diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,15575 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9997747240369452,
+  "eval_steps": 500,
+  "global_step": 2219,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0004505519261094841,
+      "grad_norm": 26.28510414143239,
+      "learning_rate": 0.0,
+      "loss": 2.0443,
+      "step": 1
+    },
+    {
+      "epoch": 0.0009011038522189682,
+      "grad_norm": 8.222095044715916,
+      "learning_rate": 2.2106472945750374e-06,
+      "loss": 1.8172,
+      "step": 2
+    },
+    {
+      "epoch": 0.0013516557783284523,
+      "grad_norm": 8.58152531674897,
+      "learning_rate": 3.5037930642221103e-06,
+      "loss": 1.8449,
+      "step": 3
+    },
+    {
+      "epoch": 0.0018022077044379365,
+      "grad_norm": 7.344261765285006,
+      "learning_rate": 4.421294589150075e-06,
+      "loss": 1.7721,
+      "step": 4
+    },
+    {
+      "epoch": 0.0022527596305474207,
+      "grad_norm": 6.058933858290448,
+      "learning_rate": 5.132964061160519e-06,
+      "loss": 1.6875,
+      "step": 5
+    },
+    {
+      "epoch": 0.0027033115566569045,
+      "grad_norm": 6.31581635219131,
+      "learning_rate": 5.714440358797148e-06,
+      "loss": 1.6827,
+      "step": 6
+    },
+    {
+      "epoch": 0.0031538634827663887,
+      "grad_norm": 4.974134087549639,
+      "learning_rate": 6.2060715633585576e-06,
+      "loss": 1.5961,
+      "step": 7
+    },
+    {
+      "epoch": 0.003604415408875873,
+      "grad_norm": 3.6875663810117874,
+      "learning_rate": 6.631941883725112e-06,
+      "loss": 1.5536,
+      "step": 8
+    },
+    {
+      "epoch": 0.004054967334985357,
+      "grad_norm": 4.4525551165323165,
+      "learning_rate": 7.007586128444221e-06,
+      "loss": 1.5576,
+      "step": 9
+    },
+    {
+      "epoch": 0.004505519261094841,
+      "grad_norm": 3.059175705061217,
+      "learning_rate": 7.343611355735557e-06,
+      "loss": 1.5529,
+      "step": 10
+    },
+    {
+      "epoch": 0.004956071187204325,
+      "grad_norm": 3.226233366612977,
+      "learning_rate": 7.647583148507886e-06,
+      "loss": 1.5459,
+      "step": 11
+    },
+    {
+      "epoch": 0.005406623113313809,
+      "grad_norm": 2.9477028598078077,
+      "learning_rate": 7.925087653372185e-06,
+      "loss": 1.4634,
+      "step": 12
+    },
+    {
+      "epoch": 0.005857175039423294,
+      "grad_norm": 2.720800057963264,
+      "learning_rate": 8.180367051646621e-06,
+      "loss": 1.5685,
+      "step": 13
+    },
+    {
+      "epoch": 0.0063077269655327775,
+      "grad_norm": 2.9145535800847022,
+      "learning_rate": 8.416718857933595e-06,
+      "loss": 1.4726,
+      "step": 14
+    },
+    {
+      "epoch": 0.006758278891642262,
+      "grad_norm": 3.1869297889189068,
+      "learning_rate": 8.636757125382629e-06,
+      "loss": 1.5608,
+      "step": 15
+    },
+    {
+      "epoch": 0.007208830817751746,
+      "grad_norm": 2.729040726576262,
+      "learning_rate": 8.84258917830015e-06,
+      "loss": 1.4937,
+      "step": 16
+    },
+    {
+      "epoch": 0.00765938274386123,
+      "grad_norm": 2.597750532724102,
+      "learning_rate": 9.03593867168606e-06,
+      "loss": 1.4803,
+      "step": 17
+    },
+    {
+      "epoch": 0.008109934669970714,
+      "grad_norm": 2.775091573536583,
+      "learning_rate": 9.218233423019256e-06,
+      "loss": 1.5027,
+      "step": 18
+    },
+    {
+      "epoch": 0.008560486596080198,
+      "grad_norm": 2.5089161733494807,
+      "learning_rate": 9.390669465144928e-06,
+      "loss": 1.5034,
+      "step": 19
+    },
+    {
+      "epoch": 0.009011038522189683,
+      "grad_norm": 2.690280429996506,
+      "learning_rate": 9.554258650310593e-06,
+      "loss": 1.4345,
+      "step": 20
+    },
+    {
+      "epoch": 0.009461590448299166,
+      "grad_norm": 2.8272167884838284,
+      "learning_rate": 9.709864627580668e-06,
+      "loss": 1.4647,
+      "step": 21
+    },
+    {
+      "epoch": 0.00991214237440865,
+      "grad_norm": 2.378855815301042,
+      "learning_rate": 9.858230443082923e-06,
+      "loss": 1.4172,
+      "step": 22
+    },
+    {
+      "epoch": 0.010362694300518135,
+      "grad_norm": 2.3805829415787887,
+      "learning_rate": 1e-05,
+      "loss": 1.4745,
+      "step": 23
+    },
+    {
+      "epoch": 0.010813246226627618,
+      "grad_norm": 2.386709390440998,
+      "learning_rate": 1e-05,
+      "loss": 1.4517,
+      "step": 24
+    },
+    {
+      "epoch": 0.011263798152737103,
+      "grad_norm": 2.518404890144143,
+      "learning_rate": 1e-05,
+      "loss": 1.4705,
+      "step": 25
+    },
+    {
+      "epoch": 0.011714350078846587,
+      "grad_norm": 2.5205110824340933,
+      "learning_rate": 1e-05,
+      "loss": 1.4409,
+      "step": 26
+    },
+    {
+      "epoch": 0.012164902004956072,
+      "grad_norm": 2.608849272769937,
+      "learning_rate": 1e-05,
+      "loss": 1.4649,
+      "step": 27
+    },
+    {
+      "epoch": 0.012615453931065555,
+      "grad_norm": 2.724386963149695,
+      "learning_rate": 1e-05,
+      "loss": 1.4349,
+      "step": 28
+    },
+    {
+      "epoch": 0.01306600585717504,
+      "grad_norm": 2.396900395164592,
+      "learning_rate": 1e-05,
+      "loss": 1.4657,
+      "step": 29
+    },
+    {
+      "epoch": 0.013516557783284524,
+      "grad_norm": 2.301795861679911,
+      "learning_rate": 1e-05,
+      "loss": 1.4494,
+      "step": 30
+    },
+    {
+      "epoch": 0.013967109709394007,
+      "grad_norm": 2.4681646640580177,
+      "learning_rate": 1e-05,
+      "loss": 1.3577,
+      "step": 31
+    },
+    {
+      "epoch": 0.014417661635503492,
+      "grad_norm": 2.2370170283767967,
+      "learning_rate": 1e-05,
+      "loss": 1.4006,
+      "step": 32
+    },
+    {
+      "epoch": 0.014868213561612977,
+      "grad_norm": 2.3618258090169224,
+      "learning_rate": 1e-05,
+      "loss": 1.4705,
+      "step": 33
+    },
+    {
+      "epoch": 0.01531876548772246,
+      "grad_norm": 2.264234680658778,
+      "learning_rate": 1e-05,
+      "loss": 1.4492,
+      "step": 34
+    },
+    {
+      "epoch": 0.015769317413831946,
+      "grad_norm": 2.6366014899470094,
+      "learning_rate": 1e-05,
+      "loss": 1.4061,
+      "step": 35
+    },
+    {
+      "epoch": 0.016219869339941427,
+      "grad_norm": 2.4059100692684936,
+      "learning_rate": 1e-05,
+      "loss": 1.4546,
+      "step": 36
+    },
+    {
+      "epoch": 0.01667042126605091,
+      "grad_norm": 2.351443360609274,
+      "learning_rate": 1e-05,
+      "loss": 1.4255,
+      "step": 37
+    },
+    {
+      "epoch": 0.017120973192160396,
+      "grad_norm": 2.2251584813515963,
+      "learning_rate": 1e-05,
+      "loss": 1.4403,
+      "step": 38
+    },
+    {
+      "epoch": 0.01757152511826988,
+      "grad_norm": 2.4666722222794655,
+      "learning_rate": 1e-05,
+      "loss": 1.4411,
+      "step": 39
+    },
+    {
+      "epoch": 0.018022077044379366,
+      "grad_norm": 2.4349781065931113,
+      "learning_rate": 1e-05,
+      "loss": 1.4366,
+      "step": 40
+    },
+    {
+      "epoch": 0.01847262897048885,
+      "grad_norm": 2.6644037834229235,
+      "learning_rate": 1e-05,
+      "loss": 1.4366,
+      "step": 41
+    },
+    {
+      "epoch": 0.01892318089659833,
+      "grad_norm": 2.3668664318822477,
+      "learning_rate": 1e-05,
+      "loss": 1.3755,
+      "step": 42
+    },
+    {
+      "epoch": 0.019373732822707816,
+      "grad_norm": 2.289846872939494,
+      "learning_rate": 1e-05,
+      "loss": 1.383,
+      "step": 43
+    },
+    {
+      "epoch": 0.0198242847488173,
+      "grad_norm": 2.3829712616178944,
+      "learning_rate": 1e-05,
+      "loss": 1.3999,
+      "step": 44
+    },
+    {
+      "epoch": 0.020274836674926786,
+      "grad_norm": 2.2758409182803225,
+      "learning_rate": 1e-05,
+      "loss": 1.413,
+      "step": 45
+    },
+    {
+      "epoch": 0.02072538860103627,
+      "grad_norm": 2.6859081702334326,
+      "learning_rate": 1e-05,
+      "loss": 1.4004,
+      "step": 46
+    },
+    {
+      "epoch": 0.021175940527145755,
+      "grad_norm": 2.6459078901377158,
+      "learning_rate": 1e-05,
+      "loss": 1.3606,
+      "step": 47
+    },
+    {
+      "epoch": 0.021626492453255236,
+      "grad_norm": 2.30611696528225,
+      "learning_rate": 1e-05,
+      "loss": 1.4375,
+      "step": 48
+    },
+    {
+      "epoch": 0.02207704437936472,
+      "grad_norm": 2.289938283664605,
+      "learning_rate": 1e-05,
+      "loss": 1.4189,
+      "step": 49
+    },
+    {
+      "epoch": 0.022527596305474205,
+      "grad_norm": 2.276199761500501,
+      "learning_rate": 1e-05,
+      "loss": 1.4109,
+      "step": 50
+    },
+    {
+      "epoch": 0.02297814823158369,
+      "grad_norm": 2.452090967696992,
+      "learning_rate": 1e-05,
+      "loss": 1.3708,
+      "step": 51
+    },
+    {
+      "epoch": 0.023428700157693175,
+      "grad_norm": 2.2807167446012806,
+      "learning_rate": 1e-05,
+      "loss": 1.3852,
+      "step": 52
+    },
+    {
+      "epoch": 0.02387925208380266,
+      "grad_norm": 2.334407850543259,
+      "learning_rate": 1e-05,
+      "loss": 1.4386,
+      "step": 53
+    },
+    {
+      "epoch": 0.024329804009912144,
+      "grad_norm": 2.5155186958910916,
+      "learning_rate": 1e-05,
+      "loss": 1.3805,
+      "step": 54
+    },
+    {
+      "epoch": 0.024780355936021625,
+      "grad_norm": 2.3936403403814515,
+      "learning_rate": 1e-05,
+      "loss": 1.3909,
+      "step": 55
+    },
+    {
+      "epoch": 0.02523090786213111,
+      "grad_norm": 2.4199229822775976,
+      "learning_rate": 1e-05,
+      "loss": 1.3892,
+      "step": 56
+    },
+    {
+      "epoch": 0.025681459788240595,
+      "grad_norm": 2.3730631304687138,
+      "learning_rate": 1e-05,
+      "loss": 1.4074,
+      "step": 57
+    },
+    {
+      "epoch": 0.02613201171435008,
+      "grad_norm": 2.2334835941143365,
+      "learning_rate": 1e-05,
+      "loss": 1.3509,
+      "step": 58
+    },
+    {
+      "epoch": 0.026582563640459564,
+      "grad_norm": 2.374174087054882,
+      "learning_rate": 1e-05,
+      "loss": 1.4301,
+      "step": 59
+    },
+    {
+      "epoch": 0.02703311556656905,
+      "grad_norm": 2.321815293727669,
+      "learning_rate": 1e-05,
+      "loss": 1.3976,
+      "step": 60
+    },
+    {
+      "epoch": 0.02748366749267853,
+      "grad_norm": 2.504847959880212,
+      "learning_rate": 1e-05,
+      "loss": 1.3677,
+      "step": 61
+    },
+    {
+      "epoch": 0.027934219418788014,
+      "grad_norm": 2.2418023299937198,
+      "learning_rate": 1e-05,
+      "loss": 1.3904,
+      "step": 62
+    },
+    {
+      "epoch": 0.0283847713448975,
+      "grad_norm": 2.2804423991950484,
+      "learning_rate": 1e-05,
+      "loss": 1.3717,
+      "step": 63
+    },
+    {
+      "epoch": 0.028835323271006984,
+      "grad_norm": 2.219781167888153,
+      "learning_rate": 1e-05,
+      "loss": 1.3992,
+      "step": 64
+    },
+    {
+      "epoch": 0.02928587519711647,
+      "grad_norm": 2.3330607239335737,
+      "learning_rate": 1e-05,
+      "loss": 1.4061,
+      "step": 65
+    },
+    {
+      "epoch": 0.029736427123225953,
+      "grad_norm": 2.4560005590707696,
+      "learning_rate": 1e-05,
+      "loss": 1.3819,
+      "step": 66
+    },
+    {
+      "epoch": 0.030186979049335434,
+      "grad_norm": 2.1318116937147025,
+      "learning_rate": 1e-05,
+      "loss": 1.3692,
+      "step": 67
+    },
+    {
+      "epoch": 0.03063753097544492,
+      "grad_norm": 2.2285877498888333,
+      "learning_rate": 1e-05,
+      "loss": 1.4126,
+      "step": 68
+    },
+    {
+      "epoch": 0.031088082901554404,
+      "grad_norm": 2.343917803713368,
+      "learning_rate": 1e-05,
+      "loss": 1.3298,
+      "step": 69
+    },
+    {
+      "epoch": 0.03153863482766389,
+      "grad_norm": 2.242840254335559,
+      "learning_rate": 1e-05,
+      "loss": 1.3887,
+      "step": 70
+    },
+    {
+      "epoch": 0.03198918675377337,
+      "grad_norm": 2.2775060074244853,
+      "learning_rate": 1e-05,
+      "loss": 1.3772,
+      "step": 71
+    },
+    {
+      "epoch": 0.032439738679882854,
+      "grad_norm": 2.0385163756818803,
+      "learning_rate": 1e-05,
+      "loss": 1.3449,
+      "step": 72
+    },
+    {
+      "epoch": 0.03289029060599234,
+      "grad_norm": 2.5082054302710666,
+      "learning_rate": 1e-05,
+      "loss": 1.3692,
+      "step": 73
+    },
+    {
+      "epoch": 0.03334084253210182,
+      "grad_norm": 2.600727120152394,
+      "learning_rate": 1e-05,
+      "loss": 1.4177,
+      "step": 74
+    },
+    {
+      "epoch": 0.03379139445821131,
+      "grad_norm": 2.2862002649869884,
+      "learning_rate": 1e-05,
+      "loss": 1.3511,
+      "step": 75
+    },
+    {
+      "epoch": 0.03424194638432079,
+      "grad_norm": 2.427908692557385,
+      "learning_rate": 1e-05,
+      "loss": 1.3903,
+      "step": 76
+    },
+    {
+      "epoch": 0.034692498310430274,
+      "grad_norm": 2.1941151719137033,
+      "learning_rate": 1e-05,
+      "loss": 1.3511,
+      "step": 77
+    },
+    {
+      "epoch": 0.03514305023653976,
+      "grad_norm": 2.030520849428415,
+      "learning_rate": 1e-05,
+      "loss": 1.3776,
+      "step": 78
+    },
+    {
+      "epoch": 0.03559360216264924,
+      "grad_norm": 2.375592708504076,
+      "learning_rate": 1e-05,
+      "loss": 1.3724,
+      "step": 79
+    },
+    {
+      "epoch": 0.03604415408875873,
+      "grad_norm": 2.2108433561033864,
+      "learning_rate": 1e-05,
+      "loss": 1.3972,
+      "step": 80
+    },
+    {
+      "epoch": 0.03649470601486821,
+      "grad_norm": 2.4539765524786334,
+      "learning_rate": 1e-05,
+      "loss": 1.3872,
+      "step": 81
+    },
+    {
+      "epoch": 0.0369452579409777,
+      "grad_norm": 2.2671405110557306,
+      "learning_rate": 1e-05,
+      "loss": 1.3848,
+      "step": 82
+    },
+    {
+      "epoch": 0.03739580986708718,
+      "grad_norm": 2.4849840452865544,
+      "learning_rate": 1e-05,
+      "loss": 1.3752,
+      "step": 83
+    },
+    {
+      "epoch": 0.03784636179319666,
+      "grad_norm": 2.172247922858377,
+      "learning_rate": 1e-05,
+      "loss": 1.3554,
+      "step": 84
+    },
+    {
+      "epoch": 0.03829691371930615,
+      "grad_norm": 2.1670772679155914,
+      "learning_rate": 1e-05,
+      "loss": 1.3743,
+      "step": 85
+    },
+    {
+      "epoch": 0.03874746564541563,
+      "grad_norm": 2.191176057713791,
+      "learning_rate": 1e-05,
+      "loss": 1.3902,
+      "step": 86
+    },
+    {
+      "epoch": 0.03919801757152512,
+      "grad_norm": 2.334687357837179,
+      "learning_rate": 1e-05,
+      "loss": 1.3638,
+      "step": 87
+    },
+    {
+      "epoch": 0.0396485694976346,
+      "grad_norm": 2.0920629315766246,
+      "learning_rate": 1e-05,
+      "loss": 1.3687,
+      "step": 88
+    },
+    {
+      "epoch": 0.04009912142374409,
+      "grad_norm": 2.2442009548950725,
+      "learning_rate": 1e-05,
+      "loss": 1.3735,
+      "step": 89
+    },
+    {
+      "epoch": 0.04054967334985357,
+      "grad_norm": 2.4644253798107654,
+      "learning_rate": 1e-05,
+      "loss": 1.3352,
+      "step": 90
+    },
+    {
+      "epoch": 0.04100022527596305,
+      "grad_norm": 3.1495591252992865,
+      "learning_rate": 1e-05,
+      "loss": 1.3492,
+      "step": 91
+    },
+    {
+      "epoch": 0.04145077720207254,
+      "grad_norm": 2.2158388490503933,
+      "learning_rate": 1e-05,
+      "loss": 1.391,
+      "step": 92
+    },
+    {
+      "epoch": 0.04190132912818202,
+      "grad_norm": 2.2643615549031253,
+      "learning_rate": 1e-05,
+      "loss": 1.3922,
+      "step": 93
+    },
+    {
+      "epoch": 0.04235188105429151,
+      "grad_norm": 2.277253984187307,
+      "learning_rate": 1e-05,
+      "loss": 1.3235,
+      "step": 94
+    },
+    {
+      "epoch": 0.04280243298040099,
+      "grad_norm": 2.4652877710696646,
+      "learning_rate": 1e-05,
+      "loss": 1.3388,
+      "step": 95
+    },
+    {
+      "epoch": 0.04325298490651047,
+      "grad_norm": 2.3757029746456726,
+      "learning_rate": 1e-05,
+      "loss": 1.393,
+      "step": 96
+    },
+    {
+      "epoch": 0.04370353683261996,
+      "grad_norm": 2.5231305121620835,
+      "learning_rate": 1e-05,
+      "loss": 1.3897,
+      "step": 97
+    },
+    {
+      "epoch": 0.04415408875872944,
+      "grad_norm": 2.1369778554629395,
+      "learning_rate": 1e-05,
+      "loss": 1.3642,
+      "step": 98
+    },
+    {
+      "epoch": 0.04460464068483893,
+      "grad_norm": 2.1260402398799743,
+      "learning_rate": 1e-05,
+      "loss": 1.3484,
+      "step": 99
+    },
+    {
+      "epoch": 0.04505519261094841,
+      "grad_norm": 2.309709263259357,
+      "learning_rate": 1e-05,
+      "loss": 1.3736,
+      "step": 100
+    },
+    {
+      "epoch": 0.0455057445370579,
+      "grad_norm": 2.1011110408842506,
+      "learning_rate": 1e-05,
+      "loss": 1.3783,
+      "step": 101
+    },
+    {
+      "epoch": 0.04595629646316738,
+      "grad_norm": 2.1914831067805336,
+      "learning_rate": 1e-05,
+      "loss": 1.3469,
+      "step": 102
+    },
+    {
+      "epoch": 0.04640684838927686,
+      "grad_norm": 2.241251638945157,
+      "learning_rate": 1e-05,
+      "loss": 1.3201,
+      "step": 103
+    },
+    {
+      "epoch": 0.04685740031538635,
+      "grad_norm": 2.209731187051675,
+      "learning_rate": 1e-05,
+      "loss": 1.343,
+      "step": 104
+    },
+    {
+      "epoch": 0.04730795224149583,
+      "grad_norm": 2.322197381391765,
+      "learning_rate": 1e-05,
+      "loss": 1.3629,
+      "step": 105
+    },
+    {
+      "epoch": 0.04775850416760532,
+      "grad_norm": 2.5964458053354282,
+      "learning_rate": 1e-05,
+      "loss": 1.3011,
+      "step": 106
+    },
+    {
+      "epoch": 0.0482090560937148,
+      "grad_norm": 2.3635572590606606,
+      "learning_rate": 1e-05,
+      "loss": 1.3734,
+      "step": 107
+    },
+    {
+      "epoch": 0.04865960801982429,
+      "grad_norm": 2.3162924552707094,
+      "learning_rate": 1e-05,
+      "loss": 1.3296,
+      "step": 108
+    },
+    {
+      "epoch": 0.04911015994593377,
+      "grad_norm": 2.280106792732187,
+      "learning_rate": 1e-05,
+      "loss": 1.386,
+      "step": 109
+    },
+    {
+      "epoch": 0.04956071187204325,
+      "grad_norm": 2.24924470423268,
+      "learning_rate": 1e-05,
+      "loss": 1.3415,
+      "step": 110
+    },
+    {
+      "epoch": 0.05001126379815274,
+      "grad_norm": 2.267342003393844,
+      "learning_rate": 1e-05,
+      "loss": 1.2756,
+      "step": 111
+    },
+    {
+      "epoch": 0.05046181572426222,
+      "grad_norm": 2.1193683158088987,
+      "learning_rate": 1e-05,
+      "loss": 1.3389,
+      "step": 112
+    },
+    {
+      "epoch": 0.05091236765037171,
+      "grad_norm": 2.187500036093264,
+      "learning_rate": 1e-05,
+      "loss": 1.3599,
+      "step": 113
+    },
+    {
+      "epoch": 0.05136291957648119,
+      "grad_norm": 2.2362627224448626,
+      "learning_rate": 1e-05,
+      "loss": 1.346,
+      "step": 114
+    },
+    {
+      "epoch": 0.05181347150259067,
+      "grad_norm": 2.189663703662425,
+      "learning_rate": 1e-05,
+      "loss": 1.3714,
+      "step": 115
+    },
+    {
+      "epoch": 0.05226402342870016,
+      "grad_norm": 2.300810784853383,
+      "learning_rate": 1e-05,
+      "loss": 1.3997,
+      "step": 116
+    },
+    {
+      "epoch": 0.05271457535480964,
+      "grad_norm": 2.3150676660413927,
+      "learning_rate": 1e-05,
+      "loss": 1.3265,
+      "step": 117
+    },
+    {
+      "epoch": 0.05316512728091913,
+      "grad_norm": 2.2842180322320575,
+      "learning_rate": 1e-05,
+      "loss": 1.37,
+      "step": 118
+    },
+    {
+      "epoch": 0.05361567920702861,
+      "grad_norm": 2.2297837308028696,
+      "learning_rate": 1e-05,
+      "loss": 1.4089,
+      "step": 119
+    },
+    {
+      "epoch": 0.0540662311331381,
+      "grad_norm": 2.402291874295927,
+      "learning_rate": 1e-05,
+      "loss": 1.4087,
+      "step": 120
+    },
+    {
+      "epoch": 0.05451678305924758,
+      "grad_norm": 2.290491505586159,
+      "learning_rate": 1e-05,
+      "loss": 1.3614,
+      "step": 121
+    },
+    {
+      "epoch": 0.05496733498535706,
+      "grad_norm": 2.065772896787654,
+      "learning_rate": 1e-05,
+      "loss": 1.3346,
+      "step": 122
+    },
+    {
+      "epoch": 0.05541788691146655,
+      "grad_norm": 2.370429850220542,
+      "learning_rate": 1e-05,
+      "loss": 1.3106,
+      "step": 123
+    },
+    {
+      "epoch": 0.05586843883757603,
+      "grad_norm": 2.2132653252379053,
+      "learning_rate": 1e-05,
+      "loss": 1.3197,
+      "step": 124
+    },
+    {
+      "epoch": 0.05631899076368552,
+      "grad_norm": 2.154390296688665,
+      "learning_rate": 1e-05,
+      "loss": 1.3043,
+      "step": 125
+    },
+    {
+      "epoch": 0.056769542689795,
+      "grad_norm": 2.304743722679258,
+      "learning_rate": 1e-05,
+      "loss": 1.3378,
+      "step": 126
+    },
+    {
+      "epoch": 0.057220094615904486,
+      "grad_norm": 1.9985607835453456,
+      "learning_rate": 1e-05,
+      "loss": 1.3635,
+      "step": 127
+    },
+    {
+      "epoch": 0.05767064654201397,
+      "grad_norm": 2.3799647674283366,
+      "learning_rate": 1e-05,
+      "loss": 1.391,
+      "step": 128
+    },
+    {
+      "epoch": 0.05812119846812345,
+      "grad_norm": 2.2473413860736375,
+      "learning_rate": 1e-05,
+      "loss": 1.351,
+      "step": 129
+    },
+    {
+      "epoch": 0.05857175039423294,
+      "grad_norm": 2.377688590024442,
+      "learning_rate": 1e-05,
+      "loss": 1.3401,
+      "step": 130
+    },
+    {
+      "epoch": 0.05902230232034242,
+      "grad_norm": 2.2073273079626836,
+      "learning_rate": 1e-05,
+      "loss": 1.3573,
+      "step": 131
+    },
+    {
+      "epoch": 0.059472854246451906,
+      "grad_norm": 2.190356460206687,
+      "learning_rate": 1e-05,
+      "loss": 1.3501,
+      "step": 132
+    },
+    {
+      "epoch": 0.05992340617256139,
+      "grad_norm": 2.0280828601230962,
+      "learning_rate": 1e-05,
+      "loss": 1.2865,
+      "step": 133
+    },
+    {
+      "epoch": 0.06037395809867087,
+      "grad_norm": 2.3849567762471136,
+      "learning_rate": 1e-05,
+      "loss": 1.3469,
+      "step": 134
+    },
+    {
+      "epoch": 0.06082451002478036,
+      "grad_norm": 2.183864843860172,
+      "learning_rate": 1e-05,
+      "loss": 1.2862,
+      "step": 135
+    },
+    {
+      "epoch": 0.06127506195088984,
+      "grad_norm": 2.1465217395570586,
+      "learning_rate": 1e-05,
+      "loss": 1.3769,
+      "step": 136
+    },
+    {
+      "epoch": 0.061725613876999326,
+      "grad_norm": 2.2304904790498217,
+      "learning_rate": 1e-05,
+      "loss": 1.3463,
+      "step": 137
+    },
+    {
+      "epoch": 0.06217616580310881,
+      "grad_norm": 2.245862844678027,
+      "learning_rate": 1e-05,
+      "loss": 1.3316,
+      "step": 138
+    },
+    {
+      "epoch": 0.0626267177292183,
+      "grad_norm": 2.0441903472041445,
+      "learning_rate": 1e-05,
+      "loss": 1.3554,
+      "step": 139
+    },
+    {
+      "epoch": 0.06307726965532778,
+      "grad_norm": 2.1541155197686557,
+      "learning_rate": 1e-05,
+      "loss": 1.3408,
+      "step": 140
+    },
+    {
+      "epoch": 0.06352782158143726,
+      "grad_norm": 2.103704298075221,
+      "learning_rate": 1e-05,
+      "loss": 1.3109,
+      "step": 141
+    },
+    {
+      "epoch": 0.06397837350754675,
+      "grad_norm": 2.143436032315168,
+      "learning_rate": 1e-05,
+      "loss": 1.344,
+      "step": 142
+    },
+    {
+      "epoch": 0.06442892543365623,
+      "grad_norm": 1.9869633772847095,
+      "learning_rate": 1e-05,
+      "loss": 1.3568,
+      "step": 143
+    },
+    {
+      "epoch": 0.06487947735976571,
+      "grad_norm": 2.0902245114076834,
+      "learning_rate": 1e-05,
+      "loss": 1.3089,
+      "step": 144
+    },
+    {
+      "epoch": 0.0653300292858752,
+      "grad_norm": 2.3943683613913453,
+      "learning_rate": 1e-05,
+      "loss": 1.3242,
+      "step": 145
+    },
+    {
+      "epoch": 0.06578058121198468,
+      "grad_norm": 2.0650060989119723,
+      "learning_rate": 1e-05,
+      "loss": 1.3943,
+      "step": 146
+    },
+    {
+      "epoch": 0.06623113313809416,
+      "grad_norm": 2.209294484404208,
+      "learning_rate": 1e-05,
+      "loss": 1.3244,
+      "step": 147
+    },
+    {
+      "epoch": 0.06668168506420365,
+      "grad_norm": 2.2369927550510846,
+      "learning_rate": 1e-05,
+      "loss": 1.3697,
+      "step": 148
+    },
+    {
+      "epoch": 0.06713223699031313,
+      "grad_norm": 2.0944369232164486,
+      "learning_rate": 1e-05,
+      "loss": 1.3656,
+      "step": 149
+    },
+    {
+      "epoch": 0.06758278891642262,
+      "grad_norm": 2.938063368949922,
+      "learning_rate": 1e-05,
+      "loss": 1.3448,
+      "step": 150
+    },
+    {
+      "epoch": 0.0680333408425321,
+      "grad_norm": 2.2018520024011092,
+      "learning_rate": 1e-05,
+      "loss": 1.3635,
+      "step": 151
+    },
+    {
+      "epoch": 0.06848389276864159,
+      "grad_norm": 2.088835022914261,
+      "learning_rate": 1e-05,
+      "loss": 1.3868,
+      "step": 152
+    },
+    {
+      "epoch": 0.06893444469475107,
+      "grad_norm": 2.27935479884058,
+      "learning_rate": 1e-05,
+      "loss": 1.3474,
+      "step": 153
+    },
+    {
+      "epoch": 0.06938499662086055,
+      "grad_norm": 2.1759713573337045,
+      "learning_rate": 1e-05,
+      "loss": 1.3319,
+      "step": 154
+    },
+    {
+      "epoch": 0.06983554854697004,
+      "grad_norm": 2.0986604325792273,
+      "learning_rate": 1e-05,
+      "loss": 1.3298,
+      "step": 155
+    },
+    {
+      "epoch": 0.07028610047307952,
+      "grad_norm": 2.0356679638693005,
+      "learning_rate": 1e-05,
+      "loss": 1.3449,
+      "step": 156
+    },
+    {
+      "epoch": 0.07073665239918901,
+      "grad_norm": 2.019473324752186,
+      "learning_rate": 1e-05,
+      "loss": 1.3056,
+      "step": 157
+    },
+    {
+      "epoch": 0.07118720432529849,
+      "grad_norm": 2.0150736309932267,
+      "learning_rate": 1e-05,
+      "loss": 1.3257,
+      "step": 158
+    },
+    {
+      "epoch": 0.07163775625140797,
+      "grad_norm": 2.0434827081661826,
+      "learning_rate": 1e-05,
+      "loss": 1.2815,
+      "step": 159
+    },
+    {
+      "epoch": 0.07208830817751746,
+      "grad_norm": 1.97423799741727,
+      "learning_rate": 1e-05,
+      "loss": 1.3446,
+      "step": 160
+    },
+    {
+      "epoch": 0.07253886010362694,
+      "grad_norm": 2.086993457301471,
+      "learning_rate": 1e-05,
+      "loss": 1.3419,
+      "step": 161
+    },
+    {
+      "epoch": 0.07298941202973643,
+      "grad_norm": 2.0583664808721416,
+      "learning_rate": 1e-05,
+      "loss": 1.3712,
+      "step": 162
+    },
+    {
+      "epoch": 0.07343996395584591,
+      "grad_norm": 1.972502617783117,
+      "learning_rate": 1e-05,
+      "loss": 1.297,
+      "step": 163
+    },
+    {
+      "epoch": 0.0738905158819554,
+      "grad_norm": 1.9403535984947597,
+      "learning_rate": 1e-05,
+      "loss": 1.2888,
+      "step": 164
+    },
+    {
+      "epoch": 0.07434106780806488,
+      "grad_norm": 2.0284573800273358,
+      "learning_rate": 1e-05,
+      "loss": 1.2949,
+      "step": 165
+    },
+    {
+      "epoch": 0.07479161973417436,
+      "grad_norm": 2.255410970147385,
+      "learning_rate": 1e-05,
+      "loss": 1.3494,
+      "step": 166
+    },
+    {
+      "epoch": 0.07524217166028385,
+      "grad_norm": 1.9373576070929324,
+      "learning_rate": 1e-05,
+      "loss": 1.3459,
+      "step": 167
+    },
+    {
+      "epoch": 0.07569272358639333,
+      "grad_norm": 2.0643336639670635,
+      "learning_rate": 1e-05,
+      "loss": 1.3367,
+      "step": 168
+    },
+    {
+      "epoch": 0.07614327551250281,
+      "grad_norm": 2.19869988638025,
+      "learning_rate": 1e-05,
+      "loss": 1.3395,
+      "step": 169
+    },
+    {
+      "epoch": 0.0765938274386123,
+      "grad_norm": 2.0235337047791413,
+      "learning_rate": 1e-05,
+      "loss": 1.2544,
+      "step": 170
+    },
+    {
+      "epoch": 0.07704437936472179,
+      "grad_norm": 2.0789686417964157,
+      "learning_rate": 1e-05,
+      "loss": 1.3147,
+      "step": 171
+    },
+    {
+      "epoch": 0.07749493129083126,
+      "grad_norm": 2.146030217715539,
+      "learning_rate": 1e-05,
+      "loss": 1.2945,
+      "step": 172
+    },
+    {
+      "epoch": 0.07794548321694075,
+      "grad_norm": 2.2702052358432585,
+      "learning_rate": 1e-05,
+      "loss": 1.3205,
+      "step": 173
+    },
+    {
+      "epoch": 0.07839603514305024,
+      "grad_norm": 2.074969369993277,
+      "learning_rate": 1e-05,
+      "loss": 1.3681,
+      "step": 174
+    },
+    {
+      "epoch": 0.07884658706915972,
+      "grad_norm": 2.0882868293617833,
+      "learning_rate": 1e-05,
+      "loss": 1.3324,
+      "step": 175
+    },
+    {
+      "epoch": 0.0792971389952692,
+      "grad_norm": 2.2428849524821275,
+      "learning_rate": 1e-05,
+      "loss": 1.3431,
+      "step": 176
+    },
+    {
+      "epoch": 0.07974769092137869,
+      "grad_norm": 1.9430898055672532,
+      "learning_rate": 1e-05,
+      "loss": 1.2792,
+      "step": 177
+    },
+    {
+      "epoch": 0.08019824284748818,
+      "grad_norm": 2.151887159652184,
+      "learning_rate": 1e-05,
+      "loss": 1.3077,
+      "step": 178
+    },
+    {
+      "epoch": 0.08064879477359765,
+      "grad_norm": 2.3642708132508754,
+      "learning_rate": 1e-05,
+      "loss": 1.3221,
+      "step": 179
+    },
+    {
+      "epoch": 0.08109934669970714,
+      "grad_norm": 2.0851971556207447,
+      "learning_rate": 1e-05,
+      "loss": 1.3298,
+      "step": 180
+    },
+    {
+      "epoch": 0.08154989862581663,
+      "grad_norm": 2.1372303530266765,
+      "learning_rate": 1e-05,
+      "loss": 1.2822,
+      "step": 181
+    },
+    {
+      "epoch": 0.0820004505519261,
+      "grad_norm": 2.025927429224001,
+      "learning_rate": 1e-05,
+      "loss": 1.3042,
+      "step": 182
+    },
+    {
+      "epoch": 0.08245100247803559,
+      "grad_norm": 2.215614207426725,
+      "learning_rate": 1e-05,
+      "loss": 1.2804,
+      "step": 183
+    },
+    {
+      "epoch": 0.08290155440414508,
+      "grad_norm": 2.147530296171878,
+      "learning_rate": 1e-05,
+      "loss": 1.3324,
+      "step": 184
+    },
+    {
+      "epoch": 0.08335210633025456,
+      "grad_norm": 2.1468751331920455,
+      "learning_rate": 1e-05,
+      "loss": 1.3324,
+      "step": 185
+    },
+    {
+      "epoch": 0.08380265825636404,
+      "grad_norm": 1.8657617259301975,
+      "learning_rate": 1e-05,
+      "loss": 1.3313,
+      "step": 186
+    },
+    {
+      "epoch": 0.08425321018247353,
+      "grad_norm": 2.1810123393988707,
+      "learning_rate": 1e-05,
+      "loss": 1.3239,
+      "step": 187
+    },
+    {
+      "epoch": 0.08470376210858302,
+      "grad_norm": 2.26292004700728,
+      "learning_rate": 1e-05,
+      "loss": 1.3161,
+      "step": 188
+    },
+    {
+      "epoch": 0.0851543140346925,
+      "grad_norm": 2.2356541214457555,
+      "learning_rate": 1e-05,
+      "loss": 1.2959,
+      "step": 189
+    },
+    {
+      "epoch": 0.08560486596080198,
+      "grad_norm": 2.340511553816989,
+      "learning_rate": 1e-05,
+      "loss": 1.3412,
+      "step": 190
+    },
+    {
+      "epoch": 0.08605541788691147,
+      "grad_norm": 1.9727356523181143,
+      "learning_rate": 1e-05,
+      "loss": 1.2852,
+      "step": 191
+    },
+    {
+      "epoch": 0.08650596981302094,
+      "grad_norm": 1.9979926064436653,
+      "learning_rate": 1e-05,
+      "loss": 1.3461,
+      "step": 192
+    },
+    {
+      "epoch": 0.08695652173913043,
+      "grad_norm": 2.088054262265907,
+      "learning_rate": 1e-05,
+      "loss": 1.3649,
+      "step": 193
+    },
+    {
+      "epoch": 0.08740707366523992,
+      "grad_norm": 2.049726673942621,
+      "learning_rate": 1e-05,
+      "loss": 1.281,
+      "step": 194
+    },
+    {
+      "epoch": 0.08785762559134941,
+      "grad_norm": 2.2460539886132325,
+      "learning_rate": 1e-05,
+      "loss": 1.2916,
+      "step": 195
+    },
+    {
+      "epoch": 0.08830817751745888,
+      "grad_norm": 2.294882839143102,
+      "learning_rate": 1e-05,
+      "loss": 1.3023,
+      "step": 196
+    },
+    {
+      "epoch": 0.08875872944356837,
+      "grad_norm": 2.0105964141212778,
+      "learning_rate": 1e-05,
+      "loss": 1.2945,
+      "step": 197
+    },
+    {
+      "epoch": 0.08920928136967786,
+      "grad_norm": 2.1261627253589355,
+      "learning_rate": 1e-05,
+      "loss": 1.3162,
+      "step": 198
+    },
+    {
+      "epoch": 0.08965983329578733,
+      "grad_norm": 2.1558255460157207,
+      "learning_rate": 1e-05,
+      "loss": 1.3119,
+      "step": 199
+    },
+    {
+      "epoch": 0.09011038522189682,
+      "grad_norm": 2.3615824846619637,
+      "learning_rate": 1e-05,
+      "loss": 1.355,
+      "step": 200
+    },
+    {
+      "epoch": 0.09056093714800631,
+      "grad_norm": 2.1010711545597682,
+      "learning_rate": 1e-05,
+      "loss": 1.2998,
+      "step": 201
+    },
+    {
+      "epoch": 0.0910114890741158,
+      "grad_norm": 2.1120748106921043,
+      "learning_rate": 1e-05,
+      "loss": 1.25,
+      "step": 202
+    },
+    {
+      "epoch": 0.09146204100022527,
+      "grad_norm": 2.1820679615413705,
+      "learning_rate": 1e-05,
+      "loss": 1.266,
+      "step": 203
+    },
+    {
+      "epoch": 0.09191259292633476,
+      "grad_norm": 2.294622936183944,
+      "learning_rate": 1e-05,
+      "loss": 1.3235,
+      "step": 204
+    },
+    {
+      "epoch": 0.09236314485244425,
+      "grad_norm": 2.154727340275725,
+      "learning_rate": 1e-05,
+      "loss": 1.2828,
+      "step": 205
+    },
+    {
+      "epoch": 0.09281369677855372,
+      "grad_norm": 1.9902518644805207,
+      "learning_rate": 1e-05,
+      "loss": 1.3093,
+      "step": 206
+    },
+    {
+      "epoch": 0.09326424870466321,
+      "grad_norm": 2.1266591160221684,
+      "learning_rate": 1e-05,
+      "loss": 1.2783,
+      "step": 207
+    },
+    {
+      "epoch": 0.0937148006307727,
+      "grad_norm": 2.10569335618675,
+      "learning_rate": 1e-05,
+      "loss": 1.3027,
+      "step": 208
+    },
+    {
+      "epoch": 0.09416535255688219,
+      "grad_norm": 2.0834789625484156,
+      "learning_rate": 1e-05,
+      "loss": 1.2972,
+      "step": 209
+    },
+    {
+      "epoch": 0.09461590448299166,
+      "grad_norm": 2.201298293871812,
+      "learning_rate": 1e-05,
+      "loss": 1.3269,
+      "step": 210
+    },
+    {
+      "epoch": 0.09506645640910115,
+      "grad_norm": 2.072133926573633,
+      "learning_rate": 1e-05,
+      "loss": 1.2411,
+      "step": 211
+    },
+    {
+      "epoch": 0.09551700833521064,
+      "grad_norm": 2.160638257283922,
+      "learning_rate": 1e-05,
+      "loss": 1.3115,
+      "step": 212
+    },
+    {
+      "epoch": 0.09596756026132011,
+      "grad_norm": 2.1756470916488544,
+      "learning_rate": 1e-05,
+      "loss": 1.3439,
+      "step": 213
+    },
+    {
+      "epoch": 0.0964181121874296,
+      "grad_norm": 1.9143146267256204,
+      "learning_rate": 1e-05,
+      "loss": 1.2647,
+      "step": 214
+    },
+    {
+      "epoch": 0.09686866411353909,
+      "grad_norm": 2.3050841717765578,
+      "learning_rate": 1e-05,
+      "loss": 1.3248,
+      "step": 215
+    },
+    {
+      "epoch": 0.09731921603964858,
+      "grad_norm": 2.4165019734863815,
+      "learning_rate": 1e-05,
+      "loss": 1.258,
+      "step": 216
+    },
+    {
+      "epoch": 0.09776976796575805,
+      "grad_norm": 2.2372311885078084,
+      "learning_rate": 1e-05,
+      "loss": 1.2905,
+      "step": 217
+    },
+    {
+      "epoch": 0.09822031989186754,
+      "grad_norm": 2.04193161203431,
+      "learning_rate": 1e-05,
+      "loss": 1.3026,
+      "step": 218
+    },
+    {
+      "epoch": 0.09867087181797703,
+      "grad_norm": 1.954613609808351,
+      "learning_rate": 1e-05,
+      "loss": 1.2869,
+      "step": 219
+    },
+    {
+      "epoch": 0.0991214237440865,
+      "grad_norm": 1.976265979622734,
+      "learning_rate": 1e-05,
+      "loss": 1.2519,
+      "step": 220
+    },
+    {
+      "epoch": 0.09957197567019599,
+      "grad_norm": 2.344636898145692,
+      "learning_rate": 1e-05,
+      "loss": 1.261,
+      "step": 221
+    },
+    {
+      "epoch": 0.10002252759630548,
+      "grad_norm": 2.079147567257483,
+      "learning_rate": 1e-05,
+      "loss": 1.3017,
+      "step": 222
+    },
+    {
+      "epoch": 0.10047307952241495,
+      "grad_norm": 1.9433808675623756,
+      "learning_rate": 1e-05,
+      "loss": 1.3133,
+      "step": 223
+    },
+    {
+      "epoch": 0.10092363144852444,
+      "grad_norm": 2.1198616346721937,
+      "learning_rate": 1e-05,
+      "loss": 1.3428,
+      "step": 224
+    },
+    {
+      "epoch": 0.10137418337463393,
+      "grad_norm": 2.2434673582395495,
+      "learning_rate": 1e-05,
+      "loss": 1.2892,
+      "step": 225
+    },
+    {
+      "epoch": 0.10182473530074342,
+      "grad_norm": 2.1693804368854783,
+      "learning_rate": 1e-05,
+      "loss": 1.3214,
+      "step": 226
+    },
+    {
+      "epoch": 0.10227528722685289,
+      "grad_norm": 2.1854057215784124,
+      "learning_rate": 1e-05,
+      "loss": 1.2451,
+      "step": 227
+    },
+    {
+      "epoch": 0.10272583915296238,
+      "grad_norm": 2.2579876090328095,
+      "learning_rate": 1e-05,
+      "loss": 1.283,
+      "step": 228
+    },
+    {
+      "epoch": 0.10317639107907187,
+      "grad_norm": 2.1231468309558963,
+      "learning_rate": 1e-05,
+      "loss": 1.355,
+      "step": 229
+    },
+    {
+      "epoch": 0.10362694300518134,
+      "grad_norm": 1.872151743679353,
+      "learning_rate": 1e-05,
+      "loss": 1.324,
+      "step": 230
+    },
+    {
+      "epoch": 0.10407749493129083,
+      "grad_norm": 2.22398898445514,
+      "learning_rate": 1e-05,
+      "loss": 1.328,
+      "step": 231
+    },
+    {
+      "epoch": 0.10452804685740032,
+      "grad_norm": 2.1507743930092404,
+      "learning_rate": 1e-05,
+      "loss": 1.3638,
+      "step": 232
+    },
+    {
+      "epoch": 0.1049785987835098,
+      "grad_norm": 2.2800866611460138,
+      "learning_rate": 1e-05,
+      "loss": 1.2881,
+      "step": 233
+    },
+    {
+      "epoch": 0.10542915070961928,
+      "grad_norm": 2.2289538826428816,
+      "learning_rate": 1e-05,
+      "loss": 1.2792,
+      "step": 234
+    },
+    {
+      "epoch": 0.10587970263572877,
+      "grad_norm": 2.009099245481224,
+      "learning_rate": 1e-05,
+      "loss": 1.3218,
+      "step": 235
+    },
+    {
+      "epoch": 0.10633025456183826,
+      "grad_norm": 2.006015485548607,
+      "learning_rate": 1e-05,
+      "loss": 1.3089,
+      "step": 236
+    },
+    {
+      "epoch": 0.10678080648794773,
+      "grad_norm": 2.1255543388528726,
+      "learning_rate": 1e-05,
+      "loss": 1.3231,
+      "step": 237
+    },
+    {
+      "epoch": 0.10723135841405722,
+      "grad_norm": 2.173061330250712,
+      "learning_rate": 1e-05,
+      "loss": 1.2384,
+      "step": 238
+    },
+    {
+      "epoch": 0.1076819103401667,
+      "grad_norm": 2.2818871812577597,
+      "learning_rate": 1e-05,
+      "loss": 1.2709,
+      "step": 239
+    },
+    {
+      "epoch": 0.1081324622662762,
+      "grad_norm": 2.3229635064658707,
+      "learning_rate": 1e-05,
+      "loss": 1.2506,
+      "step": 240
+    },
+    {
+      "epoch": 0.10858301419238567,
+      "grad_norm": 2.0566676446728005,
+      "learning_rate": 1e-05,
+      "loss": 1.292,
+      "step": 241
+    },
+    {
+      "epoch": 0.10903356611849516,
+      "grad_norm": 2.155920765024828,
+      "learning_rate": 1e-05,
+      "loss": 1.3153,
+      "step": 242
+    },
+    {
+      "epoch": 0.10948411804460464,
+      "grad_norm": 2.0841140126822943,
+      "learning_rate": 1e-05,
+      "loss": 1.2806,
+      "step": 243
+    },
+    {
+      "epoch": 0.10993466997071412,
+      "grad_norm": 2.1393721399250687,
+      "learning_rate": 1e-05,
+      "loss": 1.3261,
+      "step": 244
+    },
+    {
+      "epoch": 0.11038522189682361,
+      "grad_norm": 2.1221868581545116,
+      "learning_rate": 1e-05,
+      "loss": 1.2892,
+      "step": 245
+    },
+    {
+      "epoch": 0.1108357738229331,
+      "grad_norm": 2.1755867538401263,
+      "learning_rate": 1e-05,
+      "loss": 1.324,
+      "step": 246
+    },
+    {
+      "epoch": 0.11128632574904258,
+      "grad_norm": 2.319444842074109,
+      "learning_rate": 1e-05,
+      "loss": 1.3057,
+      "step": 247
+    },
+    {
+      "epoch": 0.11173687767515206,
+      "grad_norm": 2.1103232906021083,
+      "learning_rate": 1e-05,
+      "loss": 1.322,
+      "step": 248
+    },
+    {
+      "epoch": 0.11218742960126155,
+      "grad_norm": 2.1680871784458136,
+      "learning_rate": 1e-05,
+      "loss": 1.3126,
+      "step": 249
+    },
+    {
+      "epoch": 0.11263798152737103,
+      "grad_norm": 1.9897648507815235,
+      "learning_rate": 1e-05,
+      "loss": 1.3266,
+      "step": 250
+    },
+    {
+      "epoch": 0.11308853345348051,
+      "grad_norm": 2.1452185458575572,
+      "learning_rate": 1e-05,
+      "loss": 1.2991,
+      "step": 251
+    },
+    {
+      "epoch": 0.11353908537959,
+      "grad_norm": 2.0639404649517687,
+      "learning_rate": 1e-05,
+      "loss": 1.3345,
+      "step": 252
+    },
+    {
+      "epoch": 0.11398963730569948,
+      "grad_norm": 2.2265474509862733,
+      "learning_rate": 1e-05,
+      "loss": 1.3338,
+      "step": 253
+    },
+    {
+      "epoch": 0.11444018923180897,
+      "grad_norm": 2.1158310661962862,
+      "learning_rate": 1e-05,
+      "loss": 1.2721,
+      "step": 254
+    },
+    {
+      "epoch": 0.11489074115791845,
+      "grad_norm": 2.150280623912654,
+      "learning_rate": 1e-05,
+      "loss": 1.3154,
+      "step": 255
+    },
+    {
+      "epoch": 0.11534129308402793,
+      "grad_norm": 2.0166941783038634,
+      "learning_rate": 1e-05,
+      "loss": 1.2752,
+      "step": 256
+    },
+    {
+      "epoch": 0.11579184501013742,
+      "grad_norm": 2.108035622623124,
+      "learning_rate": 1e-05,
+      "loss": 1.2931,
+      "step": 257
+    },
+    {
+      "epoch": 0.1162423969362469,
+      "grad_norm": 2.042157781896836,
+      "learning_rate": 1e-05,
+      "loss": 1.3069,
+      "step": 258
+    },
+    {
+      "epoch": 0.11669294886235639,
+      "grad_norm": 2.0251060091348783,
+      "learning_rate": 1e-05,
+      "loss": 1.2555,
+      "step": 259
+    },
+    {
+      "epoch": 0.11714350078846587,
+      "grad_norm": 2.295329853473669,
+      "learning_rate": 1e-05,
+      "loss": 1.3083,
+      "step": 260
+    },
+    {
+      "epoch": 0.11759405271457536,
+      "grad_norm": 2.0714716088770815,
+      "learning_rate": 1e-05,
+      "loss": 1.2937,
+      "step": 261
+    },
+    {
+      "epoch": 0.11804460464068484,
+      "grad_norm": 2.396354053446123,
+      "learning_rate": 1e-05,
+      "loss": 1.2919,
+      "step": 262
+    },
+    {
+      "epoch": 0.11849515656679432,
+      "grad_norm": 2.200235892723831,
+      "learning_rate": 1e-05,
+      "loss": 1.3423,
+      "step": 263
+    },
+    {
+      "epoch": 0.11894570849290381,
+      "grad_norm": 1.9819306761172546,
+      "learning_rate": 1e-05,
+      "loss": 1.3062,
+      "step": 264
+    },
+    {
+      "epoch": 0.11939626041901329,
+      "grad_norm": 1.9841435404879328,
+      "learning_rate": 1e-05,
+      "loss": 1.2746,
+      "step": 265
+    },
+    {
+      "epoch": 0.11984681234512277,
+      "grad_norm": 2.159956082135299,
+      "learning_rate": 1e-05,
+      "loss": 1.3104,
+      "step": 266
+    },
+    {
+      "epoch": 0.12029736427123226,
+      "grad_norm": 1.9200585591152597,
+      "learning_rate": 1e-05,
+      "loss": 1.2983,
+      "step": 267
+    },
+    {
+      "epoch": 0.12074791619734174,
+      "grad_norm": 2.5259249629148592,
+      "learning_rate": 1e-05,
+      "loss": 1.2992,
+      "step": 268
+    },
+    {
+      "epoch": 0.12119846812345123,
+      "grad_norm": 1.9398891739080681,
+      "learning_rate": 1e-05,
+      "loss": 1.2928,
+      "step": 269
+    },
+    {
+      "epoch": 0.12164902004956071,
+      "grad_norm": 1.9790756436442474,
+      "learning_rate": 1e-05,
+      "loss": 1.3188,
+      "step": 270
+    },
+    {
+      "epoch": 0.1220995719756702,
+      "grad_norm": 2.242052817808025,
+      "learning_rate": 1e-05,
+      "loss": 1.2804,
+      "step": 271
+    },
+    {
+      "epoch": 0.12255012390177968,
+      "grad_norm": 2.122027788554486,
+      "learning_rate": 1e-05,
+      "loss": 1.2669,
+      "step": 272
+    },
+    {
+      "epoch": 0.12300067582788916,
+      "grad_norm": 2.1527010134914786,
+      "learning_rate": 1e-05,
+      "loss": 1.3339,
+      "step": 273
+    },
+    {
+      "epoch": 0.12345122775399865,
+      "grad_norm": 2.1764628846093816,
+      "learning_rate": 1e-05,
+      "loss": 1.2766,
+      "step": 274
+    },
+    {
+      "epoch": 0.12390177968010813,
+      "grad_norm": 1.947845094899021,
+      "learning_rate": 1e-05,
+      "loss": 1.264,
+      "step": 275
+    },
+    {
+      "epoch": 0.12435233160621761,
