536633790294611,
+      "learning_rate": 2.881953295225115e-10,
+      "loss": 0.5193,
+      "step": 1621
+    },
+    {
+      "epoch": 2.9901719901719903,
+      "grad_norm": 0.5452014032442296,
+      "learning_rate": 1.8444564877706517e-10,
+      "loss": 0.4896,
+      "step": 1622
+    },
+    {
+      "epoch": 2.9920147420147423,
+      "grad_norm": 0.5813585321533316,
+      "learning_rate": 1.0375095651160572e-10,
+      "loss": 0.497,
+      "step": 1623
+    },
+    {
+      "epoch": 2.993857493857494,
+      "grad_norm": 0.536542590436133,
+      "learning_rate": 4.6111624822375853e-11,
+      "loss": 0.5025,
+      "step": 1624
+    },
+    {
+      "epoch": 2.9957002457002457,
+      "grad_norm": 0.49431688973771876,
+      "learning_rate": 1.1527919495102347e-11,
+      "loss": 0.5243,
+      "step": 1625
+    },
+    {
+      "epoch": 2.9975429975429977,
+      "grad_norm": 0.5077162824414065,
+      "learning_rate": 0.0,
+      "loss": 0.5358,
+      "step": 1626
+    },
+    {
+      "epoch": 2.9975429975429977,
+      "step": 1626,
+      "total_flos": 5.580386298104906e+18,
+      "train_loss": 0.0,
+      "train_runtime": 6.2273,
+      "train_samples_per_second": 25095.701,
+      "train_steps_per_second": 261.107
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1626,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.580386298104906e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}