diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,36313 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.999855103962907,
+  "eval_steps": 500,
+  "global_step": 5176,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.001159168296747084,
+      "grad_norm": 2.7580204010009766,
+      "learning_rate": 4.999425287356322e-05,
+      "loss": 0.9725,
+      "step": 1
+    },
+    {
+      "epoch": 0.002318336593494168,
+      "grad_norm": 1.794647216796875,
+      "learning_rate": 4.998850574712644e-05,
+      "loss": 0.8067,
+      "step": 2
+    },
+    {
+      "epoch": 0.003477504890241252,
+      "grad_norm": 1.9236769676208496,
+      "learning_rate": 4.9982758620689654e-05,
+      "loss": 0.7883,
+      "step": 3
+    },
+    {
+      "epoch": 0.004636673186988336,
+      "grad_norm": 1.9243451356887817,
+      "learning_rate": 4.9977011494252876e-05,
+      "loss": 0.694,
+      "step": 4
+    },
+    {
+      "epoch": 0.00579584148373542,
+      "grad_norm": 2.325658082962036,
+      "learning_rate": 4.99712643678161e-05,
+      "loss": 0.658,
+      "step": 5
+    },
+    {
+      "epoch": 0.006955009780482504,
+      "grad_norm": 1.9371262788772583,
+      "learning_rate": 4.996551724137931e-05,
+      "loss": 0.6053,
+      "step": 6
+    },
+    {
+      "epoch": 0.008114178077229587,
+      "grad_norm": 2.50772762298584,
+      "learning_rate": 4.9959770114942534e-05,
+      "loss": 0.5673,
+      "step": 7
+    },
+    {
+      "epoch": 0.009273346373976673,
+      "grad_norm": 1.7069337368011475,
+      "learning_rate": 4.995402298850575e-05,
+      "loss": 0.5236,
+      "step": 8
+    },
+    {
+      "epoch": 0.010432514670723756,
+      "grad_norm": 2.124687433242798,
+      "learning_rate": 4.9948275862068964e-05,
+      "loss": 0.4998,
+      "step": 9
+    },
+    {
+      "epoch": 0.01159168296747084,
+      "grad_norm": 0.8905035257339478,
+      "learning_rate": 4.9942528735632185e-05,
+      "loss": 0.4632,
+      "step": 10
+    },
+    {
+      "epoch": 0.012750851264217924,
+      "grad_norm": 4.519129276275635,
+      "learning_rate": 4.993678160919541e-05,
+      "loss": 0.3914,
+      "step": 11
+    },
+    {
+      "epoch": 0.013910019560965008,
+      "grad_norm": 0.8016653060913086,
+      "learning_rate": 4.993103448275862e-05,
+      "loss": 0.3829,
+      "step": 12
+    },
+    {
+      "epoch": 0.015069187857712092,
+      "grad_norm": 0.8521615862846375,
+      "learning_rate": 4.9925287356321844e-05,
+      "loss": 0.3719,
+      "step": 13
+    },
+    {
+      "epoch": 0.016228356154459174,
+      "grad_norm": 0.664720356464386,
+      "learning_rate": 4.991954022988506e-05,
+      "loss": 0.3444,
+      "step": 14
+    },
+    {
+      "epoch": 0.017387524451206258,
+      "grad_norm": 1.236467719078064,
+      "learning_rate": 4.991379310344828e-05,
+      "loss": 0.3283,
+      "step": 15
+    },
+    {
+      "epoch": 0.018546692747953345,
+      "grad_norm": 1.1017507314682007,
+      "learning_rate": 4.9908045977011495e-05,
+      "loss": 0.3348,
+      "step": 16
+    },
+    {
+      "epoch": 0.01970586104470043,
+      "grad_norm": 0.70470130443573,
+      "learning_rate": 4.990229885057472e-05,
+      "loss": 0.318,
+      "step": 17
+    },
+    {
+      "epoch": 0.020865029341447513,
+      "grad_norm": 0.6223506927490234,
+      "learning_rate": 4.989655172413794e-05,
+      "loss": 0.3032,
+      "step": 18
+    },
+    {
+      "epoch": 0.022024197638194597,
+      "grad_norm": 0.7839768528938293,
+      "learning_rate": 4.989080459770115e-05,
+      "loss": 0.2854,
+      "step": 19
+    },
+    {
+      "epoch": 0.02318336593494168,
+      "grad_norm": 1.5129585266113281,
+      "learning_rate": 4.988505747126437e-05,
+      "loss": 0.2912,
+      "step": 20
+    },
+    {
+      "epoch": 0.024342534231688764,
+      "grad_norm": 0.6539391279220581,
+      "learning_rate": 4.987931034482759e-05,
+      "loss": 0.27,
+      "step": 21
+    },
+    {
+      "epoch": 0.025501702528435848,
+      "grad_norm": 0.693023145198822,
+      "learning_rate": 4.9873563218390805e-05,
+      "loss": 0.2558,
+      "step": 22
+    },
+    {
+      "epoch": 0.026660870825182932,
+      "grad_norm": 0.5335514545440674,
+      "learning_rate": 4.986781609195402e-05,
+      "loss": 0.2276,
+      "step": 23
+    },
+    {
+      "epoch": 0.027820039121930016,
+      "grad_norm": 0.6515309810638428,
+      "learning_rate": 4.986206896551724e-05,
+      "loss": 0.2594,
+      "step": 24
+    },
+    {
+      "epoch": 0.0289792074186771,
+      "grad_norm": 0.6215941905975342,
+      "learning_rate": 4.985632183908046e-05,
+      "loss": 0.255,
+      "step": 25
+    },
+    {
+      "epoch": 0.030138375715424184,
+      "grad_norm": 0.6825059056282043,
+      "learning_rate": 4.9850574712643685e-05,
+      "loss": 0.2719,
+      "step": 26
+    },
+    {
+      "epoch": 0.03129754401217127,
+      "grad_norm": 0.5802633166313171,
+      "learning_rate": 4.98448275862069e-05,
+      "loss": 0.2256,
+      "step": 27
+    },
+    {
+      "epoch": 0.03245671230891835,
+      "grad_norm": 0.5147871971130371,
+      "learning_rate": 4.9839080459770115e-05,
+      "loss": 0.214,
+      "step": 28
+    },
+    {
+      "epoch": 0.033615880605665435,
+      "grad_norm": 0.6146939396858215,
+      "learning_rate": 4.9833333333333336e-05,
+      "loss": 0.2401,
+      "step": 29
+    },
+    {
+      "epoch": 0.034775048902412516,
+      "grad_norm": 0.62993323802948,
+      "learning_rate": 4.982758620689655e-05,
+      "loss": 0.246,
+      "step": 30
+    },
+    {
+      "epoch": 0.0359342171991596,
+      "grad_norm": 0.6217732429504395,
+      "learning_rate": 4.982183908045977e-05,
+      "loss": 0.2162,
+      "step": 31
+    },
+    {
+      "epoch": 0.03709338549590669,
+      "grad_norm": 0.5851067304611206,
+      "learning_rate": 4.9816091954022994e-05,
+      "loss": 0.2226,
+      "step": 32
+    },
+    {
+      "epoch": 0.03825255379265377,
+      "grad_norm": 0.6002720594406128,
+      "learning_rate": 4.981034482758621e-05,
+      "loss": 0.226,
+      "step": 33
+    },
+    {
+      "epoch": 0.03941172208940086,
+      "grad_norm": 0.6244901418685913,
+      "learning_rate": 4.980459770114943e-05,
+      "loss": 0.2536,
+      "step": 34
+    },
+    {
+      "epoch": 0.04057089038614794,
+      "grad_norm": 0.6513347625732422,
+      "learning_rate": 4.9798850574712646e-05,
+      "loss": 0.2388,
+      "step": 35
+    },
+    {
+      "epoch": 0.041730058682895026,