+      "grad_norm": 2.368267399084212,
+      "learning_rate": 1e-05,
+      "loss": 1.3515,
+      "step": 276
+    },
+    {
+      "epoch": 0.1248028835323271,
+      "grad_norm": 2.0997819018442163,
+      "learning_rate": 1e-05,
+      "loss": 1.3069,
+      "step": 277
+    },
+    {
+      "epoch": 0.1252534354584366,
+      "grad_norm": 1.9939134555995226,
+      "learning_rate": 1e-05,
+      "loss": 1.3191,
+      "step": 278
+    },
+    {
+      "epoch": 0.12570398738454608,
+      "grad_norm": 1.9066572393040075,
+      "learning_rate": 1e-05,
+      "loss": 1.3135,
+      "step": 279
+    },
+    {
+      "epoch": 0.12615453931065557,
+      "grad_norm": 1.9057042423036947,
+      "learning_rate": 1e-05,
+      "loss": 1.3153,
+      "step": 280
+    },
+    {
+      "epoch": 0.12660509123676503,
+      "grad_norm": 2.150247021172038,
+      "learning_rate": 1e-05,
+      "loss": 1.3303,
+      "step": 281
+    },
+    {
+      "epoch": 0.12705564316287452,
+      "grad_norm": 2.0992788299613783,
+      "learning_rate": 1e-05,
+      "loss": 1.3254,
+      "step": 282
+    },
+    {
+      "epoch": 0.127506195088984,
+      "grad_norm": 2.133783599904729,
+      "learning_rate": 1e-05,
+      "loss": 1.2773,
+      "step": 283
+    },
+    {
+      "epoch": 0.1279567470150935,
+      "grad_norm": 1.9316963608148732,
+      "learning_rate": 1e-05,
+      "loss": 1.3435,
+      "step": 284
+    },
+    {
+      "epoch": 0.12840729894120298,
+      "grad_norm": 1.974806893876162,
+      "learning_rate": 1e-05,
+      "loss": 1.2523,
+      "step": 285
+    },
+    {
+      "epoch": 0.12885785086731247,
+      "grad_norm": 1.9743808573222912,
+      "learning_rate": 1e-05,
+      "loss": 1.3285,
+      "step": 286
+    },
+    {
+      "epoch": 0.12930840279342193,
+      "grad_norm": 2.103180773196452,
+      "learning_rate": 1e-05,
+      "loss": 1.2892,
+      "step": 287
+    },
+    {
+      "epoch": 0.12975895471953142,
+      "grad_norm": 1.8556401356475218,
+      "learning_rate": 1e-05,
+      "loss": 1.291,
+      "step": 288
+    },
+    {
+      "epoch": 0.1302095066456409,
+      "grad_norm": 1.9207818760870192,
+      "learning_rate": 1e-05,
+      "loss": 1.2789,
+      "step": 289
+    },
+    {
+      "epoch": 0.1306600585717504,
+      "grad_norm": 2.0033928497304627,
+      "learning_rate": 1e-05,
+      "loss": 1.2813,
+      "step": 290
+    },
+    {
+      "epoch": 0.13111061049785988,
+      "grad_norm": 1.8633066695261629,
+      "learning_rate": 1e-05,
+      "loss": 1.2738,
+      "step": 291
+    },
+    {
+      "epoch": 0.13156116242396937,
+      "grad_norm": 1.8657542972869336,
+      "learning_rate": 1e-05,
+      "loss": 1.3087,
+      "step": 292
+    },
+    {
+      "epoch": 0.13201171435007886,
+      "grad_norm": 2.1494510639786335,
+      "learning_rate": 1e-05,
+      "loss": 1.2836,
+      "step": 293
+    },
+    {
+      "epoch": 0.13246226627618832,
+      "grad_norm": 2.0505108843320503,
+      "learning_rate": 1e-05,
+      "loss": 1.3038,
+      "step": 294
+    },
+    {
+      "epoch": 0.1329128182022978,
+      "grad_norm": 2.112493335275763,
+      "learning_rate": 1e-05,
+      "loss": 1.2872,
+      "step": 295
+    },
+    {
+      "epoch": 0.1333633701284073,
+      "grad_norm": 1.924080602241478,
+      "learning_rate": 1e-05,
+      "loss": 1.2903,
+      "step": 296
+    },
+    {
+      "epoch": 0.13381392205451678,
+      "grad_norm": 2.0436320967486115,
+      "learning_rate": 1e-05,
+      "loss": 1.3554,
+      "step": 297
+    },
+    {
+      "epoch": 0.13426447398062627,
+      "grad_norm": 1.9545825121390592,
+      "learning_rate": 1e-05,
+      "loss": 1.2739,
+      "step": 298
+    },
+    {
+      "epoch": 0.13471502590673576,
+      "grad_norm": 1.9665488651762209,
+      "learning_rate": 1e-05,
+      "loss": 1.304,
+      "step": 299
+    },
+    {
+      "epoch": 0.13516557783284525,
+      "grad_norm": 1.9449462490428069,
+      "learning_rate": 1e-05,
+      "loss": 1.2848,
+      "step": 300
+    },
+    {
+      "epoch": 0.1356161297589547,
+      "grad_norm": 1.9470051383361042,
+      "learning_rate": 1e-05,
+      "loss": 1.2981,
+      "step": 301
+    },
+    {
+      "epoch": 0.1360666816850642,
+      "grad_norm": 1.9373980424275328,
+      "learning_rate": 1e-05,
+      "loss": 1.2315,
+      "step": 302
+    },
+    {
+      "epoch": 0.13651723361117368,
+      "grad_norm": 1.875270255478935,
+      "learning_rate": 1e-05,
+      "loss": 1.3268,
+      "step": 303
+    },
+    {
+      "epoch": 0.13696778553728317,
+      "grad_norm": 2.2044670214107707,
+      "learning_rate": 1e-05,
+      "loss": 1.2953,
+      "step": 304
+    },
+    {
+      "epoch": 0.13741833746339266,
+      "grad_norm": 2.075304026490976,
+      "learning_rate": 1e-05,
+      "loss": 1.288,
+      "step": 305
+    },
+    {
+      "epoch": 0.13786888938950215,
+      "grad_norm": 2.011073798276399,
+      "learning_rate": 1e-05,
+      "loss": 1.2467,
+      "step": 306
+    },
+    {
+      "epoch": 0.13831944131561164,
+      "grad_norm": 2.034679330715142,
+      "learning_rate": 1e-05,
+      "loss": 1.3292,
+      "step": 307
+    },
+    {
+      "epoch": 0.1387699932417211,
+      "grad_norm": 1.9483449655122804,
+      "learning_rate": 1e-05,
+      "loss": 1.319,
+      "step": 308
+    },
+    {
+      "epoch": 0.13922054516783058,
+      "grad_norm": 2.04678116650454,
+      "learning_rate": 1e-05,
+      "loss": 1.2547,
+      "step": 309
+    },
+    {
+      "epoch": 0.13967109709394007,
+      "grad_norm": 2.1949720256189336,
+      "learning_rate": 1e-05,
+      "loss": 1.2852,
+      "step": 310
+    },
+    {
+      "epoch": 0.14012164902004956,
+      "grad_norm": 2.0350442826126174,
+      "learning_rate": 1e-05,
+      "loss": 1.3086,
+      "step": 311
+    },
+    {
+      "epoch": 0.14057220094615905,
+      "grad_norm": 2.0258670899912095,
+      "learning_rate": 1e-05,
+      "loss": 1.271,
+      "step": 312
+    },
+    {
+      "epoch": 0.14102275287226854,
+      "grad_norm": 2.171456270634506,
+      "learning_rate": 1e-05,
+      "loss": 1.2978,
+      "step": 313
+    },
+    {
+      "epoch": 0.14147330479837802,
+      "grad_norm": 2.149145869935288,
+      "learning_rate": 1e-05,
+      "loss": 1.3287,
+      "step": 314
+    },
+    {
+      "epoch": 0.14192385672448748,
+      "grad_norm": 2.175629086835601,
+      "learning_rate": 1e-05,
+      "loss": 1.2797,
+      "step": 315
+    },
+    {
+      "epoch": 0.14237440865059697,
+      "grad_norm": 2.14936647254275,
+      "learning_rate": 1e-05,
+      "loss": 1.2917,
+      "step": 316
+    },
+    {
+      "epoch": 0.14282496057670646,
+      "grad_norm": 1.9906810415024248,
+      "learning_rate": 1e-05,
+      "loss": 1.2675,
+      "step": 317
+    },
+    {
+      "epoch": 0.14327551250281595,
+      "grad_norm": 1.988086950463775,
+      "learning_rate": 1e-05,
+      "loss": 1.2886,
+      "step": 318
+    },
+    {
+      "epoch": 0.14372606442892544,
+      "grad_norm": 1.9476086425243724,
+      "learning_rate": 1e-05,
+      "loss": 1.2644,
+      "step": 319
+    },
+    {
+      "epoch": 0.14417661635503493,
+      "grad_norm": 2.0995866280534394,
+      "learning_rate": 1e-05,
+      "loss": 1.3236,
+      "step": 320
+    },
+    {
+      "epoch": 0.1446271682811444,
+      "grad_norm": 1.996319309918401,
+      "learning_rate": 1e-05,
+      "loss": 1.2763,
+      "step": 321
+    },
+    {
+      "epoch": 0.14507772020725387,
+      "grad_norm": 2.2000731970049165,
+      "learning_rate": 1e-05,
+      "loss": 1.3132,
+      "step": 322
+    },
+    {
+      "epoch": 0.14552827213336336,
+      "grad_norm": 2.004990334407573,
+      "learning_rate": 1e-05,
+      "loss": 1.3002,
+      "step": 323
+    },
+    {
+      "epoch": 0.14597882405947285,
+      "grad_norm": 1.9688564258244992,
+      "learning_rate": 1e-05,
+      "loss": 1.2174,
+      "step": 324
+    },
+    {
+      "epoch": 0.14642937598558234,
+      "grad_norm": 2.181618588764316,
+      "learning_rate": 1e-05,
+      "loss": 1.3192,
+      "step": 325
+    },
+    {
+      "epoch": 0.14687992791169183,
+      "grad_norm": 2.069099208462894,
+      "learning_rate": 1e-05,
+      "loss": 1.3353,
+      "step": 326
+    },
+    {
+      "epoch": 0.14733047983780131,
+      "grad_norm": 2.0591506357520393,
+      "learning_rate": 1e-05,
+      "loss": 1.263,
+      "step": 327
+    },
+    {
+      "epoch": 0.1477810317639108,
+      "grad_norm": 2.041408533263358,
+      "learning_rate": 1e-05,
+      "loss": 1.3209,
+      "step": 328
+    },
+    {
+      "epoch": 0.14823158369002026,
+      "grad_norm": 1.9928222049055881,
+      "learning_rate": 1e-05,
+      "loss": 1.2589,
+      "step": 329
+    },
+    {
+      "epoch": 0.14868213561612975,
+      "grad_norm": 1.9717547921546856,
+      "learning_rate": 1e-05,
+      "loss": 1.2733,
+      "step": 330
+    },
+    {
+      "epoch": 0.14913268754223924,
+      "grad_norm": 2.0663499407777537,
+      "learning_rate": 1e-05,
+      "loss": 1.2612,
+      "step": 331
+    },
+    {
+      "epoch": 0.14958323946834873,
+      "grad_norm": 2.131614625072467,
+      "learning_rate": 1e-05,
+      "loss": 1.2867,
+      "step": 332
+    },
+    {
+      "epoch": 0.15003379139445822,
+      "grad_norm": 2.001144065765142,
+      "learning_rate": 1e-05,
+      "loss": 1.2885,
+      "step": 333
+    },
+    {
+      "epoch": 0.1504843433205677,
+      "grad_norm": 2.324747529344625,
+      "learning_rate": 1e-05,
+      "loss": 1.2727,
+      "step": 334
+    },
+    {
+      "epoch": 0.1509348952466772,
+      "grad_norm": 1.9493047619407076,
+      "learning_rate": 1e-05,
+      "loss": 1.282,
+      "step": 335
+    },
+    {
+      "epoch": 0.15138544717278665,
+      "grad_norm": 1.9618381937389244,
+      "learning_rate": 1e-05,
+      "loss": 1.2747,
+      "step": 336
+    },
+    {
+      "epoch": 0.15183599909889614,
+      "grad_norm": 2.021673157769538,
+      "learning_rate": 1e-05,
+      "loss": 1.2915,
+      "step": 337
+    },
+    {
+      "epoch": 0.15228655102500563,
+      "grad_norm": 2.0609723042172923,
+      "learning_rate": 1e-05,
+      "loss": 1.2915,
+      "step": 338
+    },
+    {
+      "epoch": 0.15273710295111512,
+      "grad_norm": 2.2411227954856856,
+      "learning_rate": 1e-05,
+      "loss": 1.311,
+      "step": 339
+    },
+    {
+      "epoch": 0.1531876548772246,
+      "grad_norm": 2.0293089438224015,
+      "learning_rate": 1e-05,
+      "loss": 1.2973,
+      "step": 340
+    },
+    {
+      "epoch": 0.1536382068033341,
+      "grad_norm": 1.8179160445288032,
+      "learning_rate": 1e-05,
+      "loss": 1.2857,
+      "step": 341
+    },
+    {
+      "epoch": 0.15408875872944358,
+      "grad_norm": 1.9623466371508445,
+      "learning_rate": 1e-05,
+      "loss": 1.2892,
+      "step": 342
+    },
+    {
+      "epoch": 0.15453931065555304,
+      "grad_norm": 1.9603857645386529,
+      "learning_rate": 1e-05,
+      "loss": 1.3028,
+      "step": 343
+    },
+    {
+      "epoch": 0.15498986258166253,
+      "grad_norm": 1.84250642856154,
+      "learning_rate": 1e-05,
+      "loss": 1.2759,
+      "step": 344
+    },
+    {
+      "epoch": 0.15544041450777202,
+      "grad_norm": 2.2444300057918998,
+      "learning_rate": 1e-05,
+      "loss": 1.2607,
+      "step": 345
+    },
+    {
+      "epoch": 0.1558909664338815,
+      "grad_norm": 1.960685051584487,
+      "learning_rate": 1e-05,
+      "loss": 1.2981,
+      "step": 346
+    },
+    {
+      "epoch": 0.156341518359991,
+      "grad_norm": 1.9720183471848456,
+      "learning_rate": 1e-05,
+      "loss": 1.2658,
+      "step": 347
+    },
+    {
+      "epoch": 0.15679207028610048,
+      "grad_norm": 2.1108681327485526,
+      "learning_rate": 1e-05,
+      "loss": 1.3252,
+      "step": 348
+    },
+    {
+      "epoch": 0.15724262221220997,
+      "grad_norm": 1.9526787609837528,
+      "learning_rate": 1e-05,
+      "loss": 1.3097,
+      "step": 349
+    },
+    {
+      "epoch": 0.15769317413831943,
+      "grad_norm": 2.102470525566385,
+      "learning_rate": 1e-05,
+      "loss": 1.2793,
+      "step": 350
+    },
+    {
+      "epoch": 0.15814372606442892,
+      "grad_norm": 1.9043399917790473,
+      "learning_rate": 1e-05,
+      "loss": 1.2901,
+      "step": 351
+    },
+    {
+      "epoch": 0.1585942779905384,
+      "grad_norm": 2.0704873691866217,
+      "learning_rate": 1e-05,
+      "loss": 1.2534,
+      "step": 352
+    },
+    {
+      "epoch": 0.1590448299166479,
+      "grad_norm": 1.8935409303536357,
+      "learning_rate": 1e-05,
+      "loss": 1.2705,
+      "step": 353
+    },
+    {
+      "epoch": 0.15949538184275738,
+      "grad_norm": 2.05329124751458,
+      "learning_rate": 1e-05,
+      "loss": 1.3187,
+      "step": 354
+    },
+    {
+      "epoch": 0.15994593376886687,
+      "grad_norm": 2.0597273322745937,
+      "learning_rate": 1e-05,
+      "loss": 1.2566,
+      "step": 355
+    },
+    {
+      "epoch": 0.16039648569497636,
+      "grad_norm": 1.994080738314847,
+      "learning_rate": 1e-05,
+      "loss": 1.2945,
+      "step": 356
+    },
+    {
+      "epoch": 0.16084703762108582,
+      "grad_norm": 1.9178614859606067,
+      "learning_rate": 1e-05,
+      "loss": 1.2729,
+      "step": 357
+    },
+    {
+      "epoch": 0.1612975895471953,
+      "grad_norm": 2.1202612219294097,
+      "learning_rate": 1e-05,
+      "loss": 1.2783,
+      "step": 358
+    },
+    {
+      "epoch": 0.1617481414733048,
+      "grad_norm": 2.0639508875751145,
+      "learning_rate": 1e-05,
+      "loss": 1.3166,
+      "step": 359
+    },
+    {
+      "epoch": 0.16219869339941428,
+      "grad_norm": 1.9219411874218848,
+      "learning_rate": 1e-05,
+      "loss": 1.2851,
+      "step": 360
+    },
+    {
+      "epoch": 0.16264924532552377,
+      "grad_norm": 1.8555787792800178,
+      "learning_rate": 1e-05,
+      "loss": 1.3075,
+      "step": 361
+    },
+    {
+      "epoch": 0.16309979725163326,
+      "grad_norm": 2.0051025638227364,
+      "learning_rate": 1e-05,
+      "loss": 1.2787,
+      "step": 362
+    },
+    {
+      "epoch": 0.16355034917774275,
+      "grad_norm": 2.071493880001985,
+      "learning_rate": 1e-05,
+      "loss": 1.3449,
+      "step": 363
+    },
+    {
+      "epoch": 0.1640009011038522,
+      "grad_norm": 2.0435140721243785,
+      "learning_rate": 1e-05,
+      "loss": 1.3004,
+      "step": 364
+    },
+    {
+      "epoch": 0.1644514530299617,
+      "grad_norm": 2.020256783371866,
+      "learning_rate": 1e-05,
+      "loss": 1.2824,
+      "step": 365
+    },
+    {
+      "epoch": 0.16490200495607119,
+      "grad_norm": 2.095838466374123,
+      "learning_rate": 1e-05,
+      "loss": 1.2231,
+      "step": 366
+    },
+    {
+      "epoch": 0.16535255688218067,
+      "grad_norm": 2.079247303020079,
+      "learning_rate": 1e-05,
+      "loss": 1.2719,
+      "step": 367
+    },
+    {
+      "epoch": 0.16580310880829016,
+      "grad_norm": 2.0004956596544683,
+      "learning_rate": 1e-05,
+      "loss": 1.3348,
+      "step": 368
+    },
+    {
+      "epoch": 0.16625366073439965,
+      "grad_norm": 2.0327189616174994,
+      "learning_rate": 1e-05,
+      "loss": 1.281,
+      "step": 369
+    },
+    {
+      "epoch": 0.1667042126605091,
+      "grad_norm": 2.0939532929360927,
+      "learning_rate": 1e-05,
+      "loss": 1.2308,
+      "step": 370
+    },
+    {
+      "epoch": 0.1671547645866186,
+      "grad_norm": 1.942640042738766,
+      "learning_rate": 1e-05,
+      "loss": 1.2309,
+      "step": 371
+    },
+    {
+      "epoch": 0.1676053165127281,
+      "grad_norm": 2.251110736450933,
+      "learning_rate": 1e-05,
+      "loss": 1.2885,
+      "step": 372
+    },
+    {
+      "epoch": 0.16805586843883757,
+      "grad_norm": 2.2861326119505048,
+      "learning_rate": 1e-05,
+      "loss": 1.272,
+      "step": 373
+    },
+    {
+      "epoch": 0.16850642036494706,
+      "grad_norm": 1.954729928516754,
+      "learning_rate": 1e-05,
+      "loss": 1.3243,
+      "step": 374
+    },
+    {
+      "epoch": 0.16895697229105655,
+      "grad_norm": 1.9117136264725754,
+      "learning_rate": 1e-05,
+      "loss": 1.3125,
+      "step": 375
+    },
+    {
+      "epoch": 0.16940752421716604,
+      "grad_norm": 2.1945864188944153,
+      "learning_rate": 1e-05,
+      "loss": 1.3087,
+      "step": 376
+    },
+    {
+      "epoch": 0.1698580761432755,
+      "grad_norm": 1.9058131074988187,
+      "learning_rate": 1e-05,
+      "loss": 1.2924,
+      "step": 377
+    },
+    {
+      "epoch": 0.170308628069385,
+      "grad_norm": 1.8407007694236421,
+      "learning_rate": 1e-05,
+      "loss": 1.2937,
+      "step": 378
+    },
+    {
+      "epoch": 0.17075917999549448,
+      "grad_norm": 2.098941464662954,
+      "learning_rate": 1e-05,
+      "loss": 1.3028,
+      "step": 379
+    },
+    {
+      "epoch": 0.17120973192160396,
+      "grad_norm": 1.9882687348840964,
+      "learning_rate": 1e-05,
+      "loss": 1.3079,
+      "step": 380
+    },
+    {
+      "epoch": 0.17166028384771345,
+      "grad_norm": 2.1011227957223704,
+      "learning_rate": 1e-05,
+      "loss": 1.2818,
+      "step": 381
+    },
+    {
+      "epoch": 0.17211083577382294,
+      "grad_norm": 2.0966958079367126,
+      "learning_rate": 1e-05,
+      "loss": 1.2357,
+      "step": 382
+    },
+    {
+      "epoch": 0.17256138769993243,
+      "grad_norm": 2.043553305436292,
+      "learning_rate": 1e-05,
+      "loss": 1.2624,
+      "step": 383
+    },
+    {
+      "epoch": 0.1730119396260419,
+      "grad_norm": 1.9157164138264542,
+      "learning_rate": 1e-05,
+      "loss": 1.2901,
+      "step": 384
+    },
+    {
+      "epoch": 0.17346249155215138,
+      "grad_norm": 2.0489774956843014,
+      "learning_rate": 1e-05,
+      "loss": 1.2954,
+      "step": 385
+    },
+    {
+      "epoch": 0.17391304347826086,
+      "grad_norm": 2.0355571212846666,
+      "learning_rate": 1e-05,
+      "loss": 1.2809,
+      "step": 386
+    },
+    {
+      "epoch": 0.17436359540437035,
+      "grad_norm": 1.7634131824701134,
+      "learning_rate": 1e-05,
+      "loss": 1.3056,
+      "step": 387
+    },
+    {
+      "epoch": 0.17481414733047984,
+      "grad_norm": 2.162318565634982,
+      "learning_rate": 1e-05,
+      "loss": 1.2765,
+      "step": 388
+    },
+    {
+      "epoch": 0.17526469925658933,
+      "grad_norm": 2.112678284922863,
+      "learning_rate": 1e-05,
+      "loss": 1.2962,
+      "step": 389
+    },
+    {
+      "epoch": 0.17571525118269882,
+      "grad_norm": 2.1494215295790795,
+      "learning_rate": 1e-05,
+      "loss": 1.2652,
+      "step": 390
+    },
+    {
+      "epoch": 0.17616580310880828,
+      "grad_norm": 2.064498142531983,
+      "learning_rate": 1e-05,
+      "loss": 1.2803,
+      "step": 391
+    },
+    {
+      "epoch": 0.17661635503491777,
+      "grad_norm": 1.8046181497392817,
+      "learning_rate": 1e-05,
+      "loss": 1.3499,
+      "step": 392
+    },
+    {
+      "epoch": 0.17706690696102725,
+      "grad_norm": 1.859711338145051,
+      "learning_rate": 1e-05,
+      "loss": 1.3156,
+      "step": 393
+    },
+    {
+      "epoch": 0.17751745888713674,
+      "grad_norm": 1.9854435645998658,
+      "learning_rate": 1e-05,
+      "loss": 1.2698,
+      "step": 394
+    },
+    {
+      "epoch": 0.17796801081324623,
+      "grad_norm": 1.917316632528049,
+      "learning_rate": 1e-05,
+      "loss": 1.3331,
+      "step": 395
+    },
+    {
+      "epoch": 0.17841856273935572,
+      "grad_norm": 2.189833730293825,
+      "learning_rate": 1e-05,
+      "loss": 1.2522,
+      "step": 396
+    },
+    {
+      "epoch": 0.1788691146654652,
+      "grad_norm": 2.016849754761407,
+      "learning_rate": 1e-05,
+      "loss": 1.2866,
+      "step": 397
+    },
+    {
+      "epoch": 0.17931966659157467,
+      "grad_norm": 2.0055991366271435,
+      "learning_rate": 1e-05,
+      "loss": 1.2886,
+      "step": 398
+    },
+    {
+      "epoch": 0.17977021851768415,
+      "grad_norm": 1.9429713181519759,
+      "learning_rate": 1e-05,
+      "loss": 1.2578,
+      "step": 399
+    },
+    {
+      "epoch": 0.18022077044379364,
+      "grad_norm": 1.9769582665952863,
+      "learning_rate": 1e-05,
+      "loss": 1.3172,
+      "step": 400
+    },
+    {
+      "epoch": 0.18067132236990313,
+      "grad_norm": 2.033065302057002,
+      "learning_rate": 1e-05,
+      "loss": 1.2717,
+      "step": 401
+    },
+    {
+      "epoch": 0.18112187429601262,
+      "grad_norm": 2.044143490422974,
+      "learning_rate": 1e-05,
+      "loss": 1.2887,
+      "step": 402
+    },
+    {
+      "epoch": 0.1815724262221221,
+      "grad_norm": 2.2312132814181664,
+      "learning_rate": 1e-05,
+      "loss": 1.2564,
+      "step": 403
+    },
+    {
+      "epoch": 0.1820229781482316,
+      "grad_norm": 1.9692640587423784,
+      "learning_rate": 1e-05,
+      "loss": 1.3192,
+      "step": 404
+    },
+    {
+      "epoch": 0.18247353007434106,
+      "grad_norm": 2.0299575879307072,
+      "learning_rate": 1e-05,
+      "loss": 1.2624,
+      "step": 405
+    },
+    {
+      "epoch": 0.18292408200045054,
+      "grad_norm": 1.8904771035708423,
+      "learning_rate": 1e-05,
+      "loss": 1.3114,
+      "step": 406
+    },
+    {
+      "epoch": 0.18337463392656003,
+      "grad_norm": 1.841332211929519,
+      "learning_rate": 1e-05,
+      "loss": 1.2567,
+      "step": 407
+    },
+    {
+      "epoch": 0.18382518585266952,
+      "grad_norm": 2.1136872115933265,
+      "learning_rate": 1e-05,
+      "loss": 1.3072,
+      "step": 408
+    },
+    {
+      "epoch": 0.184275737778779,
+      "grad_norm": 2.105056901824731,
+      "learning_rate": 1e-05,
+      "loss": 1.2333,
+      "step": 409
+    },
+    {
+      "epoch": 0.1847262897048885,
+      "grad_norm": 2.1391861806470733,
+      "learning_rate": 1e-05,
+      "loss": 1.2697,
+      "step": 410
+    },
+    {
+      "epoch": 0.18517684163099798,
+      "grad_norm": 1.9386019547928421,
+      "learning_rate": 1e-05,
+      "loss": 1.2688,
+      "step": 411
+    },
+    {
+      "epoch": 0.18562739355710745,
+      "grad_norm": 2.2075012555281837,
+      "learning_rate": 1e-05,
+      "loss": 1.2554,
+      "step": 412
+    },
+    {
+      "epoch": 0.18607794548321693,
+      "grad_norm": 1.9657658945733893,
+      "learning_rate": 1e-05,
+      "loss": 1.2818,
+      "step": 413
+    },
+    {
+      "epoch": 0.18652849740932642,
+      "grad_norm": 2.062999023916708,
+      "learning_rate": 1e-05,
+      "loss": 1.3001,
+      "step": 414
+    },
+    {
+      "epoch": 0.1869790493354359,
+      "grad_norm": 1.987774567017447,
+      "learning_rate": 1e-05,
+      "loss": 1.2666,
+      "step": 415
+    },
+    {
+      "epoch": 0.1874296012615454,
+      "grad_norm": 1.965562259197448,
+      "learning_rate": 1e-05,
+      "loss": 1.2754,
+      "step": 416
+    },
+    {
+      "epoch": 0.18788015318765489,
+      "grad_norm": 1.9874935900845858,
+      "learning_rate": 1e-05,
+      "loss": 1.298,
+      "step": 417
+    },
+    {
+      "epoch": 0.18833070511376437,
+      "grad_norm": 1.9579248305312746,
+      "learning_rate": 1e-05,
+      "loss": 1.2804,
+      "step": 418
+    },
+    {
+      "epoch": 0.18878125703987383,
+      "grad_norm": 1.9969577424329485,
+      "learning_rate": 1e-05,
+      "loss": 1.2285,
+      "step": 419
+    },
+    {
+      "epoch": 0.18923180896598332,
+      "grad_norm": 2.2398185632472356,
+      "learning_rate": 1e-05,
+      "loss": 1.3058,
+      "step": 420
+    },
+    {
+      "epoch": 0.1896823608920928,
+      "grad_norm": 1.9797098368258668,
+      "learning_rate": 1e-05,
+      "loss": 1.3245,
+      "step": 421
+    },
+    {
+      "epoch": 0.1901329128182023,
+      "grad_norm": 2.0799083824667104,
+      "learning_rate": 1e-05,
+      "loss": 1.2954,
+      "step": 422
+    },
+    {
+      "epoch": 0.1905834647443118,
+      "grad_norm": 1.9912326179789581,
+      "learning_rate": 1e-05,
+      "loss": 1.2557,
+      "step": 423
+    },
+    {
+      "epoch": 0.19103401667042127,
+      "grad_norm": 1.9822348840500932,
+      "learning_rate": 1e-05,
+      "loss": 1.2825,
+      "step": 424
+    },
+    {
+      "epoch": 0.19148456859653076,
+      "grad_norm": 1.955087556533265,
+      "learning_rate": 1e-05,
+      "loss": 1.2856,
+      "step": 425
+    },
+    {
+      "epoch": 0.19193512052264022,
+      "grad_norm": 1.8832447554381153,
+      "learning_rate": 1e-05,
+      "loss": 1.27,
+      "step": 426
+    },
+    {
+      "epoch": 0.1923856724487497,
+      "grad_norm": 2.075850607998512,
+      "learning_rate": 1e-05,
+      "loss": 1.2924,
+      "step": 427
+    },
+    {
+      "epoch": 0.1928362243748592,
+      "grad_norm": 1.9545773350054239,
+      "learning_rate": 1e-05,
+      "loss": 1.3046,
+      "step": 428
+    },
+    {
+      "epoch": 0.1932867763009687,
+      "grad_norm": 1.8106472670812626,
+      "learning_rate": 1e-05,
+      "loss": 1.241,
+      "step": 429
+    },
+    {
+      "epoch": 0.19373732822707818,
+      "grad_norm": 2.202580923848979,
+      "learning_rate": 1e-05,
+      "loss": 1.3186,
+      "step": 430
+    },
+    {
+      "epoch": 0.19418788015318766,
+      "grad_norm": 1.9802743607989126,
+      "learning_rate": 1e-05,
+      "loss": 1.2508,
+      "step": 431
+    },
+    {
+      "epoch": 0.19463843207929715,
+      "grad_norm": 2.2448484146253644,
+      "learning_rate": 1e-05,
+      "loss": 1.287,
+      "step": 432
+    },
+    {
+      "epoch": 0.1950889840054066,
+      "grad_norm": 2.0777067227076507,
+      "learning_rate": 1e-05,
+      "loss": 1.2443,
+      "step": 433
+    },
+    {
+      "epoch": 0.1955395359315161,
+      "grad_norm": 1.9836012166807184,
+      "learning_rate": 1e-05,
+      "loss": 1.2806,
+      "step": 434
+    },
+    {
+      "epoch": 0.1959900878576256,
+      "grad_norm": 1.9285909883485441,
+      "learning_rate": 1e-05,
+      "loss": 1.3033,
+      "step": 435
+    },
+    {
+      "epoch": 0.19644063978373508,
+      "grad_norm": 1.9878147349260273,
+      "learning_rate": 1e-05,
+      "loss": 1.2759,
+      "step": 436
+    },
+    {
+      "epoch": 0.19689119170984457,
+      "grad_norm": 1.9557036605383904,
+      "learning_rate": 1e-05,
+      "loss": 1.2758,
+      "step": 437
+    },
+    {
+      "epoch": 0.19734174363595405,
+      "grad_norm": 1.884658796228399,
+      "learning_rate": 1e-05,
+      "loss": 1.2771,
+      "step": 438
+    },
+    {
+      "epoch": 0.19779229556206354,
+      "grad_norm": 1.816275867047091,
+      "learning_rate": 1e-05,
+      "loss": 1.2955,
+      "step": 439
+    },
+    {
+      "epoch": 0.198242847488173,
+      "grad_norm": 2.10656333138919,
+      "learning_rate": 1e-05,
+      "loss": 1.2617,
+      "step": 440
+    },
+    {
+      "epoch": 0.1986933994142825,
+      "grad_norm": 2.0984562962870803,
+      "learning_rate": 1e-05,
+      "loss": 1.2222,
+      "step": 441
+    },
+    {
+      "epoch": 0.19914395134039198,
+      "grad_norm": 1.9348752922906438,
+      "learning_rate": 1e-05,
+      "loss": 1.2168,
+      "step": 442
+    },
+    {
+      "epoch": 0.19959450326650147,
+      "grad_norm": 1.8927559482208312,
+      "learning_rate": 1e-05,
+      "loss": 1.2447,
+      "step": 443
+    },
+    {
+      "epoch": 0.20004505519261095,
+      "grad_norm": 1.8915183032229956,
+      "learning_rate": 1e-05,
+      "loss": 1.2556,
+      "step": 444
+    },
+    {
+      "epoch": 0.20049560711872044,
+      "grad_norm": 1.8590151705324758,
+      "learning_rate": 1e-05,
+      "loss": 1.3036,
+      "step": 445
+    },
+    {
+      "epoch": 0.2009461590448299,
+      "grad_norm": 2.0188235465064834,
+      "learning_rate": 1e-05,
+      "loss": 1.2813,
+      "step": 446
+    },
+    {
+      "epoch": 0.2013967109709394,
+      "grad_norm": 1.882861465804375,
+      "learning_rate": 1e-05,
+      "loss": 1.247,
+      "step": 447
+    },
+    {
+      "epoch": 0.20184726289704888,
+      "grad_norm": 1.964160284373517,
+      "learning_rate": 1e-05,
+      "loss": 1.2808,
+      "step": 448
+    },
+    {
+      "epoch": 0.20229781482315837,
+      "grad_norm": 2.1704836512761783,
+      "learning_rate": 1e-05,
+      "loss": 1.2391,
+      "step": 449
+    },
+    {
+      "epoch": 0.20274836674926786,
+      "grad_norm": 1.7979559458234273,
+      "learning_rate": 1e-05,
+      "loss": 1.2787,
+      "step": 450
+    },
+    {
+      "epoch": 0.20319891867537734,
+      "grad_norm": 1.9531550503549668,
+      "learning_rate": 1e-05,
+      "loss": 1.3082,
+      "step": 451
+    },
+    {
+      "epoch": 0.20364947060148683,
+      "grad_norm": 1.6802653824191487,
+      "learning_rate": 1e-05,
+      "loss": 1.2592,
+      "step": 452
+    },
+    {
+      "epoch": 0.2041000225275963,
+      "grad_norm": 1.9009883595109776,
+      "learning_rate": 1e-05,
+      "loss": 1.2307,
+      "step": 453
+    },
+    {
+      "epoch": 0.20455057445370578,
+      "grad_norm": 1.924227930900898,
+      "learning_rate": 1e-05,
+      "loss": 1.2701,
+      "step": 454
+    },
+    {
+      "epoch": 0.20500112637981527,
+      "grad_norm": 2.072514033394317,
+      "learning_rate": 1e-05,
+      "loss": 1.2749,
+      "step": 455
+    },
+    {
+      "epoch": 0.20545167830592476,
+      "grad_norm": 2.1258082293123697,
+      "learning_rate": 1e-05,
+      "loss": 1.3065,
+      "step": 456
+    },
+    {
+      "epoch": 0.20590223023203424,
+      "grad_norm": 2.0719552458789545,
+      "learning_rate": 1e-05,
+      "loss": 1.2453,
+      "step": 457
+    },
+    {
+      "epoch": 0.20635278215814373,
+      "grad_norm": 1.8200768626343444,
+      "learning_rate": 1e-05,
+      "loss": 1.2943,
+      "step": 458
+    },
+    {
+      "epoch": 0.20680333408425322,
+      "grad_norm": 2.1138401076737146,
+      "learning_rate": 1e-05,
+      "loss": 1.2944,
+      "step": 459
+    },
+    {
+      "epoch": 0.20725388601036268,
+      "grad_norm": 2.2416430586143186,
+      "learning_rate": 1e-05,
+      "loss": 1.2885,
+      "step": 460
+    },
+    {
+      "epoch": 0.20770443793647217,
+      "grad_norm": 2.0189173757078067,
+      "learning_rate": 1e-05,
+      "loss": 1.2722,
+      "step": 461
+    },
+    {
+      "epoch": 0.20815498986258166,
+      "grad_norm": 1.9056409736980289,
+      "learning_rate": 1e-05,
+      "loss": 1.2481,
+      "step": 462
+    },
+    {
+      "epoch": 0.20860554178869115,
+      "grad_norm": 1.873869775109084,
+      "learning_rate": 1e-05,
+      "loss": 1.2947,
+      "step": 463
+    },
+    {
+      "epoch": 0.20905609371480063,
+      "grad_norm": 2.030085128944356,
+      "learning_rate": 1e-05,
+      "loss": 1.3246,
+      "step": 464
+    },
+    {
+      "epoch": 0.20950664564091012,
+      "grad_norm": 1.996424509377663,
+      "learning_rate": 1e-05,
+      "loss": 1.2581,
+      "step": 465
+    },
+    {
+      "epoch": 0.2099571975670196,
+      "grad_norm": 1.9704691056393633,
+      "learning_rate": 1e-05,
+      "loss": 1.2676,
+      "step": 466
+    },
+    {
+      "epoch": 0.21040774949312907,
+      "grad_norm": 2.1302562487677505,
+      "learning_rate": 1e-05,
+      "loss": 1.2801,
+      "step": 467
+    },
+    {
+      "epoch": 0.21085830141923856,
+      "grad_norm": 1.8898483209227444,
+      "learning_rate": 1e-05,
+      "loss": 1.1885,
+      "step": 468
+    },
+    {
+      "epoch": 0.21130885334534805,
+      "grad_norm": 1.9769508731754923,
+      "learning_rate": 1e-05,
+      "loss": 1.2947,
+      "step": 469
+    },
+    {
+      "epoch": 0.21175940527145753,
+      "grad_norm": 2.051397240160048,
+      "learning_rate": 1e-05,
+      "loss": 1.285,
+      "step": 470
+    },
+    {
+      "epoch": 0.21220995719756702,
+      "grad_norm": 1.827454330187214,
+      "learning_rate": 1e-05,
+      "loss": 1.2488,
+      "step": 471
+    },
+    {
+      "epoch": 0.2126605091236765,
+      "grad_norm": 1.8965671551423218,
+      "learning_rate": 1e-05,
+      "loss": 1.2383,
+      "step": 472
+    },
+    {
+      "epoch": 0.213111061049786,
+      "grad_norm": 1.9374248152180933,
+      "learning_rate": 1e-05,
+      "loss": 1.2655,
+      "step": 473
+    },
+    {
+      "epoch": 0.21356161297589546,
+      "grad_norm": 1.9152642282375247,
+      "learning_rate": 1e-05,
+      "loss": 1.269,
+      "step": 474
+    },
+    {
+      "epoch": 0.21401216490200495,
+      "grad_norm": 2.1286034986619597,
+      "learning_rate": 1e-05,
+      "loss": 1.2549,
+      "step": 475
+    },
+    {
+      "epoch": 0.21446271682811444,
+      "grad_norm": 2.1239685351695416,
+      "learning_rate": 1e-05,
+      "loss": 1.2934,
+      "step": 476
+    },
+    {
+      "epoch": 0.21491326875422392,
+      "grad_norm": 2.387986420548799,
+      "learning_rate": 1e-05,
+      "loss": 1.2209,
+      "step": 477
+    },
+    {
+      "epoch": 0.2153638206803334,
+      "grad_norm": 2.0729872120608066,
+      "learning_rate": 1e-05,
+      "loss": 1.2738,
+      "step": 478
+    },
+    {
+      "epoch": 0.2158143726064429,
+      "grad_norm": 1.9997296687320307,
+      "learning_rate": 1e-05,
+      "loss": 1.2029,
+      "step": 479
+    },
+    {
+      "epoch": 0.2162649245325524,
+      "grad_norm": 2.138097688719461,
+      "learning_rate": 1e-05,
+      "loss": 1.2721,
+      "step": 480
+    },
+    {
+      "epoch": 0.21671547645866185,
+      "grad_norm": 2.0653824446435483,
+      "learning_rate": 1e-05,
+      "loss": 1.2819,
+      "step": 481
+    },
+    {
+      "epoch": 0.21716602838477134,
+      "grad_norm": 2.072177431387959,
+      "learning_rate": 1e-05,
+      "loss": 1.2505,
+      "step": 482
+    },
+    {
+      "epoch": 0.21761658031088082,
+      "grad_norm": 1.9308751828170199,
+      "learning_rate": 1e-05,
+      "loss": 1.2837,
+      "step": 483
+    },
+    {
+      "epoch": 0.2180671322369903,
+      "grad_norm": 1.8944222007186382,
+      "learning_rate": 1e-05,
+      "loss": 1.2319,
+      "step": 484
+    },
+    {
+      "epoch": 0.2185176841630998,
+      "grad_norm": 1.9077748456798524,
+      "learning_rate": 1e-05,
+      "loss": 1.2364,
+      "step": 485
+    },
+    {
+      "epoch": 0.2189682360892093,
+      "grad_norm": 1.9631759804488607,
+      "learning_rate": 1e-05,
+      "loss": 1.2822,
+      "step": 486
+    },
+    {
+      "epoch": 0.21941878801531878,
+      "grad_norm": 1.8012739283429102,
+      "learning_rate": 1e-05,
+      "loss": 1.2737,
+      "step": 487
+    },
+    {
+      "epoch": 0.21986933994142824,
+      "grad_norm": 1.9756749334181785,
+      "learning_rate": 1e-05,
+      "loss": 1.2922,
+      "step": 488
+    },
+    {
+      "epoch": 0.22031989186753773,
+      "grad_norm": 1.9260508139187147,
+      "learning_rate": 1e-05,
+      "loss": 1.2276,
+      "step": 489
+    },
+    {
+      "epoch": 0.22077044379364721,
+      "grad_norm": 1.8435682883122477,
+      "learning_rate": 1e-05,
+      "loss": 1.2318,
+      "step": 490
+    },
+    {
+      "epoch": 0.2212209957197567,
+      "grad_norm": 1.864809666456412,
+      "learning_rate": 1e-05,
+      "loss": 1.3182,
+      "step": 491
+    },
+    {
+      "epoch": 0.2216715476458662,
+      "grad_norm": 1.8619755761060044,
+      "learning_rate": 1e-05,
+      "loss": 1.2504,
+      "step": 492
+    },
+    {
+      "epoch": 0.22212209957197568,
+      "grad_norm": 2.047961696264194,
+      "learning_rate": 1e-05,
+      "loss": 1.2295,
+      "step": 493
+    },
+    {
+      "epoch": 0.22257265149808517,
+      "grad_norm": 1.8984426277668016,
+      "learning_rate": 1e-05,
+      "loss": 1.2497,
+      "step": 494
+    },
+    {
+      "epoch": 0.22302320342419463,
+      "grad_norm": 2.0451801593017143,
+      "learning_rate": 1e-05,
+      "loss": 1.2555,
+      "step": 495
+    },
+    {
+      "epoch": 0.22347375535030412,
+      "grad_norm": 1.8921979333307704,
+      "learning_rate": 1e-05,
+      "loss": 1.2358,
+      "step": 496
+    },
+    {
+      "epoch": 0.2239243072764136,
+      "grad_norm": 1.9640954151850023,
+      "learning_rate": 1e-05,
+      "loss": 1.2156,
+      "step": 497
+    },
+    {
+      "epoch": 0.2243748592025231,
+      "grad_norm": 1.9433705241324468,
+      "learning_rate": 1e-05,
+      "loss": 1.2705,
+      "step": 498
+    },
+    {
+      "epoch": 0.22482541112863258,
+      "grad_norm": 1.932678859581744,
+      "learning_rate": 1e-05,
+      "loss": 1.2651,
+      "step": 499
+    },
+    {
+      "epoch": 0.22527596305474207,
+      "grad_norm": 1.914088641607698,
+      "learning_rate": 1e-05,
+      "loss": 1.2678,
+      "step": 500
+    },
+    {
+      "epoch": 0.22572651498085156,
+      "grad_norm": 1.9092930210655947,
+      "learning_rate": 1e-05,
+      "loss": 1.2299,
+      "step": 501
+    },
+    {
+      "epoch": 0.22617706690696102,
+      "grad_norm": 1.9149221414856479,
+      "learning_rate": 1e-05,
+      "loss": 1.2219,
+      "step": 502
+    },
+    {
+      "epoch": 0.2266276188330705,
+      "grad_norm": 2.0766417297709387,
+      "learning_rate": 1e-05,
+      "loss": 1.2767,
+      "step": 503
+    },
+    {
+      "epoch": 0.22707817075918,
+      "grad_norm": 1.928305726584179,
+      "learning_rate": 1e-05,
+      "loss": 1.2458,
+      "step": 504
+    },
+    {
+      "epoch": 0.22752872268528948,
+      "grad_norm": 2.096065242803092,
+      "learning_rate": 1e-05,
+      "loss": 1.2576,
+      "step": 505
+    },
+    {
+      "epoch": 0.22797927461139897,
+      "grad_norm": 2.0912227374818437,
+      "learning_rate": 1e-05,
+      "loss": 1.2611,
+      "step": 506
+    },
+    {
+      "epoch": 0.22842982653750846,
+      "grad_norm": 1.9815852280061006,
+      "learning_rate": 1e-05,
+      "loss": 1.2675,
+      "step": 507
+    },
+    {
+      "epoch": 0.22888037846361795,
+      "grad_norm": 2.044754812467016,
+      "learning_rate": 1e-05,
+      "loss": 1.2407,
+      "step": 508
+    },
+    {
+      "epoch": 0.2293309303897274,
+      "grad_norm": 1.829492525033062,
+      "learning_rate": 1e-05,
+      "loss": 1.3004,
+      "step": 509
+    },
+    {
+      "epoch": 0.2297814823158369,
+      "grad_norm": 2.1996988452198334,
+      "learning_rate": 1e-05,
+      "loss": 1.2411,
+      "step": 510
+    },
+    {
+      "epoch": 0.23023203424194638,
+      "grad_norm": 2.182608604414491,
+      "learning_rate": 1e-05,
+      "loss": 1.2737,
+      "step": 511
+    },
+    {
+      "epoch": 0.23068258616805587,
+      "grad_norm": 2.110776994389581,
+      "learning_rate": 1e-05,
+      "loss": 1.2685,
+      "step": 512
+    },
+    {
+      "epoch": 0.23113313809416536,
+      "grad_norm": 1.9067786503382023,
+      "learning_rate": 1e-05,
+      "loss": 1.2384,
+      "step": 513
+    },
+    {
+      "epoch": 0.23158369002027485,
+      "grad_norm": 1.929296727816064,
+      "learning_rate": 1e-05,
+      "loss": 1.2546,
+      "step": 514
+    },
+    {
+      "epoch": 0.23203424194638433,
+      "grad_norm": 2.0741880381317537,
+      "learning_rate": 1e-05,
+      "loss": 1.3031,
+      "step": 515
+    },
+    {
+      "epoch": 0.2324847938724938,
+      "grad_norm": 1.9270434416614235,
+      "learning_rate": 1e-05,
+      "loss": 1.2574,
+      "step": 516
+    },
+    {
+      "epoch": 0.23293534579860328,
+      "grad_norm": 1.9078088919166587,
+      "learning_rate": 1e-05,
+      "loss": 1.2369,
+      "step": 517
+    },
+    {
+      "epoch": 0.23338589772471277,
+      "grad_norm": 2.0310839047428293,
+      "learning_rate": 1e-05,
+      "loss": 1.22,
+      "step": 518
+    },
+    {
+      "epoch": 0.23383644965082226,
+      "grad_norm": 1.9361095982022545,
+      "learning_rate": 1e-05,
+      "loss": 1.2526,
+      "step": 519
+    },
+    {
+      "epoch": 0.23428700157693175,
+      "grad_norm": 1.8978662543954028,
+      "learning_rate": 1e-05,
+      "loss": 1.2516,
+      "step": 520
+    },
+    {
+      "epoch": 0.23473755350304124,
+      "grad_norm": 1.9683996368018823,
+      "learning_rate": 1e-05,
+      "loss": 1.2844,
+      "step": 521
+    },
+    {
+      "epoch": 0.23518810542915072,
+      "grad_norm": 1.9425722201345397,
+      "learning_rate": 1e-05,
+      "loss": 1.2705,
+      "step": 522
+    },
+    {
+      "epoch": 0.23563865735526018,
+      "grad_norm": 1.9426821596520805,
+      "learning_rate": 1e-05,
+      "loss": 1.2745,
+      "step": 523
+    },
+    {
+      "epoch": 0.23608920928136967,
+      "grad_norm": 2.081070297565642,
+      "learning_rate": 1e-05,
+      "loss": 1.3034,
+      "step": 524
+    },
+    {
+      "epoch": 0.23653976120747916,
+      "grad_norm": 1.8796976038406434,
+      "learning_rate": 1e-05,
+      "loss": 1.2481,
+      "step": 525
+    },
+    {
+      "epoch": 0.23699031313358865,
+      "grad_norm": 1.865460478581203,
+      "learning_rate": 1e-05,
+      "loss": 1.3021,
+      "step": 526
+    },
+    {
+      "epoch": 0.23744086505969814,
+      "grad_norm": 1.818970277211085,
+      "learning_rate": 1e-05,
+      "loss": 1.243,
+      "step": 527
+    },
+    {
+      "epoch": 0.23789141698580762,
+      "grad_norm": 1.9903490262035735,
+      "learning_rate": 1e-05,
+      "loss": 1.2598,
+      "step": 528
+    },
+    {
+      "epoch": 0.23834196891191708,
+      "grad_norm": 1.8917885928616347,
+      "learning_rate": 1e-05,
+      "loss": 1.3025,
+      "step": 529
+    },
+    {
+      "epoch": 0.23879252083802657,
+      "grad_norm": 1.9596289984736852,
+      "learning_rate": 1e-05,
+      "loss": 1.2954,
+      "step": 530
+    },
+    {
+      "epoch": 0.23924307276413606,
+      "grad_norm": 1.9621410635607637,
+      "learning_rate": 1e-05,
+      "loss": 1.2208,
+      "step": 531
+    },
+    {
+      "epoch": 0.23969362469024555,
+      "grad_norm": 1.922080949168734,
+      "learning_rate": 1e-05,
+      "loss": 1.2036,
+      "step": 532
+    },
+    {
+      "epoch": 0.24014417661635504,
+      "grad_norm": 2.0146667015909427,
+      "learning_rate": 1e-05,
+      "loss": 1.2538,
+      "step": 533
+    },
+    {
+      "epoch": 0.24059472854246453,
+      "grad_norm": 2.0416674478286763,
+      "learning_rate": 1e-05,
+      "loss": 1.2441,
+      "step": 534
+    },
+    {
+      "epoch": 0.241045280468574,
+      "grad_norm": 2.082711792086268,
+      "learning_rate": 1e-05,
+      "loss": 1.2083,
+      "step": 535
+    },
+    {
+      "epoch": 0.24149583239468347,
+      "grad_norm": 2.151044272905431,
+      "learning_rate": 1e-05,
+      "loss": 1.2281,
+      "step": 536
+    },
+    {
+      "epoch": 0.24194638432079296,
+      "grad_norm": 1.9625368046490017,
+      "learning_rate": 1e-05,
+      "loss": 1.275,
+      "step": 537
+    },
+    {
+      "epoch": 0.24239693624690245,
+      "grad_norm": 2.0426023983025985,
+      "learning_rate": 1e-05,
+      "loss": 1.2283,
+      "step": 538
+    },
+    {
+      "epoch": 0.24284748817301194,
+      "grad_norm": 1.853693957393131,
+      "learning_rate": 1e-05,
+      "loss": 1.2646,
+      "step": 539
+    },
+    {
+      "epoch": 0.24329804009912143,
+      "grad_norm": 2.0254715870273596,
+      "learning_rate": 1e-05,
+      "loss": 1.2516,
+      "step": 540
+    },
+    {
+      "epoch": 0.24374859202523091,
+      "grad_norm": 2.0387179440881753,
+      "learning_rate": 1e-05,
+      "loss": 1.2457,
+      "step": 541
+    },
+    {
+      "epoch": 0.2441991439513404,
+      "grad_norm": 1.9099190321454138,
+      "learning_rate": 1e-05,
+      "loss": 1.2443,
+      "step": 542
+    },
+    {
+      "epoch": 0.24464969587744986,
+      "grad_norm": 1.9176306660283398,
+      "learning_rate": 1e-05,
+      "loss": 1.2296,
+      "step": 543
+    },
+    {
+      "epoch": 0.24510024780355935,
+      "grad_norm": 2.099460935168799,
+      "learning_rate": 1e-05,
+      "loss": 1.24,
+      "step": 544
+    },
+    {
+      "epoch": 0.24555079972966884,
+      "grad_norm": 1.9843854834957728,
+      "learning_rate": 1e-05,
+      "loss": 1.2776,
+      "step": 545
+    },
+    {
+      "epoch": 0.24600135165577833,
+      "grad_norm": 2.0908098000374475,
+      "learning_rate": 1e-05,
+      "loss": 1.277,
+      "step": 546
+    },
+    {
+      "epoch": 0.24645190358188782,
+      "grad_norm": 1.857303589542034,
+      "learning_rate": 1e-05,
+      "loss": 1.2682,
+      "step": 547
+    },
+    {
+      "epoch": 0.2469024555079973,
+      "grad_norm": 1.8466377771443474,
+      "learning_rate": 1e-05,
+      "loss": 1.2519,
+      "step": 548
+    },
+    {
+      "epoch": 0.2473530074341068,
+      "grad_norm": 1.9712909163231243,
+      "learning_rate": 1e-05,
+      "loss": 1.2531,
+      "step": 549
+    },
+    {
+      "epoch": 0.24780355936021625,
+      "grad_norm": 1.7932411789570453,
+      "learning_rate": 1e-05,
+      "loss": 1.2663,
+      "step": 550
+    },
+    {
+      "epoch": 0.24825411128632574,
+      "grad_norm": 1.9020112246587428,
+      "learning_rate": 1e-05,
+      "loss": 1.2591,
+      "step": 551
+    },
+    {
+      "epoch": 0.24870466321243523,
+      "grad_norm": 1.9559681862087972,
+      "learning_rate": 1e-05,
+      "loss": 1.3088,
+      "step": 552
+    },
+    {
+      "epoch": 0.24915521513854472,
+      "grad_norm": 1.8097002149071548,
+      "learning_rate": 1e-05,
+      "loss": 1.2609,
+      "step": 553
+    },
+    {
+      "epoch": 0.2496057670646542,
+      "grad_norm": 2.0104367704997435,