+      "grad_norm": 0.45215773582458496,
+      "learning_rate": 4.979310344827586e-05,
+      "loss": 0.197,
+      "step": 36
+    },
+    {
+      "epoch": 0.042889226979642106,
+      "grad_norm": 0.45586729049682617,
+      "learning_rate": 4.978735632183908e-05,
+      "loss": 0.1992,
+      "step": 37
+    },
+    {
+      "epoch": 0.04404839527638919,
+      "grad_norm": 0.49368321895599365,
+      "learning_rate": 4.9781609195402304e-05,
+      "loss": 0.2319,
+      "step": 38
+    },
+    {
+      "epoch": 0.045207563573136274,
+      "grad_norm": 0.4354698956012726,
+      "learning_rate": 4.977586206896552e-05,
+      "loss": 0.2072,
+      "step": 39
+    },
+    {
+      "epoch": 0.04636673186988336,
+      "grad_norm": 0.40046998858451843,
+      "learning_rate": 4.977011494252874e-05,
+      "loss": 0.183,
+      "step": 40
+    },
+    {
+      "epoch": 0.04752590016663044,
+      "grad_norm": 0.41561418771743774,
+      "learning_rate": 4.9764367816091956e-05,
+      "loss": 0.196,
+      "step": 41
+    },
+    {
+      "epoch": 0.04868506846337753,
+      "grad_norm": 0.4249739646911621,
+      "learning_rate": 4.975862068965517e-05,
+      "loss": 0.1903,
+      "step": 42
+    },
+    {
+      "epoch": 0.04984423676012461,
+      "grad_norm": 0.5787274837493896,
+      "learning_rate": 4.975287356321839e-05,
+      "loss": 0.2098,
+      "step": 43
+    },
+    {
+      "epoch": 0.051003405056871697,
+      "grad_norm": 0.4138984978199005,
+      "learning_rate": 4.974712643678161e-05,
+      "loss": 0.1942,
+      "step": 44
+    },
+    {
+      "epoch": 0.05216257335361878,
+      "grad_norm": 0.4079653322696686,
+      "learning_rate": 4.9741379310344836e-05,
+      "loss": 0.2141,
+      "step": 45
+    },
+    {
+      "epoch": 0.053321741650365864,
+      "grad_norm": 0.4127311408519745,
+      "learning_rate": 4.973563218390805e-05,
+      "loss": 0.1795,
+      "step": 46
+    },
+    {
+      "epoch": 0.054480909947112945,
+      "grad_norm": 0.48196831345558167,
+      "learning_rate": 4.9729885057471265e-05,
+      "loss": 0.2008,
+      "step": 47
+    },
+    {
+      "epoch": 0.05564007824386003,
+      "grad_norm": 0.43922609090805054,
+      "learning_rate": 4.972413793103449e-05,
+      "loss": 0.1897,
+      "step": 48
+    },
+    {
+      "epoch": 0.05679924654060711,
+      "grad_norm": 0.455390065908432,
+      "learning_rate": 4.97183908045977e-05,
+      "loss": 0.1964,
+      "step": 49
+    },
+    {
+      "epoch": 0.0579584148373542,
+      "grad_norm": 0.4199633300304413,
+      "learning_rate": 4.971264367816092e-05,
+      "loss": 0.1984,
+      "step": 50
+    },
+    {
+      "epoch": 0.05911758313410128,
+      "grad_norm": 0.4313494861125946,
+      "learning_rate": 4.970689655172414e-05,
+      "loss": 0.2044,
+      "step": 51
+    },
+    {
+      "epoch": 0.06027675143084837,
+      "grad_norm": 0.431208997964859,
+      "learning_rate": 4.970114942528736e-05,
+      "loss": 0.2028,
+      "step": 52
+    },
+    {
+      "epoch": 0.06143591972759545,
+      "grad_norm": 0.4196656048297882,
+      "learning_rate": 4.969540229885058e-05,
+      "loss": 0.1892,
+      "step": 53
+    },
+    {
+      "epoch": 0.06259508802434254,
+      "grad_norm": 0.43570151925086975,
+      "learning_rate": 4.96896551724138e-05,
+      "loss": 0.2068,
+      "step": 54
+    },
+    {
+      "epoch": 0.06375425632108962,
+      "grad_norm": 0.3962918519973755,
+      "learning_rate": 4.968390804597701e-05,
+      "loss": 0.1936,
+      "step": 55
+    },
+    {
+      "epoch": 0.0649134246178367,
+      "grad_norm": 0.6284250617027283,
+      "learning_rate": 4.9678160919540233e-05,
+      "loss": 0.1999,
+      "step": 56
+    },
+    {
+      "epoch": 0.06607259291458378,
+      "grad_norm": 0.45828792452812195,
+      "learning_rate": 4.967241379310345e-05,
+      "loss": 0.1757,
+      "step": 57
+    },
+    {
+      "epoch": 0.06723176121133087,
+      "grad_norm": 0.43321043252944946,
+      "learning_rate": 4.966666666666667e-05,
+      "loss": 0.1908,
+      "step": 58
+    },
+    {
+      "epoch": 0.06839092950807796,
+      "grad_norm": 0.38716545701026917,
+      "learning_rate": 4.966091954022989e-05,
+      "loss": 0.1838,
+      "step": 59
+    },
+    {
+      "epoch": 0.06955009780482503,
+      "grad_norm": 0.5458523035049438,
+      "learning_rate": 4.9655172413793107e-05,
+      "loss": 0.1996,
+      "step": 60
+    },
+    {
+      "epoch": 0.07070926610157212,
+      "grad_norm": 0.5392327904701233,
+      "learning_rate": 4.964942528735632e-05,
+      "loss": 0.1895,
+      "step": 61
+    },
+    {
+      "epoch": 0.0718684343983192,
+      "grad_norm": 0.4704841077327728,
+      "learning_rate": 4.964367816091954e-05,
+      "loss": 0.1984,
+      "step": 62
+    },
+    {
+      "epoch": 0.0730276026950663,
+      "grad_norm": 0.3919082283973694,
+      "learning_rate": 4.963793103448276e-05,
+      "loss": 0.1642,
+      "step": 63
+    },
+    {
+      "epoch": 0.07418677099181338,
+      "grad_norm": 0.46172910928726196,
+      "learning_rate": 4.963218390804598e-05,
+      "loss": 0.1665,
+      "step": 64
+    },
+    {
+      "epoch": 0.07534593928856045,
+      "grad_norm": 0.42953500151634216,
+      "learning_rate": 4.9626436781609195e-05,
+      "loss": 0.1807,
+      "step": 65
+    },
+    {
+      "epoch": 0.07650510758530754,
+      "grad_norm": 0.4844302535057068,
+      "learning_rate": 4.9620689655172416e-05,
+      "loss": 0.1978,
+      "step": 66
+    },
+    {
+      "epoch": 0.07766427588205463,
+      "grad_norm": 0.5020731091499329,
+      "learning_rate": 4.961494252873564e-05,
+      "loss": 0.1943,
+      "step": 67
+    },
+    {
+      "epoch": 0.07882344417880172,
+      "grad_norm": 0.38344666361808777,
+      "learning_rate": 4.960919540229885e-05,
+      "loss": 0.1893,
+      "step": 68
+    },
+    {
+      "epoch": 0.07998261247554879,
+      "grad_norm": 0.46443063020706177,
+      "learning_rate": 4.960344827586207e-05,
+      "loss": 0.1896,
+      "step": 69
+    },
+    {
+      "epoch": 0.08114178077229588,
+      "grad_norm": 0.49002230167388916,
+      "learning_rate": 4.959770114942529e-05,
+      "loss": 0.1754,
+      "step": 70
+    },
+    {
+      "epoch": 0.08230094906904296,
+      "grad_norm": 0.40396377444267273,
+      "learning_rate": 4.9591954022988504e-05,
+      "loss": 0.1712,
+      "step": 71
+    },
+    {
+      "epoch": 0.08346011736579005,
+      "grad_norm": 0.44212767481803894,
+      "learning_rate": 4.9586206896551726e-05,