+      "learning_rate": 1e-05,
+      "loss": 1.2931,
+      "step": 554
+    },
+    {
+      "epoch": 0.2500563189907637,
+      "grad_norm": 1.9436277754193785,
+      "learning_rate": 1e-05,
+      "loss": 1.3028,
+      "step": 555
+    },
+    {
+      "epoch": 0.2505068709168732,
+      "grad_norm": 1.8798357276671347,
+      "learning_rate": 1e-05,
+      "loss": 1.2211,
+      "step": 556
+    },
+    {
+      "epoch": 0.25095742284298267,
+      "grad_norm": 1.9024477751630549,
+      "learning_rate": 1e-05,
+      "loss": 1.2961,
+      "step": 557
+    },
+    {
+      "epoch": 0.25140797476909216,
+      "grad_norm": 1.8153727085142268,
+      "learning_rate": 1e-05,
+      "loss": 1.2941,
+      "step": 558
+    },
+    {
+      "epoch": 0.25185852669520165,
+      "grad_norm": 2.2310326015353983,
+      "learning_rate": 1e-05,
+      "loss": 1.29,
+      "step": 559
+    },
+    {
+      "epoch": 0.25230907862131113,
+      "grad_norm": 1.8442397109048703,
+      "learning_rate": 1e-05,
+      "loss": 1.2141,
+      "step": 560
+    },
+    {
+      "epoch": 0.25275963054742057,
+      "grad_norm": 1.9019081996535443,
+      "learning_rate": 1e-05,
+      "loss": 1.2646,
+      "step": 561
+    },
+    {
+      "epoch": 0.25321018247353005,
+      "grad_norm": 1.8380479192111479,
+      "learning_rate": 1e-05,
+      "loss": 1.2558,
+      "step": 562
+    },
+    {
+      "epoch": 0.25366073439963954,
+      "grad_norm": 1.8426028450767409,
+      "learning_rate": 1e-05,
+      "loss": 1.2492,
+      "step": 563
+    },
+    {
+      "epoch": 0.25411128632574903,
+      "grad_norm": 2.010881620497572,
+      "learning_rate": 1e-05,
+      "loss": 1.2736,
+      "step": 564
+    },
+    {
+      "epoch": 0.2545618382518585,
+      "grad_norm": 1.8697749380936282,
+      "learning_rate": 1e-05,
+      "loss": 1.2506,
+      "step": 565
+    },
+    {
+      "epoch": 0.255012390177968,
+      "grad_norm": 1.8935160673483178,
+      "learning_rate": 1e-05,
+      "loss": 1.2153,
+      "step": 566
+    },
+    {
+      "epoch": 0.2554629421040775,
+      "grad_norm": 1.9153294504976563,
+      "learning_rate": 1e-05,
+      "loss": 1.2979,
+      "step": 567
+    },
+    {
+      "epoch": 0.255913494030187,
+      "grad_norm": 1.9618080034916596,
+      "learning_rate": 1e-05,
+      "loss": 1.2654,
+      "step": 568
+    },
+    {
+      "epoch": 0.25636404595629647,
+      "grad_norm": 1.993769279135054,
+      "learning_rate": 1e-05,
+      "loss": 1.3167,
+      "step": 569
+    },
+    {
+      "epoch": 0.25681459788240596,
+      "grad_norm": 1.9148075559939128,
+      "learning_rate": 1e-05,
+      "loss": 1.2521,
+      "step": 570
+    },
+    {
+      "epoch": 0.25726514980851545,
+      "grad_norm": 1.8930065639482323,
+      "learning_rate": 1e-05,
+      "loss": 1.2972,
+      "step": 571
+    },
+    {
+      "epoch": 0.25771570173462494,
+      "grad_norm": 2.135229844218331,
+      "learning_rate": 1e-05,
+      "loss": 1.2488,
+      "step": 572
+    },
+    {
+      "epoch": 0.2581662536607344,
+      "grad_norm": 2.0760526206165153,
+      "learning_rate": 1e-05,
+      "loss": 1.218,
+      "step": 573
+    },
+    {
+      "epoch": 0.25861680558684386,
+      "grad_norm": 1.938146636441883,
+      "learning_rate": 1e-05,
+      "loss": 1.257,
+      "step": 574
+    },
+    {
+      "epoch": 0.25906735751295334,
+      "grad_norm": 1.9346531249502403,
+      "learning_rate": 1e-05,
+      "loss": 1.2362,
+      "step": 575
+    },
+    {
+      "epoch": 0.25951790943906283,
+      "grad_norm": 1.8983542752101255,
+      "learning_rate": 1e-05,
+      "loss": 1.245,
+      "step": 576
+    },
+    {
+      "epoch": 0.2599684613651723,
+      "grad_norm": 1.7958759437626646,
+      "learning_rate": 1e-05,
+      "loss": 1.253,
+      "step": 577
+    },
+    {
+      "epoch": 0.2604190132912818,
+      "grad_norm": 1.8849719480056717,
+      "learning_rate": 1e-05,
+      "loss": 1.2436,
+      "step": 578
+    },
+    {
+      "epoch": 0.2608695652173913,
+      "grad_norm": 1.7589070794953543,
+      "learning_rate": 1e-05,
+      "loss": 1.2681,
+      "step": 579
+    },
+    {
+      "epoch": 0.2613201171435008,
+      "grad_norm": 1.9961918535347094,
+      "learning_rate": 1e-05,
+      "loss": 1.2522,
+      "step": 580
+    },
+    {
+      "epoch": 0.2617706690696103,
+      "grad_norm": 2.110755417506112,
+      "learning_rate": 1e-05,
+      "loss": 1.2925,
+      "step": 581
+    },
+    {
+      "epoch": 0.26222122099571976,
+      "grad_norm": 2.2152176744911043,
+      "learning_rate": 1e-05,
+      "loss": 1.2577,
+      "step": 582
+    },
+    {
+      "epoch": 0.26267177292182925,
+      "grad_norm": 1.8224419298462682,
+      "learning_rate": 1e-05,
+      "loss": 1.252,
+      "step": 583
+    },
+    {
+      "epoch": 0.26312232484793874,
+      "grad_norm": 2.1189296581041877,
+      "learning_rate": 1e-05,
+      "loss": 1.2493,
+      "step": 584
+    },
+    {
+      "epoch": 0.2635728767740482,
+      "grad_norm": 1.9324131683927712,
+      "learning_rate": 1e-05,
+      "loss": 1.2796,
+      "step": 585
+    },
+    {
+      "epoch": 0.2640234287001577,
+      "grad_norm": 2.2033632424316294,
+      "learning_rate": 1e-05,
+      "loss": 1.2669,
+      "step": 586
+    },
+    {
+      "epoch": 0.2644739806262672,
+      "grad_norm": 1.7928066330460684,
+      "learning_rate": 1e-05,
+      "loss": 1.2663,
+      "step": 587
+    },
+    {
+      "epoch": 0.26492453255237663,
+      "grad_norm": 2.0178291312075833,
+      "learning_rate": 1e-05,
+      "loss": 1.2465,
+      "step": 588
+    },
+    {
+      "epoch": 0.2653750844784861,
+      "grad_norm": 1.7942754093241515,
+      "learning_rate": 1e-05,
+      "loss": 1.2118,
+      "step": 589
+    },
+    {
+      "epoch": 0.2658256364045956,
+      "grad_norm": 2.0418669185015124,
+      "learning_rate": 1e-05,
+      "loss": 1.2664,
+      "step": 590
+    },
+    {
+      "epoch": 0.2662761883307051,
+      "grad_norm": 1.9047054777681043,
+      "learning_rate": 1e-05,
+      "loss": 1.2726,
+      "step": 591
+    },
+    {
+      "epoch": 0.2667267402568146,
+      "grad_norm": 1.8933659600328823,
+      "learning_rate": 1e-05,
+      "loss": 1.3021,
+      "step": 592
+    },
+    {
+      "epoch": 0.2671772921829241,
+      "grad_norm": 1.8504998391516625,
+      "learning_rate": 1e-05,
+      "loss": 1.2382,
+      "step": 593
+    },
+    {
+      "epoch": 0.26762784410903356,
+      "grad_norm": 1.9144625800421582,
+      "learning_rate": 1e-05,
+      "loss": 1.2684,
+      "step": 594
+    },
+    {
+      "epoch": 0.26807839603514305,
+      "grad_norm": 1.9094798541668625,
+      "learning_rate": 1e-05,
+      "loss": 1.2533,
+      "step": 595
+    },
+    {
+      "epoch": 0.26852894796125254,
+      "grad_norm": 1.8852457983238309,
+      "learning_rate": 1e-05,
+      "loss": 1.2162,
+      "step": 596
+    },
+    {
+      "epoch": 0.26897949988736203,
+      "grad_norm": 1.9097091467304366,
+      "learning_rate": 1e-05,
+      "loss": 1.3164,
+      "step": 597
+    },
+    {
+      "epoch": 0.2694300518134715,
+      "grad_norm": 1.8950963782713985,
+      "learning_rate": 1e-05,
+      "loss": 1.2247,
+      "step": 598
+    },
+    {
+      "epoch": 0.269880603739581,
+      "grad_norm": 1.6843883133171236,
+      "learning_rate": 1e-05,
+      "loss": 1.2235,
+      "step": 599
+    },
+    {
+      "epoch": 0.2703311556656905,
+      "grad_norm": 2.0418070127443153,
+      "learning_rate": 1e-05,
+      "loss": 1.2505,
+      "step": 600
+    },
+    {
+      "epoch": 0.2707817075918,
+      "grad_norm": 1.7928717250400767,
+      "learning_rate": 1e-05,
+      "loss": 1.2421,
+      "step": 601
+    },
+    {
+      "epoch": 0.2712322595179094,
+      "grad_norm": 1.9859554696363064,
+      "learning_rate": 1e-05,
+      "loss": 1.229,
+      "step": 602
+    },
+    {
+      "epoch": 0.2716828114440189,
+      "grad_norm": 2.0638907923686234,
+      "learning_rate": 1e-05,
+      "loss": 1.2687,
+      "step": 603
+    },
+    {
+      "epoch": 0.2721333633701284,
+      "grad_norm": 1.9984427057305065,
+      "learning_rate": 1e-05,
+      "loss": 1.2763,
+      "step": 604
+    },
+    {
+      "epoch": 0.2725839152962379,
+      "grad_norm": 1.830478514385362,
+      "learning_rate": 1e-05,
+      "loss": 1.2383,
+      "step": 605
+    },
+    {
+      "epoch": 0.27303446722234737,
+      "grad_norm": 1.9236051373849303,
+      "learning_rate": 1e-05,
+      "loss": 1.2394,
+      "step": 606
+    },
+    {
+      "epoch": 0.27348501914845685,
+      "grad_norm": 1.8885633057049556,
+      "learning_rate": 1e-05,
+      "loss": 1.2794,
+      "step": 607
+    },
+    {
+      "epoch": 0.27393557107456634,
+      "grad_norm": 2.0478146941614344,
+      "learning_rate": 1e-05,
+      "loss": 1.2718,
+      "step": 608
+    },
+    {
+      "epoch": 0.27438612300067583,
+      "grad_norm": 1.885228281361915,
+      "learning_rate": 1e-05,
+      "loss": 1.1898,
+      "step": 609
+    },
+    {
+      "epoch": 0.2748366749267853,
+      "grad_norm": 1.965907132172426,
+      "learning_rate": 1e-05,
+      "loss": 1.2515,
+      "step": 610
+    },
+    {
+      "epoch": 0.2752872268528948,
+      "grad_norm": 1.971709861173124,
+      "learning_rate": 1e-05,
+      "loss": 1.2287,
+      "step": 611
+    },
+    {
+      "epoch": 0.2757377787790043,
+      "grad_norm": 1.9506731322089625,
+      "learning_rate": 1e-05,
+      "loss": 1.2697,
+      "step": 612
+    },
+    {
+      "epoch": 0.2761883307051138,
+      "grad_norm": 2.037557087427492,
+      "learning_rate": 1e-05,
+      "loss": 1.2547,
+      "step": 613
+    },
+    {
+      "epoch": 0.27663888263122327,
+      "grad_norm": 1.9533614493348617,
+      "learning_rate": 1e-05,
+      "loss": 1.2732,
+      "step": 614
+    },
+    {
+      "epoch": 0.27708943455733276,
+      "grad_norm": 1.9467276392821233,
+      "learning_rate": 1e-05,
+      "loss": 1.2485,
+      "step": 615
+    },
+    {
+      "epoch": 0.2775399864834422,
+      "grad_norm": 1.8261245698686148,
+      "learning_rate": 1e-05,
+      "loss": 1.2801,
+      "step": 616
+    },
+    {
+      "epoch": 0.2779905384095517,
+      "grad_norm": 1.7677999183226603,
+      "learning_rate": 1e-05,
+      "loss": 1.2581,
+      "step": 617
+    },
+    {
+      "epoch": 0.27844109033566117,
+      "grad_norm": 1.8793767659029388,
+      "learning_rate": 1e-05,
+      "loss": 1.2155,
+      "step": 618
+    },
+    {
+      "epoch": 0.27889164226177066,
+      "grad_norm": 1.96490874825221,
+      "learning_rate": 1e-05,
+      "loss": 1.2412,
+      "step": 619
+    },
+    {
+      "epoch": 0.27934219418788014,
+      "grad_norm": 1.9862147573110902,
+      "learning_rate": 1e-05,
+      "loss": 1.2088,
+      "step": 620
+    },
+    {
+      "epoch": 0.27979274611398963,
+      "grad_norm": 1.9880527565841493,
+      "learning_rate": 1e-05,
+      "loss": 1.3268,
+      "step": 621
+    },
+    {
+      "epoch": 0.2802432980400991,
+      "grad_norm": 1.918947122409961,
+      "learning_rate": 1e-05,
+      "loss": 1.2982,
+      "step": 622
+    },
+    {
+      "epoch": 0.2806938499662086,
+      "grad_norm": 2.1011233128899054,
+      "learning_rate": 1e-05,
+      "loss": 1.2747,
+      "step": 623
+    },
+    {
+      "epoch": 0.2811444018923181,
+      "grad_norm": 2.1554619814183726,
+      "learning_rate": 1e-05,
+      "loss": 1.2303,
+      "step": 624
+    },
+    {
+      "epoch": 0.2815949538184276,
+      "grad_norm": 1.9287282770735685,
+      "learning_rate": 1e-05,
+      "loss": 1.272,
+      "step": 625
+    },
+    {
+      "epoch": 0.2820455057445371,
+      "grad_norm": 1.9136176641149933,
+      "learning_rate": 1e-05,
+      "loss": 1.2338,
+      "step": 626
+    },
+    {
+      "epoch": 0.28249605767064656,
+      "grad_norm": 1.7970486871003493,
+      "learning_rate": 1e-05,
+      "loss": 1.2632,
+      "step": 627
+    },
+    {
+      "epoch": 0.28294660959675605,
+      "grad_norm": 1.8635294906179631,
+      "learning_rate": 1e-05,
+      "loss": 1.242,
+      "step": 628
+    },
+    {
+      "epoch": 0.28339716152286554,
+      "grad_norm": 1.8187208723363777,
+      "learning_rate": 1e-05,
+      "loss": 1.2687,
+      "step": 629
+    },
+    {
+      "epoch": 0.28384771344897497,
+      "grad_norm": 1.9285995653602948,
+      "learning_rate": 1e-05,
+      "loss": 1.2532,
+      "step": 630
+    },
+    {
+      "epoch": 0.28429826537508446,
+      "grad_norm": 1.7634240721942334,
+      "learning_rate": 1e-05,
+      "loss": 1.2467,
+      "step": 631
+    },
+    {
+      "epoch": 0.28474881730119395,
+      "grad_norm": 1.9044966044859704,
+      "learning_rate": 1e-05,
+      "loss": 1.2523,
+      "step": 632
+    },
+    {
+      "epoch": 0.28519936922730343,
+      "grad_norm": 1.9045151073343214,
+      "learning_rate": 1e-05,
+      "loss": 1.3006,
+      "step": 633
+    },
+    {
+      "epoch": 0.2856499211534129,
+      "grad_norm": 1.95462789696124,
+      "learning_rate": 1e-05,
+      "loss": 1.2299,
+      "step": 634
+    },
+    {
+      "epoch": 0.2861004730795224,
+      "grad_norm": 1.772589567631728,
+      "learning_rate": 1e-05,
+      "loss": 1.2576,
+      "step": 635
+    },
+    {
+      "epoch": 0.2865510250056319,
+      "grad_norm": 1.9744864517778673,
+      "learning_rate": 1e-05,
+      "loss": 1.2105,
+      "step": 636
+    },
+    {
+      "epoch": 0.2870015769317414,
+      "grad_norm": 1.9986746713780499,
+      "learning_rate": 1e-05,
+      "loss": 1.2213,
+      "step": 637
+    },
+    {
+      "epoch": 0.2874521288578509,
+      "grad_norm": 1.8563633099604115,
+      "learning_rate": 1e-05,
+      "loss": 1.2318,
+      "step": 638
+    },
+    {
+      "epoch": 0.28790268078396036,
+      "grad_norm": 1.8947838304224478,
+      "learning_rate": 1e-05,
+      "loss": 1.2159,
+      "step": 639
+    },
+    {
+      "epoch": 0.28835323271006985,
+      "grad_norm": 1.8021751312403476,
+      "learning_rate": 1e-05,
+      "loss": 1.2135,
+      "step": 640
+    },
+    {
+      "epoch": 0.28880378463617934,
+      "grad_norm": 1.9584176887843077,
+      "learning_rate": 1e-05,
+      "loss": 1.2968,
+      "step": 641
+    },
+    {
+      "epoch": 0.2892543365622888,
+      "grad_norm": 2.1133640919058885,
+      "learning_rate": 1e-05,
+      "loss": 1.2968,
+      "step": 642
+    },
+    {
+      "epoch": 0.2897048884883983,
+      "grad_norm": 1.8296540316036876,
+      "learning_rate": 1e-05,
+      "loss": 1.2356,
+      "step": 643
+    },
+    {
+      "epoch": 0.29015544041450775,
+      "grad_norm": 1.7767733447837024,
+      "learning_rate": 1e-05,
+      "loss": 1.2734,
+      "step": 644
+    },
+    {
+      "epoch": 0.29060599234061724,
+      "grad_norm": 1.779280238237704,
+      "learning_rate": 1e-05,
+      "loss": 1.2466,
+      "step": 645
+    },
+    {
+      "epoch": 0.2910565442667267,
+      "grad_norm": 1.8168390816308468,
+      "learning_rate": 1e-05,
+      "loss": 1.2492,
+      "step": 646
+    },
+    {
+      "epoch": 0.2915070961928362,
+      "grad_norm": 2.041111946701274,
+      "learning_rate": 1e-05,
+      "loss": 1.1899,
+      "step": 647
+    },
+    {
+      "epoch": 0.2919576481189457,
+      "grad_norm": 1.9122640157995485,
+      "learning_rate": 1e-05,
+      "loss": 1.2323,
+      "step": 648
+    },
+    {
+      "epoch": 0.2924082000450552,
+      "grad_norm": 1.9426734430959265,
+      "learning_rate": 1e-05,
+      "loss": 1.2979,
+      "step": 649
+    },
+    {
+      "epoch": 0.2928587519711647,
+      "grad_norm": 2.0810089200553206,
+      "learning_rate": 1e-05,
+      "loss": 1.2241,
+      "step": 650
+    },
+    {
+      "epoch": 0.29330930389727417,
+      "grad_norm": 1.885071653308782,
+      "learning_rate": 1e-05,
+      "loss": 1.2606,
+      "step": 651
+    },
+    {
+      "epoch": 0.29375985582338365,
+      "grad_norm": 1.9440547245925526,
+      "learning_rate": 1e-05,
+      "loss": 1.2584,
+      "step": 652
+    },
+    {
+      "epoch": 0.29421040774949314,
+      "grad_norm": 1.8334532422194452,
+      "learning_rate": 1e-05,
+      "loss": 1.2687,
+      "step": 653
+    },
+    {
+      "epoch": 0.29466095967560263,
+      "grad_norm": 2.0064599716232743,
+      "learning_rate": 1e-05,
+      "loss": 1.2758,
+      "step": 654
+    },
+    {
+      "epoch": 0.2951115116017121,
+      "grad_norm": 2.0338711008764716,
+      "learning_rate": 1e-05,
+      "loss": 1.3173,
+      "step": 655
+    },
+    {
+      "epoch": 0.2955620635278216,
+      "grad_norm": 1.9613884552495555,
+      "learning_rate": 1e-05,
+      "loss": 1.2838,
+      "step": 656
+    },
+    {
+      "epoch": 0.29601261545393104,
+      "grad_norm": 1.794221735858151,
+      "learning_rate": 1e-05,
+      "loss": 1.2701,
+      "step": 657
+    },
+    {
+      "epoch": 0.2964631673800405,
+      "grad_norm": 1.7636679479293365,
+      "learning_rate": 1e-05,
+      "loss": 1.2454,
+      "step": 658
+    },
+    {
+      "epoch": 0.29691371930615,
+      "grad_norm": 1.9078463684005733,
+      "learning_rate": 1e-05,
+      "loss": 1.2439,
+      "step": 659
+    },
+    {
+      "epoch": 0.2973642712322595,
+      "grad_norm": 2.1209677317693454,
+      "learning_rate": 1e-05,
+      "loss": 1.2354,
+      "step": 660
+    },
+    {
+      "epoch": 0.297814823158369,
+      "grad_norm": 2.0008354557103623,
+      "learning_rate": 1e-05,
+      "loss": 1.2186,
+      "step": 661
+    },
+    {
+      "epoch": 0.2982653750844785,
+      "grad_norm": 1.8965814673419596,
+      "learning_rate": 1e-05,
+      "loss": 1.2603,
+      "step": 662
+    },
+    {
+      "epoch": 0.29871592701058797,
+      "grad_norm": 1.9719058474306004,
+      "learning_rate": 1e-05,
+      "loss": 1.2781,
+      "step": 663
+    },
+    {
+      "epoch": 0.29916647893669746,
+      "grad_norm": 1.7544263698970395,
+      "learning_rate": 1e-05,
+      "loss": 1.2813,
+      "step": 664
+    },
+    {
+      "epoch": 0.29961703086280694,
+      "grad_norm": 1.8711638364612417,
+      "learning_rate": 1e-05,
+      "loss": 1.2925,
+      "step": 665
+    },
+    {
+      "epoch": 0.30006758278891643,
+      "grad_norm": 2.0782600818572377,
+      "learning_rate": 1e-05,
+      "loss": 1.2416,
+      "step": 666
+    },
+    {
+      "epoch": 0.3005181347150259,
+      "grad_norm": 1.7494014354815353,
+      "learning_rate": 1e-05,
+      "loss": 1.263,
+      "step": 667
+    },
+    {
+      "epoch": 0.3009686866411354,
+      "grad_norm": 1.8595411629469847,
+      "learning_rate": 1e-05,
+      "loss": 1.2161,
+      "step": 668
+    },
+    {
+      "epoch": 0.3014192385672449,
+      "grad_norm": 1.8264571962662959,
+      "learning_rate": 1e-05,
+      "loss": 1.2698,
+      "step": 669
+    },
+    {
+      "epoch": 0.3018697904933544,
+      "grad_norm": 1.9191326189499924,
+      "learning_rate": 1e-05,
+      "loss": 1.2612,
+      "step": 670
+    },
+    {
+      "epoch": 0.3023203424194638,
+      "grad_norm": 2.160600233993354,
+      "learning_rate": 1e-05,
+      "loss": 1.2138,
+      "step": 671
+    },
+    {
+      "epoch": 0.3027708943455733,
+      "grad_norm": 1.7713948983525214,
+      "learning_rate": 1e-05,
+      "loss": 1.2791,
+      "step": 672
+    },
+    {
+      "epoch": 0.3032214462716828,
+      "grad_norm": 1.7448790476109295,
+      "learning_rate": 1e-05,
+      "loss": 1.2671,
+      "step": 673
+    },
+    {
+      "epoch": 0.3036719981977923,
+      "grad_norm": 1.8631161510984464,
+      "learning_rate": 1e-05,
+      "loss": 1.2832,
+      "step": 674
+    },
+    {
+      "epoch": 0.30412255012390177,
+      "grad_norm": 1.8206084628951194,
+      "learning_rate": 1e-05,
+      "loss": 1.2223,
+      "step": 675
+    },
+    {
+      "epoch": 0.30457310205001126,
+      "grad_norm": 1.9898347824902163,
+      "learning_rate": 1e-05,
+      "loss": 1.3051,
+      "step": 676
+    },
+    {
+      "epoch": 0.30502365397612075,
+      "grad_norm": 1.9515819844602575,
+      "learning_rate": 1e-05,
+      "loss": 1.2252,
+      "step": 677
+    },
+    {
+      "epoch": 0.30547420590223023,
+      "grad_norm": 1.9669775761642008,
+      "learning_rate": 1e-05,
+      "loss": 1.3134,
+      "step": 678
+    },
+    {
+      "epoch": 0.3059247578283397,
+      "grad_norm": 2.041871634160802,
+      "learning_rate": 1e-05,
+      "loss": 1.2395,
+      "step": 679
+    },
+    {
+      "epoch": 0.3063753097544492,
+      "grad_norm": 1.6875829319333566,
+      "learning_rate": 1e-05,
+      "loss": 1.2713,
+      "step": 680
+    },
+    {
+      "epoch": 0.3068258616805587,
+      "grad_norm": 1.8023021843830136,
+      "learning_rate": 1e-05,
+      "loss": 1.2073,
+      "step": 681
+    },
+    {
+      "epoch": 0.3072764136066682,
+      "grad_norm": 2.028430485499546,
+      "learning_rate": 1e-05,
+      "loss": 1.2381,
+      "step": 682
+    },
+    {
+      "epoch": 0.3077269655327777,
+      "grad_norm": 1.9448083800430478,
+      "learning_rate": 1e-05,
+      "loss": 1.2199,
+      "step": 683
+    },
+    {
+      "epoch": 0.30817751745888716,
+      "grad_norm": 1.820721473843506,
+      "learning_rate": 1e-05,
+      "loss": 1.2015,
+      "step": 684
+    },
+    {
+      "epoch": 0.3086280693849966,
+      "grad_norm": 1.949589011134347,
+      "learning_rate": 1e-05,
+      "loss": 1.2072,
+      "step": 685
+    },
+    {
+      "epoch": 0.3090786213111061,
+      "grad_norm": 1.8732994550966031,
+      "learning_rate": 1e-05,
+      "loss": 1.2967,
+      "step": 686
+    },
+    {
+      "epoch": 0.30952917323721557,
+      "grad_norm": 1.932829278229781,
+      "learning_rate": 1e-05,
+      "loss": 1.2673,
+      "step": 687
+    },
+    {
+      "epoch": 0.30997972516332506,
+      "grad_norm": 1.9086931128881206,
+      "learning_rate": 1e-05,
+      "loss": 1.2758,
+      "step": 688
+    },
+    {
+      "epoch": 0.31043027708943455,
+      "grad_norm": 1.760708860736313,
+      "learning_rate": 1e-05,
+      "loss": 1.2494,
+      "step": 689
+    },
+    {
+      "epoch": 0.31088082901554404,
+      "grad_norm": 1.9712662595531913,
+      "learning_rate": 1e-05,
+      "loss": 1.2472,
+      "step": 690
+    },
+    {
+      "epoch": 0.3113313809416535,
+      "grad_norm": 1.8962750227529488,
+      "learning_rate": 1e-05,
+      "loss": 1.2289,
+      "step": 691
+    },
+    {
+      "epoch": 0.311781932867763,
+      "grad_norm": 2.0300733168828597,
+      "learning_rate": 1e-05,
+      "loss": 1.2079,
+      "step": 692
+    },
+    {
+      "epoch": 0.3122324847938725,
+      "grad_norm": 1.8056562646023957,
+      "learning_rate": 1e-05,
+      "loss": 1.2212,
+      "step": 693
+    },
+    {
+      "epoch": 0.312683036719982,
+      "grad_norm": 1.7827128325999113,
+      "learning_rate": 1e-05,
+      "loss": 1.3003,
+      "step": 694
+    },
+    {
+      "epoch": 0.3131335886460915,
+      "grad_norm": 1.9333767688649806,
+      "learning_rate": 1e-05,
+      "loss": 1.2323,
+      "step": 695
+    },
+    {
+      "epoch": 0.31358414057220096,
+      "grad_norm": 1.8373807667649689,
+      "learning_rate": 1e-05,
+      "loss": 1.2484,
+      "step": 696
+    },
+    {
+      "epoch": 0.31403469249831045,
+      "grad_norm": 1.9226577869565329,
+      "learning_rate": 1e-05,
+      "loss": 1.2749,
+      "step": 697
+    },
+    {
+      "epoch": 0.31448524442441994,
+      "grad_norm": 1.984429483011316,
+      "learning_rate": 1e-05,
+      "loss": 1.2305,
+      "step": 698
+    },
+    {
+      "epoch": 0.3149357963505294,
+      "grad_norm": 1.7189193892169639,
+      "learning_rate": 1e-05,
+      "loss": 1.2669,
+      "step": 699
+    },
+    {
+      "epoch": 0.31538634827663886,
+      "grad_norm": 1.8465549129475556,
+      "learning_rate": 1e-05,
+      "loss": 1.2127,
+      "step": 700
+    },
+    {
+      "epoch": 0.31583690020274835,
+      "grad_norm": 1.8360690944861742,
+      "learning_rate": 1e-05,
+      "loss": 1.2512,
+      "step": 701
+    },
+    {
+      "epoch": 0.31628745212885784,
+      "grad_norm": 1.5944897716492685,
+      "learning_rate": 1e-05,
+      "loss": 1.2394,
+      "step": 702
+    },
+    {
+      "epoch": 0.3167380040549673,
+      "grad_norm": 1.9074999633821268,
+      "learning_rate": 1e-05,
+      "loss": 1.2171,
+      "step": 703
+    },
+    {
+      "epoch": 0.3171885559810768,
+      "grad_norm": 1.992601939946842,
+      "learning_rate": 1e-05,
+      "loss": 1.2643,
+      "step": 704
+    },
+    {
+      "epoch": 0.3176391079071863,
+      "grad_norm": 1.930457455101624,
+      "learning_rate": 1e-05,
+      "loss": 1.2193,
+      "step": 705
+    },
+    {
+      "epoch": 0.3180896598332958,
+      "grad_norm": 1.9228635882655845,
+      "learning_rate": 1e-05,
+      "loss": 1.2466,
+      "step": 706
+    },
+    {
+      "epoch": 0.3185402117594053,
+      "grad_norm": 2.042348035086456,
+      "learning_rate": 1e-05,
+      "loss": 1.2935,
+      "step": 707
+    },
+    {
+      "epoch": 0.31899076368551477,
+      "grad_norm": 1.8468885660035388,
+      "learning_rate": 1e-05,
+      "loss": 1.2542,
+      "step": 708
+    },
+    {
+      "epoch": 0.31944131561162425,
+      "grad_norm": 1.93678191805368,
+      "learning_rate": 1e-05,
+      "loss": 1.2067,
+      "step": 709
+    },
+    {
+      "epoch": 0.31989186753773374,
+      "grad_norm": 1.8568046179383066,
+      "learning_rate": 1e-05,
+      "loss": 1.2448,
+      "step": 710
+    },
+    {
+      "epoch": 0.32034241946384323,
+      "grad_norm": 1.936668150957869,
+      "learning_rate": 1e-05,
+      "loss": 1.2612,
+      "step": 711
+    },
+    {
+      "epoch": 0.3207929713899527,
+      "grad_norm": 1.7810908935731147,
+      "learning_rate": 1e-05,
+      "loss": 1.2294,
+      "step": 712
+    },
+    {
+      "epoch": 0.32124352331606215,
+      "grad_norm": 1.9363587940760691,
+      "learning_rate": 1e-05,
+      "loss": 1.2686,
+      "step": 713
+    },
+    {
+      "epoch": 0.32169407524217164,
+      "grad_norm": 1.828765552522956,
+      "learning_rate": 1e-05,
+      "loss": 1.2174,
+      "step": 714
+    },
+    {
+      "epoch": 0.32214462716828113,
+      "grad_norm": 1.9851127184045374,
+      "learning_rate": 1e-05,
+      "loss": 1.2122,
+      "step": 715
+    },
+    {
+      "epoch": 0.3225951790943906,
+      "grad_norm": 1.6979971611419993,
+      "learning_rate": 1e-05,
+      "loss": 1.2277,
+      "step": 716
+    },
+    {
+      "epoch": 0.3230457310205001,
+      "grad_norm": 1.798659712836174,
+      "learning_rate": 1e-05,
+      "loss": 1.2619,
+      "step": 717
+    },
+    {
+      "epoch": 0.3234962829466096,
+      "grad_norm": 1.9369672921677843,
+      "learning_rate": 1e-05,
+      "loss": 1.1789,
+      "step": 718
+    },
+    {
+      "epoch": 0.3239468348727191,
+      "grad_norm": 1.7233293500872506,
+      "learning_rate": 1e-05,
+      "loss": 1.2636,
+      "step": 719
+    },
+    {
+      "epoch": 0.32439738679882857,
+      "grad_norm": 1.695406821888204,
+      "learning_rate": 1e-05,
+      "loss": 1.2562,
+      "step": 720
+    },
+    {
+      "epoch": 0.32484793872493806,
+      "grad_norm": 1.7870571922328964,
+      "learning_rate": 1e-05,
+      "loss": 1.2992,
+      "step": 721
+    },
+    {
+      "epoch": 0.32529849065104754,
+      "grad_norm": 1.987769705851601,
+      "learning_rate": 1e-05,
+      "loss": 1.263,
+      "step": 722
+    },
+    {
+      "epoch": 0.32574904257715703,
+      "grad_norm": 2.073377619924158,
+      "learning_rate": 1e-05,
+      "loss": 1.2416,
+      "step": 723
+    },
+    {
+      "epoch": 0.3261995945032665,
+      "grad_norm": 1.8075389951410414,
+      "learning_rate": 1e-05,
+      "loss": 1.2239,
+      "step": 724
+    },
+    {
+      "epoch": 0.326650146429376,
+      "grad_norm": 1.8767063676009834,
+      "learning_rate": 1e-05,
+      "loss": 1.2238,
+      "step": 725
+    },
+    {
+      "epoch": 0.3271006983554855,
+      "grad_norm": 1.9335488853913598,
+      "learning_rate": 1e-05,
+      "loss": 1.2512,
+      "step": 726
+    },
+    {
+      "epoch": 0.32755125028159493,
+      "grad_norm": 1.9775223677374711,
+      "learning_rate": 1e-05,
+      "loss": 1.2708,
+      "step": 727
+    },
+    {
+      "epoch": 0.3280018022077044,
+      "grad_norm": 1.9912742641164474,
+      "learning_rate": 1e-05,
+      "loss": 1.252,
+      "step": 728
+    },
+    {
+      "epoch": 0.3284523541338139,
+      "grad_norm": 2.09177823369937,
+      "learning_rate": 1e-05,
+      "loss": 1.2249,
+      "step": 729
+    },
+    {
+      "epoch": 0.3289029060599234,
+      "grad_norm": 1.9218260164559409,
+      "learning_rate": 1e-05,
+      "loss": 1.2589,
+      "step": 730
+    },
+    {
+      "epoch": 0.3293534579860329,
+      "grad_norm": 1.816883454509941,
+      "learning_rate": 1e-05,
+      "loss": 1.2504,
+      "step": 731
+    },
+    {
+      "epoch": 0.32980400991214237,
+      "grad_norm": 1.9463305657675822,
+      "learning_rate": 1e-05,
+      "loss": 1.2332,
+      "step": 732
+    },
+    {
+      "epoch": 0.33025456183825186,
+      "grad_norm": 1.7032723184588883,
+      "learning_rate": 1e-05,
+      "loss": 1.2376,
+      "step": 733
+    },
+    {
+      "epoch": 0.33070511376436135,
+      "grad_norm": 1.991976651796448,
+      "learning_rate": 1e-05,
+      "loss": 1.2354,
+      "step": 734
+    },
+    {
+      "epoch": 0.33115566569047084,
+      "grad_norm": 1.8782791838639008,
+      "learning_rate": 1e-05,
+      "loss": 1.2649,
+      "step": 735
+    },
+    {
+      "epoch": 0.3316062176165803,
+      "grad_norm": 1.6540855870657907,
+      "learning_rate": 1e-05,
+      "loss": 1.2287,
+      "step": 736
+    },
+    {
+      "epoch": 0.3320567695426898,
+      "grad_norm": 1.9522705189092093,
+      "learning_rate": 1e-05,
+      "loss": 1.2337,
+      "step": 737
+    },
+    {
+      "epoch": 0.3325073214687993,
+      "grad_norm": 1.935043846337164,
+      "learning_rate": 1e-05,
+      "loss": 1.2327,
+      "step": 738
+    },
+    {
+      "epoch": 0.3329578733949088,
+      "grad_norm": 1.807798892625471,
+      "learning_rate": 1e-05,
+      "loss": 1.2422,
+      "step": 739
+    },
+    {
+      "epoch": 0.3334084253210182,
+      "grad_norm": 1.8889784912562422,
+      "learning_rate": 1e-05,
+      "loss": 1.2715,
+      "step": 740
+    },
+    {
+      "epoch": 0.3338589772471277,
+      "grad_norm": 1.723950690185663,
+      "learning_rate": 1e-05,
+      "loss": 1.2289,
+      "step": 741
+    },
+    {
+      "epoch": 0.3343095291732372,
+      "grad_norm": 1.5995009443472274,
+      "learning_rate": 1e-05,
+      "loss": 1.2203,
+      "step": 742
+    },
+    {
+      "epoch": 0.3347600810993467,
+      "grad_norm": 1.945758037730679,
+      "learning_rate": 1e-05,
+      "loss": 1.2116,
+      "step": 743
+    },
+    {
+      "epoch": 0.3352106330254562,
+      "grad_norm": 1.9733192591587503,
+      "learning_rate": 1e-05,
+      "loss": 1.232,
+      "step": 744
+    },
+    {
+      "epoch": 0.33566118495156566,
+      "grad_norm": 2.0633809873560884,
+      "learning_rate": 1e-05,
+      "loss": 1.2626,
+      "step": 745
+    },
+    {
+      "epoch": 0.33611173687767515,
+      "grad_norm": 2.0365944648132155,
+      "learning_rate": 1e-05,
+      "loss": 1.1862,
+      "step": 746
+    },
+    {
+      "epoch": 0.33656228880378464,
+      "grad_norm": 1.779828087983778,
+      "learning_rate": 1e-05,
+      "loss": 1.2454,
+      "step": 747
+    },
+    {
+      "epoch": 0.3370128407298941,
+      "grad_norm": 1.8123635566033534,
+      "learning_rate": 1e-05,
+      "loss": 1.2256,
+      "step": 748
+    },
+    {
+      "epoch": 0.3374633926560036,
+      "grad_norm": 2.1513724728075894,
+      "learning_rate": 1e-05,
+      "loss": 1.2773,
+      "step": 749
+    },
+    {
+      "epoch": 0.3379139445821131,
+      "grad_norm": 2.0077394950898326,
+      "learning_rate": 1e-05,
+      "loss": 1.2122,
+      "step": 750
+    },
+    {
+      "epoch": 0.3383644965082226,
+      "grad_norm": 1.9381550250884332,
+      "learning_rate": 1e-05,
+      "loss": 1.2519,
+      "step": 751
+    },
+    {
+      "epoch": 0.3388150484343321,
+      "grad_norm": 2.025361026523369,
+      "learning_rate": 1e-05,
+      "loss": 1.2078,
+      "step": 752
+    },
+    {
+      "epoch": 0.33926560036044157,
+      "grad_norm": 1.8400031398551076,
+      "learning_rate": 1e-05,
+      "loss": 1.2663,
+      "step": 753
+    },
+    {
+      "epoch": 0.339716152286551,
+      "grad_norm": 1.8809149555202167,
+      "learning_rate": 1e-05,
+      "loss": 1.2829,
+      "step": 754
+    },
+    {
+      "epoch": 0.3401667042126605,
+      "grad_norm": 1.9664266559297683,
+      "learning_rate": 1e-05,
+      "loss": 1.2397,
+      "step": 755
+    },
+    {
+      "epoch": 0.34061725613877,
+      "grad_norm": 2.041521078939078,
+      "learning_rate": 1e-05,
+      "loss": 1.2566,
+      "step": 756
+    },
+    {
+      "epoch": 0.34106780806487946,
+      "grad_norm": 1.6817834186608238,
+      "learning_rate": 1e-05,
+      "loss": 1.2897,
+      "step": 757
+    },
+    {
+      "epoch": 0.34151835999098895,
+      "grad_norm": 1.9397442643785014,
+      "learning_rate": 1e-05,
+      "loss": 1.264,
+      "step": 758
+    },
+    {
+      "epoch": 0.34196891191709844,
+      "grad_norm": 2.1191890181952964,
+      "learning_rate": 1e-05,
+      "loss": 1.184,
+      "step": 759
+    },
+    {
+      "epoch": 0.3424194638432079,
+      "grad_norm": 1.9793801207753674,
+      "learning_rate": 1e-05,
+      "loss": 1.2168,
+      "step": 760
+    },
+    {
+      "epoch": 0.3428700157693174,
+      "grad_norm": 1.9456244095521873,
+      "learning_rate": 1e-05,
+      "loss": 1.2407,
+      "step": 761
+    },
+    {
+      "epoch": 0.3433205676954269,
+      "grad_norm": 1.6993983808068778,
+      "learning_rate": 1e-05,
+      "loss": 1.2893,
+      "step": 762
+    },
+    {
+      "epoch": 0.3437711196215364,
+      "grad_norm": 1.769049718489682,
+      "learning_rate": 1e-05,
+      "loss": 1.2378,
+      "step": 763
+    },
+    {
+      "epoch": 0.3442216715476459,
+      "grad_norm": 1.9129320101600475,
+      "learning_rate": 1e-05,
+      "loss": 1.2181,
+      "step": 764
+    },
+    {
+      "epoch": 0.34467222347375537,
+      "grad_norm": 1.8656987912590273,
+      "learning_rate": 1e-05,
+      "loss": 1.1849,
+      "step": 765
+    },
+    {
+      "epoch": 0.34512277539986486,
+      "grad_norm": 2.098538830379863,
+      "learning_rate": 1e-05,
+      "loss": 1.2239,
+      "step": 766
+    },
+    {
+      "epoch": 0.34557332732597434,
+      "grad_norm": 1.8523257138169724,
+      "learning_rate": 1e-05,
+      "loss": 1.223,
+      "step": 767
+    },
+    {
+      "epoch": 0.3460238792520838,
+      "grad_norm": 1.7717813153696425,
+      "learning_rate": 1e-05,
+      "loss": 1.2795,
+      "step": 768
+    },
+    {
+      "epoch": 0.34647443117819327,
+      "grad_norm": 1.932633789733974,
+      "learning_rate": 1e-05,
+      "loss": 1.2513,
+      "step": 769
+    },
+    {
+      "epoch": 0.34692498310430275,
+      "grad_norm": 1.9188570240882774,
+      "learning_rate": 1e-05,
+      "loss": 1.2028,
+      "step": 770
+    },
+    {
+      "epoch": 0.34737553503041224,
+      "grad_norm": 1.9153912274518279,
+      "learning_rate": 1e-05,
+      "loss": 1.2137,
+      "step": 771
+    },
+    {
+      "epoch": 0.34782608695652173,
+      "grad_norm": 1.7891212005130266,
+      "learning_rate": 1e-05,
+      "loss": 1.1921,
+      "step": 772
+    },
+    {
+      "epoch": 0.3482766388826312,
+      "grad_norm": 2.1043655596657884,
+      "learning_rate": 1e-05,
+      "loss": 1.2318,
+      "step": 773
+    },
+    {
+      "epoch": 0.3487271908087407,
+      "grad_norm": 1.8048577556595926,
+      "learning_rate": 1e-05,
+      "loss": 1.2283,
+      "step": 774
+    },
+    {
+      "epoch": 0.3491777427348502,
+      "grad_norm": 1.8763088020409484,
+      "learning_rate": 1e-05,
+      "loss": 1.2978,
+      "step": 775
+    },
+    {
+      "epoch": 0.3496282946609597,
+      "grad_norm": 1.8610132874734748,
+      "learning_rate": 1e-05,
+      "loss": 1.222,
+      "step": 776
+    },
+    {
+      "epoch": 0.35007884658706917,
+      "grad_norm": 1.8498527649552017,
+      "learning_rate": 1e-05,
+      "loss": 1.2687,
+      "step": 777
+    },
+    {
+      "epoch": 0.35052939851317866,
+      "grad_norm": 1.7764009803913332,
+      "learning_rate": 1e-05,
+      "loss": 1.2407,
+      "step": 778
+    },
+    {
+      "epoch": 0.35097995043928815,
+      "grad_norm": 1.8972165021959277,
+      "learning_rate": 1e-05,
+      "loss": 1.2507,
+      "step": 779
+    },
+    {
+      "epoch": 0.35143050236539763,
+      "grad_norm": 1.8941513283311089,
+      "learning_rate": 1e-05,
+      "loss": 1.3047,
+      "step": 780
+    },
+    {
+      "epoch": 0.3518810542915071,
+      "grad_norm": 1.7516035467723052,
+      "learning_rate": 1e-05,
+      "loss": 1.2209,
+      "step": 781
+    },
+    {
+      "epoch": 0.35233160621761656,
+      "grad_norm": 1.7561059495858193,
+      "learning_rate": 1e-05,
+      "loss": 1.1997,
+      "step": 782
+    },
+    {
+      "epoch": 0.35278215814372604,
+      "grad_norm": 1.7901381312825824,
+      "learning_rate": 1e-05,
+      "loss": 1.2098,
+      "step": 783
+    },
+    {
+      "epoch": 0.35323271006983553,
+      "grad_norm": 1.7508908641532699,
+      "learning_rate": 1e-05,
+      "loss": 1.2572,
+      "step": 784
+    },
+    {
+      "epoch": 0.353683261995945,
+      "grad_norm": 1.8488980584698382,
+      "learning_rate": 1e-05,
+      "loss": 1.2055,
+      "step": 785
+    },
+    {
+      "epoch": 0.3541338139220545,
+      "grad_norm": 1.7706487550217227,
+      "learning_rate": 1e-05,
+      "loss": 1.2951,
+      "step": 786
+    },
+    {
+      "epoch": 0.354584365848164,
+      "grad_norm": 1.9049654575616393,
+      "learning_rate": 1e-05,
+      "loss": 1.2546,
+      "step": 787
+    },
+    {
+      "epoch": 0.3550349177742735,
+      "grad_norm": 1.9533455484883948,
+      "learning_rate": 1e-05,
+      "loss": 1.2211,
+      "step": 788
+    },
+    {
+      "epoch": 0.35548546970038297,
+      "grad_norm": 1.744879836527318,
+      "learning_rate": 1e-05,
+      "loss": 1.2455,
+      "step": 789
+    },
+    {
+      "epoch": 0.35593602162649246,
+      "grad_norm": 1.884398471923905,
+      "learning_rate": 1e-05,
+      "loss": 1.2628,
+      "step": 790
+    },
+    {
+      "epoch": 0.35638657355260195,
+      "grad_norm": 1.914153193744963,
+      "learning_rate": 1e-05,
+      "loss": 1.2213,
+      "step": 791
+    },
+    {
+      "epoch": 0.35683712547871144,
+      "grad_norm": 1.7549023438590918,
+      "learning_rate": 1e-05,
+      "loss": 1.2402,
+      "step": 792
+    },
+    {
+      "epoch": 0.3572876774048209,
+      "grad_norm": 1.848486134709626,
+      "learning_rate": 1e-05,
+      "loss": 1.2392,
+      "step": 793
+    },
+    {
+      "epoch": 0.3577382293309304,
+      "grad_norm": 1.7459577103976147,
+      "learning_rate": 1e-05,
+      "loss": 1.2305,
+      "step": 794
+    },
+    {
+      "epoch": 0.3581887812570399,
+      "grad_norm": 1.7715126663952638,
+      "learning_rate": 1e-05,
+      "loss": 1.2304,
+      "step": 795
+    },
+    {
+      "epoch": 0.35863933318314933,
+      "grad_norm": 1.9331485517321647,
+      "learning_rate": 1e-05,
+      "loss": 1.211,
+      "step": 796
+    },
+    {
+      "epoch": 0.3590898851092588,
+      "grad_norm": 1.9171014142037934,
+      "learning_rate": 1e-05,
+      "loss": 1.233,
+      "step": 797
+    },
+    {
+      "epoch": 0.3595404370353683,
+      "grad_norm": 1.8777048679057136,
+      "learning_rate": 1e-05,
+      "loss": 1.2483,
+      "step": 798
+    },
+    {
+      "epoch": 0.3599909889614778,
+      "grad_norm": 2.2202998663962754,
+      "learning_rate": 1e-05,
+      "loss": 1.272,
+      "step": 799
+    },
+    {
+      "epoch": 0.3604415408875873,
+      "grad_norm": 1.9769852229050389,
+      "learning_rate": 1e-05,
+      "loss": 1.236,
+      "step": 800
+    },
+    {
+      "epoch": 0.3608920928136968,
+      "grad_norm": 1.7462248980283235,
+      "learning_rate": 1e-05,
+      "loss": 1.296,
+      "step": 801
+    },
+    {
+      "epoch": 0.36134264473980626,
+      "grad_norm": 1.919523299895141,
+      "learning_rate": 1e-05,
+      "loss": 1.208,
+      "step": 802
+    },
+    {
+      "epoch": 0.36179319666591575,
+      "grad_norm": 1.7725751095521172,
+      "learning_rate": 1e-05,
+      "loss": 1.2697,
+      "step": 803
+    },
+    {
+      "epoch": 0.36224374859202524,
+      "grad_norm": 1.998043101972712,
+      "learning_rate": 1e-05,
+      "loss": 1.2168,
+      "step": 804
+    },
+    {
+      "epoch": 0.3626943005181347,
+      "grad_norm": 1.93511292404548,
+      "learning_rate": 1e-05,
+      "loss": 1.261,
+      "step": 805
+    },
+    {
+      "epoch": 0.3631448524442442,
+      "grad_norm": 1.939482584712453,
+      "learning_rate": 1e-05,
+      "loss": 1.2666,
+      "step": 806
+    },
+    {
+      "epoch": 0.3635954043703537,
+      "grad_norm": 1.9633358814141688,
+      "learning_rate": 1e-05,
+      "loss": 1.1979,
+      "step": 807
+    },
+    {
+      "epoch": 0.3640459562964632,
+      "grad_norm": 1.974552504756949,
+      "learning_rate": 1e-05,
+      "loss": 1.2228,
+      "step": 808
+    },
+    {
+      "epoch": 0.3644965082225726,
+      "grad_norm": 1.8943325796584518,
+      "learning_rate": 1e-05,
+      "loss": 1.2809,
+      "step": 809
+    },
+    {
+      "epoch": 0.3649470601486821,
+      "grad_norm": 1.8415647742821555,
+      "learning_rate": 1e-05,
+      "loss": 1.2498,
+      "step": 810
+    },
+    {
+      "epoch": 0.3653976120747916,
+      "grad_norm": 1.7198111870455757,
+      "learning_rate": 1e-05,
+      "loss": 1.2474,
+      "step": 811
+    },
+    {
+      "epoch": 0.3658481640009011,
+      "grad_norm": 1.8184528638787918,
+      "learning_rate": 1e-05,
+      "loss": 1.261,
+      "step": 812
+    },
+    {
+      "epoch": 0.3662987159270106,
+      "grad_norm": 1.908720974845513,
+      "learning_rate": 1e-05,
+      "loss": 1.2648,
+      "step": 813
+    },
+    {
+      "epoch": 0.36674926785312006,
+      "grad_norm": 1.7975656253941963,
+      "learning_rate": 1e-05,
+      "loss": 1.2426,
+      "step": 814
+    },
+    {
+      "epoch": 0.36719981977922955,
+      "grad_norm": 1.9362132595172787,
+      "learning_rate": 1e-05,
+      "loss": 1.256,
+      "step": 815
+    },
+    {
+      "epoch": 0.36765037170533904,
+      "grad_norm": 2.018637379819173,
+      "learning_rate": 1e-05,
+      "loss": 1.2344,
+      "step": 816
+    },
+    {
+      "epoch": 0.36810092363144853,
+      "grad_norm": 1.9802241053742995,
+      "learning_rate": 1e-05,
+      "loss": 1.2589,
+      "step": 817
+    },
+    {
+      "epoch": 0.368551475557558,
+      "grad_norm": 1.8099024787025881,
+      "learning_rate": 1e-05,
+      "loss": 1.2971,
+      "step": 818
+    },
+    {
+      "epoch": 0.3690020274836675,
+      "grad_norm": 2.0168776458192545,
+      "learning_rate": 1e-05,
+      "loss": 1.2287,
+      "step": 819
+    },
+    {
+      "epoch": 0.369452579409777,
+      "grad_norm": 1.8166821194359652,
+      "learning_rate": 1e-05,
+      "loss": 1.221,
+      "step": 820
+    },
+    {
+      "epoch": 0.3699031313358865,
+      "grad_norm": 1.8788222274730944,
+      "learning_rate": 1e-05,
+      "loss": 1.2146,
+      "step": 821
+    },
+    {
+      "epoch": 0.37035368326199597,
+      "grad_norm": 1.8597354689004102,
+      "learning_rate": 1e-05,
+      "loss": 1.3146,
+      "step": 822
+    },
+    {
+      "epoch": 0.3708042351881054,
+      "grad_norm": 1.8490024962659892,
+      "learning_rate": 1e-05,
+      "loss": 1.2531,
+      "step": 823
+    },
+    {
+      "epoch": 0.3712547871142149,
+      "grad_norm": 1.9932131168744172,
+      "learning_rate": 1e-05,
+      "loss": 1.2258,
+      "step": 824
+    },
+    {
+      "epoch": 0.3717053390403244,
+      "grad_norm": 1.792362296656891,
+      "learning_rate": 1e-05,
+      "loss": 1.2201,
+      "step": 825
+    },
+    {
+      "epoch": 0.37215589096643387,
+      "grad_norm": 1.938705231070029,
+      "learning_rate": 1e-05,
+      "loss": 1.2439,
+      "step": 826
+    },
+    {
+      "epoch": 0.37260644289254335,
+      "grad_norm": 1.7842902067000777,
+      "learning_rate": 1e-05,
+      "loss": 1.2847,
+      "step": 827
+    },
+    {
+      "epoch": 0.37305699481865284,
+      "grad_norm": 1.7616563146804334,
+      "learning_rate": 1e-05,
+      "loss": 1.2702,
+      "step": 828
+    },
+    {
+      "epoch": 0.37350754674476233,
+      "grad_norm": 1.7580772109432945,
+      "learning_rate": 1e-05,
+      "loss": 1.2636,
+      "step": 829
+    },
+    {
+      "epoch": 0.3739580986708718,
+      "grad_norm": 1.8708258964354627,
+      "learning_rate": 1e-05,
+      "loss": 1.2103,
+      "step": 830
+    },
+    {
+      "epoch": 0.3744086505969813,
+      "grad_norm": 2.0918050039950344,
+      "learning_rate": 1e-05,
+      "loss": 1.2447,
+      "step": 831
+    },
+    {
+      "epoch": 0.3748592025230908,
+      "grad_norm": 2.01351718562748,
+      "learning_rate": 1e-05,
+      "loss": 1.2506,
+      "step": 832
+    },
+    {
+      "epoch": 0.3753097544492003,