+      "loss": 0.1745,
+      "step": 72
+    },
+    {
+      "epoch": 0.08461928566253712,
+      "grad_norm": 0.40350252389907837,
+      "learning_rate": 4.958045977011495e-05,
+      "loss": 0.1664,
+      "step": 73
+    },
+    {
+      "epoch": 0.08577845395928421,
+      "grad_norm": 0.45109373331069946,
+      "learning_rate": 4.957471264367816e-05,
+      "loss": 0.1794,
+      "step": 74
+    },
+    {
+      "epoch": 0.0869376222560313,
+      "grad_norm": 0.4555060863494873,
+      "learning_rate": 4.9568965517241384e-05,
+      "loss": 0.1976,
+      "step": 75
+    },
+    {
+      "epoch": 0.08809679055277839,
+      "grad_norm": 0.39049214124679565,
+      "learning_rate": 4.95632183908046e-05,
+      "loss": 0.1642,
+      "step": 76
+    },
+    {
+      "epoch": 0.08925595884952546,
+      "grad_norm": 0.4325527250766754,
+      "learning_rate": 4.9557471264367814e-05,
+      "loss": 0.1735,
+      "step": 77
+    },
+    {
+      "epoch": 0.09041512714627255,
+      "grad_norm": 0.38487985730171204,
+      "learning_rate": 4.9551724137931036e-05,
+      "loss": 0.1727,
+      "step": 78
+    },
+    {
+      "epoch": 0.09157429544301963,
+      "grad_norm": 0.38345035910606384,
+      "learning_rate": 4.954597701149426e-05,
+      "loss": 0.153,
+      "step": 79
+    },
+    {
+      "epoch": 0.09273346373976672,
+      "grad_norm": 0.504005491733551,
+      "learning_rate": 4.954022988505747e-05,
+      "loss": 0.1852,
+      "step": 80
+    },
+    {
+      "epoch": 0.0938926320365138,
+      "grad_norm": 0.40111425518989563,
+      "learning_rate": 4.9534482758620694e-05,
+      "loss": 0.1834,
+      "step": 81
+    },
+    {
+      "epoch": 0.09505180033326088,
+      "grad_norm": 0.3611973226070404,
+      "learning_rate": 4.952873563218391e-05,
+      "loss": 0.1714,
+      "step": 82
+    },
+    {
+      "epoch": 0.09621096863000797,
+      "grad_norm": 0.37155240774154663,
+      "learning_rate": 4.952298850574713e-05,
+      "loss": 0.1589,
+      "step": 83
+    },
+    {
+      "epoch": 0.09737013692675506,
+      "grad_norm": 0.4023563861846924,
+      "learning_rate": 4.9517241379310346e-05,
+      "loss": 0.1751,
+      "step": 84
+    },
+    {
+      "epoch": 0.09852930522350213,
+      "grad_norm": 0.3763880431652069,
+      "learning_rate": 4.951149425287356e-05,
+      "loss": 0.1624,
+      "step": 85
+    },
+    {
+      "epoch": 0.09968847352024922,
+      "grad_norm": 0.38636812567710876,
+      "learning_rate": 4.950574712643679e-05,
+      "loss": 0.1609,
+      "step": 86
+    },
+    {
+      "epoch": 0.1008476418169963,
+      "grad_norm": 0.4320867657661438,
+      "learning_rate": 4.9500000000000004e-05,
+      "loss": 0.1945,
+      "step": 87
+    },
+    {
+      "epoch": 0.10200681011374339,
+      "grad_norm": 0.3686438798904419,
+      "learning_rate": 4.949425287356322e-05,
+      "loss": 0.167,
+      "step": 88
+    },
+    {
+      "epoch": 0.10316597841049047,
+      "grad_norm": 0.4408303201198578,
+      "learning_rate": 4.948850574712644e-05,
+      "loss": 0.1628,
+      "step": 89
+    },
+    {
+      "epoch": 0.10432514670723755,
+      "grad_norm": 0.5211153030395508,
+      "learning_rate": 4.9482758620689655e-05,
+      "loss": 0.1902,
+      "step": 90
+    },
+    {
+      "epoch": 0.10548431500398464,
+      "grad_norm": 0.3960166275501251,
+      "learning_rate": 4.947701149425288e-05,
+      "loss": 0.1615,
+      "step": 91
+    },
+    {
+      "epoch": 0.10664348330073173,
+      "grad_norm": 0.3887445628643036,
+      "learning_rate": 4.947126436781609e-05,
+      "loss": 0.1663,
+      "step": 92
+    },
+    {
+      "epoch": 0.10780265159747882,
+      "grad_norm": 0.4395574927330017,
+      "learning_rate": 4.9465517241379314e-05,
+      "loss": 0.1825,
+      "step": 93
+    },
+    {
+      "epoch": 0.10896181989422589,
+      "grad_norm": 0.4707461893558502,
+      "learning_rate": 4.9459770114942535e-05,
+      "loss": 0.1955,
+      "step": 94
+    },
+    {
+      "epoch": 0.11012098819097298,
+      "grad_norm": 0.4846711754798889,
+      "learning_rate": 4.945402298850575e-05,
+      "loss": 0.2082,
+      "step": 95
+    },
+    {
+      "epoch": 0.11128015648772006,
+      "grad_norm": 0.45903241634368896,
+      "learning_rate": 4.9448275862068965e-05,
+      "loss": 0.1856,
+      "step": 96
+    },
+    {
+      "epoch": 0.11243932478446715,
+      "grad_norm": 0.4076662063598633,
+      "learning_rate": 4.944252873563219e-05,
+      "loss": 0.1644,
+      "step": 97
+    },
+    {
+      "epoch": 0.11359849308121422,
+      "grad_norm": 0.4092468321323395,
+      "learning_rate": 4.94367816091954e-05,
+      "loss": 0.1709,
+      "step": 98
+    },
+    {
+      "epoch": 0.11475766137796131,
+      "grad_norm": 0.3405769169330597,
+      "learning_rate": 4.943103448275862e-05,
+      "loss": 0.1428,
+      "step": 99
+    },
+    {
+      "epoch": 0.1159168296747084,
+      "grad_norm": 0.38396161794662476,
+      "learning_rate": 4.9425287356321845e-05,
+      "loss": 0.1582,
+      "step": 100
+    },
+    {
+      "epoch": 0.11707599797145549,
+      "grad_norm": 0.38220831751823425,
+      "learning_rate": 4.941954022988506e-05,
+      "loss": 0.1645,
+      "step": 101
+    },
+    {
+      "epoch": 0.11823516626820256,
+      "grad_norm": 0.3981216549873352,
+      "learning_rate": 4.941379310344828e-05,
+      "loss": 0.1587,
+      "step": 102
+    },
+    {
+      "epoch": 0.11939433456494965,
+      "grad_norm": 0.3643917739391327,
+      "learning_rate": 4.9408045977011496e-05,
+      "loss": 0.1534,
+      "step": 103
+    },
+    {
+      "epoch": 0.12055350286169673,
+      "grad_norm": 0.42615723609924316,
+      "learning_rate": 4.940229885057471e-05,
+      "loss": 0.169,
+      "step": 104
+    },
+    {
+      "epoch": 0.12171267115844382,
+      "grad_norm": 0.4193412959575653,
+      "learning_rate": 4.939655172413793e-05,
+      "loss": 0.1586,
+      "step": 105
+    },
+    {
+      "epoch": 0.1228718394551909,
+      "grad_norm": 0.4014485478401184,
+      "learning_rate": 4.9390804597701155e-05,
+      "loss": 0.1523,
+      "step": 106
+    },
+    {
+      "epoch": 0.12403100775193798,
+      "grad_norm": 0.41445374488830566,
+      "learning_rate": 4.938505747126437e-05,
+      "loss": 0.1555,
+      "step": 107
+    },
+    {
+      "epoch": 0.12519017604868507,
+      "grad_norm": 0.39025428891181946,
+      "learning_rate": 4.937931034482759e-05,
+      "loss": 0.1719,
+      "step": 108
+    },
+    {
+      "epoch": 0.12634934434543216,
+      "grad_norm": 0.36380714178085327,
+      "learning_rate": 4.9373563218390806e-05,
+      "loss": 0.1539,
+      "step": 109
+    },
+    {
+      "epoch": 0.12750851264217924,
+      "grad_norm": 0.4117802679538727,
+      "learning_rate": 4.936781609195403e-05,
+      "loss": 0.162,
+      "step": 110
+    },
+    {
+      "epoch": 0.12866768093892633,
+      "grad_norm": 0.3868742883205414,
+      "learning_rate": 4.936206896551724e-05,
+      "loss": 0.1589,
+      "step": 111
+    },
+    {
+      "epoch": 0.1298268492356734,
+      "grad_norm": 0.34659501910209656,
+      "learning_rate": 4.935632183908046e-05,
+      "loss": 0.1594,
+      "step": 112
+    },
+    {
+      "epoch": 0.13098601753242048,
+      "grad_norm": 0.38265547156333923,
+      "learning_rate": 4.935057471264368e-05,
+      "loss": 0.1459,
+      "step": 113
+    },
+    {
+      "epoch": 0.13214518582916757,
+      "grad_norm": 0.39254799485206604,
+      "learning_rate": 4.93448275862069e-05,
+      "loss": 0.1608,
+      "step": 114
+    },
+    {
+      "epoch": 0.13330435412591465,
+      "grad_norm": 0.3889453411102295,
+      "learning_rate": 4.9339080459770116e-05,
+      "loss": 0.1637,
+      "step": 115
+    },
+    {
+      "epoch": 0.13446352242266174,
+      "grad_norm": 0.41454294323921204,
+      "learning_rate": 4.933333333333334e-05,
+      "loss": 0.1532,
+      "step": 116
+    },
+    {
+      "epoch": 0.13562269071940883,
+      "grad_norm": 0.4028087854385376,
+      "learning_rate": 4.932758620689655e-05,
+      "loss": 0.1569,
+      "step": 117
+    },
+    {
+      "epoch": 0.13678185901615592,
+      "grad_norm": 0.4449205994606018,
+      "learning_rate": 4.9321839080459774e-05,
+      "loss": 0.1745,
+      "step": 118
+    },
+    {
+      "epoch": 0.137941027312903,
+      "grad_norm": 0.3685709536075592,
+      "learning_rate": 4.931609195402299e-05,
+      "loss": 0.1476,
+      "step": 119
+    },
+    {
+      "epoch": 0.13910019560965006,
+      "grad_norm": 0.44437339901924133,
+      "learning_rate": 4.931034482758621e-05,
+      "loss": 0.1607,
+      "step": 120
+    },
+    {
+      "epoch": 0.14025936390639715,
+      "grad_norm": 0.5566555261611938,
+      "learning_rate": 4.930459770114943e-05,
+      "loss": 0.1943,
+      "step": 121
+    },
+    {
+      "epoch": 0.14141853220314424,
+      "grad_norm": 0.46472108364105225,
+      "learning_rate": 4.929885057471265e-05,
+      "loss": 0.1634,
+      "step": 122
+    },
+    {
+      "epoch": 0.14257770049989132,
+      "grad_norm": 0.38040968775749207,
+      "learning_rate": 4.929310344827586e-05,
+      "loss": 0.1411,
+      "step": 123
+    },
+    {
+      "epoch": 0.1437368687966384,
+      "grad_norm": 0.3785533607006073,
+      "learning_rate": 4.9287356321839084e-05,
+      "loss": 0.1515,
+      "step": 124
+    },
+    {
+      "epoch": 0.1448960370933855,
+      "grad_norm": 0.402432918548584,
+      "learning_rate": 4.92816091954023e-05,
+      "loss": 0.1529,
+      "step": 125
+    },
+    {
+      "epoch": 0.1460552053901326,
+      "grad_norm": 0.3900763690471649,
+      "learning_rate": 4.9275862068965514e-05,
+      "loss": 0.1716,
+      "step": 126
+    },
+    {
+      "epoch": 0.14721437368687967,
+      "grad_norm": 0.39079996943473816,
+      "learning_rate": 4.927011494252874e-05,
+      "loss": 0.1563,
+      "step": 127
+    },
+    {
+      "epoch": 0.14837354198362676,
+      "grad_norm": 0.38309377431869507,
+      "learning_rate": 4.926436781609196e-05,
+      "loss": 0.1602,
+      "step": 128
+    },
+    {
+      "epoch": 0.14953271028037382,
+      "grad_norm": 0.37955620884895325,
+      "learning_rate": 4.925862068965518e-05,
+      "loss": 0.1703,
+      "step": 129
+    },
+    {
+      "epoch": 0.1506918785771209,
+      "grad_norm": 0.37275955080986023,
+      "learning_rate": 4.9252873563218394e-05,
+      "loss": 0.1599,
+      "step": 130
+    },
+    {
+      "epoch": 0.151851046873868,
+      "grad_norm": 0.48944053053855896,
+      "learning_rate": 4.924712643678161e-05,
+      "loss": 0.1932,
+      "step": 131
+    },
+    {
+      "epoch": 0.15301021517061508,
+      "grad_norm": 0.4082753360271454,
+      "learning_rate": 4.924137931034483e-05,
+      "loss": 0.1674,
+      "step": 132
+    },
+    {
+      "epoch": 0.15416938346736217,
+      "grad_norm": 0.5443015098571777,
+      "learning_rate": 4.9235632183908045e-05,
+      "loss": 0.1679,
+      "step": 133
+    },
+    {
+      "epoch": 0.15532855176410926,
+      "grad_norm": 0.3231983184814453,
+      "learning_rate": 4.922988505747127e-05,
+      "loss": 0.1397,
+      "step": 134
+    },
+    {
+      "epoch": 0.15648772006085634,
+      "grad_norm": 0.384142130613327,
+      "learning_rate": 4.922413793103449e-05,
+      "loss": 0.1601,
+      "step": 135
+    },
+    {
+      "epoch": 0.15764688835760343,
+      "grad_norm": 0.4432188868522644,
+      "learning_rate": 4.9218390804597703e-05,
+      "loss": 0.1746,
+      "step": 136
+    },
+    {
+      "epoch": 0.1588060566543505,
+      "grad_norm": 0.3848440647125244,
+      "learning_rate": 4.9212643678160925e-05,
+      "loss": 0.1514,
+      "step": 137
+    },
+    {
+      "epoch": 0.15996522495109758,
+      "grad_norm": 0.40900081396102905,
+      "learning_rate": 4.920689655172414e-05,
+      "loss": 0.1728,
+      "step": 138
+    },
+    {
+      "epoch": 0.16112439324784467,
+      "grad_norm": 0.35513079166412354,
+      "learning_rate": 4.9201149425287355e-05,
+      "loss": 0.141,
+      "step": 139
+    },
+    {
+      "epoch": 0.16228356154459175,
+      "grad_norm": 0.4035043716430664,
+      "learning_rate": 4.9195402298850577e-05,
+      "loss": 0.1507,
+      "step": 140
+    },
+    {
+      "epoch": 0.16344272984133884,
+      "grad_norm": 0.4026646614074707,
+      "learning_rate": 4.91896551724138e-05,
+      "loss": 0.1629,
+      "step": 141
+    },
+    {
+      "epoch": 0.16460189813808593,
+      "grad_norm": 0.35455062985420227,
+      "learning_rate": 4.918390804597701e-05,
+      "loss": 0.1671,
+      "step": 142
+    },
+    {
+      "epoch": 0.16576106643483302,
+      "grad_norm": 0.3873222768306732,
+      "learning_rate": 4.9178160919540235e-05,
+      "loss": 0.1606,
+      "step": 143
+    },
+    {
+      "epoch": 0.1669202347315801,
+      "grad_norm": 0.34480100870132446,
+      "learning_rate": 4.917241379310345e-05,
+      "loss": 0.1446,
+      "step": 144
+    },
+    {
+      "epoch": 0.16807940302832716,