+      "grad_norm": 1.8425337248709375,
+      "learning_rate": 1e-05,
+      "loss": 1.2987,
+      "step": 833
+    },
+    {
+      "epoch": 0.37576030637530977,
+      "grad_norm": 1.8605102177776403,
+      "learning_rate": 1e-05,
+      "loss": 1.2332,
+      "step": 834
+    },
+    {
+      "epoch": 0.37621085830141926,
+      "grad_norm": 1.6946568207636692,
+      "learning_rate": 1e-05,
+      "loss": 1.2329,
+      "step": 835
+    },
+    {
+      "epoch": 0.37666141022752875,
+      "grad_norm": 1.8772757818015977,
+      "learning_rate": 1e-05,
+      "loss": 1.221,
+      "step": 836
+    },
+    {
+      "epoch": 0.3771119621536382,
+      "grad_norm": 1.842851166865783,
+      "learning_rate": 1e-05,
+      "loss": 1.2356,
+      "step": 837
+    },
+    {
+      "epoch": 0.37756251407974767,
+      "grad_norm": 1.8849523873973497,
+      "learning_rate": 1e-05,
+      "loss": 1.2147,
+      "step": 838
+    },
+    {
+      "epoch": 0.37801306600585716,
+      "grad_norm": 1.8533083831614017,
+      "learning_rate": 1e-05,
+      "loss": 1.2316,
+      "step": 839
+    },
+    {
+      "epoch": 0.37846361793196664,
+      "grad_norm": 1.9790724756108462,
+      "learning_rate": 1e-05,
+      "loss": 1.2492,
+      "step": 840
+    },
+    {
+      "epoch": 0.37891416985807613,
+      "grad_norm": 1.9415222438339077,
+      "learning_rate": 1e-05,
+      "loss": 1.2088,
+      "step": 841
+    },
+    {
+      "epoch": 0.3793647217841856,
+      "grad_norm": 2.0064179702380525,
+      "learning_rate": 1e-05,
+      "loss": 1.2054,
+      "step": 842
+    },
+    {
+      "epoch": 0.3798152737102951,
+      "grad_norm": 1.9995002419984607,
+      "learning_rate": 1e-05,
+      "loss": 1.2479,
+      "step": 843
+    },
+    {
+      "epoch": 0.3802658256364046,
+      "grad_norm": 1.746526951976324,
+      "learning_rate": 1e-05,
+      "loss": 1.2677,
+      "step": 844
+    },
+    {
+      "epoch": 0.3807163775625141,
+      "grad_norm": 1.8472340617635299,
+      "learning_rate": 1e-05,
+      "loss": 1.2229,
+      "step": 845
+    },
+    {
+      "epoch": 0.3811669294886236,
+      "grad_norm": 1.6989791803179513,
+      "learning_rate": 1e-05,
+      "loss": 1.248,
+      "step": 846
+    },
+    {
+      "epoch": 0.38161748141473306,
+      "grad_norm": 1.7861084739049955,
+      "learning_rate": 1e-05,
+      "loss": 1.2479,
+      "step": 847
+    },
+    {
+      "epoch": 0.38206803334084255,
+      "grad_norm": 1.769562122846481,
+      "learning_rate": 1e-05,
+      "loss": 1.2067,
+      "step": 848
+    },
+    {
+      "epoch": 0.38251858526695204,
+      "grad_norm": 1.884505014051771,
+      "learning_rate": 1e-05,
+      "loss": 1.2004,
+      "step": 849
+    },
+    {
+      "epoch": 0.3829691371930615,
+      "grad_norm": 1.8608952507144012,
+      "learning_rate": 1e-05,
+      "loss": 1.2203,
+      "step": 850
+    },
+    {
+      "epoch": 0.38341968911917096,
+      "grad_norm": 1.8068312924208156,
+      "learning_rate": 1e-05,
+      "loss": 1.2325,
+      "step": 851
+    },
+    {
+      "epoch": 0.38387024104528045,
+      "grad_norm": 1.9077098493511166,
+      "learning_rate": 1e-05,
+      "loss": 1.292,
+      "step": 852
+    },
+    {
+      "epoch": 0.38432079297138994,
+      "grad_norm": 1.7981833545840775,
+      "learning_rate": 1e-05,
+      "loss": 1.1984,
+      "step": 853
+    },
+    {
+      "epoch": 0.3847713448974994,
+      "grad_norm": 1.856513701528131,
+      "learning_rate": 1e-05,
+      "loss": 1.2531,
+      "step": 854
+    },
+    {
+      "epoch": 0.3852218968236089,
+      "grad_norm": 1.857015365071636,
+      "learning_rate": 1e-05,
+      "loss": 1.2707,
+      "step": 855
+    },
+    {
+      "epoch": 0.3856724487497184,
+      "grad_norm": 1.9707250335863131,
+      "learning_rate": 1e-05,
+      "loss": 1.2419,
+      "step": 856
+    },
+    {
+      "epoch": 0.3861230006758279,
+      "grad_norm": 1.9106073526456564,
+      "learning_rate": 1e-05,
+      "loss": 1.2054,
+      "step": 857
+    },
+    {
+      "epoch": 0.3865735526019374,
+      "grad_norm": 1.9501533883167215,
+      "learning_rate": 1e-05,
+      "loss": 1.2286,
+      "step": 858
+    },
+    {
+      "epoch": 0.38702410452804686,
+      "grad_norm": 1.9959109986598897,
+      "learning_rate": 1e-05,
+      "loss": 1.2468,
+      "step": 859
+    },
+    {
+      "epoch": 0.38747465645415635,
+      "grad_norm": 1.6484250314284374,
+      "learning_rate": 1e-05,
+      "loss": 1.2478,
+      "step": 860
+    },
+    {
+      "epoch": 0.38792520838026584,
+      "grad_norm": 1.9677741572608782,
+      "learning_rate": 1e-05,
+      "loss": 1.2097,
+      "step": 861
+    },
+    {
+      "epoch": 0.38837576030637533,
+      "grad_norm": 1.8559375675050855,
+      "learning_rate": 1e-05,
+      "loss": 1.2389,
+      "step": 862
+    },
+    {
+      "epoch": 0.3888263122324848,
+      "grad_norm": 1.7870988194556259,
+      "learning_rate": 1e-05,
+      "loss": 1.2223,
+      "step": 863
+    },
+    {
+      "epoch": 0.3892768641585943,
+      "grad_norm": 1.8287900417708567,
+      "learning_rate": 1e-05,
+      "loss": 1.2405,
+      "step": 864
+    },
+    {
+      "epoch": 0.38972741608470374,
+      "grad_norm": 1.951938431706165,
+      "learning_rate": 1e-05,
+      "loss": 1.2226,
+      "step": 865
+    },
+    {
+      "epoch": 0.3901779680108132,
+      "grad_norm": 1.752310475841562,
+      "learning_rate": 1e-05,
+      "loss": 1.2216,
+      "step": 866
+    },
+    {
+      "epoch": 0.3906285199369227,
+      "grad_norm": 1.9543454009939398,
+      "learning_rate": 1e-05,
+      "loss": 1.2375,
+      "step": 867
+    },
+    {
+      "epoch": 0.3910790718630322,
+      "grad_norm": 1.8219769800646242,
+      "learning_rate": 1e-05,
+      "loss": 1.2417,
+      "step": 868
+    },
+    {
+      "epoch": 0.3915296237891417,
+      "grad_norm": 1.8046726965804514,
+      "learning_rate": 1e-05,
+      "loss": 1.2429,
+      "step": 869
+    },
+    {
+      "epoch": 0.3919801757152512,
+      "grad_norm": 1.7627905770282928,
+      "learning_rate": 1e-05,
+      "loss": 1.1608,
+      "step": 870
+    },
+    {
+      "epoch": 0.39243072764136067,
+      "grad_norm": 1.815967903711212,
+      "learning_rate": 1e-05,
+      "loss": 1.26,
+      "step": 871
+    },
+    {
+      "epoch": 0.39288127956747015,
+      "grad_norm": 1.7445848284617353,
+      "learning_rate": 1e-05,
+      "loss": 1.2577,
+      "step": 872
+    },
+    {
+      "epoch": 0.39333183149357964,
+      "grad_norm": 1.868942646909875,
+      "learning_rate": 1e-05,
+      "loss": 1.214,
+      "step": 873
+    },
+    {
+      "epoch": 0.39378238341968913,
+      "grad_norm": 1.8210881389865994,
+      "learning_rate": 1e-05,
+      "loss": 1.2435,
+      "step": 874
+    },
+    {
+      "epoch": 0.3942329353457986,
+      "grad_norm": 1.705569170189639,
+      "learning_rate": 1e-05,
+      "loss": 1.241,
+      "step": 875
+    },
+    {
+      "epoch": 0.3946834872719081,
+      "grad_norm": 1.9540522749998606,
+      "learning_rate": 1e-05,
+      "loss": 1.2463,
+      "step": 876
+    },
+    {
+      "epoch": 0.3951340391980176,
+      "grad_norm": 1.8532976849033589,
+      "learning_rate": 1e-05,
+      "loss": 1.2684,
+      "step": 877
+    },
+    {
+      "epoch": 0.3955845911241271,
+      "grad_norm": 2.0267127643103375,
+      "learning_rate": 1e-05,
+      "loss": 1.2621,
+      "step": 878
+    },
+    {
+      "epoch": 0.3960351430502365,
+      "grad_norm": 1.7195863840982162,
+      "learning_rate": 1e-05,
+      "loss": 1.2528,
+      "step": 879
+    },
+    {
+      "epoch": 0.396485694976346,
+      "grad_norm": 1.9520831100294582,
+      "learning_rate": 1e-05,
+      "loss": 1.23,
+      "step": 880
+    },
+    {
+      "epoch": 0.3969362469024555,
+      "grad_norm": 1.6876082251249087,
+      "learning_rate": 1e-05,
+      "loss": 1.2124,
+      "step": 881
+    },
+    {
+      "epoch": 0.397386798828565,
+      "grad_norm": 1.9033954135692357,
+      "learning_rate": 1e-05,
+      "loss": 1.2337,
+      "step": 882
+    },
+    {
+      "epoch": 0.39783735075467447,
+      "grad_norm": 2.053902208410709,
+      "learning_rate": 1e-05,
+      "loss": 1.2482,
+      "step": 883
+    },
+    {
+      "epoch": 0.39828790268078396,
+      "grad_norm": 1.95607947734673,
+      "learning_rate": 1e-05,
+      "loss": 1.222,
+      "step": 884
+    },
+    {
+      "epoch": 0.39873845460689344,
+      "grad_norm": 1.7916894763559215,
+      "learning_rate": 1e-05,
+      "loss": 1.2135,
+      "step": 885
+    },
+    {
+      "epoch": 0.39918900653300293,
+      "grad_norm": 1.734770833440029,
+      "learning_rate": 1e-05,
+      "loss": 1.2173,
+      "step": 886
+    },
+    {
+      "epoch": 0.3996395584591124,
+      "grad_norm": 1.8631984425859183,
+      "learning_rate": 1e-05,
+      "loss": 1.1849,
+      "step": 887
+    },
+    {
+      "epoch": 0.4000901103852219,
+      "grad_norm": 1.8462058887209238,
+      "learning_rate": 1e-05,
+      "loss": 1.2135,
+      "step": 888
+    },
+    {
+      "epoch": 0.4005406623113314,
+      "grad_norm": 1.7876150498138967,
+      "learning_rate": 1e-05,
+      "loss": 1.2054,
+      "step": 889
+    },
+    {
+      "epoch": 0.4009912142374409,
+      "grad_norm": 1.854313355156316,
+      "learning_rate": 1e-05,
+      "loss": 1.2385,
+      "step": 890
+    },
+    {
+      "epoch": 0.4014417661635504,
+      "grad_norm": 1.8444446458921264,
+      "learning_rate": 1e-05,
+      "loss": 1.2009,
+      "step": 891
+    },
+    {
+      "epoch": 0.4018923180896598,
+      "grad_norm": 1.9383431386705288,
+      "learning_rate": 1e-05,
+      "loss": 1.2204,
+      "step": 892
+    },
+    {
+      "epoch": 0.4023428700157693,
+      "grad_norm": 2.1113776609966255,
+      "learning_rate": 1e-05,
+      "loss": 1.2534,
+      "step": 893
+    },
+    {
+      "epoch": 0.4027934219418788,
+      "grad_norm": 1.8768166031930706,
+      "learning_rate": 1e-05,
+      "loss": 1.2349,
+      "step": 894
+    },
+    {
+      "epoch": 0.40324397386798827,
+      "grad_norm": 1.773780616193697,
+      "learning_rate": 1e-05,
+      "loss": 1.2152,
+      "step": 895
+    },
+    {
+      "epoch": 0.40369452579409776,
+      "grad_norm": 1.6736686600690427,
+      "learning_rate": 1e-05,
+      "loss": 1.2171,
+      "step": 896
+    },
+    {
+      "epoch": 0.40414507772020725,
+      "grad_norm": 1.7515241224968952,
+      "learning_rate": 1e-05,
+      "loss": 1.2017,
+      "step": 897
+    },
+    {
+      "epoch": 0.40459562964631673,
+      "grad_norm": 1.9593202732085018,
+      "learning_rate": 1e-05,
+      "loss": 1.2439,
+      "step": 898
+    },
+    {
+      "epoch": 0.4050461815724262,
+      "grad_norm": 1.8267834579292501,
+      "learning_rate": 1e-05,
+      "loss": 1.2498,
+      "step": 899
+    },
+    {
+      "epoch": 0.4054967334985357,
+      "grad_norm": 1.8041221315512324,
+      "learning_rate": 1e-05,
+      "loss": 1.2342,
+      "step": 900
+    },
+    {
+      "epoch": 0.4059472854246452,
+      "grad_norm": 1.769381272532187,
+      "learning_rate": 1e-05,
+      "loss": 1.2233,
+      "step": 901
+    },
+    {
+      "epoch": 0.4063978373507547,
+      "grad_norm": 1.8129209057417408,
+      "learning_rate": 1e-05,
+      "loss": 1.2703,
+      "step": 902
+    },
+    {
+      "epoch": 0.4068483892768642,
+      "grad_norm": 1.9108631223024326,
+      "learning_rate": 1e-05,
+      "loss": 1.1581,
+      "step": 903
+    },
+    {
+      "epoch": 0.40729894120297366,
+      "grad_norm": 1.9726083051261702,
+      "learning_rate": 1e-05,
+      "loss": 1.2196,
+      "step": 904
+    },
+    {
+      "epoch": 0.40774949312908315,
+      "grad_norm": 1.8478866238047853,
+      "learning_rate": 1e-05,
+      "loss": 1.2491,
+      "step": 905
+    },
+    {
+      "epoch": 0.4082000450551926,
+      "grad_norm": 1.9997563451103388,
+      "learning_rate": 1e-05,
+      "loss": 1.2393,
+      "step": 906
+    },
+    {
+      "epoch": 0.40865059698130207,
+      "grad_norm": 1.8595271723897708,
+      "learning_rate": 1e-05,
+      "loss": 1.2191,
+      "step": 907
+    },
+    {
+      "epoch": 0.40910114890741156,
+      "grad_norm": 1.9314412730516308,
+      "learning_rate": 1e-05,
+      "loss": 1.2966,
+      "step": 908
+    },
+    {
+      "epoch": 0.40955170083352105,
+      "grad_norm": 1.9087951304482553,
+      "learning_rate": 1e-05,
+      "loss": 1.1789,
+      "step": 909
+    },
+    {
+      "epoch": 0.41000225275963054,
+      "grad_norm": 1.7270848651641488,
+      "learning_rate": 1e-05,
+      "loss": 1.2484,
+      "step": 910
+    },
+    {
+      "epoch": 0.41045280468574,
+      "grad_norm": 1.8133435393778934,
+      "learning_rate": 1e-05,
+      "loss": 1.2341,
+      "step": 911
+    },
+    {
+      "epoch": 0.4109033566118495,
+      "grad_norm": 1.7739710977749763,
+      "learning_rate": 1e-05,
+      "loss": 1.2591,
+      "step": 912
+    },
+    {
+      "epoch": 0.411353908537959,
+      "grad_norm": 1.8018454180699817,
+      "learning_rate": 1e-05,
+      "loss": 1.2302,
+      "step": 913
+    },
+    {
+      "epoch": 0.4118044604640685,
+      "grad_norm": 1.899747818272689,
+      "learning_rate": 1e-05,
+      "loss": 1.1914,
+      "step": 914
+    },
+    {
+      "epoch": 0.412255012390178,
+      "grad_norm": 1.6447189441511958,
+      "learning_rate": 1e-05,
+      "loss": 1.2483,
+      "step": 915
+    },
+    {
+      "epoch": 0.41270556431628747,
+      "grad_norm": 1.8335610585183395,
+      "learning_rate": 1e-05,
+      "loss": 1.2451,
+      "step": 916
+    },
+    {
+      "epoch": 0.41315611624239695,
+      "grad_norm": 1.7885413084914048,
+      "learning_rate": 1e-05,
+      "loss": 1.2423,
+      "step": 917
+    },
+    {
+      "epoch": 0.41360666816850644,
+      "grad_norm": 1.714313154932247,
+      "learning_rate": 1e-05,
+      "loss": 1.2214,
+      "step": 918
+    },
+    {
+      "epoch": 0.41405722009461593,
+      "grad_norm": 1.8999263091456016,
+      "learning_rate": 1e-05,
+      "loss": 1.211,
+      "step": 919
+    },
+    {
+      "epoch": 0.41450777202072536,
+      "grad_norm": 1.7615025789848202,
+      "learning_rate": 1e-05,
+      "loss": 1.2339,
+      "step": 920
+    },
+    {
+      "epoch": 0.41495832394683485,
+      "grad_norm": 1.984694925916507,
+      "learning_rate": 1e-05,
+      "loss": 1.2596,
+      "step": 921
+    },
+    {
+      "epoch": 0.41540887587294434,
+      "grad_norm": 1.8309947728006515,
+      "learning_rate": 1e-05,
+      "loss": 1.2188,
+      "step": 922
+    },
+    {
+      "epoch": 0.4158594277990538,
+      "grad_norm": 1.8714866019169583,
+      "learning_rate": 1e-05,
+      "loss": 1.2282,
+      "step": 923
+    },
+    {
+      "epoch": 0.4163099797251633,
+      "grad_norm": 1.6965979518458885,
+      "learning_rate": 1e-05,
+      "loss": 1.207,
+      "step": 924
+    },
+    {
+      "epoch": 0.4167605316512728,
+      "grad_norm": 1.8774421905009213,
+      "learning_rate": 1e-05,
+      "loss": 1.2191,
+      "step": 925
+    },
+    {
+      "epoch": 0.4172110835773823,
+      "grad_norm": 1.759408329157933,
+      "learning_rate": 1e-05,
+      "loss": 1.2521,
+      "step": 926
+    },
+    {
+      "epoch": 0.4176616355034918,
+      "grad_norm": 1.8325304285702253,
+      "learning_rate": 1e-05,
+      "loss": 1.2394,
+      "step": 927
+    },
+    {
+      "epoch": 0.41811218742960127,
+      "grad_norm": 1.8659079533070821,
+      "learning_rate": 1e-05,
+      "loss": 1.2526,
+      "step": 928
+    },
+    {
+      "epoch": 0.41856273935571076,
+      "grad_norm": 1.88343915219476,
+      "learning_rate": 1e-05,
+      "loss": 1.2834,
+      "step": 929
+    },
+    {
+      "epoch": 0.41901329128182024,
+      "grad_norm": 1.8107333099552414,
+      "learning_rate": 1e-05,
+      "loss": 1.2034,
+      "step": 930
+    },
+    {
+      "epoch": 0.41946384320792973,
+      "grad_norm": 1.8811280025175747,
+      "learning_rate": 1e-05,
+      "loss": 1.2346,
+      "step": 931
+    },
+    {
+      "epoch": 0.4199143951340392,
+      "grad_norm": 1.6863157009590324,
+      "learning_rate": 1e-05,
+      "loss": 1.2468,
+      "step": 932
+    },
+    {
+      "epoch": 0.4203649470601487,
+      "grad_norm": 1.6448935128144737,
+      "learning_rate": 1e-05,
+      "loss": 1.2001,
+      "step": 933
+    },
+    {
+      "epoch": 0.42081549898625814,
+      "grad_norm": 1.9123116226428942,
+      "learning_rate": 1e-05,
+      "loss": 1.2457,
+      "step": 934
+    },
+    {
+      "epoch": 0.42126605091236763,
+      "grad_norm": 1.8960168476037824,
+      "learning_rate": 1e-05,
+      "loss": 1.2395,
+      "step": 935
+    },
+    {
+      "epoch": 0.4217166028384771,
+      "grad_norm": 1.7512695559940004,
+      "learning_rate": 1e-05,
+      "loss": 1.2755,
+      "step": 936
+    },
+    {
+      "epoch": 0.4221671547645866,
+      "grad_norm": 2.112142633500385,
+      "learning_rate": 1e-05,
+      "loss": 1.2534,
+      "step": 937
+    },
+    {
+      "epoch": 0.4226177066906961,
+      "grad_norm": 1.8517229815085718,
+      "learning_rate": 1e-05,
+      "loss": 1.2409,
+      "step": 938
+    },
+    {
+      "epoch": 0.4230682586168056,
+      "grad_norm": 1.958729062600093,
+      "learning_rate": 1e-05,
+      "loss": 1.2611,
+      "step": 939
+    },
+    {
+      "epoch": 0.42351881054291507,
+      "grad_norm": 1.7826395039325846,
+      "learning_rate": 1e-05,
+      "loss": 1.2239,
+      "step": 940
+    },
+    {
+      "epoch": 0.42396936246902456,
+      "grad_norm": 1.7178891678540262,
+      "learning_rate": 1e-05,
+      "loss": 1.255,
+      "step": 941
+    },
+    {
+      "epoch": 0.42441991439513405,
+      "grad_norm": 1.7876765829669556,
+      "learning_rate": 1e-05,
+      "loss": 1.2414,
+      "step": 942
+    },
+    {
+      "epoch": 0.42487046632124353,
+      "grad_norm": 1.8427359715446618,
+      "learning_rate": 1e-05,
+      "loss": 1.2635,
+      "step": 943
+    },
+    {
+      "epoch": 0.425321018247353,
+      "grad_norm": 2.0491975947455736,
+      "learning_rate": 1e-05,
+      "loss": 1.2849,
+      "step": 944
+    },
+    {
+      "epoch": 0.4257715701734625,
+      "grad_norm": 1.6888517145569675,
+      "learning_rate": 1e-05,
+      "loss": 1.2113,
+      "step": 945
+    },
+    {
+      "epoch": 0.426222122099572,
+      "grad_norm": 1.7689491278837048,
+      "learning_rate": 1e-05,
+      "loss": 1.247,
+      "step": 946
+    },
+    {
+      "epoch": 0.4266726740256815,
+      "grad_norm": 1.796391862762361,
+      "learning_rate": 1e-05,
+      "loss": 1.2066,
+      "step": 947
+    },
+    {
+      "epoch": 0.4271232259517909,
+      "grad_norm": 1.8311369964039967,
+      "learning_rate": 1e-05,
+      "loss": 1.2364,
+      "step": 948
+    },
+    {
+      "epoch": 0.4275737778779004,
+      "grad_norm": 1.9450482355011398,
+      "learning_rate": 1e-05,
+      "loss": 1.1701,
+      "step": 949
+    },
+    {
+      "epoch": 0.4280243298040099,
+      "grad_norm": 1.8497392645254098,
+      "learning_rate": 1e-05,
+      "loss": 1.2733,
+      "step": 950
+    },
+    {
+      "epoch": 0.4284748817301194,
+      "grad_norm": 1.9061611409755275,
+      "learning_rate": 1e-05,
+      "loss": 1.2525,
+      "step": 951
+    },
+    {
+      "epoch": 0.42892543365622887,
+      "grad_norm": 1.814663803548855,
+      "learning_rate": 1e-05,
+      "loss": 1.2185,
+      "step": 952
+    },
+    {
+      "epoch": 0.42937598558233836,
+      "grad_norm": 1.7845164252468,
+      "learning_rate": 1e-05,
+      "loss": 1.2948,
+      "step": 953
+    },
+    {
+      "epoch": 0.42982653750844785,
+      "grad_norm": 1.824363044629698,
+      "learning_rate": 1e-05,
+      "loss": 1.2208,
+      "step": 954
+    },
+    {
+      "epoch": 0.43027708943455734,
+      "grad_norm": 1.86807240064721,
+      "learning_rate": 1e-05,
+      "loss": 1.233,
+      "step": 955
+    },
+    {
+      "epoch": 0.4307276413606668,
+      "grad_norm": 2.0124202223012113,
+      "learning_rate": 1e-05,
+      "loss": 1.284,
+      "step": 956
+    },
+    {
+      "epoch": 0.4311781932867763,
+      "grad_norm": 1.6962010066906617,
+      "learning_rate": 1e-05,
+      "loss": 1.2338,
+      "step": 957
+    },
+    {
+      "epoch": 0.4316287452128858,
+      "grad_norm": 1.8014128500101874,
+      "learning_rate": 1e-05,
+      "loss": 1.2282,
+      "step": 958
+    },
+    {
+      "epoch": 0.4320792971389953,
+      "grad_norm": 1.86509842034215,
+      "learning_rate": 1e-05,
+      "loss": 1.2806,
+      "step": 959
+    },
+    {
+      "epoch": 0.4325298490651048,
+      "grad_norm": 1.9463495755610225,
+      "learning_rate": 1e-05,
+      "loss": 1.2613,
+      "step": 960
+    },
+    {
+      "epoch": 0.43298040099121426,
+      "grad_norm": 2.015400634180842,
+      "learning_rate": 1e-05,
+      "loss": 1.2242,
+      "step": 961
+    },
+    {
+      "epoch": 0.4334309529173237,
+      "grad_norm": 1.7459452192089768,
+      "learning_rate": 1e-05,
+      "loss": 1.2269,
+      "step": 962
+    },
+    {
+      "epoch": 0.4338815048434332,
+      "grad_norm": 1.6682578292669217,
+      "learning_rate": 1e-05,
+      "loss": 1.2277,
+      "step": 963
+    },
+    {
+      "epoch": 0.4343320567695427,
+      "grad_norm": 1.8473342106833701,
+      "learning_rate": 1e-05,
+      "loss": 1.2663,
+      "step": 964
+    },
+    {
+      "epoch": 0.43478260869565216,
+      "grad_norm": 1.7929104274279675,
+      "learning_rate": 1e-05,
+      "loss": 1.2174,
+      "step": 965
+    },
+    {
+      "epoch": 0.43523316062176165,
+      "grad_norm": 1.9113226803614998,
+      "learning_rate": 1e-05,
+      "loss": 1.219,
+      "step": 966
+    },
+    {
+      "epoch": 0.43568371254787114,
+      "grad_norm": 1.7820315389432446,
+      "learning_rate": 1e-05,
+      "loss": 1.2093,
+      "step": 967
+    },
+    {
+      "epoch": 0.4361342644739806,
+      "grad_norm": 1.7912651903425194,
+      "learning_rate": 1e-05,
+      "loss": 1.21,
+      "step": 968
+    },
+    {
+      "epoch": 0.4365848164000901,
+      "grad_norm": 1.7021897938320005,
+      "learning_rate": 1e-05,
+      "loss": 1.2658,
+      "step": 969
+    },
+    {
+      "epoch": 0.4370353683261996,
+      "grad_norm": 1.713127972915882,
+      "learning_rate": 1e-05,
+      "loss": 1.2431,
+      "step": 970
+    },
+    {
+      "epoch": 0.4374859202523091,
+      "grad_norm": 1.976456964290821,
+      "learning_rate": 1e-05,
+      "loss": 1.2396,
+      "step": 971
+    },
+    {
+      "epoch": 0.4379364721784186,
+      "grad_norm": 1.690709484557212,
+      "learning_rate": 1e-05,
+      "loss": 1.1973,
+      "step": 972
+    },
+    {
+      "epoch": 0.43838702410452807,
+      "grad_norm": 1.7631626297677152,
+      "learning_rate": 1e-05,
+      "loss": 1.2917,
+      "step": 973
+    },
+    {
+      "epoch": 0.43883757603063756,
+      "grad_norm": 1.9321613627415042,
+      "learning_rate": 1e-05,
+      "loss": 1.2996,
+      "step": 974
+    },
+    {
+      "epoch": 0.439288127956747,
+      "grad_norm": 1.9212148960520163,
+      "learning_rate": 1e-05,
+      "loss": 1.2572,
+      "step": 975
+    },
+    {
+      "epoch": 0.4397386798828565,
+      "grad_norm": 1.8545946670818378,
+      "learning_rate": 1e-05,
+      "loss": 1.2415,
+      "step": 976
+    },
+    {
+      "epoch": 0.44018923180896596,
+      "grad_norm": 1.7357455740514371,
+      "learning_rate": 1e-05,
+      "loss": 1.2113,
+      "step": 977
+    },
+    {
+      "epoch": 0.44063978373507545,
+      "grad_norm": 1.7103721758032753,
+      "learning_rate": 1e-05,
+      "loss": 1.2573,
+      "step": 978
+    },
+    {
+      "epoch": 0.44109033566118494,
+      "grad_norm": 1.621500284952924,
+      "learning_rate": 1e-05,
+      "loss": 1.2004,
+      "step": 979
+    },
+    {
+      "epoch": 0.44154088758729443,
+      "grad_norm": 1.8143231838877647,
+      "learning_rate": 1e-05,
+      "loss": 1.2248,
+      "step": 980
+    },
+    {
+      "epoch": 0.4419914395134039,
+      "grad_norm": 1.817884639012165,
+      "learning_rate": 1e-05,
+      "loss": 1.2326,
+      "step": 981
+    },
+    {
+      "epoch": 0.4424419914395134,
+      "grad_norm": 1.8566661422575435,
+      "learning_rate": 1e-05,
+      "loss": 1.2371,
+      "step": 982
+    },
+    {
+      "epoch": 0.4428925433656229,
+      "grad_norm": 1.5932009456479728,
+      "learning_rate": 1e-05,
+      "loss": 1.243,
+      "step": 983
+    },
+    {
+      "epoch": 0.4433430952917324,
+      "grad_norm": 1.7374272520952616,
+      "learning_rate": 1e-05,
+      "loss": 1.2126,
+      "step": 984
+    },
+    {
+      "epoch": 0.44379364721784187,
+      "grad_norm": 1.8573512532324357,
+      "learning_rate": 1e-05,
+      "loss": 1.2225,
+      "step": 985
+    },
+    {
+      "epoch": 0.44424419914395136,
+      "grad_norm": 1.7423933948751205,
+      "learning_rate": 1e-05,
+      "loss": 1.2409,
+      "step": 986
+    },
+    {
+      "epoch": 0.44469475107006085,
+      "grad_norm": 1.7647060259539527,
+      "learning_rate": 1e-05,
+      "loss": 1.2877,
+      "step": 987
+    },
+    {
+      "epoch": 0.44514530299617033,
+      "grad_norm": 1.748668121574136,
+      "learning_rate": 1e-05,
+      "loss": 1.2248,
+      "step": 988
+    },
+    {
+      "epoch": 0.44559585492227977,
+      "grad_norm": 1.793432386539002,
+      "learning_rate": 1e-05,
+      "loss": 1.2415,
+      "step": 989
+    },
+    {
+      "epoch": 0.44604640684838925,
+      "grad_norm": 1.853915821574754,
+      "learning_rate": 1e-05,
+      "loss": 1.2288,
+      "step": 990
+    },
+    {
+      "epoch": 0.44649695877449874,
+      "grad_norm": 1.8560130103305665,
+      "learning_rate": 1e-05,
+      "loss": 1.1669,
+      "step": 991
+    },
+    {
+      "epoch": 0.44694751070060823,
+      "grad_norm": 1.7161372582429646,
+      "learning_rate": 1e-05,
+      "loss": 1.2325,
+      "step": 992
+    },
+    {
+      "epoch": 0.4473980626267177,
+      "grad_norm": 1.8150571727108074,
+      "learning_rate": 1e-05,
+      "loss": 1.2581,
+      "step": 993
+    },
+    {
+      "epoch": 0.4478486145528272,
+      "grad_norm": 1.7526906399784734,
+      "learning_rate": 1e-05,
+      "loss": 1.203,
+      "step": 994
+    },
+    {
+      "epoch": 0.4482991664789367,
+      "grad_norm": 1.7912002199012331,
+      "learning_rate": 1e-05,
+      "loss": 1.2332,
+      "step": 995
+    },
+    {
+      "epoch": 0.4487497184050462,
+      "grad_norm": 1.8425462851521288,
+      "learning_rate": 1e-05,
+      "loss": 1.227,
+      "step": 996
+    },
+    {
+      "epoch": 0.44920027033115567,
+      "grad_norm": 1.6377109358434168,
+      "learning_rate": 1e-05,
+      "loss": 1.2332,
+      "step": 997
+    },
+    {
+      "epoch": 0.44965082225726516,
+      "grad_norm": 2.0834025665164333,
+      "learning_rate": 1e-05,
+      "loss": 1.2755,
+      "step": 998
+    },
+    {
+      "epoch": 0.45010137418337465,
+      "grad_norm": 1.8628186553040025,
+      "learning_rate": 1e-05,
+      "loss": 1.217,
+      "step": 999
+    },
+    {
+      "epoch": 0.45055192610948414,
+      "grad_norm": 1.6906460456680688,
+      "learning_rate": 1e-05,
+      "loss": 1.2358,
+      "step": 1000
+    },
+    {
+      "epoch": 0.4510024780355936,
+      "grad_norm": 1.933170044940621,
+      "learning_rate": 1e-05,
+      "loss": 1.1983,
+      "step": 1001
+    },
+    {
+      "epoch": 0.4514530299617031,
+      "grad_norm": 1.9063631080079892,
+      "learning_rate": 1e-05,
+      "loss": 1.1888,
+      "step": 1002
+    },
+    {
+      "epoch": 0.45190358188781254,
+      "grad_norm": 2.0736812951315926,
+      "learning_rate": 1e-05,
+      "loss": 1.2588,
+      "step": 1003
+    },
+    {
+      "epoch": 0.45235413381392203,
+      "grad_norm": 1.7951054172338083,
+      "learning_rate": 1e-05,
+      "loss": 1.2045,
+      "step": 1004
+    },
+    {
+      "epoch": 0.4528046857400315,
+      "grad_norm": 1.9804276436292978,
+      "learning_rate": 1e-05,
+      "loss": 1.2217,
+      "step": 1005
+    },
+    {
+      "epoch": 0.453255237666141,
+      "grad_norm": 1.7688903889613266,
+      "learning_rate": 1e-05,
+      "loss": 1.2471,
+      "step": 1006
+    },
+    {
+      "epoch": 0.4537057895922505,
+      "grad_norm": 1.9064149035191789,
+      "learning_rate": 1e-05,
+      "loss": 1.21,
+      "step": 1007
+    },
+    {
+      "epoch": 0.45415634151836,
+      "grad_norm": 1.7500629600876403,
+      "learning_rate": 1e-05,
+      "loss": 1.2002,
+      "step": 1008
+    },
+    {
+      "epoch": 0.4546068934444695,
+      "grad_norm": 1.7935387124911324,
+      "learning_rate": 1e-05,
+      "loss": 1.2256,
+      "step": 1009
+    },
+    {
+      "epoch": 0.45505744537057896,
+      "grad_norm": 1.7428898938844228,
+      "learning_rate": 1e-05,
+      "loss": 1.1975,
+      "step": 1010
+    },
+    {
+      "epoch": 0.45550799729668845,
+      "grad_norm": 1.6232895579332256,
+      "learning_rate": 1e-05,
+      "loss": 1.212,
+      "step": 1011
+    },
+    {
+      "epoch": 0.45595854922279794,
+      "grad_norm": 1.6606790097851276,
+      "learning_rate": 1e-05,
+      "loss": 1.2307,
+      "step": 1012
+    },
+    {
+      "epoch": 0.4564091011489074,
+      "grad_norm": 1.8620402336756303,
+      "learning_rate": 1e-05,
+      "loss": 1.2091,
+      "step": 1013
+    },
+    {
+      "epoch": 0.4568596530750169,
+      "grad_norm": 1.7066063911320382,
+      "learning_rate": 1e-05,
+      "loss": 1.2301,
+      "step": 1014
+    },
+    {
+      "epoch": 0.4573102050011264,
+      "grad_norm": 2.0393744218354337,
+      "learning_rate": 1e-05,
+      "loss": 1.1975,
+      "step": 1015
+    },
+    {
+      "epoch": 0.4577607569272359,
+      "grad_norm": 1.9529652845428371,
+      "learning_rate": 1e-05,
+      "loss": 1.187,
+      "step": 1016
+    },
+    {
+      "epoch": 0.4582113088533453,
+      "grad_norm": 1.7144968408997552,
+      "learning_rate": 1e-05,
+      "loss": 1.278,
+      "step": 1017
+    },
+    {
+      "epoch": 0.4586618607794548,
+      "grad_norm": 1.8255706948280142,
+      "learning_rate": 1e-05,
+      "loss": 1.1724,
+      "step": 1018
+    },
+    {
+      "epoch": 0.4591124127055643,
+      "grad_norm": 1.7648007224714204,
+      "learning_rate": 1e-05,
+      "loss": 1.2123,
+      "step": 1019
+    },
+    {
+      "epoch": 0.4595629646316738,
+      "grad_norm": 1.822346875626478,
+      "learning_rate": 1e-05,
+      "loss": 1.2033,
+      "step": 1020
+    },
+    {
+      "epoch": 0.4600135165577833,
+      "grad_norm": 1.7709385612833495,
+      "learning_rate": 1e-05,
+      "loss": 1.2185,
+      "step": 1021
+    },
+    {
+      "epoch": 0.46046406848389276,
+      "grad_norm": 1.6569703238652924,
+      "learning_rate": 1e-05,
+      "loss": 1.2058,
+      "step": 1022
+    },
+    {
+      "epoch": 0.46091462041000225,
+      "grad_norm": 1.8915853122658959,
+      "learning_rate": 1e-05,
+      "loss": 1.2038,
+      "step": 1023
+    },
+    {
+      "epoch": 0.46136517233611174,
+      "grad_norm": 1.7120760960014065,
+      "learning_rate": 1e-05,
+      "loss": 1.2368,
+      "step": 1024
+    },
+    {
+      "epoch": 0.4618157242622212,
+      "grad_norm": 1.651758686931775,
+      "learning_rate": 1e-05,
+      "loss": 1.2043,
+      "step": 1025
+    },
+    {
+      "epoch": 0.4622662761883307,
+      "grad_norm": 1.837127284255619,
+      "learning_rate": 1e-05,
+      "loss": 1.2651,
+      "step": 1026
+    },
+    {
+      "epoch": 0.4627168281144402,
+      "grad_norm": 1.9652775493658583,
+      "learning_rate": 1e-05,
+      "loss": 1.2607,
+      "step": 1027
+    },
+    {
+      "epoch": 0.4631673800405497,
+      "grad_norm": 1.872687538220321,
+      "learning_rate": 1e-05,
+      "loss": 1.2191,
+      "step": 1028
+    },
+    {
+      "epoch": 0.4636179319666592,
+      "grad_norm": 1.671868424456358,
+      "learning_rate": 1e-05,
+      "loss": 1.2015,
+      "step": 1029
+    },
+    {
+      "epoch": 0.46406848389276867,
+      "grad_norm": 1.846576568317387,
+      "learning_rate": 1e-05,
+      "loss": 1.2195,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4645190358188781,
+      "grad_norm": 1.7524205473229308,
+      "learning_rate": 1e-05,
+      "loss": 1.1984,
+      "step": 1031
+    },
+    {
+      "epoch": 0.4649695877449876,
+      "grad_norm": 1.7582846412885695,
+      "learning_rate": 1e-05,
+      "loss": 1.267,
+      "step": 1032
+    },
+    {
+      "epoch": 0.4654201396710971,
+      "grad_norm": 1.9264915792558468,
+      "learning_rate": 1e-05,
+      "loss": 1.2354,
+      "step": 1033
+    },
+    {
+      "epoch": 0.46587069159720657,
+      "grad_norm": 1.773101550711459,
+      "learning_rate": 1e-05,
+      "loss": 1.2185,
+      "step": 1034
+    },
+    {
+      "epoch": 0.46632124352331605,
+      "grad_norm": 1.8624304126504854,
+      "learning_rate": 1e-05,
+      "loss": 1.2514,
+      "step": 1035
+    },
+    {
+      "epoch": 0.46677179544942554,
+      "grad_norm": 1.858008867057068,
+      "learning_rate": 1e-05,
+      "loss": 1.2676,
+      "step": 1036
+    },
+    {
+      "epoch": 0.46722234737553503,
+      "grad_norm": 1.8245165875314857,
+      "learning_rate": 1e-05,
+      "loss": 1.2499,
+      "step": 1037
+    },
+    {
+      "epoch": 0.4676728993016445,
+      "grad_norm": 1.9287810094534832,
+      "learning_rate": 1e-05,
+      "loss": 1.213,
+      "step": 1038
+    },
+    {
+      "epoch": 0.468123451227754,
+      "grad_norm": 1.9290438800960605,
+      "learning_rate": 1e-05,
+      "loss": 1.2774,
+      "step": 1039
+    },
+    {
+      "epoch": 0.4685740031538635,
+      "grad_norm": 1.89641334900627,
+      "learning_rate": 1e-05,
+      "loss": 1.1682,
+      "step": 1040
+    },
+    {
+      "epoch": 0.469024555079973,
+      "grad_norm": 1.8319785958476764,
+      "learning_rate": 1e-05,
+      "loss": 1.2416,
+      "step": 1041
+    },
+    {
+      "epoch": 0.46947510700608247,
+      "grad_norm": 1.842667663741529,
+      "learning_rate": 1e-05,
+      "loss": 1.173,
+      "step": 1042
+    },
+    {
+      "epoch": 0.46992565893219196,
+      "grad_norm": 1.7958175868902506,
+      "learning_rate": 1e-05,
+      "loss": 1.2279,
+      "step": 1043
+    },
+    {
+      "epoch": 0.47037621085830145,
+      "grad_norm": 1.7208189699628653,
+      "learning_rate": 1e-05,
+      "loss": 1.2433,
+      "step": 1044
+    },
+    {
+      "epoch": 0.4708267627844109,
+      "grad_norm": 1.9064596547765817,
+      "learning_rate": 1e-05,
+      "loss": 1.2874,
+      "step": 1045
+    },
+    {
+      "epoch": 0.47127731471052037,
+      "grad_norm": 1.9590432871635806,
+      "learning_rate": 1e-05,
+      "loss": 1.249,
+      "step": 1046
+    },
+    {
+      "epoch": 0.47172786663662986,
+      "grad_norm": 1.9061918225215113,
+      "learning_rate": 1e-05,
+      "loss": 1.222,
+      "step": 1047
+    },
+    {
+      "epoch": 0.47217841856273934,
+      "grad_norm": 1.8327957069073397,
+      "learning_rate": 1e-05,
+      "loss": 1.2052,
+      "step": 1048
+    },
+    {
+      "epoch": 0.47262897048884883,
+      "grad_norm": 1.889909285426708,
+      "learning_rate": 1e-05,
+      "loss": 1.2188,
+      "step": 1049
+    },
+    {
+      "epoch": 0.4730795224149583,
+      "grad_norm": 1.8888538532149346,
+      "learning_rate": 1e-05,
+      "loss": 1.2571,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4735300743410678,
+      "grad_norm": 1.7364925862968676,
+      "learning_rate": 1e-05,
+      "loss": 1.2166,
+      "step": 1051
+    },
+    {
+      "epoch": 0.4739806262671773,
+      "grad_norm": 1.940114901098188,
+      "learning_rate": 1e-05,
+      "loss": 1.1676,
+      "step": 1052
+    },
+    {
+      "epoch": 0.4744311781932868,
+      "grad_norm": 1.7824750248703818,
+      "learning_rate": 1e-05,
+      "loss": 1.1862,
+      "step": 1053
+    },
+    {
+      "epoch": 0.4748817301193963,
+      "grad_norm": 1.6632088404087289,
+      "learning_rate": 1e-05,
+      "loss": 1.2078,
+      "step": 1054
+    },
+    {
+      "epoch": 0.47533228204550576,
+      "grad_norm": 1.8535270432778796,
+      "learning_rate": 1e-05,
+      "loss": 1.1857,
+      "step": 1055
+    },
+    {
+      "epoch": 0.47578283397161525,
+      "grad_norm": 1.7490217032145572,
+      "learning_rate": 1e-05,
+      "loss": 1.2767,
+      "step": 1056
+    },
+    {
+      "epoch": 0.47623338589772474,
+      "grad_norm": 2.2784924514306497,
+      "learning_rate": 1e-05,
+      "loss": 1.2643,
+      "step": 1057
+    },
+    {
+      "epoch": 0.47668393782383417,
+      "grad_norm": 1.9290938964960718,
+      "learning_rate": 1e-05,
+      "loss": 1.2154,
+      "step": 1058
+    },
+    {
+      "epoch": 0.47713448974994366,
+      "grad_norm": 1.910082105214886,
+      "learning_rate": 1e-05,
+      "loss": 1.2332,
+      "step": 1059
+    },
+    {
+      "epoch": 0.47758504167605315,
+      "grad_norm": 1.884840709045603,
+      "learning_rate": 1e-05,
+      "loss": 1.232,
+      "step": 1060
+    },
+    {
+      "epoch": 0.47803559360216263,
+      "grad_norm": 1.7051054150373703,
+      "learning_rate": 1e-05,
+      "loss": 1.2019,
+      "step": 1061
+    },
+    {
+      "epoch": 0.4784861455282721,
+      "grad_norm": 1.7260457359308374,
+      "learning_rate": 1e-05,
+      "loss": 1.2419,
+      "step": 1062
+    },
+    {
+      "epoch": 0.4789366974543816,
+      "grad_norm": 1.7309120879574524,
+      "learning_rate": 1e-05,
+      "loss": 1.2377,
+      "step": 1063
+    },
+    {
+      "epoch": 0.4793872493804911,
+      "grad_norm": 1.9683552553648596,
+      "learning_rate": 1e-05,
+      "loss": 1.239,
+      "step": 1064
+    },
+    {
+      "epoch": 0.4798378013066006,
+      "grad_norm": 1.901677488014881,
+      "learning_rate": 1e-05,
+      "loss": 1.2173,
+      "step": 1065
+    },
+    {
+      "epoch": 0.4802883532327101,
+      "grad_norm": 1.8308295379837334,
+      "learning_rate": 1e-05,
+      "loss": 1.2389,
+      "step": 1066
+    },
+    {
+      "epoch": 0.48073890515881956,
+      "grad_norm": 1.7820508423112793,
+      "learning_rate": 1e-05,
+      "loss": 1.2076,
+      "step": 1067
+    },
+    {
+      "epoch": 0.48118945708492905,
+      "grad_norm": 1.8331685742222972,
+      "learning_rate": 1e-05,
+      "loss": 1.2598,
+      "step": 1068
+    },
+    {
+      "epoch": 0.48164000901103854,
+      "grad_norm": 1.7856887837426858,
+      "learning_rate": 1e-05,
+      "loss": 1.225,
+      "step": 1069
+    },
+    {
+      "epoch": 0.482090560937148,
+      "grad_norm": 1.9523050077246389,
+      "learning_rate": 1e-05,
+      "loss": 1.2284,
+      "step": 1070
+    },
+    {
+      "epoch": 0.4825411128632575,
+      "grad_norm": 1.7941205953769455,
+      "learning_rate": 1e-05,
+      "loss": 1.2355,
+      "step": 1071
+    },
+    {
+      "epoch": 0.48299166478936695,
+      "grad_norm": 1.977534031281802,
+      "learning_rate": 1e-05,
+      "loss": 1.2914,
+      "step": 1072
+    },
+    {
+      "epoch": 0.48344221671547644,
+      "grad_norm": 1.8023645961903285,
+      "learning_rate": 1e-05,
+      "loss": 1.2393,
+      "step": 1073
+    },
+    {
+      "epoch": 0.4838927686415859,
+      "grad_norm": 1.7954481615528401,
+      "learning_rate": 1e-05,
+      "loss": 1.205,
+      "step": 1074
+    },
+    {
+      "epoch": 0.4843433205676954,
+      "grad_norm": 1.7348221983360275,
+      "learning_rate": 1e-05,
+      "loss": 1.2209,
+      "step": 1075
+    },
+    {
+      "epoch": 0.4847938724938049,
+      "grad_norm": 1.7307336849738606,
+      "learning_rate": 1e-05,
+      "loss": 1.2444,
+      "step": 1076
+    },
+    {
+      "epoch": 0.4852444244199144,
+      "grad_norm": 1.9162431895429781,
+      "learning_rate": 1e-05,
+      "loss": 1.2518,
+      "step": 1077
+    },
+    {
+      "epoch": 0.4856949763460239,
+      "grad_norm": 1.97778966257397,
+      "learning_rate": 1e-05,
+      "loss": 1.2082,
+      "step": 1078
+    },
+    {
+      "epoch": 0.48614552827213336,
+      "grad_norm": 1.699217758383844,
+      "learning_rate": 1e-05,
+      "loss": 1.2253,
+      "step": 1079
+    },
+    {
+      "epoch": 0.48659608019824285,
+      "grad_norm": 1.6883380824264893,
+      "learning_rate": 1e-05,
+      "loss": 1.1954,
+      "step": 1080
+    },
+    {
+      "epoch": 0.48704663212435234,
+      "grad_norm": 1.8429241298457815,
+      "learning_rate": 1e-05,
+      "loss": 1.2117,
+      "step": 1081
+    },
+    {
+      "epoch": 0.48749718405046183,
+      "grad_norm": 1.8325220765257848,
+      "learning_rate": 1e-05,
+      "loss": 1.2305,
+      "step": 1082
+    },
+    {
+      "epoch": 0.4879477359765713,
+      "grad_norm": 1.8472108956272706,
+      "learning_rate": 1e-05,
+      "loss": 1.2733,
+      "step": 1083
+    },
+    {
+      "epoch": 0.4883982879026808,
+      "grad_norm": 1.8369136976527052,
+      "learning_rate": 1e-05,
+      "loss": 1.2339,
+      "step": 1084
+    },
+    {
+      "epoch": 0.4888488398287903,
+      "grad_norm": 1.6162764037207624,
+      "learning_rate": 1e-05,
+      "loss": 1.237,
+      "step": 1085
+    },
+    {
+      "epoch": 0.4892993917548997,
+      "grad_norm": 1.851187210448935,
+      "learning_rate": 1e-05,
+      "loss": 1.2331,
+      "step": 1086
+    },
+    {
+      "epoch": 0.4897499436810092,
+      "grad_norm": 1.8322504715167574,
+      "learning_rate": 1e-05,
+      "loss": 1.2485,
+      "step": 1087
+    },
+    {
+      "epoch": 0.4902004956071187,
+      "grad_norm": 1.9160880478705529,
+      "learning_rate": 1e-05,
+      "loss": 1.2001,
+      "step": 1088
+    },
+    {
+      "epoch": 0.4906510475332282,
+      "grad_norm": 1.8071171427065056,
+      "learning_rate": 1e-05,
+      "loss": 1.2349,
+      "step": 1089
+    },
+    {
+      "epoch": 0.4911015994593377,
+      "grad_norm": 1.622019523490204,
+      "learning_rate": 1e-05,
+      "loss": 1.2269,
+      "step": 1090
+    },
+    {
+      "epoch": 0.49155215138544717,
+      "grad_norm": 1.828774747448718,
+      "learning_rate": 1e-05,
+      "loss": 1.2238,
+      "step": 1091
+    },
+    {
+      "epoch": 0.49200270331155666,
+      "grad_norm": 1.6336547605350142,
+      "learning_rate": 1e-05,
+      "loss": 1.2508,
+      "step": 1092
+    },
+    {
+      "epoch": 0.49245325523766614,
+      "grad_norm": 1.9854826065315856,
+      "learning_rate": 1e-05,
+      "loss": 1.2405,
+      "step": 1093
+    },
+    {
+      "epoch": 0.49290380716377563,
+      "grad_norm": 1.9172141076014846,
+      "learning_rate": 1e-05,
+      "loss": 1.1921,
+      "step": 1094
+    },
+    {
+      "epoch": 0.4933543590898851,
+      "grad_norm": 1.7647119828471034,
+      "learning_rate": 1e-05,
+      "loss": 1.2039,
+      "step": 1095
+    },
+    {
+      "epoch": 0.4938049110159946,
+      "grad_norm": 1.7055858189962692,
+      "learning_rate": 1e-05,
+      "loss": 1.2411,
+      "step": 1096
+    },
+    {
+      "epoch": 0.4942554629421041,
+      "grad_norm": 1.794010258913513,
+      "learning_rate": 1e-05,
+      "loss": 1.2476,
+      "step": 1097
+    },
+    {
+      "epoch": 0.4947060148682136,
+      "grad_norm": 1.6278628378054147,
+      "learning_rate": 1e-05,
+      "loss": 1.2451,
+      "step": 1098
+    },
+    {
+      "epoch": 0.49515656679432307,
+      "grad_norm": 1.8243087009320678,
+      "learning_rate": 1e-05,
+      "loss": 1.2588,
+      "step": 1099
+    },
+    {
+      "epoch": 0.4956071187204325,
+      "grad_norm": 1.6759172576522992,
+      "learning_rate": 1e-05,
+      "loss": 1.2367,
+      "step": 1100
+    },
+    {
+      "epoch": 0.496057670646542,
+      "grad_norm": 1.7807547764841059,
+      "learning_rate": 1e-05,
+      "loss": 1.2052,
+      "step": 1101
+    },
+    {
+      "epoch": 0.4965082225726515,
+      "grad_norm": 1.7560604438541885,
+      "learning_rate": 1e-05,
+      "loss": 1.2343,
+      "step": 1102
+    },
+    {
+      "epoch": 0.49695877449876097,
+      "grad_norm": 1.6398485896511423,
+      "learning_rate": 1e-05,
+      "loss": 1.1882,
+      "step": 1103
+    },
+    {
+      "epoch": 0.49740932642487046,
+      "grad_norm": 1.6943249402448588,
+      "learning_rate": 1e-05,
+      "loss": 1.2074,
+      "step": 1104
+    },
+    {
+      "epoch": 0.49785987835097995,
+      "grad_norm": 1.6998993181668858,
+      "learning_rate": 1e-05,
+      "loss": 1.2035,
+      "step": 1105
+    },
+    {
+      "epoch": 0.49831043027708943,
+      "grad_norm": 1.7167357490770654,
+      "learning_rate": 1e-05,
+      "loss": 1.1755,
+      "step": 1106
+    },
+    {
+      "epoch": 0.4987609822031989,
+      "grad_norm": 1.856023947993202,
+      "learning_rate": 1e-05,
+      "loss": 1.173,
+      "step": 1107
+    },
+    {
+      "epoch": 0.4992115341293084,
+      "grad_norm": 1.7209075128988316,
+      "learning_rate": 1e-05,
+      "loss": 1.231,
+      "step": 1108
+    },
+    {
+      "epoch": 0.4996620860554179,
+      "grad_norm": 1.7321911152152958,
+      "learning_rate": 1e-05,
+      "loss": 1.2232,
+      "step": 1109
+    },
+    {
+      "epoch": 0.5001126379815274,
+      "grad_norm": 1.834703729970928,
+      "learning_rate": 1e-05,
+      "loss": 1.2282,
+      "step": 1110
+    },
+    {
+      "epoch": 0.5005631899076368,
+      "grad_norm": 1.703690752117525,
+      "learning_rate": 1e-05,
+      "loss": 1.2261,
+      "step": 1111
+    },
+    {
+      "epoch": 0.5010137418337464,
+      "grad_norm": 1.8971254709341026,
+      "learning_rate": 1e-05,
+      "loss": 1.2252,
+      "step": 1112
+    },
+    {
+      "epoch": 0.5014642937598558,
+      "grad_norm": 1.6885863371613574,
+      "learning_rate": 1e-05,
+      "loss": 1.1882,
+      "step": 1113
+    },
+    {
+      "epoch": 0.5019148456859653,
+      "grad_norm": 1.7867795820634051,
+      "learning_rate": 1e-05,
+      "loss": 1.2161,
+      "step": 1114
+    },
+    {
+      "epoch": 0.5023653976120748,
+      "grad_norm": 1.7866472158501792,
+      "learning_rate": 1e-05,
+      "loss": 1.2407,
+      "step": 1115
+    },
+    {
+      "epoch": 0.5028159495381843,
+      "grad_norm": 1.7145118086058329,
+      "learning_rate": 1e-05,
+      "loss": 1.2003,
+      "step": 1116
+    },
+    {
+      "epoch": 0.5032665014642937,
+      "grad_norm": 1.8420656895369183,
+      "learning_rate": 1e-05,
+      "loss": 1.245,
+      "step": 1117
+    },
+    {
+      "epoch": 0.5037170533904033,
+      "grad_norm": 1.8881987897318155,
+      "learning_rate": 1e-05,
+      "loss": 1.2149,
+      "step": 1118
+    },
+    {
+      "epoch": 0.5041676053165127,
+      "grad_norm": 1.662823224734846,
+      "learning_rate": 1e-05,
+      "loss": 1.2788,
+      "step": 1119
+    },
+    {
+      "epoch": 0.5046181572426223,
+      "grad_norm": 1.6455901767468586,
+      "learning_rate": 1e-05,
+      "loss": 1.2202,
+      "step": 1120
+    },
+    {
+      "epoch": 0.5050687091687317,
+      "grad_norm": 1.6179584810828676,
+      "learning_rate": 1e-05,
+      "loss": 1.2427,
+      "step": 1121
+    },
+    {
+      "epoch": 0.5055192610948411,
+      "grad_norm": 1.7161125239457964,
+      "learning_rate": 1e-05,
+      "loss": 1.2381,
+      "step": 1122
+    },
+    {
+      "epoch": 0.5059698130209507,
+      "grad_norm": 1.7815519260922321,
+      "learning_rate": 1e-05,
+      "loss": 1.2181,
+      "step": 1123
+    },
+    {
+      "epoch": 0.5064203649470601,
+      "grad_norm": 1.8429736829187442,
+      "learning_rate": 1e-05,
+      "loss": 1.2138,
+      "step": 1124
+    },
+    {
+      "epoch": 0.5068709168731697,
+      "grad_norm": 1.7426923228641775,
+      "learning_rate": 1e-05,
+      "loss": 1.2327,
+      "step": 1125
+    },
+    {
+      "epoch": 0.5073214687992791,
+      "grad_norm": 1.8345535027242015,
+      "learning_rate": 1e-05,
+      "loss": 1.184,
+      "step": 1126
+    },
+    {
+      "epoch": 0.5077720207253886,
+      "grad_norm": 1.7373886015772988,
+      "learning_rate": 1e-05,
+      "loss": 1.2015,
+      "step": 1127
+    },
+    {
+      "epoch": 0.5082225726514981,
+      "grad_norm": 1.7368499412838856,
+      "learning_rate": 1e-05,
+      "loss": 1.2256,
+      "step": 1128
+    },
+    {
+      "epoch": 0.5086731245776076,
+      "grad_norm": 1.666050196887123,
+      "learning_rate": 1e-05,
+      "loss": 1.1866,
+      "step": 1129
+    },
+    {
+      "epoch": 0.509123676503717,
+      "grad_norm": 1.7223799725798914,
+      "learning_rate": 1e-05,
+      "loss": 1.2122,
+      "step": 1130
+    },
+    {
+      "epoch": 0.5095742284298266,
+      "grad_norm": 1.784934201113948,
+      "learning_rate": 1e-05,
+      "loss": 1.1619,
+      "step": 1131
+    },
+    {
+      "epoch": 0.510024780355936,
+      "grad_norm": 1.7354950600530021,
+      "learning_rate": 1e-05,
+      "loss": 1.2096,
+      "step": 1132
+    },
+    {
+      "epoch": 0.5104753322820456,
+      "grad_norm": 1.7779168631616875,
+      "learning_rate": 1e-05,
+      "loss": 1.2107,
+      "step": 1133
+    },
+    {
+      "epoch": 0.510925884208155,
+      "grad_norm": 1.8849253874183658,
+      "learning_rate": 1e-05,
+      "loss": 1.2027,
+      "step": 1134
+    },
+    {
+      "epoch": 0.5113764361342644,
+      "grad_norm": 1.8102302267850547,
+      "learning_rate": 1e-05,
+      "loss": 1.2296,
+      "step": 1135
+    },
+    {
+      "epoch": 0.511826988060374,
+      "grad_norm": 1.7416638130497022,
+      "learning_rate": 1e-05,
+      "loss": 1.2782,
+      "step": 1136
+    },
+    {
+      "epoch": 0.5122775399864834,
+      "grad_norm": 1.945864303499097,
+      "learning_rate": 1e-05,
+      "loss": 1.2051,
+      "step": 1137
+    },
+    {
+      "epoch": 0.5127280919125929,
+      "grad_norm": 1.7873013100729083,
+      "learning_rate": 1e-05,
+      "loss": 1.2439,
+      "step": 1138
+    },
+    {
+      "epoch": 0.5131786438387024,
+      "grad_norm": 1.7613293131158412,
+      "learning_rate": 1e-05,
+      "loss": 1.2109,
+      "step": 1139
+    },
+    {
+      "epoch": 0.5136291957648119,
+      "grad_norm": 1.6975318487518078,
+      "learning_rate": 1e-05,
+      "loss": 1.2473,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5140797476909214,
+      "grad_norm": 1.7674609411831723,
+      "learning_rate": 1e-05,
+      "loss": 1.239,
+      "step": 1141
+    },
+    {
+      "epoch": 0.5145302996170309,
+      "grad_norm": 1.7714493294198033,
+      "learning_rate": 1e-05,
+      "loss": 1.2214,
+      "step": 1142
+    },
+    {
+      "epoch": 0.5149808515431403,
+      "grad_norm": 1.9020951604324543,
+      "learning_rate": 1e-05,
+      "loss": 1.2285,
+      "step": 1143
+    },
+    {
+      "epoch": 0.5154314034692499,
+      "grad_norm": 1.7262916022035897,
+      "learning_rate": 1e-05,
+      "loss": 1.1955,
+      "step": 1144
+    },
+    {
+      "epoch": 0.5158819553953593,
+      "grad_norm": 1.8065375077357888,
+      "learning_rate": 1e-05,
+      "loss": 1.2188,
+      "step": 1145
+    },
+    {
+      "epoch": 0.5163325073214688,
+      "grad_norm": 1.9000422236133612,
+      "learning_rate": 1e-05,
+      "loss": 1.2332,
+      "step": 1146
+    },
+    {
+      "epoch": 0.5167830592475783,
+      "grad_norm": 1.820127159138276,
+      "learning_rate": 1e-05,
+      "loss": 1.1912,
+      "step": 1147
+    },
+    {
+      "epoch": 0.5172336111736877,
+      "grad_norm": 2.0478804185280914,
+      "learning_rate": 1e-05,
+      "loss": 1.1894,
+      "step": 1148
+    },
+    {
+      "epoch": 0.5176841630997973,
+      "grad_norm": 1.6483893814399748,
+      "learning_rate": 1e-05,
+      "loss": 1.2214,
+      "step": 1149
+    },
+    {
+      "epoch": 0.5181347150259067,
+      "grad_norm": 1.8585701430807764,
+      "learning_rate": 1e-05,
+      "loss": 1.2955,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5185852669520162,
+      "grad_norm": 1.7839142062682194,
+      "learning_rate": 1e-05,
+      "loss": 1.2145,
+      "step": 1151
+    },
+    {
+      "epoch": 0.5190358188781257,
+      "grad_norm": 1.8390477603191218,
+      "learning_rate": 1e-05,
+      "loss": 1.2206,
+      "step": 1152
+    },
+    {
+      "epoch": 0.5194863708042352,
+      "grad_norm": 1.8242871766456614,
+      "learning_rate": 1e-05,
+      "loss": 1.2778,
+      "step": 1153
+    },
+    {
+      "epoch": 0.5199369227303446,
+      "grad_norm": 2.0086407164587268,
+      "learning_rate": 1e-05,
+      "loss": 1.2485,
+      "step": 1154
+    },
+    {
+      "epoch": 0.5203874746564542,
+      "grad_norm": 1.6880513466993963,
+      "learning_rate": 1e-05,
+      "loss": 1.251,
+      "step": 1155
+    },
+    {
+      "epoch": 0.5208380265825636,
+      "grad_norm": 1.7131540209069966,
+      "learning_rate": 1e-05,
+      "loss": 1.2352,
+      "step": 1156
+    },
+    {
+      "epoch": 0.5212885785086732,
+      "grad_norm": 1.77424723685144,
+      "learning_rate": 1e-05,
+      "loss": 1.196,
+      "step": 1157
+    },
+    {
+      "epoch": 0.5217391304347826,
+      "grad_norm": 1.694871403898286,
+      "learning_rate": 1e-05,
+      "loss": 1.1821,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5221896823608921,
+      "grad_norm": 1.5785506005919077,
+      "learning_rate": 1e-05,
+      "loss": 1.218,
+      "step": 1159
+    },
+    {
+      "epoch": 0.5226402342870016,
+      "grad_norm": 1.8796378901257582,
+      "learning_rate": 1e-05,
+      "loss": 1.191,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5230907862131111,
+      "grad_norm": 1.7924233055693763,
+      "learning_rate": 1e-05,
+      "loss": 1.2295,
+      "step": 1161
+    },
+    {
+      "epoch": 0.5235413381392205,
+      "grad_norm": 1.7781927608468,
+      "learning_rate": 1e-05,
+      "loss": 1.2309,
+      "step": 1162
+    },
+    {
+      "epoch": 0.52399189006533,
+      "grad_norm": 1.799427621383034,
+      "learning_rate": 1e-05,
+      "loss": 1.1959,
+      "step": 1163
+    },
+    {
+      "epoch": 0.5244424419914395,
+      "grad_norm": 1.8834615193453594,
+      "learning_rate": 1e-05,
+      "loss": 1.2077,
+      "step": 1164
+    },
+    {
+      "epoch": 0.524892993917549,
+      "grad_norm": 1.8016925557595358,
+      "learning_rate": 1e-05,
+      "loss": 1.1745,
+      "step": 1165
+    },
+    {
+      "epoch": 0.5253435458436585,
+      "grad_norm": 1.705402400130282,
+      "learning_rate": 1e-05,
+      "loss": 1.227,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5257940977697679,
+      "grad_norm": 1.7686687798590577,
+      "learning_rate": 1e-05,
+      "loss": 1.2233,
+      "step": 1167
+    },
+    {
+      "epoch": 0.5262446496958775,
+      "grad_norm": 1.658258877352761,
+      "learning_rate": 1e-05,
+      "loss": 1.2014,
+      "step": 1168
+    },
+    {
+      "epoch": 0.5266952016219869,
+      "grad_norm": 1.8191890012375032,
+      "learning_rate": 1e-05,
+      "loss": 1.2677,
+      "step": 1169
+    },
+    {
+      "epoch": 0.5271457535480965,
+      "grad_norm": 1.7701142620729842,
+      "learning_rate": 1e-05,
+      "loss": 1.214,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5275963054742059,
+      "grad_norm": 1.8133893275054933,
+      "learning_rate": 1e-05,
+      "loss": 1.2402,
+      "step": 1171
+    },
+    {
+      "epoch": 0.5280468574003154,
+      "grad_norm": 1.740931963668351,
+      "learning_rate": 1e-05,
+      "loss": 1.1761,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5284974093264249,
+      "grad_norm": 1.7818969524312216,
+      "learning_rate": 1e-05,
+      "loss": 1.2553,
+      "step": 1173
+    },
+    {
+      "epoch": 0.5289479612525344,
+      "grad_norm": 1.7237927032769915,
+      "learning_rate": 1e-05,
+      "loss": 1.2491,
+      "step": 1174
+    },
+    {
+      "epoch": 0.5293985131786438,
+      "grad_norm": 1.7559928322729434,
+      "learning_rate": 1e-05,
+      "loss": 1.2524,
+      "step": 1175
+    },
+    {
+      "epoch": 0.5298490651047533,
+      "grad_norm": 1.832205300455528,
+      "learning_rate": 1e-05,
+      "loss": 1.2073,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5302996170308628,
+      "grad_norm": 2.0093659961338277,
+      "learning_rate": 1e-05,
+      "loss": 1.2001,
+      "step": 1177
+    },
+    {
+      "epoch": 0.5307501689569722,
+      "grad_norm": 1.9178053907781756,
+      "learning_rate": 1e-05,
+      "loss": 1.1835,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5312007208830818,
+      "grad_norm": 1.7443974234038635,
+      "learning_rate": 1e-05,
+      "loss": 1.1814,
+      "step": 1179
+    },
+    {
+      "epoch": 0.5316512728091912,
+      "grad_norm": 1.9130309444097395,
+      "learning_rate": 1e-05,
+      "loss": 1.2044,
+      "step": 1180
+    },
+    {
+      "epoch": 0.5321018247353008,
+      "grad_norm": 2.0645138856559093,
+      "learning_rate": 1e-05,
+      "loss": 1.2481,
+      "step": 1181
+    },
+    {
+      "epoch": 0.5325523766614102,
+      "grad_norm": 1.9165903238442665,
+      "learning_rate": 1e-05,
+      "loss": 1.2731,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5330029285875197,
+      "grad_norm": 1.7745230817996194,
+      "learning_rate": 1e-05,
+      "loss": 1.2351,
+      "step": 1183
+    },
+    {
+      "epoch": 0.5334534805136292,
+      "grad_norm": 1.779111603058722,
+      "learning_rate": 1e-05,
+      "loss": 1.2212,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5339040324397387,
+      "grad_norm": 1.7484601964668258,
+      "learning_rate": 1e-05,
+      "loss": 1.192,
+      "step": 1185
+    },
+    {
+      "epoch": 0.5343545843658482,
+      "grad_norm": 1.6836885724980257,
+      "learning_rate": 1e-05,
+      "loss": 1.2735,
+      "step": 1186
+    },
+    {
+      "epoch": 0.5348051362919577,
+      "grad_norm": 1.8114416538391223,
+      "learning_rate": 1e-05,
+      "loss": 1.2181,
+      "step": 1187
+    },
+    {
+      "epoch": 0.5352556882180671,
+      "grad_norm": 1.8448999435632383,
+      "learning_rate": 1e-05,
+      "loss": 1.2197,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5357062401441767,
+      "grad_norm": 1.9435339007338701,
+      "learning_rate": 1e-05,
+      "loss": 1.2194,
+      "step": 1189
+    },
+    {
+      "epoch": 0.5361567920702861,
+      "grad_norm": 1.8025109235583145,
+      "learning_rate": 1e-05,
+      "loss": 1.2042,
+      "step": 1190
+    },
+    {
+      "epoch": 0.5366073439963955,
+      "grad_norm": 1.6700090053309786,
+      "learning_rate": 1e-05,
+      "loss": 1.2033,
+      "step": 1191
+    },
+    {
+      "epoch": 0.5370578959225051,
+      "grad_norm": 1.6849205350953573,
+      "learning_rate": 1e-05,
+      "loss": 1.2215,
+      "step": 1192
+    },
+    {
+      "epoch": 0.5375084478486145,
+      "grad_norm": 1.7147161960304929,
+      "learning_rate": 1e-05,
+      "loss": 1.2341,
+      "step": 1193
+    },
+    {
+      "epoch": 0.5379589997747241,
+      "grad_norm": 1.7746421868361555,
+      "learning_rate": 1e-05,
+      "loss": 1.1747,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5384095517008335,
+      "grad_norm": 1.8599995128441753,
+      "learning_rate": 1e-05,
+      "loss": 1.2487,
+      "step": 1195
+    },
+    {
+      "epoch": 0.538860103626943,
+      "grad_norm": 1.7399761480508251,
+      "learning_rate": 1e-05,
+      "loss": 1.1832,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5393106555530525,
+      "grad_norm": 1.799614681377465,
+      "learning_rate": 1e-05,
+      "loss": 1.1554,
+      "step": 1197
+    },
+    {
+      "epoch": 0.539761207479162,
+      "grad_norm": 1.892989010192148,
+      "learning_rate": 1e-05,
+      "loss": 1.2031,
+      "step": 1198
+    },
+    {
+      "epoch": 0.5402117594052714,
+      "grad_norm": 1.7929467713485536,
+      "learning_rate": 1e-05,
+      "loss": 1.202,
+      "step": 1199
+    },
+    {
+      "epoch": 0.540662311331381,
+      "grad_norm": 1.655986074624804,
+      "learning_rate": 1e-05,
+      "loss": 1.2022,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5411128632574904,
+      "grad_norm": 1.8235445374371355,
+      "learning_rate": 1e-05,
+      "loss": 1.2826,
+      "step": 1201
+    },
+    {
+      "epoch": 0.5415634151836,
+      "grad_norm": 1.6988588552090702,
+      "learning_rate": 1e-05,
+      "loss": 1.2316,
+      "step": 1202
+    },
+    {
+      "epoch": 0.5420139671097094,
+      "grad_norm": 1.8175071133701286,
+      "learning_rate": 1e-05,
+      "loss": 1.1956,
+      "step": 1203
+    },
+    {
+      "epoch": 0.5424645190358188,
+      "grad_norm": 1.8050236092770742,
+      "learning_rate": 1e-05,
+      "loss": 1.2068,
+      "step": 1204
+    },
+    {
+      "epoch": 0.5429150709619284,
+      "grad_norm": 1.869406592824594,
+      "learning_rate": 1e-05,
+      "loss": 1.2322,
+      "step": 1205
+    },
+    {
+      "epoch": 0.5433656228880378,
+      "grad_norm": 1.7853907899983779,
+      "learning_rate": 1e-05,
+      "loss": 1.2501,
+      "step": 1206
+    },
+    {
+      "epoch": 0.5438161748141473,
+      "grad_norm": 1.8810947527108874,
+      "learning_rate": 1e-05,
+      "loss": 1.1695,
+      "step": 1207
+    },
+    {
+      "epoch": 0.5442667267402568,
+      "grad_norm": 1.945647569363387,
+      "learning_rate": 1e-05,
+      "loss": 1.1909,
+      "step": 1208
+    },
+    {
+      "epoch": 0.5447172786663663,
+      "grad_norm": 1.9087443366683496,
+      "learning_rate": 1e-05,
+      "loss": 1.2084,
+      "step": 1209
+    },
+    {
+      "epoch": 0.5451678305924758,
+      "grad_norm": 1.8357196765209838,
+      "learning_rate": 1e-05,
+      "loss": 1.2312,
+      "step": 1210
+    },
+    {
+      "epoch": 0.5456183825185853,
+      "grad_norm": 1.8138651667673715,
+      "learning_rate": 1e-05,
+      "loss": 1.2034,
+      "step": 1211
+    },
+    {
+      "epoch": 0.5460689344446947,
+      "grad_norm": 1.8099605933670535,
+      "learning_rate": 1e-05,
+      "loss": 1.2638,
+      "step": 1212
+    },
+    {
+      "epoch": 0.5465194863708043,
+      "grad_norm": 1.773024385547567,
+      "learning_rate": 1e-05,
+      "loss": 1.1959,
+      "step": 1213
+    },
+    {
+      "epoch": 0.5469700382969137,
+      "grad_norm": 1.7383391718814192,
+      "learning_rate": 1e-05,
+      "loss": 1.2083,
+      "step": 1214
+    },
+    {
+      "epoch": 0.5474205902230233,
+      "grad_norm": 1.9103185288324496,
+      "learning_rate": 1e-05,
+      "loss": 1.1993,
+      "step": 1215
+    },
+    {
+      "epoch": 0.5478711421491327,
+      "grad_norm": 1.700385208833561,
+      "learning_rate": 1e-05,
+      "loss": 1.2476,
+      "step": 1216
+    },
+    {
+      "epoch": 0.5483216940752421,
+      "grad_norm": 1.6383806331975956,
+      "learning_rate": 1e-05,
+      "loss": 1.2314,
+      "step": 1217
+    },
+    {
+      "epoch": 0.5487722460013517,
+      "grad_norm": 1.6895621880809684,
+      "learning_rate": 1e-05,
+      "loss": 1.2214,
+      "step": 1218
+    },
+    {
+      "epoch": 0.5492227979274611,
+      "grad_norm": 1.7873286157278356,
+      "learning_rate": 1e-05,
+      "loss": 1.2299,
+      "step": 1219
+    },
+    {
+      "epoch": 0.5496733498535706,
+      "grad_norm": 1.930894859148649,
+      "learning_rate": 1e-05,
+      "loss": 1.2487,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5501239017796801,
+      "grad_norm": 2.0135423824053675,
+      "learning_rate": 1e-05,
+      "loss": 1.2151,
+      "step": 1221
+    },
+    {
+      "epoch": 0.5505744537057896,
+      "grad_norm": 1.6520001512756237,
+      "learning_rate": 1e-05,
+      "loss": 1.2274,
+      "step": 1222
+    },
+    {
+      "epoch": 0.551025005631899,
+      "grad_norm": 1.7675149095183413,
+      "learning_rate": 1e-05,
+      "loss": 1.1628,
+      "step": 1223
+    },
+    {
+      "epoch": 0.5514755575580086,
+      "grad_norm": 1.670533487591758,
+      "learning_rate": 1e-05,
+      "loss": 1.2193,
+      "step": 1224
+    },
+    {
+      "epoch": 0.551926109484118,
+      "grad_norm": 1.7402222912391112,
+      "learning_rate": 1e-05,
+      "loss": 1.1725,
+      "step": 1225
+    },
+    {
+      "epoch": 0.5523766614102276,
+      "grad_norm": 1.8433878045228402,
+      "learning_rate": 1e-05,
+      "loss": 1.2262,
+      "step": 1226
+    },
+    {
+      "epoch": 0.552827213336337,
+      "grad_norm": 1.7083694572915074,
+      "learning_rate": 1e-05,
+      "loss": 1.1985,
+      "step": 1227
+    },
+    {
+      "epoch": 0.5532777652624465,
+      "grad_norm": 1.749289550207512,
+      "learning_rate": 1e-05,
+      "loss": 1.1765,
+      "step": 1228
+    },
+    {
+      "epoch": 0.553728317188556,
+      "grad_norm": 1.790532022586267,
+      "learning_rate": 1e-05,
+      "loss": 1.2446,
+      "step": 1229
+    },
+    {
+      "epoch": 0.5541788691146655,
+      "grad_norm": 1.8473192179714735,
+      "learning_rate": 1e-05,
+      "loss": 1.2218,
+      "step": 1230
+    },
+    {
+      "epoch": 0.554629421040775,
+      "grad_norm": 1.765691891742622,
+      "learning_rate": 1e-05,
+      "loss": 1.2279,
+      "step": 1231
+    },
+    {
+      "epoch": 0.5550799729668844,
+      "grad_norm": 1.7761654789595203,
+      "learning_rate": 1e-05,
+      "loss": 1.2273,
+      "step": 1232
+    },
+    {
+      "epoch": 0.5555305248929939,
+      "grad_norm": 1.808037518854502,
+      "learning_rate": 1e-05,
+      "loss": 1.2546,
+      "step": 1233
+    },
+    {
+      "epoch": 0.5559810768191034,
+      "grad_norm": 1.7323127037119805,
+      "learning_rate": 1e-05,
+      "loss": 1.251,
+      "step": 1234
+    },
+    {
+      "epoch": 0.5564316287452129,
+      "grad_norm": 1.7872621678328413,
+      "learning_rate": 1e-05,
+      "loss": 1.2239,
+      "step": 1235
+    },
+    {
+      "epoch": 0.5568821806713223,
+      "grad_norm": 1.5816400521913458,
+      "learning_rate": 1e-05,
+      "loss": 1.2246,
+      "step": 1236
+    },
+    {
+      "epoch": 0.5573327325974319,
+      "grad_norm": 1.78128066934941,
+      "learning_rate": 1e-05,
+      "loss": 1.214,
+      "step": 1237
+    },
+    {
+      "epoch": 0.5577832845235413,
+      "grad_norm": 1.6569491612432181,
+      "learning_rate": 1e-05,
+      "loss": 1.2065,
+      "step": 1238
+    },
+    {
+      "epoch": 0.5582338364496509,
+      "grad_norm": 1.8776190258672327,
+      "learning_rate": 1e-05,
+      "loss": 1.1834,
+      "step": 1239
+    },
+    {
+      "epoch": 0.5586843883757603,
+      "grad_norm": 1.7696075196857524,
+      "learning_rate": 1e-05,
+      "loss": 1.2107,
+      "step": 1240
+    },
+    {
+      "epoch": 0.5591349403018698,
+      "grad_norm": 1.729313440288939,
+      "learning_rate": 1e-05,
+      "loss": 1.2318,
+      "step": 1241
+    },
+    {
+      "epoch": 0.5595854922279793,
+      "grad_norm": 1.6447216627865269,
+      "learning_rate": 1e-05,
+      "loss": 1.1784,
+      "step": 1242
+    },
+    {
+      "epoch": 0.5600360441540888,
+      "grad_norm": 1.6745547082296826,
+      "learning_rate": 1e-05,
+      "loss": 1.1845,
+      "step": 1243
+    },
+    {
+      "epoch": 0.5604865960801982,
+      "grad_norm": 1.8037442589388843,
+      "learning_rate": 1e-05,
+      "loss": 1.1784,
+      "step": 1244
+    },
+    {
+      "epoch": 0.5609371480063077,
+      "grad_norm": 1.82565372827476,
+      "learning_rate": 1e-05,
+      "loss": 1.2405,
+      "step": 1245
+    },
+    {
+      "epoch": 0.5613876999324172,
+      "grad_norm": 1.8098475115447121,
+      "learning_rate": 1e-05,
+      "loss": 1.2163,
+      "step": 1246
+    },
+    {
+      "epoch": 0.5618382518585266,
+      "grad_norm": 1.9220636481752578,
+      "learning_rate": 1e-05,
+      "loss": 1.2288,
+      "step": 1247
+    },
+    {
+      "epoch": 0.5622888037846362,
+      "grad_norm": 1.710731088857779,
+      "learning_rate": 1e-05,
+      "loss": 1.2183,
+      "step": 1248
+    },
+    {
+      "epoch": 0.5627393557107456,
+      "grad_norm": 1.8052982521525502,
+      "learning_rate": 1e-05,
+      "loss": 1.1942,
+      "step": 1249
+    },
+    {
+      "epoch": 0.5631899076368552,
+      "grad_norm": 1.640205511214467,
+      "learning_rate": 1e-05,
+      "loss": 1.2008,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5636404595629646,
+      "grad_norm": 2.0393672038821564,
+      "learning_rate": 1e-05,
+      "loss": 1.2977,
+      "step": 1251
+    },
+    {
+      "epoch": 0.5640910114890741,
+      "grad_norm": 1.8026799358887735,
+      "learning_rate": 1e-05,
+      "loss": 1.1981,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5645415634151836,
+      "grad_norm": 1.718751835030151,
+      "learning_rate": 1e-05,
+      "loss": 1.2208,
+      "step": 1253
+    },
+    {
+      "epoch": 0.5649921153412931,
+      "grad_norm": 1.6918767310612728,
+      "learning_rate": 1e-05,
+      "loss": 1.2205,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5654426672674026,
+      "grad_norm": 1.6717442792677766,
+      "learning_rate": 1e-05,
+      "loss": 1.1739,
+      "step": 1255
+    },
+    {
+      "epoch": 0.5658932191935121,
+      "grad_norm": 1.961048303687513,
+      "learning_rate": 1e-05,
+      "loss": 1.206,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5663437711196215,
+      "grad_norm": 1.8056887914479824,
+      "learning_rate": 1e-05,
+      "loss": 1.2089,
+      "step": 1257
+    },
+    {
+      "epoch": 0.5667943230457311,
+      "grad_norm": 1.6984295897919814,
+      "learning_rate": 1e-05,
+      "loss": 1.2279,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5672448749718405,
+      "grad_norm": 1.7622743101852951,
+      "learning_rate": 1e-05,
+      "loss": 1.1912,
+      "step": 1259
+    },
+    {
+      "epoch": 0.5676954268979499,
+      "grad_norm": 1.583640334650597,
+      "learning_rate": 1e-05,
+      "loss": 1.2112,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5681459788240595,
+      "grad_norm": 1.6079332863360147,
+      "learning_rate": 1e-05,
+      "loss": 1.1877,
+      "step": 1261
+    },
+    {
+      "epoch": 0.5685965307501689,
+      "grad_norm": 1.6952757868493882,
+      "learning_rate": 1e-05,
+      "loss": 1.1768,
+      "step": 1262
+    },
+    {
+      "epoch": 0.5690470826762785,
+      "grad_norm": 1.668674496114196,
+      "learning_rate": 1e-05,
+      "loss": 1.174,
+      "step": 1263
+    },
+    {
+      "epoch": 0.5694976346023879,
+      "grad_norm": 1.9683106986964076,
+      "learning_rate": 1e-05,
+      "loss": 1.229,
+      "step": 1264
+    },
+    {
+      "epoch": 0.5699481865284974,
+      "grad_norm": 1.739019753022858,
+      "learning_rate": 1e-05,
+      "loss": 1.17,
+      "step": 1265
+    },
+    {
+      "epoch": 0.5703987384546069,
+      "grad_norm": 1.692161131470966,
+      "learning_rate": 1e-05,
+      "loss": 1.1942,
+      "step": 1266
+    },
+    {
+      "epoch": 0.5708492903807164,
+      "grad_norm": 1.7293550537323032,
+      "learning_rate": 1e-05,
+      "loss": 1.1882,
+      "step": 1267
+    },
+    {
+      "epoch": 0.5712998423068258,
+      "grad_norm": 1.7932383980869326,
+      "learning_rate": 1e-05,
+      "loss": 1.2103,
+      "step": 1268
+    },
+    {
+      "epoch": 0.5717503942329354,
+      "grad_norm": 1.7337813641116824,
+      "learning_rate": 1e-05,
+      "loss": 1.1794,
+      "step": 1269
+    },
+    {
+      "epoch": 0.5722009461590448,
+      "grad_norm": 1.7602788271157532,
+      "learning_rate": 1e-05,
+      "loss": 1.2374,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5726514980851544,
+      "grad_norm": 1.663773026864633,
+      "learning_rate": 1e-05,
+      "loss": 1.2071,
+      "step": 1271
+    },
+    {
+      "epoch": 0.5731020500112638,
+      "grad_norm": 1.806661906427446,
+      "learning_rate": 1e-05,
+      "loss": 1.1879,
+      "step": 1272
+    },
+    {
+      "epoch": 0.5735526019373732,
+      "grad_norm": 1.722384599755511,
+      "learning_rate": 1e-05,
+      "loss": 1.1938,
+      "step": 1273
+    },
+    {
+      "epoch": 0.5740031538634828,
+      "grad_norm": 1.8231291007726889,
+      "learning_rate": 1e-05,
+      "loss": 1.2121,
+      "step": 1274
+    },
+    {
+      "epoch": 0.5744537057895922,
+      "grad_norm": 1.8167541605452744,
+      "learning_rate": 1e-05,
+      "loss": 1.2475,
+      "step": 1275
+    },
+    {
+      "epoch": 0.5749042577157017,
+      "grad_norm": 1.7818831527774315,
+      "learning_rate": 1e-05,
+      "loss": 1.2115,
+      "step": 1276
+    },
+    {
+      "epoch": 0.5753548096418112,
+      "grad_norm": 1.6617095262867325,
+      "learning_rate": 1e-05,
+      "loss": 1.227,
+      "step": 1277
+    },
+    {
+      "epoch": 0.5758053615679207,
+      "grad_norm": 1.7007240432875168,
+      "learning_rate": 1e-05,
+      "loss": 1.2356,
+      "step": 1278
+    },
+    {
+      "epoch": 0.5762559134940302,
+      "grad_norm": 1.76925981886769,
+      "learning_rate": 1e-05,
+      "loss": 1.1294,
+      "step": 1279
+    },
+    {
+      "epoch": 0.5767064654201397,
+      "grad_norm": 1.8815257385659285,
+      "learning_rate": 1e-05,
+      "loss": 1.1888,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5771570173462491,
+      "grad_norm": 1.5160018182898973,
+      "learning_rate": 1e-05,
+      "loss": 1.2336,
+      "step": 1281
+    },
+    {
+      "epoch": 0.5776075692723587,
+      "grad_norm": 1.6549581185662503,
+      "learning_rate": 1e-05,
+      "loss": 1.251,
+      "step": 1282
+    },
+    {
+      "epoch": 0.5780581211984681,
+      "grad_norm": 1.756618270675133,
+      "learning_rate": 1e-05,
+      "loss": 1.232,
+      "step": 1283
+    },
+    {
+      "epoch": 0.5785086731245777,
+      "grad_norm": 1.7419052722691228,
+      "learning_rate": 1e-05,
+      "loss": 1.17,
+      "step": 1284
+    },
+    {
+      "epoch": 0.5789592250506871,
+      "grad_norm": 1.7160979101151757,
+      "learning_rate": 1e-05,
+      "loss": 1.2386,
+      "step": 1285
+    },
+    {
+      "epoch": 0.5794097769767966,
+      "grad_norm": 1.8243826066374753,
+      "learning_rate": 1e-05,
+      "loss": 1.2472,
+      "step": 1286
+    },
+    {
+      "epoch": 0.5798603289029061,
+      "grad_norm": 1.933479040541196,
+      "learning_rate": 1e-05,
+      "loss": 1.2354,
+      "step": 1287
+    },
+    {
+      "epoch": 0.5803108808290155,
+      "grad_norm": 1.719961458326181,
+      "learning_rate": 1e-05,
+      "loss": 1.1676,
+      "step": 1288
+    },
+    {
+      "epoch": 0.580761432755125,
+      "grad_norm": 1.627151067117838,
+      "learning_rate": 1e-05,
+      "loss": 1.1455,
+      "step": 1289
+    },
+    {
+      "epoch": 0.5812119846812345,
+      "grad_norm": 1.9107495551066946,
+      "learning_rate": 1e-05,
+      "loss": 1.2433,
+      "step": 1290
+    },
+    {
+      "epoch": 0.581662536607344,
+      "grad_norm": 1.7825032477174885,
+      "learning_rate": 1e-05,
+      "loss": 1.2024,
+      "step": 1291
+    },
+    {
+      "epoch": 0.5821130885334534,
+      "grad_norm": 1.6348409962935258,
+      "learning_rate": 1e-05,
+      "loss": 1.2242,
+      "step": 1292
+    },
+    {
+      "epoch": 0.582563640459563,
+      "grad_norm": 1.699968988824632,
+      "learning_rate": 1e-05,
+      "loss": 1.2582,
+      "step": 1293
+    },
+    {
+      "epoch": 0.5830141923856724,
+      "grad_norm": 1.8974440062338904,
+      "learning_rate": 1e-05,
+      "loss": 1.1932,
+      "step": 1294
+    },
+    {
+      "epoch": 0.583464744311782,
+      "grad_norm": 1.8263445611498854,
+      "learning_rate": 1e-05,
+      "loss": 1.1732,
+      "step": 1295
+    },
+    {
+      "epoch": 0.5839152962378914,
+      "grad_norm": 1.8527801966410529,
+      "learning_rate": 1e-05,
+      "loss": 1.2289,
+      "step": 1296
+    },
+    {
+      "epoch": 0.584365848164001,
+      "grad_norm": 1.7644064170863516,
+      "learning_rate": 1e-05,
+      "loss": 1.2403,
+      "step": 1297
+    },
+    {
+      "epoch": 0.5848164000901104,
+      "grad_norm": 1.686109827756735,
+      "learning_rate": 1e-05,
+      "loss": 1.2233,
+      "step": 1298
+    },
+    {
+      "epoch": 0.5852669520162199,
+      "grad_norm": 1.6650733956443948,
+      "learning_rate": 1e-05,
+      "loss": 1.1874,
+      "step": 1299
+    },
+    {
+      "epoch": 0.5857175039423294,
+      "grad_norm": 1.871982358828454,
+      "learning_rate": 1e-05,
+      "loss": 1.188,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5861680558684388,
+      "grad_norm": 1.8230592097279792,
+      "learning_rate": 1e-05,
+      "loss": 1.2099,
+      "step": 1301
+    },
+    {
+      "epoch": 0.5866186077945483,
+      "grad_norm": 1.6857220231101702,
+      "learning_rate": 1e-05,
+      "loss": 1.2375,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5870691597206578,
+      "grad_norm": 1.8959198732393088,
+      "learning_rate": 1e-05,
+      "loss": 1.1937,
+      "step": 1303
+    },
+    {
+      "epoch": 0.5875197116467673,
+      "grad_norm": 1.7160428014338607,
+      "learning_rate": 1e-05,
+      "loss": 1.1922,
+      "step": 1304
+    },
+    {
+      "epoch": 0.5879702635728767,
+      "grad_norm": 1.7226545455972886,
+      "learning_rate": 1e-05,
+      "loss": 1.2047,
+      "step": 1305
+    },
+    {
+      "epoch": 0.5884208154989863,
+      "grad_norm": 1.8027359299917531,
+      "learning_rate": 1e-05,
+      "loss": 1.2287,
+      "step": 1306
+    },
+    {
+      "epoch": 0.5888713674250957,
+      "grad_norm": 1.7292176768544731,
+      "learning_rate": 1e-05,
+      "loss": 1.2721,
+      "step": 1307
+    },
+    {
+      "epoch": 0.5893219193512053,
+      "grad_norm": 1.776100854875579,
+      "learning_rate": 1e-05,
+      "loss": 1.2539,
+      "step": 1308
+    },
+    {
+      "epoch": 0.5897724712773147,
+      "grad_norm": 1.81697377707163,
+      "learning_rate": 1e-05,
+      "loss": 1.2325,
+      "step": 1309
+    },
+    {
+      "epoch": 0.5902230232034242,
+      "grad_norm": 1.8478715783483286,
+      "learning_rate": 1e-05,
+      "loss": 1.1742,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5906735751295337,
+      "grad_norm": 1.6954920926484498,
+      "learning_rate": 1e-05,
+      "loss": 1.1995,
+      "step": 1311
+    },
+    {
+      "epoch": 0.5911241270556432,
+      "grad_norm": 1.7772856974838311,
+      "learning_rate": 1e-05,
+      "loss": 1.2106,
+      "step": 1312
+    },
+    {
+      "epoch": 0.5915746789817526,
+      "grad_norm": 1.790739411389188,
+      "learning_rate": 1e-05,
+      "loss": 1.1941,
+      "step": 1313
+    },
+    {
+      "epoch": 0.5920252309078621,
+      "grad_norm": 1.6940522084311165,
+      "learning_rate": 1e-05,
+      "loss": 1.2367,
+      "step": 1314
+    },
+    {
+      "epoch": 0.5924757828339716,
+      "grad_norm": 1.8268985226553285,
+      "learning_rate": 1e-05,
+      "loss": 1.2083,
+      "step": 1315
+    },
+    {
+      "epoch": 0.592926334760081,
+      "grad_norm": 1.9954520134048521,
+      "learning_rate": 1e-05,
+      "loss": 1.179,
+      "step": 1316
+    },
+    {
+      "epoch": 0.5933768866861906,
+      "grad_norm": 1.8629824076550818,
+      "learning_rate": 1e-05,
+      "loss": 1.2344,
+      "step": 1317
+    },
+    {
+      "epoch": 0.5938274386123,
+      "grad_norm": 2.0103774181218332,
+      "learning_rate": 1e-05,
+      "loss": 1.2594,
+      "step": 1318
+    },
+    {
+      "epoch": 0.5942779905384096,
+      "grad_norm": 1.6660420981044828,
+      "learning_rate": 1e-05,
+      "loss": 1.1565,
+      "step": 1319
+    },
+    {
+      "epoch": 0.594728542464519,
+      "grad_norm": 1.6853788872552433,
+      "learning_rate": 1e-05,
+      "loss": 1.1985,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5951790943906285,
+      "grad_norm": 1.8487969481917361,
+      "learning_rate": 1e-05,
+      "loss": 1.2616,
+      "step": 1321
+    },
+    {
+      "epoch": 0.595629646316738,
+      "grad_norm": 1.8469994164982355,
+      "learning_rate": 1e-05,
+      "loss": 1.2165,
+      "step": 1322
+    },
+    {
+      "epoch": 0.5960801982428475,
+      "grad_norm": 1.7808612583191794,
+      "learning_rate": 1e-05,
+      "loss": 1.196,
+      "step": 1323
+    },
+    {
+      "epoch": 0.596530750168957,
+      "grad_norm": 1.6874331263721494,
+      "learning_rate": 1e-05,
+      "loss": 1.2272,
+      "step": 1324
+    },
+    {
+      "epoch": 0.5969813020950665,
+      "grad_norm": 1.7758083439810735,
+      "learning_rate": 1e-05,
+      "loss": 1.1748,
+      "step": 1325
+    },
+    {
+      "epoch": 0.5974318540211759,
+      "grad_norm": 1.6678835969274108,
+      "learning_rate": 1e-05,
+      "loss": 1.2267,
+      "step": 1326
+    },
+    {
+      "epoch": 0.5978824059472855,
+      "grad_norm": 1.803673440473812,
+      "learning_rate": 1e-05,
+      "loss": 1.1627,
+      "step": 1327
+    },
+    {
+      "epoch": 0.5983329578733949,
+      "grad_norm": 1.8984680335014743,
+      "learning_rate": 1e-05,
+      "loss": 1.2117,
+      "step": 1328
+    },
+    {
+      "epoch": 0.5987835097995043,
+      "grad_norm": 1.8927002548152252,
+      "learning_rate": 1e-05,
+      "loss": 1.2069,
+      "step": 1329
+    },
+    {
+      "epoch": 0.5992340617256139,
+      "grad_norm": 1.7466479887423474,
+      "learning_rate": 1e-05,
+      "loss": 1.2336,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5996846136517233,
+      "grad_norm": 1.8992416636445528,
+      "learning_rate": 1e-05,
+      "loss": 1.1989,
+      "step": 1331
+    },
+    {
+      "epoch": 0.6001351655778329,
+      "grad_norm": 1.8491010305275333,
+      "learning_rate": 1e-05,
+      "loss": 1.2014,
+      "step": 1332
+    },
+    {
+      "epoch": 0.6005857175039423,
+      "grad_norm": 1.819825114115824,
+      "learning_rate": 1e-05,
+      "loss": 1.2468,
+      "step": 1333
+    },
+    {
+      "epoch": 0.6010362694300518,
+      "grad_norm": 1.7991712243380846,
+      "learning_rate": 1e-05,
+      "loss": 1.1827,
+      "step": 1334
+    },
+    {
+      "epoch": 0.6014868213561613,
+      "grad_norm": 1.7071817076402331,
+      "learning_rate": 1e-05,
+      "loss": 1.2202,
+      "step": 1335
+    },
+    {
+      "epoch": 0.6019373732822708,
+      "grad_norm": 1.7303007334112557,
+      "learning_rate": 1e-05,
+      "loss": 1.2097,
+      "step": 1336
+    },
+    {
+      "epoch": 0.6023879252083802,
+      "grad_norm": 1.5340971873303832,
+      "learning_rate": 1e-05,
+      "loss": 1.1909,
+      "step": 1337
+    },
+    {
+      "epoch": 0.6028384771344898,
+      "grad_norm": 1.6591955553042324,
+      "learning_rate": 1e-05,
+      "loss": 1.2635,
+      "step": 1338
+    },
+    {
+      "epoch": 0.6032890290605992,
+      "grad_norm": 1.8169846626883783,
+      "learning_rate": 1e-05,
+      "loss": 1.2285,
+      "step": 1339
+    },
+    {
+      "epoch": 0.6037395809867088,
+      "grad_norm": 2.0689419123732202,
+      "learning_rate": 1e-05,
+      "loss": 1.2503,
+      "step": 1340
+    },
+    {
+      "epoch": 0.6041901329128182,
+      "grad_norm": 1.8044095506809605,
+      "learning_rate": 1e-05,
+      "loss": 1.2008,
+      "step": 1341
+    },
+    {
+      "epoch": 0.6046406848389276,
+      "grad_norm": 1.7016093429282768,
+      "learning_rate": 1e-05,
+      "loss": 1.2318,
+      "step": 1342
+    },
+    {
+      "epoch": 0.6050912367650372,
+      "grad_norm": 1.6926471157882403,
+      "learning_rate": 1e-05,
+      "loss": 1.1924,
+      "step": 1343
+    },
+    {
+      "epoch": 0.6055417886911466,
+      "grad_norm": 1.726036101180517,
+      "learning_rate": 1e-05,
+      "loss": 1.2183,
+      "step": 1344
+    },
+    {
+      "epoch": 0.6059923406172562,
+      "grad_norm": 1.6284686389088292,
+      "learning_rate": 1e-05,
+      "loss": 1.2263,
+      "step": 1345
+    },
+    {
+      "epoch": 0.6064428925433656,
+      "grad_norm": 1.8219348319849917,
+      "learning_rate": 1e-05,
+      "loss": 1.2356,
+      "step": 1346
+    },
+    {
+      "epoch": 0.6068934444694751,
+      "grad_norm": 1.867304570054251,
+      "learning_rate": 1e-05,
+      "loss": 1.2246,
+      "step": 1347
+    },
+    {
+      "epoch": 0.6073439963955846,
+      "grad_norm": 1.7241172819179993,
+      "learning_rate": 1e-05,
+      "loss": 1.1936,
+      "step": 1348
+    },
+    {
+      "epoch": 0.6077945483216941,
+      "grad_norm": 1.6534720493893311,
+      "learning_rate": 1e-05,
+      "loss": 1.2811,
+      "step": 1349
+    },
+    {
+      "epoch": 0.6082451002478035,
+      "grad_norm": 1.6786794751334155,
+      "learning_rate": 1e-05,
+      "loss": 1.1915,
+      "step": 1350
+    },
+    {
+      "epoch": 0.6086956521739131,
+      "grad_norm": 1.7585178814796325,
+      "learning_rate": 1e-05,
+      "loss": 1.1843,
+      "step": 1351
+    },
+    {
+      "epoch": 0.6091462041000225,
+      "grad_norm": 1.7960034529195434,
+      "learning_rate": 1e-05,
+      "loss": 1.1596,
+      "step": 1352
+    },
+    {
+      "epoch": 0.6095967560261321,
+      "grad_norm": 1.6927853300950366,
+      "learning_rate": 1e-05,
+      "loss": 1.2017,
+      "step": 1353
+    },
+    {
+      "epoch": 0.6100473079522415,
+      "grad_norm": 1.7275520080961777,
+      "learning_rate": 1e-05,
+      "loss": 1.2036,
+      "step": 1354
+    },
+    {
+      "epoch": 0.610497859878351,
+      "grad_norm": 1.9279648803221345,
+      "learning_rate": 1e-05,
+      "loss": 1.2357,
+      "step": 1355
+    },
+    {
+      "epoch": 0.6109484118044605,
+      "grad_norm": 1.716209317173508,
+      "learning_rate": 1e-05,
+      "loss": 1.2058,
+      "step": 1356
+    },
+    {
+      "epoch": 0.6113989637305699,
+      "grad_norm": 1.7036585157626531,
+      "learning_rate": 1e-05,
+      "loss": 1.1875,
+      "step": 1357
+    },
+    {
+      "epoch": 0.6118495156566794,
+      "grad_norm": 1.6337918035391776,
+      "learning_rate": 1e-05,
+      "loss": 1.2264,
+      "step": 1358
+    },
+    {
+      "epoch": 0.6123000675827889,
+      "grad_norm": 1.5820117076593503,
+      "learning_rate": 1e-05,
+      "loss": 1.2066,
+      "step": 1359
+    },
+    {
+      "epoch": 0.6127506195088984,
+      "grad_norm": 1.930044225209007,
+      "learning_rate": 1e-05,
+      "loss": 1.2437,
+      "step": 1360
+    },
+    {
+      "epoch": 0.6132011714350079,
+      "grad_norm": 1.6566093920031004,
+      "learning_rate": 1e-05,
+      "loss": 1.1883,
+      "step": 1361
+    },
+    {
+      "epoch": 0.6136517233611174,
+      "grad_norm": 1.779881594851804,
+      "learning_rate": 1e-05,
+      "loss": 1.2089,
+      "step": 1362
+    },
+    {
+      "epoch": 0.6141022752872268,
+      "grad_norm": 1.764450046969482,
+      "learning_rate": 1e-05,
+      "loss": 1.2103,
+      "step": 1363
+    },
+    {
+      "epoch": 0.6145528272133364,
+      "grad_norm": 1.7755722026363077,
+      "learning_rate": 1e-05,
+      "loss": 1.2131,
+      "step": 1364
+    },
+    {
+      "epoch": 0.6150033791394458,
+      "grad_norm": 1.7162557771636968,
+      "learning_rate": 1e-05,
+      "loss": 1.2231,
+      "step": 1365
+    },
+    {
+      "epoch": 0.6154539310655553,
+      "grad_norm": 1.7411428223191674,
+      "learning_rate": 1e-05,
+      "loss": 1.1688,
+      "step": 1366
+    },
+    {
+      "epoch": 0.6159044829916648,
+      "grad_norm": 1.5915237847294443,
+      "learning_rate": 1e-05,
+      "loss": 1.1696,
+      "step": 1367
+    },
+    {
+      "epoch": 0.6163550349177743,
+      "grad_norm": 1.6602744896219306,
+      "learning_rate": 1e-05,
+      "loss": 1.203,
+      "step": 1368
+    },
+    {
+      "epoch": 0.6168055868438838,
+      "grad_norm": 1.6811181986810635,
+      "learning_rate": 1e-05,
+      "loss": 1.1846,
+      "step": 1369
+    },
+    {
+      "epoch": 0.6172561387699932,
+      "grad_norm": 1.8880690213174565,
+      "learning_rate": 1e-05,
+      "loss": 1.2611,
+      "step": 1370
+    },
+    {
+      "epoch": 0.6177066906961027,
+      "grad_norm": 1.6852043893589732,
+      "learning_rate": 1e-05,
+      "loss": 1.2065,
+      "step": 1371
+    },
+    {
+      "epoch": 0.6181572426222122,
+      "grad_norm": 1.7370535118725405,
+      "learning_rate": 1e-05,
+      "loss": 1.2133,
+      "step": 1372
+    },
+    {
+      "epoch": 0.6186077945483217,
+      "grad_norm": 1.8432188590958254,
+      "learning_rate": 1e-05,
+      "loss": 1.1533,
+      "step": 1373
+    },
+    {
+      "epoch": 0.6190583464744311,
+      "grad_norm": 1.7090225875378293,
+      "learning_rate": 1e-05,
+      "loss": 1.2126,
+      "step": 1374
+    },
+    {
+      "epoch": 0.6195088984005407,
+      "grad_norm": 1.7471266346091723,
+      "learning_rate": 1e-05,
+      "loss": 1.229,
+      "step": 1375
+    },
+    {
+      "epoch": 0.6199594503266501,
+      "grad_norm": 1.6751556261213432,
+      "learning_rate": 1e-05,
+      "loss": 1.1679,
+      "step": 1376
+    },
+    {
+      "epoch": 0.6204100022527597,
+      "grad_norm": 1.840837168317006,
+      "learning_rate": 1e-05,
+      "loss": 1.2597,
+      "step": 1377
+    },
+    {
+      "epoch": 0.6208605541788691,
+      "grad_norm": 1.7050669339799445,
+      "learning_rate": 1e-05,
+      "loss": 1.1779,
+      "step": 1378
+    },
+    {
+      "epoch": 0.6213111061049786,
+      "grad_norm": 1.675799407147007,
+      "learning_rate": 1e-05,
+      "loss": 1.1792,
+      "step": 1379
+    },
+    {
+      "epoch": 0.6217616580310881,
+      "grad_norm": 1.8471267408877814,
+      "learning_rate": 1e-05,
+      "loss": 1.1939,
+      "step": 1380
+    },
+    {
+      "epoch": 0.6222122099571976,
+      "grad_norm": 1.7646840616526538,
+      "learning_rate": 1e-05,
+      "loss": 1.1809,
+      "step": 1381
+    },
+    {
+      "epoch": 0.622662761883307,
+      "grad_norm": 1.613532346023576,
+      "learning_rate": 1e-05,
+      "loss": 1.2278,
+      "step": 1382
+    },
+    {
+      "epoch": 0.6231133138094165,
+      "grad_norm": 1.7524354574585752,
+      "learning_rate": 1e-05,
+      "loss": 1.1992,
+      "step": 1383
+    },
+    {
+      "epoch": 0.623563865735526,
+      "grad_norm": 1.8946756153547877,
+      "learning_rate": 1e-05,
+      "loss": 1.218,
+      "step": 1384
+    },
+    {
+      "epoch": 0.6240144176616355,
+      "grad_norm": 1.8070652923583572,
+      "learning_rate": 1e-05,
+      "loss": 1.2454,
+      "step": 1385
+    },
+    {
+      "epoch": 0.624464969587745,
+      "grad_norm": 1.934785501668537,
+      "learning_rate": 1e-05,
+      "loss": 1.1521,
+      "step": 1386
+    },
+    {
+      "epoch": 0.6249155215138544,
+      "grad_norm": 1.7400549394661604,
+      "learning_rate": 1e-05,
+      "loss": 1.196,
+      "step": 1387
+    },
+    {
+      "epoch": 0.625366073439964,
+      "grad_norm": 1.607968603657251,
+      "learning_rate": 1e-05,
+      "loss": 1.199,
+      "step": 1388
+    },
+    {
+      "epoch": 0.6258166253660734,
+      "grad_norm": 1.8044675154447942,