+      "grad_norm": 0.3599451184272766,
+      "learning_rate": 4.9166666666666665e-05,
+      "loss": 0.1432,
+      "step": 145
+    },
+    {
+      "epoch": 0.16923857132507425,
+      "grad_norm": 0.38096269965171814,
+      "learning_rate": 4.9160919540229886e-05,
+      "loss": 0.1726,
+      "step": 146
+    },
+    {
+      "epoch": 0.17039773962182134,
+      "grad_norm": 0.384888619184494,
+      "learning_rate": 4.915517241379311e-05,
+      "loss": 0.1579,
+      "step": 147
+    },
+    {
+      "epoch": 0.17155690791856842,
+      "grad_norm": 0.3715001344680786,
+      "learning_rate": 4.914942528735633e-05,
+      "loss": 0.1734,
+      "step": 148
+    },
+    {
+      "epoch": 0.1727160762153155,
+      "grad_norm": 0.4220834970474243,
+      "learning_rate": 4.9143678160919545e-05,
+      "loss": 0.1752,
+      "step": 149
+    },
+    {
+      "epoch": 0.1738752445120626,
+      "grad_norm": 0.3289235830307007,
+      "learning_rate": 4.913793103448276e-05,
+      "loss": 0.1441,
+      "step": 150
+    },
+    {
+      "epoch": 0.1750344128088097,
+      "grad_norm": 0.3469524085521698,
+      "learning_rate": 4.913218390804598e-05,
+      "loss": 0.1416,
+      "step": 151
+    },
+    {
+      "epoch": 0.17619358110555677,
+      "grad_norm": 0.38370615243911743,
+      "learning_rate": 4.9126436781609196e-05,
+      "loss": 0.1604,
+      "step": 152
+    },
+    {
+      "epoch": 0.17735274940230383,
+      "grad_norm": 0.33676958084106445,
+      "learning_rate": 4.912068965517241e-05,
+      "loss": 0.1585,
+      "step": 153
+    },
+    {
+      "epoch": 0.17851191769905092,
+      "grad_norm": 0.3484872281551361,
+      "learning_rate": 4.911494252873563e-05,
+      "loss": 0.1516,
+      "step": 154
+    },
+    {
+      "epoch": 0.179671085995798,
+      "grad_norm": 0.3425162434577942,
+      "learning_rate": 4.9109195402298854e-05,
+      "loss": 0.1468,
+      "step": 155
+    },
+    {
+      "epoch": 0.1808302542925451,
+      "grad_norm": 0.3834597170352936,
+      "learning_rate": 4.9103448275862076e-05,
+      "loss": 0.1902,
+      "step": 156
+    },
+    {
+      "epoch": 0.18198942258929218,
+      "grad_norm": 0.4861818253993988,
+      "learning_rate": 4.909770114942529e-05,
+      "loss": 0.1666,
+      "step": 157
+    },
+    {
+      "epoch": 0.18314859088603927,
+      "grad_norm": 0.4293661415576935,
+      "learning_rate": 4.9091954022988506e-05,
+      "loss": 0.1476,
+      "step": 158
+    },
+    {
+      "epoch": 0.18430775918278636,
+      "grad_norm": 0.4156044125556946,
+      "learning_rate": 4.908620689655173e-05,
+      "loss": 0.1427,
+      "step": 159
+    },
+    {
+      "epoch": 0.18546692747953344,
+      "grad_norm": 0.3734503388404846,
+      "learning_rate": 4.908045977011494e-05,
+      "loss": 0.1461,
+      "step": 160
+    },
+    {
+      "epoch": 0.18662609577628053,
+      "grad_norm": 0.4116658568382263,
+      "learning_rate": 4.9074712643678164e-05,
+      "loss": 0.1538,
+      "step": 161
+    },
+    {
+      "epoch": 0.1877852640730276,
+      "grad_norm": 0.40862905979156494,
+      "learning_rate": 4.9068965517241386e-05,
+      "loss": 0.1567,
+      "step": 162
+    },
+    {
+      "epoch": 0.18894443236977468,
+      "grad_norm": 0.3499276041984558,
+      "learning_rate": 4.90632183908046e-05,
+      "loss": 0.1413,
+      "step": 163
+    },
+    {
+      "epoch": 0.19010360066652177,
+      "grad_norm": 0.40991586446762085,
+      "learning_rate": 4.9057471264367816e-05,
+      "loss": 0.1686,
+      "step": 164
+    },
+    {
+      "epoch": 0.19126276896326885,
+      "grad_norm": 0.3891771733760834,
+      "learning_rate": 4.905172413793104e-05,
+      "loss": 0.157,
+      "step": 165
+    },
+    {
+      "epoch": 0.19242193726001594,
+      "grad_norm": 0.3548631966114044,
+      "learning_rate": 4.904597701149425e-05,
+      "loss": 0.1384,
+      "step": 166
+    },
+    {
+      "epoch": 0.19358110555676303,
+      "grad_norm": 0.36914005875587463,
+      "learning_rate": 4.9040229885057474e-05,
+      "loss": 0.1409,
+      "step": 167
+    },
+    {
+      "epoch": 0.19474027385351012,
+      "grad_norm": 0.36561641097068787,
+      "learning_rate": 4.9034482758620695e-05,
+      "loss": 0.155,
+      "step": 168
+    },
+    {
+      "epoch": 0.1958994421502572,
+      "grad_norm": 0.38047492504119873,
+      "learning_rate": 4.902873563218391e-05,
+      "loss": 0.153,
+      "step": 169
+    },
+    {
+      "epoch": 0.19705861044700426,
+      "grad_norm": 0.3930132985115051,
+      "learning_rate": 4.902298850574713e-05,
+      "loss": 0.168,
+      "step": 170
+    },
+    {
+      "epoch": 0.19821777874375135,
+      "grad_norm": 0.4097788333892822,
+      "learning_rate": 4.901724137931035e-05,
+      "loss": 0.1684,
+      "step": 171
+    },
+    {
+      "epoch": 0.19937694704049844,
+      "grad_norm": 0.3460334837436676,
+      "learning_rate": 4.901149425287356e-05,
+      "loss": 0.1477,
+      "step": 172
+    },
+    {
+      "epoch": 0.20053611533724552,
+      "grad_norm": 0.31552115082740784,
+      "learning_rate": 4.9005747126436784e-05,
+      "loss": 0.1382,
+      "step": 173
+    },
+    {
+      "epoch": 0.2016952836339926,
+      "grad_norm": 0.36335986852645874,
+      "learning_rate": 4.9e-05,
+      "loss": 0.1429,
+      "step": 174
+    },
+    {
+      "epoch": 0.2028544519307397,
+      "grad_norm": 0.32661738991737366,
+      "learning_rate": 4.899425287356322e-05,
+      "loss": 0.1515,
+      "step": 175
+    },
+    {
+      "epoch": 0.20401362022748679,
+      "grad_norm": 0.386250764131546,
+      "learning_rate": 4.898850574712644e-05,
+      "loss": 0.1558,
+      "step": 176
+    },
+    {
+      "epoch": 0.20517278852423387,
+      "grad_norm": 0.4087134599685669,
+      "learning_rate": 4.898275862068966e-05,
+      "loss": 0.1639,
+      "step": 177
+    },
+    {
+      "epoch": 0.20633195682098093,
+      "grad_norm": 0.39643535017967224,
+      "learning_rate": 4.897701149425288e-05,
+      "loss": 0.1487,
+      "step": 178
+    },
+    {
+      "epoch": 0.20749112511772802,
+      "grad_norm": 0.36634111404418945,
+      "learning_rate": 4.897126436781609e-05,
+      "loss": 0.1383,
+      "step": 179
+    },
+    {
+      "epoch": 0.2086502934144751,
+      "grad_norm": 0.32414987683296204,
+      "learning_rate": 4.896551724137931e-05,
+      "loss": 0.1345,
+      "step": 180
+    },
+    {
+      "epoch": 0.2098094617112222,
+      "grad_norm": 0.3729398548603058,
+      "learning_rate": 4.895977011494253e-05,