+      "learning_rate": 1e-05,
+      "loss": 1.1741,
+      "step": 1389
+    },
+    {
+      "epoch": 0.626267177292183,
+      "grad_norm": 1.7688196483248884,
+      "learning_rate": 1e-05,
+      "loss": 1.1747,
+      "step": 1390
+    },
+    {
+      "epoch": 0.6267177292182924,
+      "grad_norm": 1.759460497999574,
+      "learning_rate": 1e-05,
+      "loss": 1.1521,
+      "step": 1391
+    },
+    {
+      "epoch": 0.6271682811444019,
+      "grad_norm": 1.7143078451595086,
+      "learning_rate": 1e-05,
+      "loss": 1.2089,
+      "step": 1392
+    },
+    {
+      "epoch": 0.6276188330705114,
+      "grad_norm": 1.7577585746100481,
+      "learning_rate": 1e-05,
+      "loss": 1.2127,
+      "step": 1393
+    },
+    {
+      "epoch": 0.6280693849966209,
+      "grad_norm": 1.7184496518733532,
+      "learning_rate": 1e-05,
+      "loss": 1.2616,
+      "step": 1394
+    },
+    {
+      "epoch": 0.6285199369227303,
+      "grad_norm": 1.8030703326828137,
+      "learning_rate": 1e-05,
+      "loss": 1.2213,
+      "step": 1395
+    },
+    {
+      "epoch": 0.6289704888488399,
+      "grad_norm": 1.7208734506499097,
+      "learning_rate": 1e-05,
+      "loss": 1.2436,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6294210407749493,
+      "grad_norm": 1.8016050010759244,
+      "learning_rate": 1e-05,
+      "loss": 1.2276,
+      "step": 1397
+    },
+    {
+      "epoch": 0.6298715927010587,
+      "grad_norm": 1.828897332337839,
+      "learning_rate": 1e-05,
+      "loss": 1.1914,
+      "step": 1398
+    },
+    {
+      "epoch": 0.6303221446271683,
+      "grad_norm": 1.7219910585566085,
+      "learning_rate": 1e-05,
+      "loss": 1.1792,
+      "step": 1399
+    },
+    {
+      "epoch": 0.6307726965532777,
+      "grad_norm": 1.8449535380865347,
+      "learning_rate": 1e-05,
+      "loss": 1.2105,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6312232484793873,
+      "grad_norm": 1.6087786279962788,
+      "learning_rate": 1e-05,
+      "loss": 1.1846,
+      "step": 1401
+    },
+    {
+      "epoch": 0.6316738004054967,
+      "grad_norm": 2.0102078883332144,
+      "learning_rate": 1e-05,
+      "loss": 1.2488,
+      "step": 1402
+    },
+    {
+      "epoch": 0.6321243523316062,
+      "grad_norm": 1.7119388385471392,
+      "learning_rate": 1e-05,
+      "loss": 1.2099,
+      "step": 1403
+    },
+    {
+      "epoch": 0.6325749042577157,
+      "grad_norm": 1.6487861408950628,
+      "learning_rate": 1e-05,
+      "loss": 1.1942,
+      "step": 1404
+    },
+    {
+      "epoch": 0.6330254561838252,
+      "grad_norm": 1.6649600775876516,
+      "learning_rate": 1e-05,
+      "loss": 1.2256,
+      "step": 1405
+    },
+    {
+      "epoch": 0.6334760081099347,
+      "grad_norm": 1.782388209434633,
+      "learning_rate": 1e-05,
+      "loss": 1.2136,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6339265600360442,
+      "grad_norm": 1.7316483411168198,
+      "learning_rate": 1e-05,
+      "loss": 1.2479,
+      "step": 1407
+    },
+    {
+      "epoch": 0.6343771119621536,
+      "grad_norm": 1.8474748436342696,
+      "learning_rate": 1e-05,
+      "loss": 1.2448,
+      "step": 1408
+    },
+    {
+      "epoch": 0.6348276638882632,
+      "grad_norm": 1.8881771017116113,
+      "learning_rate": 1e-05,
+      "loss": 1.2296,
+      "step": 1409
+    },
+    {
+      "epoch": 0.6352782158143726,
+      "grad_norm": 1.6815920585030426,
+      "learning_rate": 1e-05,
+      "loss": 1.2309,
+      "step": 1410
+    },
+    {
+      "epoch": 0.635728767740482,
+      "grad_norm": 1.7368229873136294,
+      "learning_rate": 1e-05,
+      "loss": 1.1924,
+      "step": 1411
+    },
+    {
+      "epoch": 0.6361793196665916,
+      "grad_norm": 1.622168896559517,
+      "learning_rate": 1e-05,
+      "loss": 1.2089,
+      "step": 1412
+    },
+    {
+      "epoch": 0.636629871592701,
+      "grad_norm": 1.912293481955647,
+      "learning_rate": 1e-05,
+      "loss": 1.1906,
+      "step": 1413
+    },
+    {
+      "epoch": 0.6370804235188106,
+      "grad_norm": 1.6915846270575778,
+      "learning_rate": 1e-05,
+      "loss": 1.179,
+      "step": 1414
+    },
+    {
+      "epoch": 0.63753097544492,
+      "grad_norm": 1.7556753029844074,
+      "learning_rate": 1e-05,
+      "loss": 1.1796,
+      "step": 1415
+    },
+    {
+      "epoch": 0.6379815273710295,
+      "grad_norm": 1.6918747056678867,
+      "learning_rate": 1e-05,
+      "loss": 1.2222,
+      "step": 1416
+    },
+    {
+      "epoch": 0.638432079297139,
+      "grad_norm": 1.7310795539635673,
+      "learning_rate": 1e-05,
+      "loss": 1.1973,
+      "step": 1417
+    },
+    {
+      "epoch": 0.6388826312232485,
+      "grad_norm": 1.6894866324396298,
+      "learning_rate": 1e-05,
+      "loss": 1.2472,
+      "step": 1418
+    },
+    {
+      "epoch": 0.6393331831493579,
+      "grad_norm": 1.7382727610437598,
+      "learning_rate": 1e-05,
+      "loss": 1.2301,
+      "step": 1419
+    },
+    {
+      "epoch": 0.6397837350754675,
+      "grad_norm": 1.943319888398003,
+      "learning_rate": 1e-05,
+      "loss": 1.1965,
+      "step": 1420
+    },
+    {
+      "epoch": 0.6402342870015769,
+      "grad_norm": 1.696952940528892,
+      "learning_rate": 1e-05,
+      "loss": 1.2014,
+      "step": 1421
+    },
+    {
+      "epoch": 0.6406848389276865,
+      "grad_norm": 1.714810552322869,
+      "learning_rate": 1e-05,
+      "loss": 1.183,
+      "step": 1422
+    },
+    {
+      "epoch": 0.6411353908537959,
+      "grad_norm": 1.7048297703948376,
+      "learning_rate": 1e-05,
+      "loss": 1.2197,
+      "step": 1423
+    },
+    {
+      "epoch": 0.6415859427799054,
+      "grad_norm": 1.7726722279557274,
+      "learning_rate": 1e-05,
+      "loss": 1.2121,
+      "step": 1424
+    },
+    {
+      "epoch": 0.6420364947060149,
+      "grad_norm": 1.6887525653748252,
+      "learning_rate": 1e-05,
+      "loss": 1.2374,
+      "step": 1425
+    },
+    {
+      "epoch": 0.6424870466321243,
+      "grad_norm": 1.8340053037918147,
+      "learning_rate": 1e-05,
+      "loss": 1.2096,
+      "step": 1426
+    },
+    {
+      "epoch": 0.6429375985582338,
+      "grad_norm": 1.7779256927355376,
+      "learning_rate": 1e-05,
+      "loss": 1.1675,
+      "step": 1427
+    },
+    {
+      "epoch": 0.6433881504843433,
+      "grad_norm": 1.7405618803764125,
+      "learning_rate": 1e-05,
+      "loss": 1.1953,
+      "step": 1428
+    },
+    {
+      "epoch": 0.6438387024104528,
+      "grad_norm": 1.7971599220737207,
+      "learning_rate": 1e-05,
+      "loss": 1.2136,
+      "step": 1429
+    },
+    {
+      "epoch": 0.6442892543365623,
+      "grad_norm": 1.7114769331843658,
+      "learning_rate": 1e-05,
+      "loss": 1.2055,
+      "step": 1430
+    },
+    {
+      "epoch": 0.6447398062626718,
+      "grad_norm": 1.7760274627193031,
+      "learning_rate": 1e-05,
+      "loss": 1.156,
+      "step": 1431
+    },
+    {
+      "epoch": 0.6451903581887812,
+      "grad_norm": 1.7056122994068905,
+      "learning_rate": 1e-05,
+      "loss": 1.2215,
+      "step": 1432
+    },
+    {
+      "epoch": 0.6456409101148908,
+      "grad_norm": 1.7839219911420403,
+      "learning_rate": 1e-05,
+      "loss": 1.2541,
+      "step": 1433
+    },
+    {
+      "epoch": 0.6460914620410002,
+      "grad_norm": 1.8744821096312974,
+      "learning_rate": 1e-05,
+      "loss": 1.2339,
+      "step": 1434
+    },
+    {
+      "epoch": 0.6465420139671098,
+      "grad_norm": 1.839865376865044,
+      "learning_rate": 1e-05,
+      "loss": 1.2439,
+      "step": 1435
+    },
+    {
+      "epoch": 0.6469925658932192,
+      "grad_norm": 1.8639431131353121,
+      "learning_rate": 1e-05,
+      "loss": 1.1954,
+      "step": 1436
+    },
+    {
+      "epoch": 0.6474431178193287,
+      "grad_norm": 1.8293807546531986,
+      "learning_rate": 1e-05,
+      "loss": 1.175,
+      "step": 1437
+    },
+    {
+      "epoch": 0.6478936697454382,
+      "grad_norm": 1.6694526478693845,
+      "learning_rate": 1e-05,
+      "loss": 1.1901,
+      "step": 1438
+    },
+    {
+      "epoch": 0.6483442216715476,
+      "grad_norm": 1.8084879536612948,
+      "learning_rate": 1e-05,
+      "loss": 1.2469,
+      "step": 1439
+    },
+    {
+      "epoch": 0.6487947735976571,
+      "grad_norm": 1.7432929325836355,
+      "learning_rate": 1e-05,
+      "loss": 1.223,
+      "step": 1440
+    },
+    {
+      "epoch": 0.6492453255237666,
+      "grad_norm": 1.5605053909073308,
+      "learning_rate": 1e-05,
+      "loss": 1.2234,
+      "step": 1441
+    },
+    {
+      "epoch": 0.6496958774498761,
+      "grad_norm": 1.6995774197072129,
+      "learning_rate": 1e-05,
+      "loss": 1.2038,
+      "step": 1442
+    },
+    {
+      "epoch": 0.6501464293759855,
+      "grad_norm": 1.649924736851965,
+      "learning_rate": 1e-05,
+      "loss": 1.2179,
+      "step": 1443
+    },
+    {
+      "epoch": 0.6505969813020951,
+      "grad_norm": 1.677375810646424,
+      "learning_rate": 1e-05,
+      "loss": 1.215,
+      "step": 1444
+    },
+    {
+      "epoch": 0.6510475332282045,
+      "grad_norm": 1.687153477900548,
+      "learning_rate": 1e-05,
+      "loss": 1.2227,
+      "step": 1445
+    },
+    {
+      "epoch": 0.6514980851543141,
+      "grad_norm": 1.7842664158579138,
+      "learning_rate": 1e-05,
+      "loss": 1.2258,
+      "step": 1446
+    },
+    {
+      "epoch": 0.6519486370804235,
+      "grad_norm": 1.7699069004158208,
+      "learning_rate": 1e-05,
+      "loss": 1.2028,
+      "step": 1447
+    },
+    {
+      "epoch": 0.652399189006533,
+      "grad_norm": 1.7048564627062228,
+      "learning_rate": 1e-05,
+      "loss": 1.1685,
+      "step": 1448
+    },
+    {
+      "epoch": 0.6528497409326425,
+      "grad_norm": 1.7209083290062486,
+      "learning_rate": 1e-05,
+      "loss": 1.2508,
+      "step": 1449
+    },
+    {
+      "epoch": 0.653300292858752,
+      "grad_norm": 1.7614080078327725,
+      "learning_rate": 1e-05,
+      "loss": 1.1635,
+      "step": 1450
+    },
+    {
+      "epoch": 0.6537508447848615,
+      "grad_norm": 1.9896284472985386,
+      "learning_rate": 1e-05,
+      "loss": 1.1742,
+      "step": 1451
+    },
+    {
+      "epoch": 0.654201396710971,
+      "grad_norm": 2.0460215106674284,
+      "learning_rate": 1e-05,
+      "loss": 1.2027,
+      "step": 1452
+    },
+    {
+      "epoch": 0.6546519486370804,
+      "grad_norm": 1.7851444286720592,
+      "learning_rate": 1e-05,
+      "loss": 1.1954,
+      "step": 1453
+    },
+    {
+      "epoch": 0.6551025005631899,
+      "grad_norm": 1.875687692273614,
+      "learning_rate": 1e-05,
+      "loss": 1.2338,
+      "step": 1454
+    },
+    {
+      "epoch": 0.6555530524892994,
+      "grad_norm": 1.8123153895280242,
+      "learning_rate": 1e-05,
+      "loss": 1.2024,
+      "step": 1455
+    },
+    {
+      "epoch": 0.6560036044154088,
+      "grad_norm": 1.6178100607967243,
+      "learning_rate": 1e-05,
+      "loss": 1.1982,
+      "step": 1456
+    },
+    {
+      "epoch": 0.6564541563415184,
+      "grad_norm": 1.7615096838359687,
+      "learning_rate": 1e-05,
+      "loss": 1.1939,
+      "step": 1457
+    },
+    {
+      "epoch": 0.6569047082676278,
+      "grad_norm": 1.8122650299160075,
+      "learning_rate": 1e-05,
+      "loss": 1.1995,
+      "step": 1458
+    },
+    {
+      "epoch": 0.6573552601937374,
+      "grad_norm": 1.7898039749755277,
+      "learning_rate": 1e-05,
+      "loss": 1.2364,
+      "step": 1459
+    },
+    {
+      "epoch": 0.6578058121198468,
+      "grad_norm": 1.7004419153173866,
+      "learning_rate": 1e-05,
+      "loss": 1.1836,
+      "step": 1460
+    },
+    {
+      "epoch": 0.6582563640459563,
+      "grad_norm": 1.6632662674579606,
+      "learning_rate": 1e-05,
+      "loss": 1.2021,
+      "step": 1461
+    },
+    {
+      "epoch": 0.6587069159720658,
+      "grad_norm": 1.744403489143352,
+      "learning_rate": 1e-05,
+      "loss": 1.1767,
+      "step": 1462
+    },
+    {
+      "epoch": 0.6591574678981753,
+      "grad_norm": 1.8525400652515043,
+      "learning_rate": 1e-05,
+      "loss": 1.1795,
+      "step": 1463
+    },
+    {
+      "epoch": 0.6596080198242847,
+      "grad_norm": 1.788612318743038,
+      "learning_rate": 1e-05,
+      "loss": 1.2182,
+      "step": 1464
+    },
+    {
+      "epoch": 0.6600585717503943,
+      "grad_norm": 1.8299636065907605,
+      "learning_rate": 1e-05,
+      "loss": 1.2462,
+      "step": 1465
+    },
+    {
+      "epoch": 0.6605091236765037,
+      "grad_norm": 1.8145114327639338,
+      "learning_rate": 1e-05,
+      "loss": 1.2028,
+      "step": 1466
+    },
+    {
+      "epoch": 0.6609596756026132,
+      "grad_norm": 1.7717768279354946,
+      "learning_rate": 1e-05,
+      "loss": 1.2048,
+      "step": 1467
+    },
+    {
+      "epoch": 0.6614102275287227,
+      "grad_norm": 1.7837418410858035,
+      "learning_rate": 1e-05,
+      "loss": 1.1785,
+      "step": 1468
+    },
+    {
+      "epoch": 0.6618607794548321,
+      "grad_norm": 1.8712209185517494,
+      "learning_rate": 1e-05,
+      "loss": 1.2264,
+      "step": 1469
+    },
+    {
+      "epoch": 0.6623113313809417,
+      "grad_norm": 1.7860580091025848,
+      "learning_rate": 1e-05,
+      "loss": 1.1879,
+      "step": 1470
+    },
+    {
+      "epoch": 0.6627618833070511,
+      "grad_norm": 1.6606100128063292,
+      "learning_rate": 1e-05,
+      "loss": 1.1584,
+      "step": 1471
+    },
+    {
+      "epoch": 0.6632124352331606,
+      "grad_norm": 1.6797327456630582,
+      "learning_rate": 1e-05,
+      "loss": 1.2382,
+      "step": 1472
+    },
+    {
+      "epoch": 0.6636629871592701,
+      "grad_norm": 1.779156003997943,
+      "learning_rate": 1e-05,
+      "loss": 1.2082,
+      "step": 1473
+    },
+    {
+      "epoch": 0.6641135390853796,
+      "grad_norm": 1.8478302534625182,
+      "learning_rate": 1e-05,
+      "loss": 1.2254,
+      "step": 1474
+    },
+    {
+      "epoch": 0.664564091011489,
+      "grad_norm": 1.6148156185454428,
+      "learning_rate": 1e-05,
+      "loss": 1.249,
+      "step": 1475
+    },
+    {
+      "epoch": 0.6650146429375986,
+      "grad_norm": 1.75522802249122,
+      "learning_rate": 1e-05,
+      "loss": 1.2592,
+      "step": 1476
+    },
+    {
+      "epoch": 0.665465194863708,
+      "grad_norm": 1.5213446285248269,
+      "learning_rate": 1e-05,
+      "loss": 1.2248,
+      "step": 1477
+    },
+    {
+      "epoch": 0.6659157467898176,
+      "grad_norm": 1.7710285474424141,
+      "learning_rate": 1e-05,
+      "loss": 1.2387,
+      "step": 1478
+    },
+    {
+      "epoch": 0.666366298715927,
+      "grad_norm": 1.680337020579166,
+      "learning_rate": 1e-05,
+      "loss": 1.289,
+      "step": 1479
+    },
+    {
+      "epoch": 0.6668168506420364,
+      "grad_norm": 1.7109187930443421,
+      "learning_rate": 1e-05,
+      "loss": 1.194,
+      "step": 1480
+    },
+    {
+      "epoch": 0.667267402568146,
+      "grad_norm": 1.8027228771531587,
+      "learning_rate": 1e-05,
+      "loss": 1.2099,
+      "step": 1481
+    },
+    {
+      "epoch": 0.6677179544942554,
+      "grad_norm": 1.6876214039504391,
+      "learning_rate": 1e-05,
+      "loss": 1.1938,
+      "step": 1482
+    },
+    {
+      "epoch": 0.668168506420365,
+      "grad_norm": 1.8428577669555128,
+      "learning_rate": 1e-05,
+      "loss": 1.1474,
+      "step": 1483
+    },
+    {
+      "epoch": 0.6686190583464744,
+      "grad_norm": 1.6964020418470995,
+      "learning_rate": 1e-05,
+      "loss": 1.1333,
+      "step": 1484
+    },
+    {
+      "epoch": 0.6690696102725839,
+      "grad_norm": 1.7276792301617687,
+      "learning_rate": 1e-05,
+      "loss": 1.2064,
+      "step": 1485
+    },
+    {
+      "epoch": 0.6695201621986934,
+      "grad_norm": 1.7431273311944988,
+      "learning_rate": 1e-05,
+      "loss": 1.1688,
+      "step": 1486
+    },
+    {
+      "epoch": 0.6699707141248029,
+      "grad_norm": 1.8501026168048027,
+      "learning_rate": 1e-05,
+      "loss": 1.2096,
+      "step": 1487
+    },
+    {
+      "epoch": 0.6704212660509123,
+      "grad_norm": 1.7677780866844144,
+      "learning_rate": 1e-05,
+      "loss": 1.2035,
+      "step": 1488
+    },
+    {
+      "epoch": 0.6708718179770219,
+      "grad_norm": 1.7150756049457243,
+      "learning_rate": 1e-05,
+      "loss": 1.2254,
+      "step": 1489
+    },
+    {
+      "epoch": 0.6713223699031313,
+      "grad_norm": 1.7671950790182738,
+      "learning_rate": 1e-05,
+      "loss": 1.1698,
+      "step": 1490
+    },
+    {
+      "epoch": 0.6717729218292409,
+      "grad_norm": 1.5858226806327806,
+      "learning_rate": 1e-05,
+      "loss": 1.2184,
+      "step": 1491
+    },
+    {
+      "epoch": 0.6722234737553503,
+      "grad_norm": 1.717479513832325,
+      "learning_rate": 1e-05,
+      "loss": 1.2017,
+      "step": 1492
+    },
+    {
+      "epoch": 0.6726740256814598,
+      "grad_norm": 1.7178760535041833,
+      "learning_rate": 1e-05,
+      "loss": 1.1896,
+      "step": 1493
+    },
+    {
+      "epoch": 0.6731245776075693,
+      "grad_norm": 1.7378252367998972,
+      "learning_rate": 1e-05,
+      "loss": 1.1888,
+      "step": 1494
+    },
+    {
+      "epoch": 0.6735751295336787,
+      "grad_norm": 1.686646531763602,
+      "learning_rate": 1e-05,
+      "loss": 1.2322,
+      "step": 1495
+    },
+    {
+      "epoch": 0.6740256814597883,
+      "grad_norm": 1.7785805981710803,
+      "learning_rate": 1e-05,
+      "loss": 1.1652,
+      "step": 1496
+    },
+    {
+      "epoch": 0.6744762333858977,
+      "grad_norm": 1.863415067381876,
+      "learning_rate": 1e-05,
+      "loss": 1.1878,
+      "step": 1497
+    },
+    {
+      "epoch": 0.6749267853120072,
+      "grad_norm": 1.7432798280237887,
+      "learning_rate": 1e-05,
+      "loss": 1.2158,
+      "step": 1498
+    },
+    {
+      "epoch": 0.6753773372381167,
+      "grad_norm": 1.6729316504931013,
+      "learning_rate": 1e-05,
+      "loss": 1.2433,
+      "step": 1499
+    },
+    {
+      "epoch": 0.6758278891642262,
+      "grad_norm": 1.6629313114328055,
+      "learning_rate": 1e-05,
+      "loss": 1.1942,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6762784410903356,
+      "grad_norm": 1.6065777404365544,
+      "learning_rate": 1e-05,
+      "loss": 1.2327,
+      "step": 1501
+    },
+    {
+      "epoch": 0.6767289930164452,
+      "grad_norm": 1.5557977992450556,
+      "learning_rate": 1e-05,
+      "loss": 1.1825,
+      "step": 1502
+    },
+    {
+      "epoch": 0.6771795449425546,
+      "grad_norm": 1.9016853376062381,
+      "learning_rate": 1e-05,
+      "loss": 1.1943,
+      "step": 1503
+    },
+    {
+      "epoch": 0.6776300968686642,
+      "grad_norm": 1.590043710512899,
+      "learning_rate": 1e-05,
+      "loss": 1.2031,
+      "step": 1504
+    },
+    {
+      "epoch": 0.6780806487947736,
+      "grad_norm": 1.7796113694185258,
+      "learning_rate": 1e-05,
+      "loss": 1.2222,
+      "step": 1505
+    },
+    {
+      "epoch": 0.6785312007208831,
+      "grad_norm": 1.7367622130464204,
+      "learning_rate": 1e-05,
+      "loss": 1.2028,
+      "step": 1506
+    },
+    {
+      "epoch": 0.6789817526469926,
+      "grad_norm": 1.5468093833959076,
+      "learning_rate": 1e-05,
+      "loss": 1.1449,
+      "step": 1507
+    },
+    {
+      "epoch": 0.679432304573102,
+      "grad_norm": 1.615544002799072,
+      "learning_rate": 1e-05,
+      "loss": 1.181,
+      "step": 1508
+    },
+    {
+      "epoch": 0.6798828564992115,
+      "grad_norm": 1.6654340696921206,
+      "learning_rate": 1e-05,
+      "loss": 1.1835,
+      "step": 1509
+    },
+    {
+      "epoch": 0.680333408425321,
+      "grad_norm": 1.8111265166768526,
+      "learning_rate": 1e-05,
+      "loss": 1.2335,
+      "step": 1510
+    },
+    {
+      "epoch": 0.6807839603514305,
+      "grad_norm": 1.9126157729140019,
+      "learning_rate": 1e-05,
+      "loss": 1.2202,
+      "step": 1511
+    },
+    {
+      "epoch": 0.68123451227754,
+      "grad_norm": 1.8063762771293985,
+      "learning_rate": 1e-05,
+      "loss": 1.1887,
+      "step": 1512
+    },
+    {
+      "epoch": 0.6816850642036495,
+      "grad_norm": 1.808564622464707,
+      "learning_rate": 1e-05,
+      "loss": 1.1837,
+      "step": 1513
+    },
+    {
+      "epoch": 0.6821356161297589,
+      "grad_norm": 1.7637059867029303,
+      "learning_rate": 1e-05,
+      "loss": 1.2166,
+      "step": 1514
+    },
+    {
+      "epoch": 0.6825861680558685,
+      "grad_norm": 1.6100181389970576,
+      "learning_rate": 1e-05,
+      "loss": 1.2362,
+      "step": 1515
+    },
+    {
+      "epoch": 0.6830367199819779,
+      "grad_norm": 1.8426325458109705,
+      "learning_rate": 1e-05,
+      "loss": 1.2488,
+      "step": 1516
+    },
+    {
+      "epoch": 0.6834872719080874,
+      "grad_norm": 1.829455745565599,
+      "learning_rate": 1e-05,
+      "loss": 1.2223,
+      "step": 1517
+    },
+    {
+      "epoch": 0.6839378238341969,
+      "grad_norm": 1.679105198586097,
+      "learning_rate": 1e-05,
+      "loss": 1.1963,
+      "step": 1518
+    },
+    {
+      "epoch": 0.6843883757603064,
+      "grad_norm": 1.8167196778392032,
+      "learning_rate": 1e-05,
+      "loss": 1.234,
+      "step": 1519
+    },
+    {
+      "epoch": 0.6848389276864159,
+      "grad_norm": 1.8156765386085207,
+      "learning_rate": 1e-05,
+      "loss": 1.2101,
+      "step": 1520
+    },
+    {
+      "epoch": 0.6852894796125254,
+      "grad_norm": 1.7612340227331704,
+      "learning_rate": 1e-05,
+      "loss": 1.2334,
+      "step": 1521
+    },
+    {
+      "epoch": 0.6857400315386348,
+      "grad_norm": 1.7988044156380567,
+      "learning_rate": 1e-05,
+      "loss": 1.2129,
+      "step": 1522
+    },
+    {
+      "epoch": 0.6861905834647443,
+      "grad_norm": 1.9904012936293098,
+      "learning_rate": 1e-05,
+      "loss": 1.2339,
+      "step": 1523
+    },
+    {
+      "epoch": 0.6866411353908538,
+      "grad_norm": 1.8835220745766155,
+      "learning_rate": 1e-05,
+      "loss": 1.2215,
+      "step": 1524
+    },
+    {
+      "epoch": 0.6870916873169632,
+      "grad_norm": 1.6911101505931265,
+      "learning_rate": 1e-05,
+      "loss": 1.2243,
+      "step": 1525
+    },
+    {
+      "epoch": 0.6875422392430728,
+      "grad_norm": 1.9621820059570811,
+      "learning_rate": 1e-05,
+      "loss": 1.2103,
+      "step": 1526
+    },
+    {
+      "epoch": 0.6879927911691822,
+      "grad_norm": 1.5896032578241706,
+      "learning_rate": 1e-05,
+      "loss": 1.2022,
+      "step": 1527
+    },
+    {
+      "epoch": 0.6884433430952918,
+      "grad_norm": 1.6167331916388736,
+      "learning_rate": 1e-05,
+      "loss": 1.2181,
+      "step": 1528
+    },
+    {
+      "epoch": 0.6888938950214012,
+      "grad_norm": 1.8187292950298952,
+      "learning_rate": 1e-05,
+      "loss": 1.2165,
+      "step": 1529
+    },
+    {
+      "epoch": 0.6893444469475107,
+      "grad_norm": 1.8082302021678878,
+      "learning_rate": 1e-05,
+      "loss": 1.1962,
+      "step": 1530
+    },
+    {
+      "epoch": 0.6897949988736202,
+      "grad_norm": 1.7671311925432198,
+      "learning_rate": 1e-05,
+      "loss": 1.1896,
+      "step": 1531
+    },
+    {
+      "epoch": 0.6902455507997297,
+      "grad_norm": 1.5678329258738315,
+      "learning_rate": 1e-05,
+      "loss": 1.1907,
+      "step": 1532
+    },
+    {
+      "epoch": 0.6906961027258391,
+      "grad_norm": 1.695281139458708,
+      "learning_rate": 1e-05,
+      "loss": 1.2393,
+      "step": 1533
+    },
+    {
+      "epoch": 0.6911466546519487,
+      "grad_norm": 1.7283322631447566,
+      "learning_rate": 1e-05,
+      "loss": 1.2076,
+      "step": 1534
+    },
+    {
+      "epoch": 0.6915972065780581,
+      "grad_norm": 1.8763264466537393,
+      "learning_rate": 1e-05,
+      "loss": 1.2439,
+      "step": 1535
+    },
+    {
+      "epoch": 0.6920477585041676,
+      "grad_norm": 1.6627011146386719,
+      "learning_rate": 1e-05,
+      "loss": 1.2044,
+      "step": 1536
+    },
+    {
+      "epoch": 0.6924983104302771,
+      "grad_norm": 1.7034196511170738,
+      "learning_rate": 1e-05,
+      "loss": 1.2172,
+      "step": 1537
+    },
+    {
+      "epoch": 0.6929488623563865,
+      "grad_norm": 1.6600270926222587,
+      "learning_rate": 1e-05,
+      "loss": 1.224,
+      "step": 1538
+    },
+    {
+      "epoch": 0.6933994142824961,
+      "grad_norm": 1.7470572585648798,
+      "learning_rate": 1e-05,
+      "loss": 1.2131,
+      "step": 1539
+    },
+    {
+      "epoch": 0.6938499662086055,
+      "grad_norm": 1.6756058848466726,
+      "learning_rate": 1e-05,
+      "loss": 1.2199,
+      "step": 1540
+    },
+    {
+      "epoch": 0.694300518134715,
+      "grad_norm": 1.6866150389106547,
+      "learning_rate": 1e-05,
+      "loss": 1.1937,
+      "step": 1541
+    },
+    {
+      "epoch": 0.6947510700608245,
+      "grad_norm": 1.8182803750396543,
+      "learning_rate": 1e-05,
+      "loss": 1.2329,
+      "step": 1542
+    },
+    {
+      "epoch": 0.695201621986934,
+      "grad_norm": 1.8179083545886239,
+      "learning_rate": 1e-05,
+      "loss": 1.2352,
+      "step": 1543
+    },
+    {
+      "epoch": 0.6956521739130435,
+      "grad_norm": 1.7826450014663868,
+      "learning_rate": 1e-05,
+      "loss": 1.2025,
+      "step": 1544
+    },
+    {
+      "epoch": 0.696102725839153,
+      "grad_norm": 1.6686981124730706,
+      "learning_rate": 1e-05,
+      "loss": 1.1848,
+      "step": 1545
+    },
+    {
+      "epoch": 0.6965532777652624,
+      "grad_norm": 1.8117545255777887,
+      "learning_rate": 1e-05,
+      "loss": 1.1969,
+      "step": 1546
+    },
+    {
+      "epoch": 0.697003829691372,
+      "grad_norm": 1.6142263149647902,
+      "learning_rate": 1e-05,
+      "loss": 1.1976,
+      "step": 1547
+    },
+    {
+      "epoch": 0.6974543816174814,
+      "grad_norm": 1.6610024951944695,
+      "learning_rate": 1e-05,
+      "loss": 1.2529,
+      "step": 1548
+    },
+    {
+      "epoch": 0.6979049335435908,
+      "grad_norm": 1.6013470250591175,
+      "learning_rate": 1e-05,
+      "loss": 1.1785,
+      "step": 1549
+    },
+    {
+      "epoch": 0.6983554854697004,
+      "grad_norm": 1.7140287273065915,
+      "learning_rate": 1e-05,
+      "loss": 1.1858,
+      "step": 1550
+    },
+    {
+      "epoch": 0.6988060373958098,
+      "grad_norm": 1.7758993147566013,
+      "learning_rate": 1e-05,
+      "loss": 1.2053,
+      "step": 1551
+    },
+    {
+      "epoch": 0.6992565893219194,
+      "grad_norm": 1.797803196846316,
+      "learning_rate": 1e-05,
+      "loss": 1.2303,
+      "step": 1552
+    },
+    {
+      "epoch": 0.6997071412480288,
+      "grad_norm": 1.8656479863941466,
+      "learning_rate": 1e-05,
+      "loss": 1.2259,
+      "step": 1553
+    },
+    {
+      "epoch": 0.7001576931741383,
+      "grad_norm": 1.726345010700787,
+      "learning_rate": 1e-05,
+      "loss": 1.1792,
+      "step": 1554
+    },
+    {
+      "epoch": 0.7006082451002478,
+      "grad_norm": 1.7547131217994272,
+      "learning_rate": 1e-05,
+      "loss": 1.2178,
+      "step": 1555
+    },
+    {
+      "epoch": 0.7010587970263573,
+      "grad_norm": 1.5810118037989145,
+      "learning_rate": 1e-05,
+      "loss": 1.1964,
+      "step": 1556
+    },
+    {
+      "epoch": 0.7015093489524667,
+      "grad_norm": 1.6349651639810352,
+      "learning_rate": 1e-05,
+      "loss": 1.1872,
+      "step": 1557
+    },
+    {
+      "epoch": 0.7019599008785763,
+      "grad_norm": 1.8351718999767024,
+      "learning_rate": 1e-05,
+      "loss": 1.2411,
+      "step": 1558
+    },
+    {
+      "epoch": 0.7024104528046857,
+      "grad_norm": 1.79325340556643,
+      "learning_rate": 1e-05,
+      "loss": 1.1546,
+      "step": 1559
+    },
+    {
+      "epoch": 0.7028610047307953,
+      "grad_norm": 1.6932223721377095,
+      "learning_rate": 1e-05,
+      "loss": 1.216,
+      "step": 1560
+    },
+    {
+      "epoch": 0.7033115566569047,
+      "grad_norm": 1.7572509603512143,
+      "learning_rate": 1e-05,
+      "loss": 1.1863,
+      "step": 1561
+    },
+    {
+      "epoch": 0.7037621085830142,
+      "grad_norm": 1.7345565420160527,
+      "learning_rate": 1e-05,
+      "loss": 1.1564,
+      "step": 1562
+    },
+    {
+      "epoch": 0.7042126605091237,
+      "grad_norm": 1.642208909407649,
+      "learning_rate": 1e-05,
+      "loss": 1.1407,
+      "step": 1563
+    },
+    {
+      "epoch": 0.7046632124352331,
+      "grad_norm": 1.7821873058298203,
+      "learning_rate": 1e-05,
+      "loss": 1.1878,
+      "step": 1564
+    },
+    {
+      "epoch": 0.7051137643613427,
+      "grad_norm": 1.6436486535365389,
+      "learning_rate": 1e-05,
+      "loss": 1.1418,
+      "step": 1565
+    },
+    {
+      "epoch": 0.7055643162874521,
+      "grad_norm": 1.7879936125811284,
+      "learning_rate": 1e-05,
+      "loss": 1.2328,
+      "step": 1566
+    },
+    {
+      "epoch": 0.7060148682135616,
+      "grad_norm": 1.7849417260891618,
+      "learning_rate": 1e-05,
+      "loss": 1.1732,
+      "step": 1567
+    },
+    {
+      "epoch": 0.7064654201396711,
+      "grad_norm": 1.5640189291154702,
+      "learning_rate": 1e-05,
+      "loss": 1.1912,
+      "step": 1568
+    },
+    {
+      "epoch": 0.7069159720657806,
+      "grad_norm": 1.847230070236875,
+      "learning_rate": 1e-05,
+      "loss": 1.2405,
+      "step": 1569
+    },
+    {
+      "epoch": 0.70736652399189,
+      "grad_norm": 1.7915035073553303,
+      "learning_rate": 1e-05,
+      "loss": 1.248,
+      "step": 1570
+    },
+    {
+      "epoch": 0.7078170759179996,
+      "grad_norm": 1.669729647831837,
+      "learning_rate": 1e-05,
+      "loss": 1.228,
+      "step": 1571
+    },
+    {
+      "epoch": 0.708267627844109,
+      "grad_norm": 1.753022180495031,
+      "learning_rate": 1e-05,
+      "loss": 1.195,
+      "step": 1572
+    },
+    {
+      "epoch": 0.7087181797702186,
+      "grad_norm": 1.8953343083160659,
+      "learning_rate": 1e-05,
+      "loss": 1.2168,
+      "step": 1573
+    },
+    {
+      "epoch": 0.709168731696328,
+      "grad_norm": 1.7374549036280298,
+      "learning_rate": 1e-05,
+      "loss": 1.1432,
+      "step": 1574
+    },
+    {
+      "epoch": 0.7096192836224375,
+      "grad_norm": 1.8178560560782415,
+      "learning_rate": 1e-05,
+      "loss": 1.2117,
+      "step": 1575
+    },
+    {
+      "epoch": 0.710069835548547,
+      "grad_norm": 1.7536064319434517,
+      "learning_rate": 1e-05,
+      "loss": 1.188,
+      "step": 1576
+    },
+    {
+      "epoch": 0.7105203874746564,
+      "grad_norm": 1.8131392516496259,
+      "learning_rate": 1e-05,
+      "loss": 1.1563,
+      "step": 1577
+    },
+    {
+      "epoch": 0.7109709394007659,
+      "grad_norm": 1.8091956809767324,
+      "learning_rate": 1e-05,
+      "loss": 1.1893,
+      "step": 1578
+    },
+    {
+      "epoch": 0.7114214913268754,
+      "grad_norm": 1.6431862046099364,
+      "learning_rate": 1e-05,
+      "loss": 1.2216,
+      "step": 1579
+    },
+    {
+      "epoch": 0.7118720432529849,
+      "grad_norm": 1.7559439281770257,
+      "learning_rate": 1e-05,
+      "loss": 1.1981,
+      "step": 1580
+    },
+    {
+      "epoch": 0.7123225951790944,
+      "grad_norm": 1.8701009634372734,
+      "learning_rate": 1e-05,
+      "loss": 1.1879,
+      "step": 1581
+    },
+    {
+      "epoch": 0.7127731471052039,
+      "grad_norm": 1.7142600163138828,
+      "learning_rate": 1e-05,
+      "loss": 1.2113,
+      "step": 1582
+    },
+    {
+      "epoch": 0.7132236990313133,
+      "grad_norm": 1.9174721675261706,
+      "learning_rate": 1e-05,
+      "loss": 1.2255,
+      "step": 1583
+    },
+    {
+      "epoch": 0.7136742509574229,
+      "grad_norm": 1.7559177222668094,
+      "learning_rate": 1e-05,
+      "loss": 1.1863,
+      "step": 1584
+    },
+    {
+      "epoch": 0.7141248028835323,
+      "grad_norm": 1.748516334016906,
+      "learning_rate": 1e-05,
+      "loss": 1.2024,
+      "step": 1585
+    },
+    {
+      "epoch": 0.7145753548096418,
+      "grad_norm": 1.7933589848387923,
+      "learning_rate": 1e-05,
+      "loss": 1.1801,
+      "step": 1586
+    },
+    {
+      "epoch": 0.7150259067357513,
+      "grad_norm": 1.9270765133493692,
+      "learning_rate": 1e-05,
+      "loss": 1.1939,
+      "step": 1587
+    },
+    {
+      "epoch": 0.7154764586618608,
+      "grad_norm": 1.7346018389981674,
+      "learning_rate": 1e-05,
+      "loss": 1.2283,
+      "step": 1588
+    },
+    {
+      "epoch": 0.7159270105879703,
+      "grad_norm": 1.7325242832750585,
+      "learning_rate": 1e-05,
+      "loss": 1.1963,
+      "step": 1589
+    },
+    {
+      "epoch": 0.7163775625140798,
+      "grad_norm": 1.7444000627612912,
+      "learning_rate": 1e-05,
+      "loss": 1.1531,
+      "step": 1590
+    },
+    {
+      "epoch": 0.7168281144401892,
+      "grad_norm": 1.8912072864723126,
+      "learning_rate": 1e-05,
+      "loss": 1.1566,
+      "step": 1591
+    },
+    {
+      "epoch": 0.7172786663662987,
+      "grad_norm": 1.733590476278565,
+      "learning_rate": 1e-05,
+      "loss": 1.2231,
+      "step": 1592
+    },
+    {
+      "epoch": 0.7177292182924082,
+      "grad_norm": 1.749969276816314,
+      "learning_rate": 1e-05,
+      "loss": 1.2064,
+      "step": 1593
+    },
+    {
+      "epoch": 0.7181797702185176,
+      "grad_norm": 1.7989817552283915,
+      "learning_rate": 1e-05,
+      "loss": 1.2495,
+      "step": 1594
+    },
+    {
+      "epoch": 0.7186303221446272,
+      "grad_norm": 1.6821062511170501,
+      "learning_rate": 1e-05,
+      "loss": 1.2158,
+      "step": 1595
+    },
+    {
+      "epoch": 0.7190808740707366,
+      "grad_norm": 1.724992081467261,
+      "learning_rate": 1e-05,
+      "loss": 1.1243,
+      "step": 1596
+    },
+    {
+      "epoch": 0.7195314259968462,
+      "grad_norm": 1.5558985298158818,
+      "learning_rate": 1e-05,
+      "loss": 1.1994,
+      "step": 1597
+    },
+    {
+      "epoch": 0.7199819779229556,
+      "grad_norm": 1.6563535657227137,
+      "learning_rate": 1e-05,
+      "loss": 1.2341,
+      "step": 1598
+    },
+    {
+      "epoch": 0.7204325298490651,
+      "grad_norm": 1.710169957463851,
+      "learning_rate": 1e-05,
+      "loss": 1.2425,
+      "step": 1599
+    },
+    {
+      "epoch": 0.7208830817751746,
+      "grad_norm": 1.7955061952535722,
+      "learning_rate": 1e-05,
+      "loss": 1.217,
+      "step": 1600
+    },
+    {
+      "epoch": 0.7213336337012841,
+      "grad_norm": 1.963471144681431,
+      "learning_rate": 1e-05,
+      "loss": 1.1847,
+      "step": 1601
+    },
+    {
+      "epoch": 0.7217841856273935,
+      "grad_norm": 1.7220523584163132,
+      "learning_rate": 1e-05,
+      "loss": 1.2386,
+      "step": 1602
+    },
+    {
+      "epoch": 0.7222347375535031,
+      "grad_norm": 1.6971781187080373,
+      "learning_rate": 1e-05,
+      "loss": 1.2152,
+      "step": 1603
+    },
+    {
+      "epoch": 0.7226852894796125,
+      "grad_norm": 1.7770582851966428,
+      "learning_rate": 1e-05,
+      "loss": 1.1843,
+      "step": 1604
+    },
+    {
+      "epoch": 0.723135841405722,
+      "grad_norm": 1.79805152863676,
+      "learning_rate": 1e-05,
+      "loss": 1.2518,
+      "step": 1605
+    },
+    {
+      "epoch": 0.7235863933318315,
+      "grad_norm": 1.9077231530916592,
+      "learning_rate": 1e-05,
+      "loss": 1.1668,
+      "step": 1606
+    },
+    {
+      "epoch": 0.7240369452579409,
+      "grad_norm": 1.5492100134600801,
+      "learning_rate": 1e-05,
+      "loss": 1.2113,
+      "step": 1607
+    },
+    {
+      "epoch": 0.7244874971840505,
+      "grad_norm": 1.9142045705248958,
+      "learning_rate": 1e-05,
+      "loss": 1.2466,
+      "step": 1608
+    },
+    {
+      "epoch": 0.7249380491101599,
+      "grad_norm": 1.6101992830507001,
+      "learning_rate": 1e-05,
+      "loss": 1.1995,
+      "step": 1609
+    },
+    {
+      "epoch": 0.7253886010362695,
+      "grad_norm": 1.7515548027062005,
+      "learning_rate": 1e-05,
+      "loss": 1.1555,
+      "step": 1610
+    },
+    {
+      "epoch": 0.7258391529623789,
+      "grad_norm": 1.6762276735393804,
+      "learning_rate": 1e-05,
+      "loss": 1.229,
+      "step": 1611
+    },
+    {
+      "epoch": 0.7262897048884884,
+      "grad_norm": 1.627656753569499,
+      "learning_rate": 1e-05,
+      "loss": 1.1657,
+      "step": 1612
+    },
+    {
+      "epoch": 0.7267402568145979,
+      "grad_norm": 1.7496013678598379,
+      "learning_rate": 1e-05,
+      "loss": 1.2081,
+      "step": 1613
+    },
+    {
+      "epoch": 0.7271908087407074,
+      "grad_norm": 1.8545784833141947,
+      "learning_rate": 1e-05,
+      "loss": 1.2059,
+      "step": 1614
+    },
+    {
+      "epoch": 0.7276413606668168,
+      "grad_norm": 2.00188202430048,
+      "learning_rate": 1e-05,
+      "loss": 1.2155,
+      "step": 1615
+    },
+    {
+      "epoch": 0.7280919125929264,
+      "grad_norm": 1.9014407802443913,
+      "learning_rate": 1e-05,
+      "loss": 1.2485,
+      "step": 1616
+    },
+    {
+      "epoch": 0.7285424645190358,
+      "grad_norm": 1.6826040941086706,
+      "learning_rate": 1e-05,
+      "loss": 1.2039,
+      "step": 1617
+    },
+    {
+      "epoch": 0.7289930164451452,
+      "grad_norm": 1.7676232423354608,
+      "learning_rate": 1e-05,
+      "loss": 1.2278,
+      "step": 1618
+    },
+    {
+      "epoch": 0.7294435683712548,
+      "grad_norm": 1.7829262719988204,
+      "learning_rate": 1e-05,
+      "loss": 1.1751,
+      "step": 1619
+    },
+    {
+      "epoch": 0.7298941202973642,
+      "grad_norm": 1.6967225202117673,
+      "learning_rate": 1e-05,
+      "loss": 1.2246,
+      "step": 1620
+    },
+    {
+      "epoch": 0.7303446722234738,
+      "grad_norm": 1.7824532206264245,
+      "learning_rate": 1e-05,
+      "loss": 1.1883,
+      "step": 1621
+    },
+    {
+      "epoch": 0.7307952241495832,
+      "grad_norm": 1.8185538924815645,
+      "learning_rate": 1e-05,
+      "loss": 1.2015,
+      "step": 1622
+    },
+    {
+      "epoch": 0.7312457760756927,
+      "grad_norm": 1.6654605627242565,
+      "learning_rate": 1e-05,
+      "loss": 1.1439,
+      "step": 1623
+    },
+    {
+      "epoch": 0.7316963280018022,
+      "grad_norm": 1.8029330897676907,
+      "learning_rate": 1e-05,
+      "loss": 1.2079,
+      "step": 1624
+    },
+    {
+      "epoch": 0.7321468799279117,
+      "grad_norm": 1.7215814233611655,
+      "learning_rate": 1e-05,
+      "loss": 1.1704,
+      "step": 1625
+    },
+    {
+      "epoch": 0.7325974318540212,
+      "grad_norm": 1.6040166139979504,
+      "learning_rate": 1e-05,
+      "loss": 1.1985,
+      "step": 1626
+    },
+    {
+      "epoch": 0.7330479837801307,
+      "grad_norm": 1.7937428268112714,
+      "learning_rate": 1e-05,
+      "loss": 1.2049,
+      "step": 1627
+    },
+    {
+      "epoch": 0.7334985357062401,
+      "grad_norm": 1.9822342995484614,
+      "learning_rate": 1e-05,
+      "loss": 1.1894,
+      "step": 1628
+    },
+    {
+      "epoch": 0.7339490876323497,
+      "grad_norm": 1.7672706967543768,
+      "learning_rate": 1e-05,
+      "loss": 1.2336,
+      "step": 1629
+    },
+    {
+      "epoch": 0.7343996395584591,
+      "grad_norm": 1.631220283803437,
+      "learning_rate": 1e-05,
+      "loss": 1.2061,
+      "step": 1630
+    },
+    {
+      "epoch": 0.7348501914845686,
+      "grad_norm": 1.7624571913302236,
+      "learning_rate": 1e-05,
+      "loss": 1.2216,
+      "step": 1631
+    },
+    {
+      "epoch": 0.7353007434106781,
+      "grad_norm": 1.7053116058692035,
+      "learning_rate": 1e-05,
+      "loss": 1.1874,
+      "step": 1632
+    },
+    {
+      "epoch": 0.7357512953367875,
+      "grad_norm": 1.783535417096981,
+      "learning_rate": 1e-05,
+      "loss": 1.2203,
+      "step": 1633
+    },
+    {
+      "epoch": 0.7362018472628971,
+      "grad_norm": 1.847392777457146,
+      "learning_rate": 1e-05,
+      "loss": 1.2229,
+      "step": 1634
+    },
+    {
+      "epoch": 0.7366523991890065,
+      "grad_norm": 1.9496620431006855,
+      "learning_rate": 1e-05,
+      "loss": 1.2311,
+      "step": 1635
+    },
+    {
+      "epoch": 0.737102951115116,
+      "grad_norm": 1.8380697829735286,
+      "learning_rate": 1e-05,
+      "loss": 1.2176,
+      "step": 1636
+    },
+    {
+      "epoch": 0.7375535030412255,
+      "grad_norm": 1.5333611622857826,
+      "learning_rate": 1e-05,
+      "loss": 1.1909,
+      "step": 1637
+    },
+    {
+      "epoch": 0.738004054967335,
+      "grad_norm": 1.9704628520280674,
+      "learning_rate": 1e-05,
+      "loss": 1.1882,
+      "step": 1638
+    },
+    {
+      "epoch": 0.7384546068934444,
+      "grad_norm": 1.7125255731444782,
+      "learning_rate": 1e-05,
+      "loss": 1.1701,
+      "step": 1639
+    },
+    {
+      "epoch": 0.738905158819554,
+      "grad_norm": 1.7479134783417707,
+      "learning_rate": 1e-05,
+      "loss": 1.2084,
+      "step": 1640
+    },
+    {
+      "epoch": 0.7393557107456634,
+      "grad_norm": 1.6900244842691479,
+      "learning_rate": 1e-05,
+      "loss": 1.1688,
+      "step": 1641
+    },
+    {
+      "epoch": 0.739806262671773,
+      "grad_norm": 1.7304043413819028,
+      "learning_rate": 1e-05,
+      "loss": 1.2006,
+      "step": 1642
+    },
+    {
+      "epoch": 0.7402568145978824,
+      "grad_norm": 1.7264034979823348,
+      "learning_rate": 1e-05,
+      "loss": 1.2405,
+      "step": 1643
+    },
+    {
+      "epoch": 0.7407073665239919,
+      "grad_norm": 1.7754883669138675,
+      "learning_rate": 1e-05,
+      "loss": 1.1632,
+      "step": 1644
+    },
+    {
+      "epoch": 0.7411579184501014,
+      "grad_norm": 1.6539240626012026,
+      "learning_rate": 1e-05,
+      "loss": 1.2114,
+      "step": 1645
+    },
+    {
+      "epoch": 0.7416084703762108,
+      "grad_norm": 1.7747504452166638,
+      "learning_rate": 1e-05,
+      "loss": 1.1517,
+      "step": 1646
+    },
+    {
+      "epoch": 0.7420590223023203,
+      "grad_norm": 1.6843163661125733,
+      "learning_rate": 1e-05,
+      "loss": 1.1739,
+      "step": 1647
+    },
+    {
+      "epoch": 0.7425095742284298,
+      "grad_norm": 1.654873351842606,
+      "learning_rate": 1e-05,
+      "loss": 1.2159,
+      "step": 1648
+    },
+    {
+      "epoch": 0.7429601261545393,
+      "grad_norm": 1.7765851776191028,
+      "learning_rate": 1e-05,
+      "loss": 1.1683,
+      "step": 1649
+    },
+    {
+      "epoch": 0.7434106780806488,
+      "grad_norm": 1.766473346190109,
+      "learning_rate": 1e-05,
+      "loss": 1.1727,
+      "step": 1650
+    },
+    {
+      "epoch": 0.7438612300067583,
+      "grad_norm": 1.7732215218572276,
+      "learning_rate": 1e-05,
+      "loss": 1.1981,
+      "step": 1651
+    },
+    {
+      "epoch": 0.7443117819328677,
+      "grad_norm": 1.765700001263744,
+      "learning_rate": 1e-05,
+      "loss": 1.199,
+      "step": 1652
+    },
+    {
+      "epoch": 0.7447623338589773,
+      "grad_norm": 1.6173817287431251,
+      "learning_rate": 1e-05,
+      "loss": 1.1757,
+      "step": 1653
+    },
+    {
+      "epoch": 0.7452128857850867,
+      "grad_norm": 1.7428494378215253,
+      "learning_rate": 1e-05,
+      "loss": 1.215,
+      "step": 1654
+    },
+    {
+      "epoch": 0.7456634377111963,
+      "grad_norm": 1.814149946273443,
+      "learning_rate": 1e-05,
+      "loss": 1.173,
+      "step": 1655
+    },
+    {
+      "epoch": 0.7461139896373057,
+      "grad_norm": 1.7816463532898834,
+      "learning_rate": 1e-05,
+      "loss": 1.1542,
+      "step": 1656
+    },
+    {
+      "epoch": 0.7465645415634152,
+      "grad_norm": 1.6724681088945228,
+      "learning_rate": 1e-05,
+      "loss": 1.151,
+      "step": 1657
+    },
+    {
+      "epoch": 0.7470150934895247,
+      "grad_norm": 1.7304730114747535,
+      "learning_rate": 1e-05,
+      "loss": 1.1957,
+      "step": 1658
+    },
+    {
+      "epoch": 0.7474656454156342,
+      "grad_norm": 1.9411514227103825,
+      "learning_rate": 1e-05,
+      "loss": 1.1814,
+      "step": 1659
+    },
+    {
+      "epoch": 0.7479161973417436,
+      "grad_norm": 1.568979236716191,
+      "learning_rate": 1e-05,
+      "loss": 1.2002,
+      "step": 1660
+    },
+    {
+      "epoch": 0.7483667492678531,
+      "grad_norm": 1.6982277329516287,
+      "learning_rate": 1e-05,
+      "loss": 1.1841,
+      "step": 1661
+    },
+    {
+      "epoch": 0.7488173011939626,
+      "grad_norm": 1.6759893107103216,
+      "learning_rate": 1e-05,
+      "loss": 1.2325,
+      "step": 1662
+    },
+    {
+      "epoch": 0.749267853120072,
+      "grad_norm": 1.6620514013752155,
+      "learning_rate": 1e-05,
+      "loss": 1.1932,
+      "step": 1663
+    },
+    {
+      "epoch": 0.7497184050461816,
+      "grad_norm": 1.8423705266471422,
+      "learning_rate": 1e-05,
+      "loss": 1.2196,
+      "step": 1664
+    },
+    {
+      "epoch": 0.750168956972291,
+      "grad_norm": 1.708307386102823,
+      "learning_rate": 1e-05,
+      "loss": 1.1776,
+      "step": 1665
+    },
+    {
+      "epoch": 0.7506195088984006,
+      "grad_norm": 1.6022053767174338,
+      "learning_rate": 1e-05,
+      "loss": 1.2495,
+      "step": 1666
+    },
+    {
+      "epoch": 0.75107006082451,
+      "grad_norm": 1.561987979726797,
+      "learning_rate": 1e-05,
+      "loss": 1.1684,
+      "step": 1667
+    },
+    {
+      "epoch": 0.7515206127506195,
+      "grad_norm": 1.8992134971270394,
+      "learning_rate": 1e-05,
+      "loss": 1.1751,
+      "step": 1668
+    },
+    {
+      "epoch": 0.751971164676729,
+      "grad_norm": 1.8111560222561016,
+      "learning_rate": 1e-05,
+      "loss": 1.1868,
+      "step": 1669
+    },
+    {
+      "epoch": 0.7524217166028385,
+      "grad_norm": 1.8190244202436154,
+      "learning_rate": 1e-05,
+      "loss": 1.2032,
+      "step": 1670
+    },
+    {
+      "epoch": 0.752872268528948,
+      "grad_norm": 1.6469750299329777,
+      "learning_rate": 1e-05,
+      "loss": 1.2034,
+      "step": 1671
+    },
+    {
+      "epoch": 0.7533228204550575,
+      "grad_norm": 1.8926937557156034,
+      "learning_rate": 1e-05,
+      "loss": 1.2443,
+      "step": 1672
+    },
+    {
+      "epoch": 0.7537733723811669,
+      "grad_norm": 1.8153985153002195,
+      "learning_rate": 1e-05,
+      "loss": 1.2156,
+      "step": 1673
+    },
+    {
+      "epoch": 0.7542239243072764,
+      "grad_norm": 1.7348525698050452,
+      "learning_rate": 1e-05,
+      "loss": 1.2106,
+      "step": 1674
+    },
+    {
+      "epoch": 0.7546744762333859,
+      "grad_norm": 1.661136537982251,
+      "learning_rate": 1e-05,
+      "loss": 1.169,
+      "step": 1675
+    },
+    {
+      "epoch": 0.7551250281594953,
+      "grad_norm": 1.6169169082935835,
+      "learning_rate": 1e-05,
+      "loss": 1.2337,
+      "step": 1676
+    },
+    {
+      "epoch": 0.7555755800856049,
+      "grad_norm": 1.7406684017522813,
+      "learning_rate": 1e-05,
+      "loss": 1.2192,
+      "step": 1677
+    },
+    {
+      "epoch": 0.7560261320117143,
+      "grad_norm": 1.8106982172244952,
+      "learning_rate": 1e-05,
+      "loss": 1.1707,
+      "step": 1678
+    },
+    {
+      "epoch": 0.7564766839378239,
+      "grad_norm": 1.6359043165430858,
+      "learning_rate": 1e-05,
+      "loss": 1.2168,
+      "step": 1679
+    },
+    {
+      "epoch": 0.7569272358639333,
+      "grad_norm": 1.7071707918397376,
+      "learning_rate": 1e-05,
+      "loss": 1.2194,
+      "step": 1680
+    },
+    {
+      "epoch": 0.7573777877900428,
+      "grad_norm": 1.8100793765227274,
+      "learning_rate": 1e-05,
+      "loss": 1.1894,
+      "step": 1681
+    },
+    {
+      "epoch": 0.7578283397161523,
+      "grad_norm": 1.838337039847543,
+      "learning_rate": 1e-05,
+      "loss": 1.2753,
+      "step": 1682
+    },
+    {
+      "epoch": 0.7582788916422618,
+      "grad_norm": 1.6451626473007042,
+      "learning_rate": 1e-05,
+      "loss": 1.1912,
+      "step": 1683
+    },
+    {
+      "epoch": 0.7587294435683712,
+      "grad_norm": 1.7967639278107814,
+      "learning_rate": 1e-05,
+      "loss": 1.2071,
+      "step": 1684
+    },
+    {
+      "epoch": 0.7591799954944808,
+      "grad_norm": 1.6660682047340702,
+      "learning_rate": 1e-05,
+      "loss": 1.1666,
+      "step": 1685
+    },
+    {
+      "epoch": 0.7596305474205902,
+      "grad_norm": 1.644060554767546,
+      "learning_rate": 1e-05,
+      "loss": 1.2144,
+      "step": 1686
+    },
+    {
+      "epoch": 0.7600810993466998,
+      "grad_norm": 1.7529047759013408,
+      "learning_rate": 1e-05,
+      "loss": 1.1907,
+      "step": 1687
+    },
+    {
+      "epoch": 0.7605316512728092,
+      "grad_norm": 1.7081032194106698,
+      "learning_rate": 1e-05,
+      "loss": 1.205,
+      "step": 1688
+    },
+    {
+      "epoch": 0.7609822031989186,
+      "grad_norm": 1.6534657086617957,
+      "learning_rate": 1e-05,
+      "loss": 1.1923,
+      "step": 1689
+    },
+    {
+      "epoch": 0.7614327551250282,
+      "grad_norm": 1.7220403657284369,
+      "learning_rate": 1e-05,
+      "loss": 1.1939,
+      "step": 1690
+    },
+    {
+      "epoch": 0.7618833070511376,
+      "grad_norm": 1.6483255299122568,
+      "learning_rate": 1e-05,
+      "loss": 1.1956,
+      "step": 1691
+    },
+    {
+      "epoch": 0.7623338589772471,
+      "grad_norm": 1.796404617306694,
+      "learning_rate": 1e-05,
+      "loss": 1.2375,
+      "step": 1692
+    },
+    {
+      "epoch": 0.7627844109033566,
+      "grad_norm": 1.7868456616778812,
+      "learning_rate": 1e-05,
+      "loss": 1.1456,
+      "step": 1693
+    },
+    {
+      "epoch": 0.7632349628294661,
+      "grad_norm": 1.7165716057910696,
+      "learning_rate": 1e-05,
+      "loss": 1.1772,
+      "step": 1694
+    },
+    {
+      "epoch": 0.7636855147555756,
+      "grad_norm": 1.7780125742011994,
+      "learning_rate": 1e-05,
+      "loss": 1.1915,
+      "step": 1695
+    },
+    {
+      "epoch": 0.7641360666816851,
+      "grad_norm": 1.6989972107664026,
+      "learning_rate": 1e-05,
+      "loss": 1.2096,
+      "step": 1696
+    },
+    {
+      "epoch": 0.7645866186077945,
+      "grad_norm": 1.6383750155560448,
+      "learning_rate": 1e-05,
+      "loss": 1.235,
+      "step": 1697
+    },
+    {
+      "epoch": 0.7650371705339041,
+      "grad_norm": 1.8560501612834974,
+      "learning_rate": 1e-05,
+      "loss": 1.2285,
+      "step": 1698
+    },
+    {
+      "epoch": 0.7654877224600135,
+      "grad_norm": 1.909815939253733,
+      "learning_rate": 1e-05,
+      "loss": 1.2085,
+      "step": 1699
+    },
+    {
+      "epoch": 0.765938274386123,
+      "grad_norm": 1.7146072974709152,
+      "learning_rate": 1e-05,
+      "loss": 1.1551,
+      "step": 1700
+    },
+    {
+      "epoch": 0.7663888263122325,
+      "grad_norm": 1.7092825861043606,
+      "learning_rate": 1e-05,
+      "loss": 1.1833,
+      "step": 1701
+    },
+    {
+      "epoch": 0.7668393782383419,
+      "grad_norm": 1.9280825836498556,
+      "learning_rate": 1e-05,
+      "loss": 1.1698,
+      "step": 1702
+    },
+    {
+      "epoch": 0.7672899301644515,
+      "grad_norm": 1.5595640210310495,
+      "learning_rate": 1e-05,
+      "loss": 1.2188,
+      "step": 1703
+    },
+    {
+      "epoch": 0.7677404820905609,
+      "grad_norm": 1.5818772470951465,
+      "learning_rate": 1e-05,
+      "loss": 1.1521,
+      "step": 1704
+    },
+    {
+      "epoch": 0.7681910340166704,
+      "grad_norm": 1.7331777117877314,
+      "learning_rate": 1e-05,
+      "loss": 1.2049,
+      "step": 1705
+    },
+    {
+      "epoch": 0.7686415859427799,
+      "grad_norm": 1.7539450858019312,
+      "learning_rate": 1e-05,
+      "loss": 1.2063,
+      "step": 1706
+    },
+    {
+      "epoch": 0.7690921378688894,
+      "grad_norm": 1.7403663376766787,
+      "learning_rate": 1e-05,
+      "loss": 1.1681,
+      "step": 1707
+    },
+    {
+      "epoch": 0.7695426897949988,
+      "grad_norm": 1.7345378343179203,
+      "learning_rate": 1e-05,
+      "loss": 1.1693,
+      "step": 1708
+    },
+    {
+      "epoch": 0.7699932417211084,
+      "grad_norm": 1.81637035399044,
+      "learning_rate": 1e-05,
+      "loss": 1.2041,
+      "step": 1709
+    },
+    {
+      "epoch": 0.7704437936472178,
+      "grad_norm": 1.6417819735404637,
+      "learning_rate": 1e-05,
+      "loss": 1.1634,
+      "step": 1710
+    },
+    {
+      "epoch": 0.7708943455733274,
+      "grad_norm": 1.8461910039021008,
+      "learning_rate": 1e-05,
+      "loss": 1.1926,
+      "step": 1711
+    },
+    {
+      "epoch": 0.7713448974994368,
+      "grad_norm": 1.6440018690387175,
+      "learning_rate": 1e-05,
+      "loss": 1.1833,
+      "step": 1712
+    },
+    {
+      "epoch": 0.7717954494255463,
+      "grad_norm": 1.8508994498444773,
+      "learning_rate": 1e-05,
+      "loss": 1.2303,
+      "step": 1713
+    },
+    {
+      "epoch": 0.7722460013516558,
+      "grad_norm": 1.7443234713383082,
+      "learning_rate": 1e-05,
+      "loss": 1.2142,
+      "step": 1714
+    },
+    {
+      "epoch": 0.7726965532777652,
+      "grad_norm": 1.7168353461896915,
+      "learning_rate": 1e-05,
+      "loss": 1.2048,
+      "step": 1715
+    },
+    {
+      "epoch": 0.7731471052038748,
+      "grad_norm": 1.6190431946454633,
+      "learning_rate": 1e-05,
+      "loss": 1.1731,
+      "step": 1716
+    },
+    {
+      "epoch": 0.7735976571299842,
+      "grad_norm": 1.819388151551019,
+      "learning_rate": 1e-05,
+      "loss": 1.199,
+      "step": 1717
+    },
+    {
+      "epoch": 0.7740482090560937,
+      "grad_norm": 1.7030116397503874,
+      "learning_rate": 1e-05,
+      "loss": 1.1851,
+      "step": 1718
+    },
+    {
+      "epoch": 0.7744987609822032,
+      "grad_norm": 1.7229972907895592,
+      "learning_rate": 1e-05,
+      "loss": 1.2051,
+      "step": 1719
+    },
+    {
+      "epoch": 0.7749493129083127,
+      "grad_norm": 1.7504320739342736,
+      "learning_rate": 1e-05,
+      "loss": 1.2057,
+      "step": 1720
+    },
+    {
+      "epoch": 0.7753998648344221,
+      "grad_norm": 1.6323365236334266,
+      "learning_rate": 1e-05,
+      "loss": 1.1922,
+      "step": 1721
+    },
+    {
+      "epoch": 0.7758504167605317,
+      "grad_norm": 1.7472841784426592,
+      "learning_rate": 1e-05,
+      "loss": 1.1621,
+      "step": 1722
+    },
+    {
+      "epoch": 0.7763009686866411,
+      "grad_norm": 1.6314516314515433,
+      "learning_rate": 1e-05,
+      "loss": 1.1741,
+      "step": 1723
+    },
+    {
+      "epoch": 0.7767515206127507,
+      "grad_norm": 1.7208094873956943,
+      "learning_rate": 1e-05,
+      "loss": 1.2228,
+      "step": 1724
+    },
+    {
+      "epoch": 0.7772020725388601,
+      "grad_norm": 1.6848426955294054,
+      "learning_rate": 1e-05,
+      "loss": 1.1896,
+      "step": 1725
+    },
+    {
+      "epoch": 0.7776526244649696,
+      "grad_norm": 1.866021427559413,
+      "learning_rate": 1e-05,
+      "loss": 1.2043,
+      "step": 1726
+    },
+    {
+      "epoch": 0.7781031763910791,
+      "grad_norm": 1.8912442275828933,
+      "learning_rate": 1e-05,
+      "loss": 1.1582,
+      "step": 1727
+    },
+    {
+      "epoch": 0.7785537283171886,
+      "grad_norm": 1.71422552420474,
+      "learning_rate": 1e-05,
+      "loss": 1.1946,
+      "step": 1728
+    },
+    {
+      "epoch": 0.779004280243298,
+      "grad_norm": 1.6427767378374356,
+      "learning_rate": 1e-05,
+      "loss": 1.2069,
+      "step": 1729
+    },
+    {
+      "epoch": 0.7794548321694075,
+      "grad_norm": 1.6066707395149764,
+      "learning_rate": 1e-05,
+      "loss": 1.1584,
+      "step": 1730
+    },
+    {
+      "epoch": 0.779905384095517,
+      "grad_norm": 1.7181703507041595,
+      "learning_rate": 1e-05,
+      "loss": 1.1939,
+      "step": 1731
+    },
+    {
+      "epoch": 0.7803559360216265,
+      "grad_norm": 1.7829887511924638,
+      "learning_rate": 1e-05,
+      "loss": 1.1991,
+      "step": 1732
+    },
+    {
+      "epoch": 0.780806487947736,
+      "grad_norm": 1.673584325888017,
+      "learning_rate": 1e-05,
+      "loss": 1.2075,
+      "step": 1733
+    },
+    {
+      "epoch": 0.7812570398738454,
+      "grad_norm": 1.757262862429782,
+      "learning_rate": 1e-05,
+      "loss": 1.243,
+      "step": 1734
+    },
+    {
+      "epoch": 0.781707591799955,
+      "grad_norm": 1.8297059704504663,
+      "learning_rate": 1e-05,
+      "loss": 1.2154,
+      "step": 1735
+    },
+    {
+      "epoch": 0.7821581437260644,
+      "grad_norm": 1.7922089407986124,
+      "learning_rate": 1e-05,
+      "loss": 1.2178,
+      "step": 1736
+    },
+    {
+      "epoch": 0.782608695652174,
+      "grad_norm": 1.699372079542919,
+      "learning_rate": 1e-05,
+      "loss": 1.1847,
+      "step": 1737
+    },
+    {
+      "epoch": 0.7830592475782834,
+      "grad_norm": 1.812467744918672,
+      "learning_rate": 1e-05,
+      "loss": 1.1993,
+      "step": 1738
+    },
+    {
+      "epoch": 0.7835097995043929,
+      "grad_norm": 1.806682015541317,
+      "learning_rate": 1e-05,
+      "loss": 1.2317,
+      "step": 1739
+    },
+    {
+      "epoch": 0.7839603514305024,
+      "grad_norm": 2.0428325187686154,
+      "learning_rate": 1e-05,
+      "loss": 1.22,
+      "step": 1740
+    },
+    {
+      "epoch": 0.7844109033566119,
+      "grad_norm": 1.6579826323082347,
+      "learning_rate": 1e-05,
+      "loss": 1.152,
+      "step": 1741
+    },
+    {
+      "epoch": 0.7848614552827213,
+      "grad_norm": 1.6688203186414754,
+      "learning_rate": 1e-05,
+      "loss": 1.2005,
+      "step": 1742
+    },
+    {
+      "epoch": 0.7853120072088308,
+      "grad_norm": 1.646632420536205,
+      "learning_rate": 1e-05,
+      "loss": 1.1693,
+      "step": 1743
+    },
+    {
+      "epoch": 0.7857625591349403,
+      "grad_norm": 1.7542126824558943,
+      "learning_rate": 1e-05,
+      "loss": 1.1658,
+      "step": 1744
+    },
+    {
+      "epoch": 0.7862131110610497,
+      "grad_norm": 1.9774026203805404,
+      "learning_rate": 1e-05,
+      "loss": 1.1946,
+      "step": 1745
+    },
+    {
+      "epoch": 0.7866636629871593,
+      "grad_norm": 1.9150771509861109,
+      "learning_rate": 1e-05,
+      "loss": 1.1811,
+      "step": 1746
+    },
+    {
+      "epoch": 0.7871142149132687,
+      "grad_norm": 1.9746730698587291,
+      "learning_rate": 1e-05,
+      "loss": 1.2175,
+      "step": 1747
+    },
+    {
+      "epoch": 0.7875647668393783,
+      "grad_norm": 1.8540179428694414,
+      "learning_rate": 1e-05,
+      "loss": 1.2127,
+      "step": 1748
+    },
+    {
+      "epoch": 0.7880153187654877,
+      "grad_norm": 1.7878971194265023,
+      "learning_rate": 1e-05,
+      "loss": 1.1843,
+      "step": 1749
+    },
+    {
+      "epoch": 0.7884658706915972,
+      "grad_norm": 1.7091001177940246,
+      "learning_rate": 1e-05,
+      "loss": 1.2027,
+      "step": 1750
+    },
+    {
+      "epoch": 0.7889164226177067,
+      "grad_norm": 1.6117635444294602,
+      "learning_rate": 1e-05,
+      "loss": 1.2083,
+      "step": 1751
+    },
+    {
+      "epoch": 0.7893669745438162,
+      "grad_norm": 1.8820711179069227,
+      "learning_rate": 1e-05,
+      "loss": 1.2228,
+      "step": 1752
+    },
+    {
+      "epoch": 0.7898175264699256,
+      "grad_norm": 1.6934011531025193,
+      "learning_rate": 1e-05,
+      "loss": 1.1865,
+      "step": 1753
+    },
+    {
+      "epoch": 0.7902680783960352,
+      "grad_norm": 1.74142677650193,
+      "learning_rate": 1e-05,
+      "loss": 1.2068,
+      "step": 1754
+    },
+    {
+      "epoch": 0.7907186303221446,
+      "grad_norm": 1.8215817246840316,
+      "learning_rate": 1e-05,
+      "loss": 1.1876,
+      "step": 1755
+    },
+    {
+      "epoch": 0.7911691822482542,
+      "grad_norm": 1.7720699713860941,
+      "learning_rate": 1e-05,
+      "loss": 1.2015,
+      "step": 1756
+    },
+    {
+      "epoch": 0.7916197341743636,
+      "grad_norm": 1.7469765363624727,
+      "learning_rate": 1e-05,
+      "loss": 1.1414,
+      "step": 1757
+    },
+    {
+      "epoch": 0.792070286100473,
+      "grad_norm": 1.8180964693581891,
+      "learning_rate": 1e-05,
+      "loss": 1.2038,
+      "step": 1758
+    },
+    {
+      "epoch": 0.7925208380265826,
+      "grad_norm": 1.6796050239341567,
+      "learning_rate": 1e-05,
+      "loss": 1.2388,
+      "step": 1759
+    },
+    {
+      "epoch": 0.792971389952692,
+      "grad_norm": 1.8410353568960462,
+      "learning_rate": 1e-05,
+      "loss": 1.1842,
+      "step": 1760
+    },
+    {
+      "epoch": 0.7934219418788016,
+      "grad_norm": 1.7180786562319967,
+      "learning_rate": 1e-05,
+      "loss": 1.2057,
+      "step": 1761
+    },
+    {
+      "epoch": 0.793872493804911,
+      "grad_norm": 1.7048127655518062,
+      "learning_rate": 1e-05,
+      "loss": 1.1976,
+      "step": 1762
+    },
+    {
+      "epoch": 0.7943230457310205,
+      "grad_norm": 1.875493721146123,
+      "learning_rate": 1e-05,
+      "loss": 1.2299,
+      "step": 1763
+    },
+    {
+      "epoch": 0.79477359765713,
+      "grad_norm": 1.6894023148868924,
+      "learning_rate": 1e-05,
+      "loss": 1.1831,
+      "step": 1764
+    },
+    {
+      "epoch": 0.7952241495832395,
+      "grad_norm": 1.7007836557399247,
+      "learning_rate": 1e-05,
+      "loss": 1.1997,
+      "step": 1765
+    },
+    {
+      "epoch": 0.7956747015093489,
+      "grad_norm": 1.7234616576952273,
+      "learning_rate": 1e-05,
+      "loss": 1.1705,
+      "step": 1766
+    },
+    {
+      "epoch": 0.7961252534354585,
+      "grad_norm": 1.6785410428689458,
+      "learning_rate": 1e-05,
+      "loss": 1.2251,
+      "step": 1767
+    },
+    {
+      "epoch": 0.7965758053615679,
+      "grad_norm": 1.7915367124382264,
+      "learning_rate": 1e-05,
+      "loss": 1.2132,
+      "step": 1768
+    },
+    {
+      "epoch": 0.7970263572876775,
+      "grad_norm": 1.7408723133985373,
+      "learning_rate": 1e-05,
+      "loss": 1.18,
+      "step": 1769
+    },
+    {
+      "epoch": 0.7974769092137869,
+      "grad_norm": 1.579411643669021,
+      "learning_rate": 1e-05,
+      "loss": 1.1705,
+      "step": 1770
+    },
+    {
+      "epoch": 0.7979274611398963,
+      "grad_norm": 1.7059046806169196,
+      "learning_rate": 1e-05,
+      "loss": 1.2293,
+      "step": 1771
+    },
+    {
+      "epoch": 0.7983780130660059,
+      "grad_norm": 1.556289164485916,
+      "learning_rate": 1e-05,
+      "loss": 1.2469,
+      "step": 1772
+    },
+    {
+      "epoch": 0.7988285649921153,
+      "grad_norm": 1.80268197301653,
+      "learning_rate": 1e-05,
+      "loss": 1.2354,
+      "step": 1773
+    },
+    {
+      "epoch": 0.7992791169182248,
+      "grad_norm": 1.6570194827647189,
+      "learning_rate": 1e-05,
+      "loss": 1.2282,
+      "step": 1774
+    },
+    {
+      "epoch": 0.7997296688443343,
+      "grad_norm": 1.6781340182826867,
+      "learning_rate": 1e-05,
+      "loss": 1.1614,
+      "step": 1775
+    },
+    {
+      "epoch": 0.8001802207704438,
+      "grad_norm": 1.9247040424846475,
+      "learning_rate": 1e-05,
+      "loss": 1.179,
+      "step": 1776
+    },
+    {
+      "epoch": 0.8006307726965533,
+      "grad_norm": 1.7275158966315673,
+      "learning_rate": 1e-05,
+      "loss": 1.2007,
+      "step": 1777
+    },
+    {
+      "epoch": 0.8010813246226628,
+      "grad_norm": 1.752167686702428,
+      "learning_rate": 1e-05,
+      "loss": 1.1741,
+      "step": 1778
+    },
+    {
+      "epoch": 0.8015318765487722,
+      "grad_norm": 1.6709785896835017,
+      "learning_rate": 1e-05,
+      "loss": 1.2334,
+      "step": 1779
+    },
+    {
+      "epoch": 0.8019824284748818,
+      "grad_norm": 1.7216061976974615,
+      "learning_rate": 1e-05,
+      "loss": 1.1788,
+      "step": 1780
+    },
+    {
+      "epoch": 0.8024329804009912,
+      "grad_norm": 1.659990474406928,
+      "learning_rate": 1e-05,
+      "loss": 1.1738,
+      "step": 1781
+    },
+    {
+      "epoch": 0.8028835323271007,
+      "grad_norm": 1.7641067294313082,
+      "learning_rate": 1e-05,
+      "loss": 1.1689,
+      "step": 1782
+    },
+    {
+      "epoch": 0.8033340842532102,
+      "grad_norm": 1.8687111423180383,
+      "learning_rate": 1e-05,
+      "loss": 1.2167,
+      "step": 1783
+    },
+    {
+      "epoch": 0.8037846361793196,
+      "grad_norm": 1.6467592276784964,
+      "learning_rate": 1e-05,
+      "loss": 1.1728,
+      "step": 1784
+    },
+    {
+      "epoch": 0.8042351881054292,
+      "grad_norm": 1.8845041751538107,
+      "learning_rate": 1e-05,
+      "loss": 1.1963,
+      "step": 1785
+    },
+    {
+      "epoch": 0.8046857400315386,
+      "grad_norm": 1.6606200852666957,
+      "learning_rate": 1e-05,
+      "loss": 1.2419,
+      "step": 1786
+    },
+    {
+      "epoch": 0.8051362919576481,
+      "grad_norm": 1.7195875080928165,
+      "learning_rate": 1e-05,
+      "loss": 1.2053,
+      "step": 1787
+    },
+    {
+      "epoch": 0.8055868438837576,
+      "grad_norm": 1.7450042789170206,
+      "learning_rate": 1e-05,
+      "loss": 1.1653,
+      "step": 1788
+    },
+    {
+      "epoch": 0.8060373958098671,
+      "grad_norm": 1.607903307986601,
+      "learning_rate": 1e-05,
+      "loss": 1.2252,
+      "step": 1789
+    },
+    {
+      "epoch": 0.8064879477359765,
+      "grad_norm": 1.6291198311827304,
+      "learning_rate": 1e-05,
+      "loss": 1.1807,
+      "step": 1790
+    },
+    {
+      "epoch": 0.8069384996620861,
+      "grad_norm": 1.7570954418935782,
+      "learning_rate": 1e-05,
+      "loss": 1.1372,
+      "step": 1791
+    },
+    {
+      "epoch": 0.8073890515881955,
+      "grad_norm": 1.6940183248513656,
+      "learning_rate": 1e-05,
+      "loss": 1.1764,
+      "step": 1792
+    },
+    {
+      "epoch": 0.8078396035143051,
+      "grad_norm": 1.6715246144432852,
+      "learning_rate": 1e-05,
+      "loss": 1.1682,
+      "step": 1793
+    },
+    {
+      "epoch": 0.8082901554404145,
+      "grad_norm": 1.622579390147315,
+      "learning_rate": 1e-05,
+      "loss": 1.1824,
+      "step": 1794
+    },
+    {
+      "epoch": 0.808740707366524,
+      "grad_norm": 1.6122444971448244,
+      "learning_rate": 1e-05,
+      "loss": 1.1938,
+      "step": 1795
+    },
+    {
+      "epoch": 0.8091912592926335,
+      "grad_norm": 1.673915226874395,
+      "learning_rate": 1e-05,
+      "loss": 1.2029,
+      "step": 1796
+    },
+    {
+      "epoch": 0.809641811218743,
+      "grad_norm": 1.6289651786498598,
+      "learning_rate": 1e-05,
+      "loss": 1.221,
+      "step": 1797
+    },
+    {
+      "epoch": 0.8100923631448524,
+      "grad_norm": 1.8574225182558701,
+      "learning_rate": 1e-05,
+      "loss": 1.1884,
+      "step": 1798
+    },
+    {
+      "epoch": 0.8105429150709619,
+      "grad_norm": 1.7759255504996503,
+      "learning_rate": 1e-05,
+      "loss": 1.1806,
+      "step": 1799
+    },
+    {
+      "epoch": 0.8109934669970714,
+      "grad_norm": 1.6065268073103804,
+      "learning_rate": 1e-05,
+      "loss": 1.1875,
+      "step": 1800
+    },
+    {
+      "epoch": 0.8114440189231809,
+      "grad_norm": 1.8073045942786299,
+      "learning_rate": 1e-05,
+      "loss": 1.1745,
+      "step": 1801
+    },
+    {
+      "epoch": 0.8118945708492904,
+      "grad_norm": 1.5086722747406716,
+      "learning_rate": 1e-05,
+      "loss": 1.206,
+      "step": 1802
+    },
+    {
+      "epoch": 0.8123451227753998,
+      "grad_norm": 1.6410734376734943,
+      "learning_rate": 1e-05,
+      "loss": 1.2089,
+      "step": 1803
+    },
+    {
+      "epoch": 0.8127956747015094,
+      "grad_norm": 1.6696031782373022,
+      "learning_rate": 1e-05,
+      "loss": 1.2102,
+      "step": 1804
+    },
+    {
+      "epoch": 0.8132462266276188,
+      "grad_norm": 1.72933695131258,
+      "learning_rate": 1e-05,
+      "loss": 1.1674,
+      "step": 1805
+    },
+    {
+      "epoch": 0.8136967785537284,
+      "grad_norm": 1.6713287790566425,
+      "learning_rate": 1e-05,
+      "loss": 1.1421,
+      "step": 1806
+    },
+    {
+      "epoch": 0.8141473304798378,
+      "grad_norm": 1.555689526971806,
+      "learning_rate": 1e-05,
+      "loss": 1.2045,
+      "step": 1807
+    },
+    {
+      "epoch": 0.8145978824059473,
+      "grad_norm": 1.7107816419209407,
+      "learning_rate": 1e-05,
+      "loss": 1.254,
+      "step": 1808
+    },
+    {
+      "epoch": 0.8150484343320568,
+      "grad_norm": 1.6978797539741441,
+      "learning_rate": 1e-05,
+      "loss": 1.1999,
+      "step": 1809
+    },
+    {
+      "epoch": 0.8154989862581663,
+      "grad_norm": 1.7342072656013985,
+      "learning_rate": 1e-05,
+      "loss": 1.161,
+      "step": 1810
+    },
+    {
+      "epoch": 0.8159495381842757,
+      "grad_norm": 1.7215478015529795,
+      "learning_rate": 1e-05,
+      "loss": 1.1938,
+      "step": 1811
+    },
+    {
+      "epoch": 0.8164000901103852,
+      "grad_norm": 1.8168631125275996,
+      "learning_rate": 1e-05,
+      "loss": 1.1559,
+      "step": 1812
+    },
+    {
+      "epoch": 0.8168506420364947,
+      "grad_norm": 1.6730864612404082,
+      "learning_rate": 1e-05,
+      "loss": 1.2148,
+      "step": 1813
+    },
+    {
+      "epoch": 0.8173011939626041,
+      "grad_norm": 1.794971257052246,
+      "learning_rate": 1e-05,
+      "loss": 1.1674,
+      "step": 1814
+    },
+    {
+      "epoch": 0.8177517458887137,
+      "grad_norm": 1.833276009223717,
+      "learning_rate": 1e-05,
+      "loss": 1.1529,
+      "step": 1815
+    },
+    {
+      "epoch": 0.8182022978148231,
+      "grad_norm": 1.755511955138208,
+      "learning_rate": 1e-05,
+      "loss": 1.1837,
+      "step": 1816
+    },
+    {
+      "epoch": 0.8186528497409327,
+      "grad_norm": 1.8568036299566444,
+      "learning_rate": 1e-05,
+      "loss": 1.2397,
+      "step": 1817
+    },
+    {
+      "epoch": 0.8191034016670421,
+      "grad_norm": 1.730612575135526,
+      "learning_rate": 1e-05,
+      "loss": 1.2339,
+      "step": 1818
+    },
+    {
+      "epoch": 0.8195539535931516,
+      "grad_norm": 1.8010708713586632,
+      "learning_rate": 1e-05,
+      "loss": 1.2136,
+      "step": 1819
+    },
+    {
+      "epoch": 0.8200045055192611,
+      "grad_norm": 1.7522398396456647,
+      "learning_rate": 1e-05,
+      "loss": 1.1606,
+      "step": 1820
+    },
+    {
+      "epoch": 0.8204550574453706,
+      "grad_norm": 1.7871145554348764,
+      "learning_rate": 1e-05,
+      "loss": 1.2302,
+      "step": 1821
+    },
+    {
+      "epoch": 0.82090560937148,
+      "grad_norm": 1.5881068278749262,
+      "learning_rate": 1e-05,
+      "loss": 1.2138,
+      "step": 1822
+    },
+    {
+      "epoch": 0.8213561612975896,
+      "grad_norm": 1.8361277896093844,
+      "learning_rate": 1e-05,
+      "loss": 1.1689,
+      "step": 1823
+    },
+    {
+      "epoch": 0.821806713223699,
+      "grad_norm": 1.5772982242765807,
+      "learning_rate": 1e-05,
+      "loss": 1.2147,
+      "step": 1824
+    },
+    {
+      "epoch": 0.8222572651498086,
+      "grad_norm": 1.6907674511280488,
+      "learning_rate": 1e-05,
+      "loss": 1.1599,
+      "step": 1825
+    },
+    {
+      "epoch": 0.822707817075918,
+      "grad_norm": 1.6915559112820235,
+      "learning_rate": 1e-05,
+      "loss": 1.2331,
+      "step": 1826
+    },
+    {
+      "epoch": 0.8231583690020274,
+      "grad_norm": 1.7271988450470883,
+      "learning_rate": 1e-05,
+      "loss": 1.2356,
+      "step": 1827
+    },
+    {
+      "epoch": 0.823608920928137,
+      "grad_norm": 1.6495148178297072,
+      "learning_rate": 1e-05,
+      "loss": 1.1727,
+      "step": 1828
+    },
+    {
+      "epoch": 0.8240594728542464,
+      "grad_norm": 1.5429531334169466,
+      "learning_rate": 1e-05,
+      "loss": 1.2308,
+      "step": 1829
+    },
+    {
+      "epoch": 0.824510024780356,
+      "grad_norm": 1.6528167550186859,
+      "learning_rate": 1e-05,
+      "loss": 1.221,
+      "step": 1830
+    },
+    {
+      "epoch": 0.8249605767064654,
+      "grad_norm": 1.7028016391985052,
+      "learning_rate": 1e-05,
+      "loss": 1.2303,
+      "step": 1831
+    },
+    {
+      "epoch": 0.8254111286325749,
+      "grad_norm": 1.6519648763476065,
+      "learning_rate": 1e-05,
+      "loss": 1.1734,
+      "step": 1832
+    },
+    {
+      "epoch": 0.8258616805586844,
+      "grad_norm": 1.7831930073112774,
+      "learning_rate": 1e-05,
+      "loss": 1.1996,
+      "step": 1833
+    },
+    {
+      "epoch": 0.8263122324847939,
+      "grad_norm": 1.569943737291928,
+      "learning_rate": 1e-05,
+      "loss": 1.166,
+      "step": 1834
+    },
+    {
+      "epoch": 0.8267627844109033,
+      "grad_norm": 1.749828893584674,
+      "learning_rate": 1e-05,
+      "loss": 1.1314,
+      "step": 1835
+    },
+    {
+      "epoch": 0.8272133363370129,
+      "grad_norm": 1.49333625744596,
+      "learning_rate": 1e-05,
+      "loss": 1.1943,
+      "step": 1836
+    },
+    {
+      "epoch": 0.8276638882631223,
+      "grad_norm": 1.701130962185549,
+      "learning_rate": 1e-05,
+      "loss": 1.1847,
+      "step": 1837
+    },
+    {
+      "epoch": 0.8281144401892319,
+      "grad_norm": 1.7597771845463768,
+      "learning_rate": 1e-05,
+      "loss": 1.2266,
+      "step": 1838
+    },
+    {
+      "epoch": 0.8285649921153413,
+      "grad_norm": 1.6822295737976576,
+      "learning_rate": 1e-05,
+      "loss": 1.1882,
+      "step": 1839
+    },
+    {
+      "epoch": 0.8290155440414507,
+      "grad_norm": 1.7578620391435773,
+      "learning_rate": 1e-05,
+      "loss": 1.1696,
+      "step": 1840
+    },
+    {
+      "epoch": 0.8294660959675603,
+      "grad_norm": 1.7515912207578863,
+      "learning_rate": 1e-05,
+      "loss": 1.1676,
+      "step": 1841
+    },
+    {
+      "epoch": 0.8299166478936697,
+      "grad_norm": 1.7140724085890104,
+      "learning_rate": 1e-05,
+      "loss": 1.1911,
+      "step": 1842
+    },
+    {
+      "epoch": 0.8303671998197792,
+      "grad_norm": 1.639741990912705,
+      "learning_rate": 1e-05,
+      "loss": 1.1805,
+      "step": 1843
+    },
+    {
+      "epoch": 0.8308177517458887,
+      "grad_norm": 1.5237143397840152,
+      "learning_rate": 1e-05,
+      "loss": 1.1468,
+      "step": 1844
+    },
+    {
+      "epoch": 0.8312683036719982,
+      "grad_norm": 1.8327144880859152,
+      "learning_rate": 1e-05,
+      "loss": 1.2012,
+      "step": 1845
+    },
+    {
+      "epoch": 0.8317188555981077,
+      "grad_norm": 1.784939482353893,
+      "learning_rate": 1e-05,
+      "loss": 1.2006,
+      "step": 1846
+    },
+    {
+      "epoch": 0.8321694075242172,
+      "grad_norm": 1.6765228375047296,
+      "learning_rate": 1e-05,
+      "loss": 1.1382,
+      "step": 1847
+    },
+    {
+      "epoch": 0.8326199594503266,
+      "grad_norm": 1.6807484329690845,
+      "learning_rate": 1e-05,
+      "loss": 1.1825,
+      "step": 1848
+    },
+    {
+      "epoch": 0.8330705113764362,
+      "grad_norm": 1.6556503576432688,
+      "learning_rate": 1e-05,
+      "loss": 1.1642,
+      "step": 1849
+    },
+    {
+      "epoch": 0.8335210633025456,
+      "grad_norm": 1.8457718442378077,
+      "learning_rate": 1e-05,
+      "loss": 1.2018,
+      "step": 1850
+    },
+    {
+      "epoch": 0.8339716152286551,
+      "grad_norm": 1.7386324171860397,
+      "learning_rate": 1e-05,
+      "loss": 1.2086,
+      "step": 1851
+    },
+    {
+      "epoch": 0.8344221671547646,
+      "grad_norm": 1.7834582941382333,
+      "learning_rate": 1e-05,
+      "loss": 1.1935,
+      "step": 1852
+    },
+    {
+      "epoch": 0.8348727190808741,
+      "grad_norm": 1.5578726204866538,
+      "learning_rate": 1e-05,
+      "loss": 1.1852,
+      "step": 1853
+    },
+    {
+      "epoch": 0.8353232710069836,
+      "grad_norm": 1.645561699294569,
+      "learning_rate": 1e-05,
+      "loss": 1.1927,
+      "step": 1854
+    },
+    {
+      "epoch": 0.835773822933093,
+      "grad_norm": 1.628572224310932,
+      "learning_rate": 1e-05,
+      "loss": 1.2055,
+      "step": 1855
+    },
+    {
+      "epoch": 0.8362243748592025,
+      "grad_norm": 1.7569072711235298,
+      "learning_rate": 1e-05,
+      "loss": 1.2308,
+      "step": 1856
+    },
+    {
+      "epoch": 0.836674926785312,
+      "grad_norm": 1.6861628602830974,
+      "learning_rate": 1e-05,
+      "loss": 1.187,
+      "step": 1857
+    },
+    {
+      "epoch": 0.8371254787114215,
+      "grad_norm": 1.6393350502863893,
+      "learning_rate": 1e-05,
+      "loss": 1.19,
+      "step": 1858
+    },
+    {
+      "epoch": 0.8375760306375309,
+      "grad_norm": 1.7432904377757557,
+      "learning_rate": 1e-05,
+      "loss": 1.1788,
+      "step": 1859
+    },
+    {
+      "epoch": 0.8380265825636405,
+      "grad_norm": 1.6497139581484863,
+      "learning_rate": 1e-05,
+      "loss": 1.1959,
+      "step": 1860
+    },
+    {
+      "epoch": 0.8384771344897499,
+      "grad_norm": 1.7551824550788737,
+      "learning_rate": 1e-05,
+      "loss": 1.2106,
+      "step": 1861
+    },
+    {
+      "epoch": 0.8389276864158595,
+      "grad_norm": 1.7720142131386385,
+      "learning_rate": 1e-05,
+      "loss": 1.1487,
+      "step": 1862
+    },
+    {
+      "epoch": 0.8393782383419689,
+      "grad_norm": 1.6155334459556048,
+      "learning_rate": 1e-05,
+      "loss": 1.2256,
+      "step": 1863
+    },
+    {
+      "epoch": 0.8398287902680784,
+      "grad_norm": 1.7381417026213979,
+      "learning_rate": 1e-05,
+      "loss": 1.1875,
+      "step": 1864
+    },
+    {
+      "epoch": 0.8402793421941879,
+      "grad_norm": 1.6518693993382525,
+      "learning_rate": 1e-05,
+      "loss": 1.187,
+      "step": 1865
+    },
+    {
+      "epoch": 0.8407298941202974,
+      "grad_norm": 1.7581564397885485,
+      "learning_rate": 1e-05,
+      "loss": 1.1662,
+      "step": 1866
+    },
+    {
+      "epoch": 0.8411804460464068,
+      "grad_norm": 1.7407070017452284,
+      "learning_rate": 1e-05,
+      "loss": 1.1695,
+      "step": 1867
+    },
+    {
+      "epoch": 0.8416309979725163,
+      "grad_norm": 1.696193545487891,
+      "learning_rate": 1e-05,
+      "loss": 1.2007,
+      "step": 1868
+    },
+    {
+      "epoch": 0.8420815498986258,
+      "grad_norm": 1.675694788012286,
+      "learning_rate": 1e-05,
+      "loss": 1.2231,
+      "step": 1869
+    },
+    {
+      "epoch": 0.8425321018247353,
+      "grad_norm": 1.732119679262739,
+      "learning_rate": 1e-05,
+      "loss": 1.1746,
+      "step": 1870
+    },
+    {
+      "epoch": 0.8429826537508448,
+      "grad_norm": 1.7359533865764607,
+      "learning_rate": 1e-05,
+      "loss": 1.2073,
+      "step": 1871
+    },
+    {
+      "epoch": 0.8434332056769542,
+      "grad_norm": 1.6255239440036602,
+      "learning_rate": 1e-05,
+      "loss": 1.2126,
+      "step": 1872
+    },
+    {
+      "epoch": 0.8438837576030638,
+      "grad_norm": 1.603652875682749,
+      "learning_rate": 1e-05,
+      "loss": 1.1704,
+      "step": 1873
+    },
+    {
+      "epoch": 0.8443343095291732,
+      "grad_norm": 1.7171999769504531,
+      "learning_rate": 1e-05,
+      "loss": 1.1826,
+      "step": 1874
+    },
+    {
+      "epoch": 0.8447848614552828,
+      "grad_norm": 1.7695406001085834,
+      "learning_rate": 1e-05,
+      "loss": 1.1859,
+      "step": 1875
+    },
+    {
+      "epoch": 0.8452354133813922,
+      "grad_norm": 1.608752510331531,
+      "learning_rate": 1e-05,
+      "loss": 1.1943,
+      "step": 1876
+    },
+    {
+      "epoch": 0.8456859653075017,
+      "grad_norm": 1.7423124752559593,
+      "learning_rate": 1e-05,
+      "loss": 1.2367,
+      "step": 1877
+    },
+    {
+      "epoch": 0.8461365172336112,
+      "grad_norm": 1.68481610871632,
+      "learning_rate": 1e-05,
+      "loss": 1.1829,
+      "step": 1878
+    },
+    {
+      "epoch": 0.8465870691597207,
+      "grad_norm": 1.8527501439577547,
+      "learning_rate": 1e-05,
+      "loss": 1.2476,
+      "step": 1879
+    },
+    {
+      "epoch": 0.8470376210858301,
+      "grad_norm": 1.9968473859003448,
+      "learning_rate": 1e-05,
+      "loss": 1.1167,
+      "step": 1880
+    },
+    {
+      "epoch": 0.8474881730119396,
+      "grad_norm": 1.7678137412578787,
+      "learning_rate": 1e-05,
+      "loss": 1.1633,
+      "step": 1881
+    },
+    {
+      "epoch": 0.8479387249380491,
+      "grad_norm": 1.556477379139996,
+      "learning_rate": 1e-05,
+      "loss": 1.1773,
+      "step": 1882
+    },
+    {
+      "epoch": 0.8483892768641585,
+      "grad_norm": 1.7924479162860214,
+      "learning_rate": 1e-05,
+      "loss": 1.1828,
+      "step": 1883
+    },
+    {
+      "epoch": 0.8488398287902681,
+      "grad_norm": 1.6292880584988436,
+      "learning_rate": 1e-05,
+      "loss": 1.1717,
+      "step": 1884
+    },
+    {
+      "epoch": 0.8492903807163775,
+      "grad_norm": 1.8370736351623214,
+      "learning_rate": 1e-05,
+      "loss": 1.1136,
+      "step": 1885
+    },
+    {
+      "epoch": 0.8497409326424871,
+      "grad_norm": 1.712098595924913,
+      "learning_rate": 1e-05,
+      "loss": 1.2015,
+      "step": 1886
+    },
+    {
+      "epoch": 0.8501914845685965,
+      "grad_norm": 1.732302685864171,
+      "learning_rate": 1e-05,
+      "loss": 1.1663,
+      "step": 1887
+    },
+    {
+      "epoch": 0.850642036494706,
+      "grad_norm": 1.7428759852946527,
+      "learning_rate": 1e-05,
+      "loss": 1.157,
+      "step": 1888
+    },
+    {
+      "epoch": 0.8510925884208155,
+      "grad_norm": 1.6311039887003997,
+      "learning_rate": 1e-05,
+      "loss": 1.1947,
+      "step": 1889
+    },
+    {
+      "epoch": 0.851543140346925,
+      "grad_norm": 1.6693782203489504,
+      "learning_rate": 1e-05,
+      "loss": 1.1194,
+      "step": 1890
+    },
+    {
+      "epoch": 0.8519936922730345,
+      "grad_norm": 1.6151122944453822,
+      "learning_rate": 1e-05,
+      "loss": 1.2245,
+      "step": 1891
+    },
+    {
+      "epoch": 0.852444244199144,
+      "grad_norm": 1.6666911048541473,
+      "learning_rate": 1e-05,
+      "loss": 1.1455,
+      "step": 1892
+    },
+    {
+      "epoch": 0.8528947961252534,
+      "grad_norm": 1.7271702493835965,
+      "learning_rate": 1e-05,
+      "loss": 1.2418,
+      "step": 1893
+    },
+    {
+      "epoch": 0.853345348051363,
+      "grad_norm": 1.625278036644014,
+      "learning_rate": 1e-05,
+      "loss": 1.1439,
+      "step": 1894
+    },
+    {
+      "epoch": 0.8537958999774724,
+      "grad_norm": 1.8266926896290014,
+      "learning_rate": 1e-05,
+      "loss": 1.1757,
+      "step": 1895
+    },
+    {
+      "epoch": 0.8542464519035818,
+      "grad_norm": 1.8088659008337555,
+      "learning_rate": 1e-05,
+      "loss": 1.2177,
+      "step": 1896
+    },
+    {
+      "epoch": 0.8546970038296914,
+      "grad_norm": 1.5188719773527812,
+      "learning_rate": 1e-05,
+      "loss": 1.1786,
+      "step": 1897
+    },
+    {
+      "epoch": 0.8551475557558008,
+      "grad_norm": 1.698235192975051,
+      "learning_rate": 1e-05,
+      "loss": 1.1502,
+      "step": 1898
+    },
+    {
+      "epoch": 0.8555981076819104,
+      "grad_norm": 1.682321920659013,
+      "learning_rate": 1e-05,
+      "loss": 1.1755,
+      "step": 1899
+    },
+    {
+      "epoch": 0.8560486596080198,
+      "grad_norm": 1.7211441240079381,
+      "learning_rate": 1e-05,
+      "loss": 1.1948,
+      "step": 1900
+    },
+    {
+      "epoch": 0.8564992115341293,
+      "grad_norm": 1.5671613014454864,
+      "learning_rate": 1e-05,
+      "loss": 1.1532,
+      "step": 1901
+    },
+    {
+      "epoch": 0.8569497634602388,
+      "grad_norm": 1.7570383061221853,
+      "learning_rate": 1e-05,
+      "loss": 1.1508,
+      "step": 1902
+    },
+    {
+      "epoch": 0.8574003153863483,
+      "grad_norm": 1.8218014930265243,
+      "learning_rate": 1e-05,
+      "loss": 1.1678,
+      "step": 1903
+    },
+    {
+      "epoch": 0.8578508673124577,
+      "grad_norm": 1.5796342577148133,
+      "learning_rate": 1e-05,
+      "loss": 1.1813,
+      "step": 1904
+    },
+    {
+      "epoch": 0.8583014192385673,
+      "grad_norm": 1.6305157420347662,
+      "learning_rate": 1e-05,
+      "loss": 1.2015,
+      "step": 1905
+    },
+    {
+      "epoch": 0.8587519711646767,
+      "grad_norm": 1.5123328622529004,
+      "learning_rate": 1e-05,
+      "loss": 1.1888,
+      "step": 1906
+    },
+    {
+      "epoch": 0.8592025230907863,
+      "grad_norm": 1.7615806652394799,
+      "learning_rate": 1e-05,
+      "loss": 1.1527,
+      "step": 1907
+    },
+    {
+      "epoch": 0.8596530750168957,
+      "grad_norm": 1.8080875330863884,
+      "learning_rate": 1e-05,
+      "loss": 1.1483,
+      "step": 1908
+    },
+    {
+      "epoch": 0.8601036269430051,
+      "grad_norm": 1.6531915110793447,
+      "learning_rate": 1e-05,
+      "loss": 1.1528,
+      "step": 1909
+    },
+    {
+      "epoch": 0.8605541788691147,
+      "grad_norm": 1.7109784395736596,
+      "learning_rate": 1e-05,
+      "loss": 1.1782,
+      "step": 1910
+    },
+    {
+      "epoch": 0.8610047307952241,
+      "grad_norm": 1.7292321924749698,
+      "learning_rate": 1e-05,
+      "loss": 1.1707,
+      "step": 1911
+    },
+    {
+      "epoch": 0.8614552827213336,
+      "grad_norm": 1.8983091050254104,
+      "learning_rate": 1e-05,
+      "loss": 1.2272,
+      "step": 1912
+    },
+    {
+      "epoch": 0.8619058346474431,
+      "grad_norm": 1.6990088023494063,
+      "learning_rate": 1e-05,
+      "loss": 1.1365,
+      "step": 1913
+    },
+    {
+      "epoch": 0.8623563865735526,
+      "grad_norm": 1.8009079667610153,
+      "learning_rate": 1e-05,
+      "loss": 1.1821,
+      "step": 1914
+    },
+    {
+      "epoch": 0.8628069384996621,
+      "grad_norm": 1.6029370701232601,
+      "learning_rate": 1e-05,
+      "loss": 1.1858,
+      "step": 1915
+    },
+    {
+      "epoch": 0.8632574904257716,
+      "grad_norm": 1.736786043826039,
+      "learning_rate": 1e-05,
+      "loss": 1.174,
+      "step": 1916
+    },
+    {
+      "epoch": 0.863708042351881,
+      "grad_norm": 1.4943056336840133,
+      "learning_rate": 1e-05,
+      "loss": 1.1687,
+      "step": 1917
+    },
+    {
+      "epoch": 0.8641585942779906,
+      "grad_norm": 1.7782797122098448,
+      "learning_rate": 1e-05,
+      "loss": 1.137,
+      "step": 1918
+    },
+    {
+      "epoch": 0.8646091462041,
+      "grad_norm": 1.5722811399895673,
+      "learning_rate": 1e-05,
+      "loss": 1.224,
+      "step": 1919
+    },
+    {
+      "epoch": 0.8650596981302096,
+      "grad_norm": 1.8196960961003943,
+      "learning_rate": 1e-05,
+      "loss": 1.1641,
+      "step": 1920
+    },
+    {
+      "epoch": 0.865510250056319,
+      "grad_norm": 1.5014356478018795,
+      "learning_rate": 1e-05,
+      "loss": 1.2079,
+      "step": 1921
+    },
+    {
+      "epoch": 0.8659608019824285,
+      "grad_norm": 1.7155412730644062,
+      "learning_rate": 1e-05,
+      "loss": 1.1497,
+      "step": 1922
+    },
+    {
+      "epoch": 0.866411353908538,
+      "grad_norm": 1.7479774465805296,
+      "learning_rate": 1e-05,
+      "loss": 1.201,
+      "step": 1923
+    },
+    {
+      "epoch": 0.8668619058346474,
+      "grad_norm": 1.8635908641176946,
+      "learning_rate": 1e-05,
+      "loss": 1.2052,
+      "step": 1924
+    },
+    {
+      "epoch": 0.8673124577607569,
+      "grad_norm": 1.6981131344398972,
+      "learning_rate": 1e-05,
+      "loss": 1.1979,
+      "step": 1925
+    },
+    {
+      "epoch": 0.8677630096868664,
+      "grad_norm": 1.8298095483126138,
+      "learning_rate": 1e-05,
+      "loss": 1.1667,
+      "step": 1926
+    },
+    {
+      "epoch": 0.8682135616129759,
+      "grad_norm": 1.7194418701797443,
+      "learning_rate": 1e-05,
+      "loss": 1.2701,
+      "step": 1927
+    },
+    {
+      "epoch": 0.8686641135390853,
+      "grad_norm": 1.7527047305719925,
+      "learning_rate": 1e-05,
+      "loss": 1.1983,
+      "step": 1928
+    },
+    {
+      "epoch": 0.8691146654651949,
+      "grad_norm": 1.7420697895766035,
+      "learning_rate": 1e-05,
+      "loss": 1.1632,
+      "step": 1929
+    },
+    {
+      "epoch": 0.8695652173913043,
+      "grad_norm": 1.8830249678980415,
+      "learning_rate": 1e-05,
+      "loss": 1.2274,
+      "step": 1930
+    },
+    {
+      "epoch": 0.8700157693174139,
+      "grad_norm": 1.7902124544077,
+      "learning_rate": 1e-05,
+      "loss": 1.1519,
+      "step": 1931
+    },
+    {
+      "epoch": 0.8704663212435233,
+      "grad_norm": 1.8125679907386705,
+      "learning_rate": 1e-05,
+      "loss": 1.2235,
+      "step": 1932
+    },
+    {
+      "epoch": 0.8709168731696328,
+      "grad_norm": 1.5364302166060644,
+      "learning_rate": 1e-05,
+      "loss": 1.1675,
+      "step": 1933
+    },
+    {
+      "epoch": 0.8713674250957423,
+      "grad_norm": 1.71309437243926,
+      "learning_rate": 1e-05,
+      "loss": 1.1971,
+      "step": 1934
+    },
+    {
+      "epoch": 0.8718179770218518,
+      "grad_norm": 1.8382299502817083,
+      "learning_rate": 1e-05,
+      "loss": 1.1889,
+      "step": 1935
+    },
+    {
+      "epoch": 0.8722685289479613,
+      "grad_norm": 1.4800146563735435,
+      "learning_rate": 1e-05,
+      "loss": 1.1834,
+      "step": 1936
+    },
+    {
+      "epoch": 0.8727190808740707,
+      "grad_norm": 1.8419305600469726,
+      "learning_rate": 1e-05,
+      "loss": 1.1682,
+      "step": 1937
+    },
+    {
+      "epoch": 0.8731696328001802,
+      "grad_norm": 1.6402860515722093,
+      "learning_rate": 1e-05,
+      "loss": 1.191,
+      "step": 1938
+    },
+    {
+      "epoch": 0.8736201847262897,
+      "grad_norm": 1.5750853324413905,
+      "learning_rate": 1e-05,
+      "loss": 1.1931,
+      "step": 1939
+    },
+    {
+      "epoch": 0.8740707366523992,
+      "grad_norm": 1.7122071346392769,
+      "learning_rate": 1e-05,
+      "loss": 1.2012,
+      "step": 1940
+    },
+    {
+      "epoch": 0.8745212885785086,
+      "grad_norm": 1.962267954992774,
+      "learning_rate": 1e-05,
+      "loss": 1.2298,
+      "step": 1941
+    },
+    {
+      "epoch": 0.8749718405046182,
+      "grad_norm": 1.6470105267146145,
+      "learning_rate": 1e-05,
+      "loss": 1.1909,
+      "step": 1942
+    },
+    {
+      "epoch": 0.8754223924307276,
+      "grad_norm": 1.696760791120632,
+      "learning_rate": 1e-05,
+      "loss": 1.1283,
+      "step": 1943
+    },
+    {
+      "epoch": 0.8758729443568372,
+      "grad_norm": 1.6512854525210154,
+      "learning_rate": 1e-05,
+      "loss": 1.1498,
+      "step": 1944
+    },
+    {