+      "loss": 0.14,
+      "step": 181
+    },
+    {
+      "epoch": 0.21096863000796928,
+      "grad_norm": 0.32442769408226013,
+      "learning_rate": 4.895402298850575e-05,
+      "loss": 0.1573,
+      "step": 182
+    },
+    {
+      "epoch": 0.21212779830471637,
+      "grad_norm": 0.33038130402565,
+      "learning_rate": 4.8948275862068966e-05,
+      "loss": 0.1539,
+      "step": 183
+    },
+    {
+      "epoch": 0.21328696660146346,
+      "grad_norm": 0.3526158630847931,
+      "learning_rate": 4.894252873563219e-05,
+      "loss": 0.1471,
+      "step": 184
+    },
+    {
+      "epoch": 0.21444613489821054,
+      "grad_norm": 0.3698391020298004,
+      "learning_rate": 4.89367816091954e-05,
+      "loss": 0.155,
+      "step": 185
+    },
+    {
+      "epoch": 0.21560530319495763,
+      "grad_norm": 0.35763153433799744,
+      "learning_rate": 4.8931034482758625e-05,
+      "loss": 0.145,
+      "step": 186
+    },
+    {
+      "epoch": 0.2167644714917047,
+      "grad_norm": 0.396771639585495,
+      "learning_rate": 4.892528735632184e-05,
+      "loss": 0.1616,
+      "step": 187
+    },
+    {
+      "epoch": 0.21792363978845178,
+      "grad_norm": 0.3490588963031769,
+      "learning_rate": 4.891954022988506e-05,
+      "loss": 0.1491,
+      "step": 188
+    },
+    {
+      "epoch": 0.21908280808519887,
+      "grad_norm": 0.3624797761440277,
+      "learning_rate": 4.891379310344828e-05,
+      "loss": 0.1731,
+      "step": 189
+    },
+    {
+      "epoch": 0.22024197638194595,
+      "grad_norm": 0.45168307423591614,
+      "learning_rate": 4.89080459770115e-05,
+      "loss": 0.1598,
+      "step": 190
+    },
+    {
+      "epoch": 0.22140114467869304,
+      "grad_norm": 0.35063374042510986,
+      "learning_rate": 4.890229885057471e-05,
+      "loss": 0.1479,
+      "step": 191
+    },
+    {
+      "epoch": 0.22256031297544013,
+      "grad_norm": 0.317920982837677,
+      "learning_rate": 4.8896551724137934e-05,
+      "loss": 0.1432,
+      "step": 192
+    },
+    {
+      "epoch": 0.22371948127218722,
+      "grad_norm": 0.31059324741363525,
+      "learning_rate": 4.889080459770115e-05,
+      "loss": 0.1419,
+      "step": 193
+    },
+    {
+      "epoch": 0.2248786495689343,
+      "grad_norm": 0.3705626428127289,
+      "learning_rate": 4.888505747126437e-05,
+      "loss": 0.1583,
+      "step": 194
+    },
+    {
+      "epoch": 0.22603781786568136,
+      "grad_norm": 0.3402438759803772,
+      "learning_rate": 4.8879310344827586e-05,
+      "loss": 0.138,
+      "step": 195
+    },
+    {
+      "epoch": 0.22719698616242845,
+      "grad_norm": 0.3639499247074127,
+      "learning_rate": 4.887356321839081e-05,
+      "loss": 0.154,
+      "step": 196
+    },
+    {
+      "epoch": 0.22835615445917554,
+      "grad_norm": 0.33627521991729736,
+      "learning_rate": 4.886781609195403e-05,
+      "loss": 0.1435,
+      "step": 197
+    },
+    {
+      "epoch": 0.22951532275592262,
+      "grad_norm": 0.35936155915260315,
+      "learning_rate": 4.8862068965517244e-05,
+      "loss": 0.1567,
+      "step": 198
+    },
+    {
+      "epoch": 0.2306744910526697,
+      "grad_norm": 0.4516816735267639,
+      "learning_rate": 4.885632183908046e-05,
+      "loss": 0.14,
+      "step": 199
+    },
+    {
+      "epoch": 0.2318336593494168,
+      "grad_norm": 0.3906489610671997,
+      "learning_rate": 4.885057471264368e-05,
+      "loss": 0.1405,
+      "step": 200
+    },
+    {
+      "epoch": 0.23299282764616389,
+      "grad_norm": 0.34470584988594055,
+      "learning_rate": 4.8844827586206896e-05,
+      "loss": 0.1519,
+      "step": 201
+    },
+    {
+      "epoch": 0.23415199594291097,
+      "grad_norm": 0.3606179654598236,
+      "learning_rate": 4.883908045977012e-05,
+      "loss": 0.1514,
+      "step": 202
+    },
+    {
+      "epoch": 0.23531116423965803,
+      "grad_norm": 0.31473562121391296,
+      "learning_rate": 4.883333333333334e-05,
+      "loss": 0.1472,
+      "step": 203
+    },
+    {
+      "epoch": 0.23647033253640512,
+      "grad_norm": 0.3649226725101471,
+      "learning_rate": 4.8827586206896554e-05,
+      "loss": 0.1565,
+      "step": 204
+    },
+    {
+      "epoch": 0.2376295008331522,
+      "grad_norm": 0.3161579370498657,
+      "learning_rate": 4.8821839080459776e-05,
+      "loss": 0.1357,
+      "step": 205
+    },
+    {
+      "epoch": 0.2387886691298993,
+      "grad_norm": 0.3359610140323639,
+      "learning_rate": 4.881609195402299e-05,
+      "loss": 0.1486,
+      "step": 206
+    },
+    {
+      "epoch": 0.23994783742664638,
+      "grad_norm": 0.38040071725845337,
+      "learning_rate": 4.8810344827586205e-05,
+      "loss": 0.1607,
+      "step": 207
+    },
+    {
+      "epoch": 0.24110700572339347,
+      "grad_norm": 0.37179407477378845,
+      "learning_rate": 4.880459770114943e-05,
+      "loss": 0.1472,
+      "step": 208
+    },
+    {
+      "epoch": 0.24226617402014056,
+      "grad_norm": 0.3661664128303528,
+      "learning_rate": 4.879885057471265e-05,
+      "loss": 0.1528,
+      "step": 209
+    },
+    {
+      "epoch": 0.24342534231688764,
+      "grad_norm": 0.39830392599105835,
+      "learning_rate": 4.8793103448275864e-05,
+      "loss": 0.1465,
+      "step": 210
+    },
+    {
+      "epoch": 0.2445845106136347,
+      "grad_norm": 0.4218502938747406,
+      "learning_rate": 4.8787356321839085e-05,
+      "loss": 0.1552,
+      "step": 211
+    },
+    {
+      "epoch": 0.2457436789103818,
+      "grad_norm": 0.45591163635253906,
+      "learning_rate": 4.87816091954023e-05,
+      "loss": 0.1648,
+      "step": 212
+    },
+    {
+      "epoch": 0.24690284720712888,
+      "grad_norm": 0.33883577585220337,
+      "learning_rate": 4.877586206896552e-05,
+      "loss": 0.1471,
+      "step": 213
+    },
+    {
+      "epoch": 0.24806201550387597,
+      "grad_norm": 0.4139156937599182,
+      "learning_rate": 4.877011494252874e-05,
+      "loss": 0.1603,
+      "step": 214
+    },
+    {
+      "epoch": 0.24922118380062305,
+      "grad_norm": 0.4006992280483246,
+      "learning_rate": 4.876436781609195e-05,
+      "loss": 0.1698,
+      "step": 215
+    },
+    {
+      "epoch": 0.25038035209737014,
+      "grad_norm": 0.3272416889667511,
+      "learning_rate": 4.875862068965517e-05,
+      "loss": 0.1271,
+      "step": 216
+    },
+    {
+      "epoch": 0.2515395203941172,
+      "grad_norm": 0.3487521708011627,
+      "learning_rate": 4.8752873563218395e-05,
+      "loss": 0.1528,
+      "step": 217