+      "epoch": 0.8763234962829466,
+      "grad_norm": 1.7346141415486804,
+      "learning_rate": 1e-05,
+      "loss": 1.182,
+      "step": 1945
+    },
+    {
+      "epoch": 0.8767740482090561,
+      "grad_norm": 1.7333886670705998,
+      "learning_rate": 1e-05,
+      "loss": 1.175,
+      "step": 1946
+    },
+    {
+      "epoch": 0.8772246001351656,
+      "grad_norm": 1.7197986361204711,
+      "learning_rate": 1e-05,
+      "loss": 1.189,
+      "step": 1947
+    },
+    {
+      "epoch": 0.8776751520612751,
+      "grad_norm": 1.5322161639690512,
+      "learning_rate": 1e-05,
+      "loss": 1.2137,
+      "step": 1948
+    },
+    {
+      "epoch": 0.8781257039873845,
+      "grad_norm": 1.8152703348833092,
+      "learning_rate": 1e-05,
+      "loss": 1.2056,
+      "step": 1949
+    },
+    {
+      "epoch": 0.878576255913494,
+      "grad_norm": 1.572175310710983,
+      "learning_rate": 1e-05,
+      "loss": 1.2318,
+      "step": 1950
+    },
+    {
+      "epoch": 0.8790268078396035,
+      "grad_norm": 1.8533395139877775,
+      "learning_rate": 1e-05,
+      "loss": 1.2218,
+      "step": 1951
+    },
+    {
+      "epoch": 0.879477359765713,
+      "grad_norm": 1.7114207143773543,
+      "learning_rate": 1e-05,
+      "loss": 1.1862,
+      "step": 1952
+    },
+    {
+      "epoch": 0.8799279116918225,
+      "grad_norm": 2.0850669987105075,
+      "learning_rate": 1e-05,
+      "loss": 1.1992,
+      "step": 1953
+    },
+    {
+      "epoch": 0.8803784636179319,
+      "grad_norm": 1.684177597482306,
+      "learning_rate": 1e-05,
+      "loss": 1.254,
+      "step": 1954
+    },
+    {
+      "epoch": 0.8808290155440415,
+      "grad_norm": 1.5762065907916154,
+      "learning_rate": 1e-05,
+      "loss": 1.1994,
+      "step": 1955
+    },
+    {
+      "epoch": 0.8812795674701509,
+      "grad_norm": 1.791877464246084,
+      "learning_rate": 1e-05,
+      "loss": 1.1817,
+      "step": 1956
+    },
+    {
+      "epoch": 0.8817301193962604,
+      "grad_norm": 1.619852262764895,
+      "learning_rate": 1e-05,
+      "loss": 1.1919,
+      "step": 1957
+    },
+    {
+      "epoch": 0.8821806713223699,
+      "grad_norm": 1.7205647996126916,
+      "learning_rate": 1e-05,
+      "loss": 1.1754,
+      "step": 1958
+    },
+    {
+      "epoch": 0.8826312232484794,
+      "grad_norm": 1.5313646610580463,
+      "learning_rate": 1e-05,
+      "loss": 1.236,
+      "step": 1959
+    },
+    {
+      "epoch": 0.8830817751745889,
+      "grad_norm": 1.5825029652532525,
+      "learning_rate": 1e-05,
+      "loss": 1.1857,
+      "step": 1960
+    },
+    {
+      "epoch": 0.8835323271006984,
+      "grad_norm": 1.6849573240870548,
+      "learning_rate": 1e-05,
+      "loss": 1.1486,
+      "step": 1961
+    },
+    {
+      "epoch": 0.8839828790268078,
+      "grad_norm": 1.8040464344018217,
+      "learning_rate": 1e-05,
+      "loss": 1.1794,
+      "step": 1962
+    },
+    {
+      "epoch": 0.8844334309529174,
+      "grad_norm": 1.6890087972359196,
+      "learning_rate": 1e-05,
+      "loss": 1.1859,
+      "step": 1963
+    },
+    {
+      "epoch": 0.8848839828790268,
+      "grad_norm": 1.617560696442198,
+      "learning_rate": 1e-05,
+      "loss": 1.2189,
+      "step": 1964
+    },
+    {
+      "epoch": 0.8853345348051362,
+      "grad_norm": 1.6811332514125268,
+      "learning_rate": 1e-05,
+      "loss": 1.2115,
+      "step": 1965
+    },
+    {
+      "epoch": 0.8857850867312458,
+      "grad_norm": 1.7534007222902268,
+      "learning_rate": 1e-05,
+      "loss": 1.1822,
+      "step": 1966
+    },
+    {
+      "epoch": 0.8862356386573552,
+      "grad_norm": 1.5847852679799652,
+      "learning_rate": 1e-05,
+      "loss": 1.2162,
+      "step": 1967
+    },
+    {
+      "epoch": 0.8866861905834648,
+      "grad_norm": 1.7632898753820936,
+      "learning_rate": 1e-05,
+      "loss": 1.2327,
+      "step": 1968
+    },
+    {
+      "epoch": 0.8871367425095742,
+      "grad_norm": 1.6310639286552508,
+      "learning_rate": 1e-05,
+      "loss": 1.179,
+      "step": 1969
+    },
+    {
+      "epoch": 0.8875872944356837,
+      "grad_norm": 1.843730565121264,
+      "learning_rate": 1e-05,
+      "loss": 1.1949,
+      "step": 1970
+    },
+    {
+      "epoch": 0.8880378463617932,
+      "grad_norm": 1.781736616332847,
+      "learning_rate": 1e-05,
+      "loss": 1.1743,
+      "step": 1971
+    },
+    {
+      "epoch": 0.8884883982879027,
+      "grad_norm": 1.5943169480925572,
+      "learning_rate": 1e-05,
+      "loss": 1.2062,
+      "step": 1972
+    },
+    {
+      "epoch": 0.8889389502140121,
+      "grad_norm": 1.4993920792787943,
+      "learning_rate": 1e-05,
+      "loss": 1.1956,
+      "step": 1973
+    },
+    {
+      "epoch": 0.8893895021401217,
+      "grad_norm": 1.540824279950419,
+      "learning_rate": 1e-05,
+      "loss": 1.2087,
+      "step": 1974
+    },
+    {
+      "epoch": 0.8898400540662311,
+      "grad_norm": 1.7252658755827255,
+      "learning_rate": 1e-05,
+      "loss": 1.1728,
+      "step": 1975
+    },
+    {
+      "epoch": 0.8902906059923407,
+      "grad_norm": 1.7052017798554122,
+      "learning_rate": 1e-05,
+      "loss": 1.1743,
+      "step": 1976
+    },
+    {
+      "epoch": 0.8907411579184501,
+      "grad_norm": 1.6612625789667403,
+      "learning_rate": 1e-05,
+      "loss": 1.2041,
+      "step": 1977
+    },
+    {
+      "epoch": 0.8911917098445595,
+      "grad_norm": 1.6587744402696805,
+      "learning_rate": 1e-05,
+      "loss": 1.2129,
+      "step": 1978
+    },
+    {
+      "epoch": 0.8916422617706691,
+      "grad_norm": 1.6447097169325238,
+      "learning_rate": 1e-05,
+      "loss": 1.1925,
+      "step": 1979
+    },
+    {
+      "epoch": 0.8920928136967785,
+      "grad_norm": 1.65524150981378,
+      "learning_rate": 1e-05,
+      "loss": 1.1487,
+      "step": 1980
+    },
+    {
+      "epoch": 0.892543365622888,
+      "grad_norm": 1.7112097802046498,
+      "learning_rate": 1e-05,
+      "loss": 1.2032,
+      "step": 1981
+    },
+    {
+      "epoch": 0.8929939175489975,
+      "grad_norm": 1.7547125950361586,
+      "learning_rate": 1e-05,
+      "loss": 1.1676,
+      "step": 1982
+    },
+    {
+      "epoch": 0.893444469475107,
+      "grad_norm": 1.6425869215805582,
+      "learning_rate": 1e-05,
+      "loss": 1.1941,
+      "step": 1983
+    },
+    {
+      "epoch": 0.8938950214012165,
+      "grad_norm": 1.4451572493199791,
+      "learning_rate": 1e-05,
+      "loss": 1.1587,
+      "step": 1984
+    },
+    {
+      "epoch": 0.894345573327326,
+      "grad_norm": 1.572571629382231,
+      "learning_rate": 1e-05,
+      "loss": 1.1649,
+      "step": 1985
+    },
+    {
+      "epoch": 0.8947961252534354,
+      "grad_norm": 1.7864696673674751,
+      "learning_rate": 1e-05,
+      "loss": 1.182,
+      "step": 1986
+    },
+    {
+      "epoch": 0.895246677179545,
+      "grad_norm": 1.6268030280618402,
+      "learning_rate": 1e-05,
+      "loss": 1.1916,
+      "step": 1987
+    },
+    {
+      "epoch": 0.8956972291056544,
+      "grad_norm": 1.6707602189816138,
+      "learning_rate": 1e-05,
+      "loss": 1.2154,
+      "step": 1988
+    },
+    {
+      "epoch": 0.896147781031764,
+      "grad_norm": 1.6520039500127561,
+      "learning_rate": 1e-05,
+      "loss": 1.2046,
+      "step": 1989
+    },
+    {
+      "epoch": 0.8965983329578734,
+      "grad_norm": 1.8866840650686438,
+      "learning_rate": 1e-05,
+      "loss": 1.1683,
+      "step": 1990
+    },
+    {
+      "epoch": 0.8970488848839829,
+      "grad_norm": 1.6635248858057377,
+      "learning_rate": 1e-05,
+      "loss": 1.1416,
+      "step": 1991
+    },
+    {
+      "epoch": 0.8974994368100924,
+      "grad_norm": 1.6801657472532312,
+      "learning_rate": 1e-05,
+      "loss": 1.2457,
+      "step": 1992
+    },
+    {
+      "epoch": 0.8979499887362018,
+      "grad_norm": 1.7513150745030448,
+      "learning_rate": 1e-05,
+      "loss": 1.1839,
+      "step": 1993
+    },
+    {
+      "epoch": 0.8984005406623113,
+      "grad_norm": 1.7926390263455134,
+      "learning_rate": 1e-05,
+      "loss": 1.1908,
+      "step": 1994
+    },
+    {
+      "epoch": 0.8988510925884208,
+      "grad_norm": 1.7440011648485394,
+      "learning_rate": 1e-05,
+      "loss": 1.1758,
+      "step": 1995
+    },
+    {
+      "epoch": 0.8993016445145303,
+      "grad_norm": 1.802874400960212,
+      "learning_rate": 1e-05,
+      "loss": 1.153,
+      "step": 1996
+    },
+    {
+      "epoch": 0.8997521964406398,
+      "grad_norm": 1.674275456851461,
+      "learning_rate": 1e-05,
+      "loss": 1.2032,
+      "step": 1997
+    },
+    {
+      "epoch": 0.9002027483667493,
+      "grad_norm": 1.70597562331596,
+      "learning_rate": 1e-05,
+      "loss": 1.2026,
+      "step": 1998
+    },
+    {
+      "epoch": 0.9006533002928587,
+      "grad_norm": 1.6645508955826498,
+      "learning_rate": 1e-05,
+      "loss": 1.204,
+      "step": 1999
+    },
+    {
+      "epoch": 0.9011038522189683,
+      "grad_norm": 1.6437466582223337,
+      "learning_rate": 1e-05,
+      "loss": 1.1853,
+      "step": 2000
+    },
+    {
+      "epoch": 0.9015544041450777,
+      "grad_norm": 1.752087077429046,
+      "learning_rate": 1e-05,
+      "loss": 1.2628,
+      "step": 2001
+    },
+    {
+      "epoch": 0.9020049560711872,
+      "grad_norm": 1.669099591859999,
+      "learning_rate": 1e-05,
+      "loss": 1.2362,
+      "step": 2002
+    },
+    {
+      "epoch": 0.9024555079972967,
+      "grad_norm": 1.6895858952402134,
+      "learning_rate": 1e-05,
+      "loss": 1.1824,
+      "step": 2003
+    },
+    {
+      "epoch": 0.9029060599234062,
+      "grad_norm": 1.6481183332774054,
+      "learning_rate": 1e-05,
+      "loss": 1.1734,
+      "step": 2004
+    },
+    {
+      "epoch": 0.9033566118495157,
+      "grad_norm": 1.8295713923594479,
+      "learning_rate": 1e-05,
+      "loss": 1.1994,
+      "step": 2005
+    },
+    {
+      "epoch": 0.9038071637756251,
+      "grad_norm": 1.8460703606205773,
+      "learning_rate": 1e-05,
+      "loss": 1.1777,
+      "step": 2006
+    },
+    {
+      "epoch": 0.9042577157017346,
+      "grad_norm": 1.6241536868747677,
+      "learning_rate": 1e-05,
+      "loss": 1.1913,
+      "step": 2007
+    },
+    {
+      "epoch": 0.9047082676278441,
+      "grad_norm": 1.7175802542448775,
+      "learning_rate": 1e-05,
+      "loss": 1.1814,
+      "step": 2008
+    },
+    {
+      "epoch": 0.9051588195539536,
+      "grad_norm": 1.5828181067077758,
+      "learning_rate": 1e-05,
+      "loss": 1.1803,
+      "step": 2009
+    },
+    {
+      "epoch": 0.905609371480063,
+      "grad_norm": 1.8423204404690332,
+      "learning_rate": 1e-05,
+      "loss": 1.1417,
+      "step": 2010
+    },
+    {
+      "epoch": 0.9060599234061726,
+      "grad_norm": 1.683197073295049,
+      "learning_rate": 1e-05,
+      "loss": 1.1936,
+      "step": 2011
+    },
+    {
+      "epoch": 0.906510475332282,
+      "grad_norm": 1.8219461398772028,
+      "learning_rate": 1e-05,
+      "loss": 1.1845,
+      "step": 2012
+    },
+    {
+      "epoch": 0.9069610272583916,
+      "grad_norm": 1.7365303973164858,
+      "learning_rate": 1e-05,
+      "loss": 1.1968,
+      "step": 2013
+    },
+    {
+      "epoch": 0.907411579184501,
+      "grad_norm": 1.7252487536901362,
+      "learning_rate": 1e-05,
+      "loss": 1.1837,
+      "step": 2014
+    },
+    {
+      "epoch": 0.9078621311106105,
+      "grad_norm": 1.7973801355067949,
+      "learning_rate": 1e-05,
+      "loss": 1.217,
+      "step": 2015
+    },
+    {
+      "epoch": 0.90831268303672,
+      "grad_norm": 1.7309101052129412,
+      "learning_rate": 1e-05,
+      "loss": 1.1768,
+      "step": 2016
+    },
+    {
+      "epoch": 0.9087632349628295,
+      "grad_norm": 1.651613975910536,
+      "learning_rate": 1e-05,
+      "loss": 1.1767,
+      "step": 2017
+    },
+    {
+      "epoch": 0.909213786888939,
+      "grad_norm": 1.5932297428511177,
+      "learning_rate": 1e-05,
+      "loss": 1.1776,
+      "step": 2018
+    },
+    {
+      "epoch": 0.9096643388150484,
+      "grad_norm": 1.7249736812948915,
+      "learning_rate": 1e-05,
+      "loss": 1.2391,
+      "step": 2019
+    },
+    {
+      "epoch": 0.9101148907411579,
+      "grad_norm": 1.7592791830484773,
+      "learning_rate": 1e-05,
+      "loss": 1.1678,
+      "step": 2020
+    },
+    {
+      "epoch": 0.9105654426672674,
+      "grad_norm": 1.8001511327251414,
+      "learning_rate": 1e-05,
+      "loss": 1.1881,
+      "step": 2021
+    },
+    {
+      "epoch": 0.9110159945933769,
+      "grad_norm": 1.5958915091678874,
+      "learning_rate": 1e-05,
+      "loss": 1.1829,
+      "step": 2022
+    },
+    {
+      "epoch": 0.9114665465194863,
+      "grad_norm": 1.7053854208063959,
+      "learning_rate": 1e-05,
+      "loss": 1.1706,
+      "step": 2023
+    },
+    {
+      "epoch": 0.9119170984455959,
+      "grad_norm": 1.603121726297471,
+      "learning_rate": 1e-05,
+      "loss": 1.157,
+      "step": 2024
+    },
+    {
+      "epoch": 0.9123676503717053,
+      "grad_norm": 1.741748960234155,
+      "learning_rate": 1e-05,
+      "loss": 1.1776,
+      "step": 2025
+    },
+    {
+      "epoch": 0.9128182022978149,
+      "grad_norm": 1.6666156009125006,
+      "learning_rate": 1e-05,
+      "loss": 1.2061,
+      "step": 2026
+    },
+    {
+      "epoch": 0.9132687542239243,
+      "grad_norm": 1.6999991012738433,
+      "learning_rate": 1e-05,
+      "loss": 1.1603,
+      "step": 2027
+    },
+    {
+      "epoch": 0.9137193061500338,
+      "grad_norm": 1.5071174447484084,
+      "learning_rate": 1e-05,
+      "loss": 1.157,
+      "step": 2028
+    },
+    {
+      "epoch": 0.9141698580761433,
+      "grad_norm": 1.5934489986133855,
+      "learning_rate": 1e-05,
+      "loss": 1.1652,
+      "step": 2029
+    },
+    {
+      "epoch": 0.9146204100022528,
+      "grad_norm": 1.7691389722709743,
+      "learning_rate": 1e-05,
+      "loss": 1.2329,
+      "step": 2030
+    },
+    {
+      "epoch": 0.9150709619283622,
+      "grad_norm": 1.612525557681105,
+      "learning_rate": 1e-05,
+      "loss": 1.2585,
+      "step": 2031
+    },
+    {
+      "epoch": 0.9155215138544718,
+      "grad_norm": 1.823896119381473,
+      "learning_rate": 1e-05,
+      "loss": 1.1784,
+      "step": 2032
+    },
+    {
+      "epoch": 0.9159720657805812,
+      "grad_norm": 1.6001182819638253,
+      "learning_rate": 1e-05,
+      "loss": 1.1729,
+      "step": 2033
+    },
+    {
+      "epoch": 0.9164226177066906,
+      "grad_norm": 1.7114437153741997,
+      "learning_rate": 1e-05,
+      "loss": 1.2322,
+      "step": 2034
+    },
+    {
+      "epoch": 0.9168731696328002,
+      "grad_norm": 1.7578588566571445,
+      "learning_rate": 1e-05,
+      "loss": 1.1919,
+      "step": 2035
+    },
+    {
+      "epoch": 0.9173237215589096,
+      "grad_norm": 1.684811816484171,
+      "learning_rate": 1e-05,
+      "loss": 1.1927,
+      "step": 2036
+    },
+    {
+      "epoch": 0.9177742734850192,
+      "grad_norm": 1.6758870836191506,
+      "learning_rate": 1e-05,
+      "loss": 1.1887,
+      "step": 2037
+    },
+    {
+      "epoch": 0.9182248254111286,
+      "grad_norm": 1.784155879996892,
+      "learning_rate": 1e-05,
+      "loss": 1.1696,
+      "step": 2038
+    },
+    {
+      "epoch": 0.9186753773372381,
+      "grad_norm": 1.7001875941313467,
+      "learning_rate": 1e-05,
+      "loss": 1.2048,
+      "step": 2039
+    },
+    {
+      "epoch": 0.9191259292633476,
+      "grad_norm": 1.5193718812259038,
+      "learning_rate": 1e-05,
+      "loss": 1.14,
+      "step": 2040
+    },
+    {
+      "epoch": 0.9195764811894571,
+      "grad_norm": 1.7140562171701816,
+      "learning_rate": 1e-05,
+      "loss": 1.2036,
+      "step": 2041
+    },
+    {
+      "epoch": 0.9200270331155666,
+      "grad_norm": 1.6658501846960208,
+      "learning_rate": 1e-05,
+      "loss": 1.2263,
+      "step": 2042
+    },
+    {
+      "epoch": 0.9204775850416761,
+      "grad_norm": 1.6245115612898635,
+      "learning_rate": 1e-05,
+      "loss": 1.1818,
+      "step": 2043
+    },
+    {
+      "epoch": 0.9209281369677855,
+      "grad_norm": 1.7707988391102771,
+      "learning_rate": 1e-05,
+      "loss": 1.1845,
+      "step": 2044
+    },
+    {
+      "epoch": 0.9213786888938951,
+      "grad_norm": 1.700605838458128,
+      "learning_rate": 1e-05,
+      "loss": 1.2123,
+      "step": 2045
+    },
+    {
+      "epoch": 0.9218292408200045,
+      "grad_norm": 1.8486779226587806,
+      "learning_rate": 1e-05,
+      "loss": 1.1775,
+      "step": 2046
+    },
+    {
+      "epoch": 0.9222797927461139,
+      "grad_norm": 1.8102808308048854,
+      "learning_rate": 1e-05,
+      "loss": 1.1988,
+      "step": 2047
+    },
+    {
+      "epoch": 0.9227303446722235,
+      "grad_norm": 1.5288101087416415,
+      "learning_rate": 1e-05,
+      "loss": 1.1526,
+      "step": 2048
+    },
+    {
+      "epoch": 0.9231808965983329,
+      "grad_norm": 1.604554857518699,
+      "learning_rate": 1e-05,
+      "loss": 1.1632,
+      "step": 2049
+    },
+    {
+      "epoch": 0.9236314485244425,
+      "grad_norm": 1.8100618174923913,
+      "learning_rate": 1e-05,
+      "loss": 1.1515,
+      "step": 2050
+    },
+    {
+      "epoch": 0.9240820004505519,
+      "grad_norm": 1.6319685863655795,
+      "learning_rate": 1e-05,
+      "loss": 1.1523,
+      "step": 2051
+    },
+    {
+      "epoch": 0.9245325523766614,
+      "grad_norm": 1.632142134315281,
+      "learning_rate": 1e-05,
+      "loss": 1.2029,
+      "step": 2052
+    },
+    {
+      "epoch": 0.9249831043027709,
+      "grad_norm": 1.8187454131617482,
+      "learning_rate": 1e-05,
+      "loss": 1.1761,
+      "step": 2053
+    },
+    {
+      "epoch": 0.9254336562288804,
+      "grad_norm": 1.6826786299296643,
+      "learning_rate": 1e-05,
+      "loss": 1.202,
+      "step": 2054
+    },
+    {
+      "epoch": 0.9258842081549898,
+      "grad_norm": 1.7098050761568442,
+      "learning_rate": 1e-05,
+      "loss": 1.2148,
+      "step": 2055
+    },
+    {
+      "epoch": 0.9263347600810994,
+      "grad_norm": 1.639004737641357,
+      "learning_rate": 1e-05,
+      "loss": 1.171,
+      "step": 2056
+    },
+    {
+      "epoch": 0.9267853120072088,
+      "grad_norm": 1.623859948439882,
+      "learning_rate": 1e-05,
+      "loss": 1.1859,
+      "step": 2057
+    },
+    {
+      "epoch": 0.9272358639333184,
+      "grad_norm": 1.7021487486732494,
+      "learning_rate": 1e-05,
+      "loss": 1.2176,
+      "step": 2058
+    },
+    {
+      "epoch": 0.9276864158594278,
+      "grad_norm": 1.7717027001668448,
+      "learning_rate": 1e-05,
+      "loss": 1.2052,
+      "step": 2059
+    },
+    {
+      "epoch": 0.9281369677855373,
+      "grad_norm": 1.797295137865806,
+      "learning_rate": 1e-05,
+      "loss": 1.1753,
+      "step": 2060
+    },
+    {
+      "epoch": 0.9285875197116468,
+      "grad_norm": 1.6138084929771288,
+      "learning_rate": 1e-05,
+      "loss": 1.1809,
+      "step": 2061
+    },
+    {
+      "epoch": 0.9290380716377562,
+      "grad_norm": 1.698916134145838,
+      "learning_rate": 1e-05,
+      "loss": 1.1883,
+      "step": 2062
+    },
+    {
+      "epoch": 0.9294886235638657,
+      "grad_norm": 1.8018354499014477,
+      "learning_rate": 1e-05,
+      "loss": 1.1626,
+      "step": 2063
+    },
+    {
+      "epoch": 0.9299391754899752,
+      "grad_norm": 1.7321854344599588,
+      "learning_rate": 1e-05,
+      "loss": 1.2136,
+      "step": 2064
+    },
+    {
+      "epoch": 0.9303897274160847,
+      "grad_norm": 1.634004679541849,
+      "learning_rate": 1e-05,
+      "loss": 1.149,
+      "step": 2065
+    },
+    {
+      "epoch": 0.9308402793421942,
+      "grad_norm": 1.7033993770183842,
+      "learning_rate": 1e-05,
+      "loss": 1.1627,
+      "step": 2066
+    },
+    {
+      "epoch": 0.9312908312683037,
+      "grad_norm": 1.888189431846741,
+      "learning_rate": 1e-05,
+      "loss": 1.1643,
+      "step": 2067
+    },
+    {
+      "epoch": 0.9317413831944131,
+      "grad_norm": 1.705533244923517,
+      "learning_rate": 1e-05,
+      "loss": 1.1602,
+      "step": 2068
+    },
+    {
+      "epoch": 0.9321919351205227,
+      "grad_norm": 1.7096701473629878,
+      "learning_rate": 1e-05,
+      "loss": 1.1988,
+      "step": 2069
+    },
+    {
+      "epoch": 0.9326424870466321,
+      "grad_norm": 1.630666376640028,
+      "learning_rate": 1e-05,
+      "loss": 1.1713,
+      "step": 2070
+    },
+    {
+      "epoch": 0.9330930389727417,
+      "grad_norm": 1.8493121915990847,
+      "learning_rate": 1e-05,
+      "loss": 1.207,
+      "step": 2071
+    },
+    {
+      "epoch": 0.9335435908988511,
+      "grad_norm": 1.815624099029328,
+      "learning_rate": 1e-05,
+      "loss": 1.182,
+      "step": 2072
+    },
+    {
+      "epoch": 0.9339941428249606,
+      "grad_norm": 1.6384160771080598,
+      "learning_rate": 1e-05,
+      "loss": 1.1903,
+      "step": 2073
+    },
+    {
+      "epoch": 0.9344446947510701,
+      "grad_norm": 1.6288305079143965,
+      "learning_rate": 1e-05,
+      "loss": 1.1821,
+      "step": 2074
+    },
+    {
+      "epoch": 0.9348952466771795,
+      "grad_norm": 1.6490190090104768,
+      "learning_rate": 1e-05,
+      "loss": 1.1727,
+      "step": 2075
+    },
+    {
+      "epoch": 0.935345798603289,
+      "grad_norm": 1.6223468525759248,
+      "learning_rate": 1e-05,
+      "loss": 1.179,
+      "step": 2076
+    },
+    {
+      "epoch": 0.9357963505293985,
+      "grad_norm": 1.7946445207666113,
+      "learning_rate": 1e-05,
+      "loss": 1.2144,
+      "step": 2077
+    },
+    {
+      "epoch": 0.936246902455508,
+      "grad_norm": 1.6897767509695476,
+      "learning_rate": 1e-05,
+      "loss": 1.2137,
+      "step": 2078
+    },
+    {
+      "epoch": 0.9366974543816174,
+      "grad_norm": 1.6188498752212164,
+      "learning_rate": 1e-05,
+      "loss": 1.1902,
+      "step": 2079
+    },
+    {
+      "epoch": 0.937148006307727,
+      "grad_norm": 1.681733067879976,
+      "learning_rate": 1e-05,
+      "loss": 1.2483,
+      "step": 2080
+    },
+    {
+      "epoch": 0.9375985582338364,
+      "grad_norm": 1.725178267652548,
+      "learning_rate": 1e-05,
+      "loss": 1.1524,
+      "step": 2081
+    },
+    {
+      "epoch": 0.938049110159946,
+      "grad_norm": 1.7196779773962776,
+      "learning_rate": 1e-05,
+      "loss": 1.2441,
+      "step": 2082
+    },
+    {
+      "epoch": 0.9384996620860554,
+      "grad_norm": 1.72948940896615,
+      "learning_rate": 1e-05,
+      "loss": 1.1142,
+      "step": 2083
+    },
+    {
+      "epoch": 0.9389502140121649,
+      "grad_norm": 1.7511010335092747,
+      "learning_rate": 1e-05,
+      "loss": 1.2238,
+      "step": 2084
+    },
+    {
+      "epoch": 0.9394007659382744,
+      "grad_norm": 1.838057472974213,
+      "learning_rate": 1e-05,
+      "loss": 1.1773,
+      "step": 2085
+    },
+    {
+      "epoch": 0.9398513178643839,
+      "grad_norm": 1.6591311122470425,
+      "learning_rate": 1e-05,
+      "loss": 1.1552,
+      "step": 2086
+    },
+    {
+      "epoch": 0.9403018697904933,
+      "grad_norm": 1.5791720221435517,
+      "learning_rate": 1e-05,
+      "loss": 1.1942,
+      "step": 2087
+    },
+    {
+      "epoch": 0.9407524217166029,
+      "grad_norm": 1.6260408967438869,
+      "learning_rate": 1e-05,
+      "loss": 1.1934,
+      "step": 2088
+    },
+    {
+      "epoch": 0.9412029736427123,
+      "grad_norm": 1.8301203454759956,
+      "learning_rate": 1e-05,
+      "loss": 1.1426,
+      "step": 2089
+    },
+    {
+      "epoch": 0.9416535255688218,
+      "grad_norm": 1.6393692809665887,
+      "learning_rate": 1e-05,
+      "loss": 1.1764,
+      "step": 2090
+    },
+    {
+      "epoch": 0.9421040774949313,
+      "grad_norm": 1.6165840529847568,
+      "learning_rate": 1e-05,
+      "loss": 1.1811,
+      "step": 2091
+    },
+    {
+      "epoch": 0.9425546294210407,
+      "grad_norm": 1.5798081012769836,
+      "learning_rate": 1e-05,
+      "loss": 1.2356,
+      "step": 2092
+    },
+    {
+      "epoch": 0.9430051813471503,
+      "grad_norm": 1.727700337892747,
+      "learning_rate": 1e-05,
+      "loss": 1.1749,
+      "step": 2093
+    },
+    {
+      "epoch": 0.9434557332732597,
+      "grad_norm": 1.9714736909888164,
+      "learning_rate": 1e-05,
+      "loss": 1.1896,
+      "step": 2094
+    },
+    {
+      "epoch": 0.9439062851993693,
+      "grad_norm": 1.6580452596611326,
+      "learning_rate": 1e-05,
+      "loss": 1.1921,
+      "step": 2095
+    },
+    {
+      "epoch": 0.9443568371254787,
+      "grad_norm": 1.866349427049117,
+      "learning_rate": 1e-05,
+      "loss": 1.2059,
+      "step": 2096
+    },
+    {
+      "epoch": 0.9448073890515882,
+      "grad_norm": 1.7313573957932022,
+      "learning_rate": 1e-05,
+      "loss": 1.175,
+      "step": 2097
+    },
+    {
+      "epoch": 0.9452579409776977,
+      "grad_norm": 1.7221346323183881,
+      "learning_rate": 1e-05,
+      "loss": 1.1439,
+      "step": 2098
+    },
+    {
+      "epoch": 0.9457084929038072,
+      "grad_norm": 1.6381714881988407,
+      "learning_rate": 1e-05,
+      "loss": 1.2122,
+      "step": 2099
+    },
+    {
+      "epoch": 0.9461590448299166,
+      "grad_norm": 1.6447047547046818,
+      "learning_rate": 1e-05,
+      "loss": 1.2059,
+      "step": 2100
+    },
+    {
+      "epoch": 0.9466095967560262,
+      "grad_norm": 1.6758634414894065,
+      "learning_rate": 1e-05,
+      "loss": 1.0905,
+      "step": 2101
+    },
+    {
+      "epoch": 0.9470601486821356,
+      "grad_norm": 1.624740500489032,
+      "learning_rate": 1e-05,
+      "loss": 1.1452,
+      "step": 2102
+    },
+    {
+      "epoch": 0.947510700608245,
+      "grad_norm": 1.634954570703128,
+      "learning_rate": 1e-05,
+      "loss": 1.1753,
+      "step": 2103
+    },
+    {
+      "epoch": 0.9479612525343546,
+      "grad_norm": 1.8032264648997465,
+      "learning_rate": 1e-05,
+      "loss": 1.1978,
+      "step": 2104
+    },
+    {
+      "epoch": 0.948411804460464,
+      "grad_norm": 1.7215681096454558,
+      "learning_rate": 1e-05,
+      "loss": 1.177,
+      "step": 2105
+    },
+    {
+      "epoch": 0.9488623563865736,
+      "grad_norm": 1.6686611213810127,
+      "learning_rate": 1e-05,
+      "loss": 1.1717,
+      "step": 2106
+    },
+    {
+      "epoch": 0.949312908312683,
+      "grad_norm": 1.881143362237432,
+      "learning_rate": 1e-05,
+      "loss": 1.1908,
+      "step": 2107
+    },
+    {
+      "epoch": 0.9497634602387925,
+      "grad_norm": 1.6451020286185194,
+      "learning_rate": 1e-05,
+      "loss": 1.1849,
+      "step": 2108
+    },
+    {
+      "epoch": 0.950214012164902,
+      "grad_norm": 1.754167561683326,
+      "learning_rate": 1e-05,
+      "loss": 1.1952,
+      "step": 2109
+    },
+    {
+      "epoch": 0.9506645640910115,
+      "grad_norm": 1.7556968199681056,
+      "learning_rate": 1e-05,
+      "loss": 1.1717,
+      "step": 2110
+    },
+    {
+      "epoch": 0.951115116017121,
+      "grad_norm": 1.7095415110303946,
+      "learning_rate": 1e-05,
+      "loss": 1.2258,
+      "step": 2111
+    },
+    {
+      "epoch": 0.9515656679432305,
+      "grad_norm": 1.706848125511313,
+      "learning_rate": 1e-05,
+      "loss": 1.2049,
+      "step": 2112
+    },
+    {
+      "epoch": 0.9520162198693399,
+      "grad_norm": 1.5703031012611752,
+      "learning_rate": 1e-05,
+      "loss": 1.2062,
+      "step": 2113
+    },
+    {
+      "epoch": 0.9524667717954495,
+      "grad_norm": 1.5547854742449227,
+      "learning_rate": 1e-05,
+      "loss": 1.1723,
+      "step": 2114
+    },
+    {
+      "epoch": 0.9529173237215589,
+      "grad_norm": 1.5813108924809067,
+      "learning_rate": 1e-05,
+      "loss": 1.1692,
+      "step": 2115
+    },
+    {
+      "epoch": 0.9533678756476683,
+      "grad_norm": 1.6400217382575737,
+      "learning_rate": 1e-05,
+      "loss": 1.2242,
+      "step": 2116
+    },
+    {
+      "epoch": 0.9538184275737779,
+      "grad_norm": 1.719910875150272,
+      "learning_rate": 1e-05,
+      "loss": 1.2065,
+      "step": 2117
+    },
+    {
+      "epoch": 0.9542689794998873,
+      "grad_norm": 1.7754453365847451,
+      "learning_rate": 1e-05,
+      "loss": 1.2161,
+      "step": 2118
+    },
+    {
+      "epoch": 0.9547195314259969,
+      "grad_norm": 1.5679928263707885,
+      "learning_rate": 1e-05,
+      "loss": 1.1943,
+      "step": 2119
+    },
+    {
+      "epoch": 0.9551700833521063,
+      "grad_norm": 1.5326901408842268,
+      "learning_rate": 1e-05,
+      "loss": 1.1702,
+      "step": 2120
+    },
+    {
+      "epoch": 0.9556206352782158,
+      "grad_norm": 1.7934219244055345,
+      "learning_rate": 1e-05,
+      "loss": 1.1612,
+      "step": 2121
+    },
+    {
+      "epoch": 0.9560711872043253,
+      "grad_norm": 1.5903967174339735,
+      "learning_rate": 1e-05,
+      "loss": 1.223,
+      "step": 2122
+    },
+    {
+      "epoch": 0.9565217391304348,
+      "grad_norm": 1.8576688324700052,
+      "learning_rate": 1e-05,
+      "loss": 1.1855,
+      "step": 2123
+    },
+    {
+      "epoch": 0.9569722910565442,
+      "grad_norm": 1.5024626501001406,
+      "learning_rate": 1e-05,
+      "loss": 1.1583,
+      "step": 2124
+    },
+    {
+      "epoch": 0.9574228429826538,
+      "grad_norm": 1.913175515389359,
+      "learning_rate": 1e-05,
+      "loss": 1.1555,
+      "step": 2125
+    },
+    {
+      "epoch": 0.9578733949087632,
+      "grad_norm": 1.6565160630402702,
+      "learning_rate": 1e-05,
+      "loss": 1.1382,
+      "step": 2126
+    },
+    {
+      "epoch": 0.9583239468348728,
+      "grad_norm": 1.7612841877743974,
+      "learning_rate": 1e-05,
+      "loss": 1.1882,
+      "step": 2127
+    },
+    {
+      "epoch": 0.9587744987609822,
+      "grad_norm": 1.6999960214638121,
+      "learning_rate": 1e-05,
+      "loss": 1.127,
+      "step": 2128
+    },
+    {
+      "epoch": 0.9592250506870917,
+      "grad_norm": 1.7865687508854,
+      "learning_rate": 1e-05,
+      "loss": 1.2001,
+      "step": 2129
+    },
+    {
+      "epoch": 0.9596756026132012,
+      "grad_norm": 1.7229040927458896,
+      "learning_rate": 1e-05,
+      "loss": 1.1457,
+      "step": 2130
+    },
+    {
+      "epoch": 0.9601261545393106,
+      "grad_norm": 1.6649741455846447,
+      "learning_rate": 1e-05,
+      "loss": 1.138,
+      "step": 2131
+    },
+    {
+      "epoch": 0.9605767064654201,
+      "grad_norm": 1.6951738897843966,
+      "learning_rate": 1e-05,
+      "loss": 1.202,
+      "step": 2132
+    },
+    {
+      "epoch": 0.9610272583915296,
+      "grad_norm": 1.6004134883352765,
+      "learning_rate": 1e-05,
+      "loss": 1.2104,
+      "step": 2133
+    },
+    {
+      "epoch": 0.9614778103176391,
+      "grad_norm": 1.6271378214931058,
+      "learning_rate": 1e-05,
+      "loss": 1.1665,
+      "step": 2134
+    },
+    {
+      "epoch": 0.9619283622437486,
+      "grad_norm": 1.5995941621214527,
+      "learning_rate": 1e-05,
+      "loss": 1.1781,
+      "step": 2135
+    },
+    {
+      "epoch": 0.9623789141698581,
+      "grad_norm": 1.6152784854202955,
+      "learning_rate": 1e-05,
+      "loss": 1.2213,
+      "step": 2136
+    },
+    {
+      "epoch": 0.9628294660959675,
+      "grad_norm": 1.8385052853015975,
+      "learning_rate": 1e-05,
+      "loss": 1.1412,
+      "step": 2137
+    },
+    {
+      "epoch": 0.9632800180220771,
+      "grad_norm": 1.7068694440237158,
+      "learning_rate": 1e-05,
+      "loss": 1.1898,
+      "step": 2138
+    },
+    {
+      "epoch": 0.9637305699481865,
+      "grad_norm": 1.8055622353281682,
+      "learning_rate": 1e-05,
+      "loss": 1.2078,
+      "step": 2139
+    },
+    {
+      "epoch": 0.964181121874296,
+      "grad_norm": 1.5892068342229684,
+      "learning_rate": 1e-05,
+      "loss": 1.215,
+      "step": 2140
+    },
+    {
+      "epoch": 0.9646316738004055,
+      "grad_norm": 1.6626282035716575,
+      "learning_rate": 1e-05,
+      "loss": 1.1684,
+      "step": 2141
+    },
+    {
+      "epoch": 0.965082225726515,
+      "grad_norm": 1.5815313004686862,
+      "learning_rate": 1e-05,
+      "loss": 1.1426,
+      "step": 2142
+    },
+    {
+      "epoch": 0.9655327776526245,
+      "grad_norm": 1.6104144911672769,
+      "learning_rate": 1e-05,
+      "loss": 1.2065,
+      "step": 2143
+    },
+    {
+      "epoch": 0.9659833295787339,
+      "grad_norm": 1.7223066918002348,
+      "learning_rate": 1e-05,
+      "loss": 1.216,
+      "step": 2144
+    },
+    {
+      "epoch": 0.9664338815048434,
+      "grad_norm": 1.626978120340767,
+      "learning_rate": 1e-05,
+      "loss": 1.1555,
+      "step": 2145
+    },
+    {
+      "epoch": 0.9668844334309529,
+      "grad_norm": 1.5977128449160958,
+      "learning_rate": 1e-05,
+      "loss": 1.1722,
+      "step": 2146
+    },
+    {
+      "epoch": 0.9673349853570624,
+      "grad_norm": 1.7382985565577844,
+      "learning_rate": 1e-05,
+      "loss": 1.1813,
+      "step": 2147
+    },
+    {
+      "epoch": 0.9677855372831718,
+      "grad_norm": 1.7100089095382642,
+      "learning_rate": 1e-05,
+      "loss": 1.186,
+      "step": 2148
+    },
+    {
+      "epoch": 0.9682360892092814,
+      "grad_norm": 1.5140761682833617,
+      "learning_rate": 1e-05,
+      "loss": 1.2091,
+      "step": 2149
+    },
+    {
+      "epoch": 0.9686866411353908,
+      "grad_norm": 1.631458054037132,
+      "learning_rate": 1e-05,
+      "loss": 1.1977,
+      "step": 2150
+    },
+    {
+      "epoch": 0.9691371930615004,
+      "grad_norm": 1.9407101557307815,
+      "learning_rate": 1e-05,
+      "loss": 1.244,
+      "step": 2151
+    },
+    {
+      "epoch": 0.9695877449876098,
+      "grad_norm": 1.644705809226952,
+      "learning_rate": 1e-05,
+      "loss": 1.1926,
+      "step": 2152
+    },
+    {
+      "epoch": 0.9700382969137193,
+      "grad_norm": 1.7738926775892274,
+      "learning_rate": 1e-05,
+      "loss": 1.1801,
+      "step": 2153
+    },
+    {
+      "epoch": 0.9704888488398288,
+      "grad_norm": 1.632866466849786,
+      "learning_rate": 1e-05,
+      "loss": 1.1703,
+      "step": 2154
+    },
+    {
+      "epoch": 0.9709394007659383,
+      "grad_norm": 1.6489079715720707,
+      "learning_rate": 1e-05,
+      "loss": 1.1688,
+      "step": 2155
+    },
+    {
+      "epoch": 0.9713899526920478,
+      "grad_norm": 1.6991057253621062,
+      "learning_rate": 1e-05,
+      "loss": 1.1681,
+      "step": 2156
+    },
+    {
+      "epoch": 0.9718405046181573,
+      "grad_norm": 1.6331345480206434,
+      "learning_rate": 1e-05,
+      "loss": 1.1734,
+      "step": 2157
+    },
+    {
+      "epoch": 0.9722910565442667,
+      "grad_norm": 1.6750211080976438,
+      "learning_rate": 1e-05,
+      "loss": 1.2016,
+      "step": 2158
+    },
+    {
+      "epoch": 0.9727416084703762,
+      "grad_norm": 1.750153664716623,
+      "learning_rate": 1e-05,
+      "loss": 1.1745,
+      "step": 2159
+    },
+    {
+      "epoch": 0.9731921603964857,
+      "grad_norm": 1.6629057903262463,
+      "learning_rate": 1e-05,
+      "loss": 1.1462,
+      "step": 2160
+    },
+    {
+      "epoch": 0.9736427123225951,
+      "grad_norm": 1.633481623623148,
+      "learning_rate": 1e-05,
+      "loss": 1.212,
+      "step": 2161
+    },
+    {
+      "epoch": 0.9740932642487047,
+      "grad_norm": 1.6677123407725638,
+      "learning_rate": 1e-05,
+      "loss": 1.1675,
+      "step": 2162
+    },
+    {
+      "epoch": 0.9745438161748141,
+      "grad_norm": 1.6603950206096891,
+      "learning_rate": 1e-05,
+      "loss": 1.1501,
+      "step": 2163
+    },
+    {
+      "epoch": 0.9749943681009237,
+      "grad_norm": 1.5383013282675635,
+      "learning_rate": 1e-05,
+      "loss": 1.167,
+      "step": 2164
+    },
+    {
+      "epoch": 0.9754449200270331,
+      "grad_norm": 1.6339350842028757,
+      "learning_rate": 1e-05,
+      "loss": 1.1703,
+      "step": 2165
+    },
+    {
+      "epoch": 0.9758954719531426,
+      "grad_norm": 1.6419327750701314,
+      "learning_rate": 1e-05,
+      "loss": 1.2046,
+      "step": 2166
+    },
+    {
+      "epoch": 0.9763460238792521,
+      "grad_norm": 1.6071485373938563,
+      "learning_rate": 1e-05,
+      "loss": 1.1564,
+      "step": 2167
+    },
+    {
+      "epoch": 0.9767965758053616,
+      "grad_norm": 1.6645829201081386,
+      "learning_rate": 1e-05,
+      "loss": 1.1567,
+      "step": 2168
+    },
+    {
+      "epoch": 0.977247127731471,
+      "grad_norm": 1.834241571711109,
+      "learning_rate": 1e-05,
+      "loss": 1.1959,
+      "step": 2169
+    },
+    {
+      "epoch": 0.9776976796575806,
+      "grad_norm": 1.7851782257705597,
+      "learning_rate": 1e-05,
+      "loss": 1.1073,
+      "step": 2170
+    },
+    {
+      "epoch": 0.97814823158369,
+      "grad_norm": 1.6662785700861493,
+      "learning_rate": 1e-05,
+      "loss": 1.2193,
+      "step": 2171
+    },
+    {
+      "epoch": 0.9785987835097995,
+      "grad_norm": 1.6137357355336408,
+      "learning_rate": 1e-05,
+      "loss": 1.1989,
+      "step": 2172
+    },
+    {
+      "epoch": 0.979049335435909,
+      "grad_norm": 1.769977579305047,
+      "learning_rate": 1e-05,
+      "loss": 1.1943,
+      "step": 2173
+    },
+    {
+      "epoch": 0.9794998873620184,
+      "grad_norm": 1.5440600304095895,
+      "learning_rate": 1e-05,
+      "loss": 1.2057,
+      "step": 2174
+    },
+    {
+      "epoch": 0.979950439288128,
+      "grad_norm": 1.5650860905655992,
+      "learning_rate": 1e-05,
+      "loss": 1.163,
+      "step": 2175
+    },
+    {
+      "epoch": 0.9804009912142374,
+      "grad_norm": 1.683105123119004,
+      "learning_rate": 1e-05,
+      "loss": 1.1673,
+      "step": 2176
+    },
+    {
+      "epoch": 0.980851543140347,
+      "grad_norm": 1.6772024747381555,
+      "learning_rate": 1e-05,
+      "loss": 1.2483,
+      "step": 2177
+    },
+    {
+      "epoch": 0.9813020950664564,
+      "grad_norm": 1.9429115939809587,
+      "learning_rate": 1e-05,
+      "loss": 1.1323,
+      "step": 2178
+    },
+    {
+      "epoch": 0.9817526469925659,
+      "grad_norm": 1.7168014980135007,
+      "learning_rate": 1e-05,
+      "loss": 1.1686,
+      "step": 2179
+    },
+    {
+      "epoch": 0.9822031989186754,
+      "grad_norm": 1.627958580173763,
+      "learning_rate": 1e-05,
+      "loss": 1.1377,
+      "step": 2180
+    },
+    {
+      "epoch": 0.9826537508447849,
+      "grad_norm": 1.8630592315550862,
+      "learning_rate": 1e-05,
+      "loss": 1.197,
+      "step": 2181
+    },
+    {
+      "epoch": 0.9831043027708943,
+      "grad_norm": 1.5226363816071424,
+      "learning_rate": 1e-05,
+      "loss": 1.2748,
+      "step": 2182
+    },
+    {
+      "epoch": 0.9835548546970039,
+      "grad_norm": 1.640663627306306,
+      "learning_rate": 1e-05,
+      "loss": 1.1931,
+      "step": 2183
+    },
+    {
+      "epoch": 0.9840054066231133,
+      "grad_norm": 1.792672713213748,
+      "learning_rate": 1e-05,
+      "loss": 1.1694,
+      "step": 2184
+    },
+    {
+      "epoch": 0.9844559585492227,
+      "grad_norm": 1.657565748047637,
+      "learning_rate": 1e-05,
+      "loss": 1.1532,
+      "step": 2185
+    },
+    {
+      "epoch": 0.9849065104753323,
+      "grad_norm": 1.599593629920567,
+      "learning_rate": 1e-05,
+      "loss": 1.2105,
+      "step": 2186
+    },
+    {
+      "epoch": 0.9853570624014417,
+      "grad_norm": 1.6978249473374794,
+      "learning_rate": 1e-05,
+      "loss": 1.1582,
+      "step": 2187
+    },
+    {
+      "epoch": 0.9858076143275513,
+      "grad_norm": 1.6437621469181267,
+      "learning_rate": 1e-05,
+      "loss": 1.1904,
+      "step": 2188
+    },
+    {
+      "epoch": 0.9862581662536607,
+      "grad_norm": 1.6908522774335355,
+      "learning_rate": 1e-05,
+      "loss": 1.1329,
+      "step": 2189
+    },
+    {
+      "epoch": 0.9867087181797702,
+      "grad_norm": 1.7108282497456526,
+      "learning_rate": 1e-05,
+      "loss": 1.1667,
+      "step": 2190
+    },
+    {
+      "epoch": 0.9871592701058797,
+      "grad_norm": 1.650745417574159,
+      "learning_rate": 1e-05,
+      "loss": 1.1542,
+      "step": 2191
+    },
+    {
+      "epoch": 0.9876098220319892,
+      "grad_norm": 1.8415193285070817,
+      "learning_rate": 1e-05,
+      "loss": 1.1436,
+      "step": 2192
+    },
+    {
+      "epoch": 0.9880603739580986,
+      "grad_norm": 1.7265003600000148,
+      "learning_rate": 1e-05,
+      "loss": 1.1802,
+      "step": 2193
+    },
+    {
+      "epoch": 0.9885109258842082,
+      "grad_norm": 1.809638358823917,
+      "learning_rate": 1e-05,
+      "loss": 1.1701,
+      "step": 2194
+    },
+    {
+      "epoch": 0.9889614778103176,
+      "grad_norm": 1.6772766425201677,
+      "learning_rate": 1e-05,
+      "loss": 1.1852,
+      "step": 2195
+    },
+    {
+      "epoch": 0.9894120297364272,
+      "grad_norm": 1.7234560256340692,
+      "learning_rate": 1e-05,
+      "loss": 1.2143,
+      "step": 2196
+    },
+    {
+      "epoch": 0.9898625816625366,
+      "grad_norm": 1.7230785087089597,
+      "learning_rate": 1e-05,
+      "loss": 1.1305,
+      "step": 2197
+    },
+    {
+      "epoch": 0.9903131335886461,
+      "grad_norm": 1.5677372020042342,
+      "learning_rate": 1e-05,
+      "loss": 1.2346,
+      "step": 2198
+    },
+    {
+      "epoch": 0.9907636855147556,
+      "grad_norm": 1.6534627672302962,
+      "learning_rate": 1e-05,
+      "loss": 1.1358,
+      "step": 2199
+    },
+    {
+      "epoch": 0.991214237440865,
+      "grad_norm": 1.6677962601647158,
+      "learning_rate": 1e-05,
+      "loss": 1.2025,
+      "step": 2200
+    },
+    {
+      "epoch": 0.9916647893669746,
+      "grad_norm": 1.5957431691240571,
+      "learning_rate": 1e-05,
+      "loss": 1.2021,
+      "step": 2201
+    },
+    {
+      "epoch": 0.992115341293084,
+      "grad_norm": 1.6825396781316948,
+      "learning_rate": 1e-05,
+      "loss": 1.1906,
+      "step": 2202
+    },
+    {
+      "epoch": 0.9925658932191935,
+      "grad_norm": 1.631478160156966,
+      "learning_rate": 1e-05,
+      "loss": 1.2176,
+      "step": 2203
+    },
+    {
+      "epoch": 0.993016445145303,
+      "grad_norm": 1.6181064708610031,
+      "learning_rate": 1e-05,
+      "loss": 1.1497,
+      "step": 2204
+    },
+    {
+      "epoch": 0.9934669970714125,
+      "grad_norm": 1.6731227448613548,
+      "learning_rate": 1e-05,
+      "loss": 1.2529,
+      "step": 2205
+    },
+    {
+      "epoch": 0.9939175489975219,
+      "grad_norm": 1.7203610581692894,
+      "learning_rate": 1e-05,
+      "loss": 1.2136,
+      "step": 2206
+    },
+    {
+      "epoch": 0.9943681009236315,
+      "grad_norm": 1.641928282303174,
+      "learning_rate": 1e-05,
+      "loss": 1.2041,
+      "step": 2207
+    },
+    {
+      "epoch": 0.9948186528497409,
+      "grad_norm": 1.753504116694314,
+      "learning_rate": 1e-05,
+      "loss": 1.2071,
+      "step": 2208
+    },
+    {
+      "epoch": 0.9952692047758505,
+      "grad_norm": 1.7634227391080495,
+      "learning_rate": 1e-05,
+      "loss": 1.1856,
+      "step": 2209
+    },
+    {
+      "epoch": 0.9957197567019599,
+      "grad_norm": 1.6307678361862732,
+      "learning_rate": 1e-05,
+      "loss": 1.1747,
+      "step": 2210
+    },
+    {
+      "epoch": 0.9961703086280694,
+      "grad_norm": 1.676694512275526,
+      "learning_rate": 1e-05,
+      "loss": 1.1732,
+      "step": 2211
+    },
+    {
+      "epoch": 0.9966208605541789,
+      "grad_norm": 1.5468525592889175,
+      "learning_rate": 1e-05,
+      "loss": 1.1757,
+      "step": 2212
+    },
+    {
+      "epoch": 0.9970714124802883,
+      "grad_norm": 1.762742264188603,
+      "learning_rate": 1e-05,
+      "loss": 1.1747,
+      "step": 2213
+    },
+    {
+      "epoch": 0.9975219644063978,
+      "grad_norm": 1.606944223414691,
+      "learning_rate": 1e-05,
+      "loss": 1.1967,
+      "step": 2214
+    },
+    {
+      "epoch": 0.9979725163325073,
+      "grad_norm": 1.603067947399285,
+      "learning_rate": 1e-05,
+      "loss": 1.1813,
+      "step": 2215
+    },
+    {
+      "epoch": 0.9984230682586168,
+      "grad_norm": 1.6802311418144311,
+      "learning_rate": 1e-05,
+      "loss": 1.2049,
+      "step": 2216
+    },
+    {
+      "epoch": 0.9988736201847263,
+      "grad_norm": 1.4778117214314126,
+      "learning_rate": 1e-05,
+      "loss": 1.1737,
+      "step": 2217
+    },
+    {
+      "epoch": 0.9993241721108358,
+      "grad_norm": 1.6302785209744532,
+      "learning_rate": 1e-05,
+      "loss": 1.1906,
+      "step": 2218
+    },
+    {
+      "epoch": 0.9997747240369452,
+      "grad_norm": 1.568318632001129,
+      "learning_rate": 1e-05,
+      "loss": 1.191,
+      "step": 2219
+    },
+    {
+      "epoch": 0.9997747240369452,
+      "step": 2219,
+      "total_flos": 3591718242877440.0,
+      "train_loss": 1.239272991082001,
+      "train_runtime": 334031.5377,
+      "train_samples_per_second": 0.851,
+      "train_steps_per_second": 0.007
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2219,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 10,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3591718242877440.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}