+    },
+    {
+      "epoch": 0.2526986886908643,
+      "grad_norm": 0.39113104343414307,
+      "learning_rate": 4.874712643678161e-05,
+      "loss": 0.1763,
+      "step": 218
+    },
+    {
+      "epoch": 0.2538578569876114,
+      "grad_norm": 0.32122161984443665,
+      "learning_rate": 4.874137931034483e-05,
+      "loss": 0.1324,
+      "step": 219
+    },
+    {
+      "epoch": 0.2550170252843585,
+      "grad_norm": 0.32524266839027405,
+      "learning_rate": 4.8735632183908047e-05,
+      "loss": 0.1483,
+      "step": 220
+    },
+    {
+      "epoch": 0.25617619358110555,
+      "grad_norm": 0.3443804085254669,
+      "learning_rate": 4.872988505747126e-05,
+      "loss": 0.1656,
+      "step": 221
+    },
+    {
+      "epoch": 0.25733536187785266,
+      "grad_norm": 0.33968833088874817,
+      "learning_rate": 4.872413793103448e-05,
+      "loss": 0.1425,
+      "step": 222
+    },
+    {
+      "epoch": 0.2584945301745997,
+      "grad_norm": 0.29967644810676575,
+      "learning_rate": 4.8718390804597705e-05,
+      "loss": 0.1458,
+      "step": 223
+    },
+    {
+      "epoch": 0.2596536984713468,
+      "grad_norm": 0.36407482624053955,
+      "learning_rate": 4.8712643678160926e-05,
+      "loss": 0.1467,
+      "step": 224
+    },
+    {
+      "epoch": 0.2608128667680939,
+      "grad_norm": 0.42614683508872986,
+      "learning_rate": 4.870689655172414e-05,
+      "loss": 0.1665,
+      "step": 225
+    },
+    {
+      "epoch": 0.26197203506484096,
+      "grad_norm": 0.3307810425758362,
+      "learning_rate": 4.8701149425287356e-05,
+      "loss": 0.1389,
+      "step": 226
+    },
+    {
+      "epoch": 0.2631312033615881,
+      "grad_norm": 0.3416149616241455,
+      "learning_rate": 4.869540229885058e-05,
+      "loss": 0.1616,
+      "step": 227
+    },
+    {
+      "epoch": 0.26429037165833513,
+      "grad_norm": 0.3303369879722595,
+      "learning_rate": 4.868965517241379e-05,
+      "loss": 0.1472,
+      "step": 228
+    },
+    {
+      "epoch": 0.26544953995508225,
+      "grad_norm": 0.31898778676986694,
+      "learning_rate": 4.8683908045977015e-05,
+      "loss": 0.1374,
+      "step": 229
+    },
+    {
+      "epoch": 0.2666087082518293,
+      "grad_norm": 0.33968180418014526,
+      "learning_rate": 4.8678160919540236e-05,
+      "loss": 0.1549,
+      "step": 230
+    },
+    {
+      "epoch": 0.2677678765485764,
+      "grad_norm": 0.3462338149547577,
+      "learning_rate": 4.867241379310345e-05,
+      "loss": 0.1548,
+      "step": 231
+    },
+    {
+      "epoch": 0.2689270448453235,
+      "grad_norm": 0.27784019708633423,
+      "learning_rate": 4.866666666666667e-05,
+      "loss": 0.1323,
+      "step": 232
+    },
+    {
+      "epoch": 0.27008621314207054,
+      "grad_norm": 0.3605106770992279,
+      "learning_rate": 4.866091954022989e-05,
+      "loss": 0.149,
+      "step": 233
+    },
+    {
+      "epoch": 0.27124538143881766,
+      "grad_norm": 0.39858150482177734,
+      "learning_rate": 4.86551724137931e-05,
+      "loss": 0.1625,
+      "step": 234
+    },
+    {
+      "epoch": 0.2724045497355647,
+      "grad_norm": 0.34399569034576416,
+      "learning_rate": 4.8649425287356324e-05,
+      "loss": 0.1536,
+      "step": 235
+    },
+    {
+      "epoch": 0.27356371803231183,
+      "grad_norm": 0.33953142166137695,
+      "learning_rate": 4.864367816091954e-05,
+      "loss": 0.1471,
+      "step": 236
+    },
+    {
+      "epoch": 0.2747228863290589,
+      "grad_norm": 0.3053218722343445,
+      "learning_rate": 4.863793103448276e-05,
+      "loss": 0.1288,
+      "step": 237
+    },
+    {
+      "epoch": 0.275882054625806,
+      "grad_norm": 0.40773364901542664,
+      "learning_rate": 4.863218390804598e-05,
+      "loss": 0.1814,
+      "step": 238
+    },
+    {
+      "epoch": 0.27704122292255307,
+      "grad_norm": 0.35415759682655334,
+      "learning_rate": 4.86264367816092e-05,
+      "loss": 0.1579,
+      "step": 239
+    },
+    {
+      "epoch": 0.2782003912193001,
+      "grad_norm": 0.39274096488952637,
+      "learning_rate": 4.862068965517241e-05,
+      "loss": 0.1678,
+      "step": 240
+    },
+    {
+      "epoch": 0.27935955951604724,
+      "grad_norm": 0.3034732937812805,
+      "learning_rate": 4.8614942528735634e-05,
+      "loss": 0.1498,
+      "step": 241
+    },
+    {
+      "epoch": 0.2805187278127943,
+      "grad_norm": 0.3109908998012543,
+      "learning_rate": 4.860919540229885e-05,
+      "loss": 0.149,
+      "step": 242
+    },
+    {
+      "epoch": 0.2816778961095414,
+      "grad_norm": 0.31770509481430054,
+      "learning_rate": 4.860344827586207e-05,
+      "loss": 0.1479,
+      "step": 243
+    },
+    {
+      "epoch": 0.2828370644062885,
+      "grad_norm": 0.3324630856513977,
+      "learning_rate": 4.859770114942529e-05,
+      "loss": 0.15,
+      "step": 244
+    },
+    {
+      "epoch": 0.2839962327030356,
+      "grad_norm": 0.30650344491004944,
+      "learning_rate": 4.859195402298851e-05,
+      "loss": 0.1328,
+      "step": 245
+    },
+    {
+      "epoch": 0.28515540099978265,
+      "grad_norm": 0.3292284309864044,
+      "learning_rate": 4.858620689655173e-05,
+      "loss": 0.1432,
+      "step": 246
+    },
+    {
+      "epoch": 0.28631456929652976,
+      "grad_norm": 0.3721246123313904,
+      "learning_rate": 4.8580459770114944e-05,
+      "loss": 0.1695,
+      "step": 247
+    },
+    {
+      "epoch": 0.2874737375932768,
+      "grad_norm": 0.39138564467430115,
+      "learning_rate": 4.857471264367816e-05,
+      "loss": 0.1595,
+      "step": 248
+    },
+    {
+      "epoch": 0.2886329058900239,
+      "grad_norm": 0.3113502860069275,
+      "learning_rate": 4.856896551724138e-05,
+      "loss": 0.1474,
+      "step": 249
+    },
+    {
+      "epoch": 0.289792074186771,
+      "grad_norm": 0.3272615373134613,
+      "learning_rate": 4.85632183908046e-05,
+      "loss": 0.1557,
+      "step": 250
+    },
+    {
+      "epoch": 0.29095124248351806,
+      "grad_norm": 0.32985273003578186,
+      "learning_rate": 4.8557471264367824e-05,
+      "loss": 0.1543,
+      "step": 251
+    },
+    {
+      "epoch": 0.2921104107802652,
+      "grad_norm": 0.4038926362991333,
+      "learning_rate": 4.855172413793104e-05,
+      "loss": 0.1762,
+      "step": 252
+    },
+    {
+      "epoch": 0.29326957907701223,
+      "grad_norm": 0.5201258063316345,
+      "learning_rate": 4.8545977011494253e-05,
+      "loss": 0.1435,
+      "step": 253
+    },
+    {
+      "epoch": 0.29442874737375935,
+      "grad_norm": 0.43428415060043335,
+      "learning_rate": 4.8540229885057475e-05,
+      "loss": 0.1461,
+      "step": 254
+    },
+    {
+      "epoch": 0.2955879156705064,
+      "grad_norm": 0.29131442308425903,
+      "learning_rate": 4.853448275862069e-05,
+      "loss": 0.1384,
+      "step": 255
+    },
+    {
+      "epoch": 0.2967470839672535,
+      "grad_norm": 0.40190252661705017,
+      "learning_rate": 4.8528735632183905e-05,
+      "loss": 0.1433,
+      "step": 256
+    },
+    {
+      "epoch": 0.2979062522640006,
+      "grad_norm": 0.34535473585128784,
+      "learning_rate": 4.8522988505747133e-05,
+      "loss": 0.147,
+      "step": 257
+    },
+    {
+      "epoch": 0.29906542056074764,
+      "grad_norm": 0.3541038930416107,
+      "learning_rate": 4.851724137931035e-05,
+      "loss": 0.1658,
+      "step": 258
+    },
+    {
+      "epoch": 0.30022458885749476,
+      "grad_norm": 0.39234408736228943,
+      "learning_rate": 4.851149425287357e-05,
+      "loss": 0.1711,
+      "step": 259
+    },
+    {
+      "epoch": 0.3013837571542418,
+      "grad_norm": 0.37314096093177795,
+      "learning_rate": 4.8505747126436785e-05,
+      "loss": 0.1475,
+      "step": 260
+    },
+    {
+      "epoch": 0.30254292545098893,
+      "grad_norm": 0.3385733366012573,
+      "learning_rate": 4.85e-05,
+      "loss": 0.1406,
+      "step": 261
+    },
+    {
+      "epoch": 0.303702093747736,
+      "grad_norm": 0.6898816823959351,
+      "learning_rate": 4.849425287356322e-05,
+      "loss": 0.1584,
+      "step": 262
+    },
+    {
+      "epoch": 0.3048612620444831,
+      "grad_norm": 0.30045434832572937,
+      "learning_rate": 4.8488505747126436e-05,
+      "loss": 0.1329,
+      "step": 263
+    },
+    {
+      "epoch": 0.30602043034123017,
+      "grad_norm": 0.40219223499298096,
+      "learning_rate": 4.848275862068966e-05,
+      "loss": 0.1529,
+      "step": 264
+    },
+    {
+      "epoch": 0.3071795986379772,
+      "grad_norm": 0.30509233474731445,
+      "learning_rate": 4.847701149425288e-05,
+      "loss": 0.1591,
+      "step": 265
+    },
+    {
+      "epoch": 0.30833876693472434,
+      "grad_norm": 0.3983840048313141,
+      "learning_rate": 4.8471264367816095e-05,
+      "loss": 0.1598,
+      "step": 266
+    },
+    {
+      "epoch": 0.3094979352314714,
+      "grad_norm": 0.39158061146736145,
+      "learning_rate": 4.846551724137931e-05,
+      "loss": 0.164,
+      "step": 267
+    },
+    {
+      "epoch": 0.3106571035282185,
+      "grad_norm": 0.3619145452976227,
+      "learning_rate": 4.845977011494253e-05,
+      "loss": 0.152,
+      "step": 268
+    },
+    {
+      "epoch": 0.3118162718249656,
+      "grad_norm": 0.34043577313423157,
+      "learning_rate": 4.8454022988505746e-05,
+      "loss": 0.1465,
+      "step": 269
+    },
+    {
+      "epoch": 0.3129754401217127,
+      "grad_norm": 0.38417112827301025,
+      "learning_rate": 4.844827586206897e-05,
+      "loss": 0.1398,
+      "step": 270
+    },
+    {
+      "epoch": 0.31413460841845975,
+      "grad_norm": 0.33110880851745605,
+      "learning_rate": 4.844252873563219e-05,
+      "loss": 0.1435,
+      "step": 271
+    },
+    {
+      "epoch": 0.31529377671520686,
+      "grad_norm": 0.3095364272594452,
+      "learning_rate": 4.8436781609195404e-05,
+      "loss": 0.1474,
+      "step": 272
+    },
+    {
+      "epoch": 0.3164529450119539,
+      "grad_norm": 0.34774062037467957,
+      "learning_rate": 4.8431034482758626e-05,
+      "loss": 0.156,
+      "step": 273
+    },
+    {
+      "epoch": 0.317612113308701,
+      "grad_norm": 0.3605648875236511,
+      "learning_rate": 4.842528735632184e-05,
+      "loss": 0.1545,
+      "step": 274
+    },
+    {
+      "epoch": 0.3187712816054481,
+      "grad_norm": 0.32547488808631897,
+      "learning_rate": 4.8419540229885056e-05,
+      "loss": 0.1448,
+      "step": 275
+    },
+    {
+      "epoch": 0.31993044990219516,
+      "grad_norm": 0.4434783458709717,
+      "learning_rate": 4.841379310344828e-05,
+      "loss": 0.1837,
+      "step": 276
+    },
+    {
+      "epoch": 0.3210896181989423,
+      "grad_norm": 0.3950771391391754,
+      "learning_rate": 4.840804597701149e-05,
+      "loss": 0.1667,
+      "step": 277
+    },
+    {
+      "epoch": 0.32224878649568933,
+      "grad_norm": 0.27604764699935913,
+      "learning_rate": 4.840229885057472e-05,
+      "loss": 0.1311,
+      "step": 278
+    },
+    {
+      "epoch": 0.32340795479243645,
+      "grad_norm": 0.29781222343444824,
+      "learning_rate": 4.8396551724137936e-05,
+      "loss": 0.1507,
+      "step": 279
+    },
+    {
+      "epoch": 0.3245671230891835,
+      "grad_norm": 0.300820529460907,
+      "learning_rate": 4.839080459770115e-05,
+      "loss": 0.1395,
+      "step": 280
+    },
+    {
+      "epoch": 0.3257262913859306,
+      "grad_norm": 0.3563086688518524,
+      "learning_rate": 4.838505747126437e-05,
+      "loss": 0.1542,
+      "step": 281
+    },
+    {
+      "epoch": 0.3268854596826777,
+      "grad_norm": 0.3500363826751709,
+      "learning_rate": 4.837931034482759e-05,
+      "loss": 0.1531,
+      "step": 282
+    },
+    {
+      "epoch": 0.32804462797942474,
+      "grad_norm": 0.340751975774765,
+      "learning_rate": 4.83735632183908e-05,
+      "loss": 0.154,
+      "step": 283
+    },
+    {
+      "epoch": 0.32920379627617186,
+      "grad_norm": 0.3770763874053955,
+      "learning_rate": 4.8367816091954024e-05,
+      "loss": 0.1632,
+      "step": 284
+    },
+    {
+      "epoch": 0.3303629645729189,
+      "grad_norm": 0.3394136130809784,
+      "learning_rate": 4.8362068965517246e-05,
+      "loss": 0.1575,
+      "step": 285
+    },
+    {
+      "epoch": 0.33152213286966603,
+      "grad_norm": 0.3633858263492584,
+      "learning_rate": 4.835632183908046e-05,
+      "loss": 0.1511,
+      "step": 286
+    },
+    {
+      "epoch": 0.3326813011664131,
+      "grad_norm": 0.27468162775039673,
+      "learning_rate": 4.835057471264368e-05,
+      "loss": 0.1445,
+      "step": 287
+    },
+    {
+      "epoch": 0.3338404694631602,
+      "grad_norm": 0.29763033986091614,
+      "learning_rate": 4.83448275862069e-05,
+      "loss": 0.1462,
+      "step": 288
+    },
+    {
+      "epoch": 0.33499963775990726,
+      "grad_norm": 0.2883772850036621,
+      "learning_rate": 4.833908045977012e-05,
+      "loss": 0.135,
+      "step": 289
+    },
+    {
+      "epoch": 0.3361588060566543,
+      "grad_norm": 0.3238728642463684,
+      "learning_rate": 4.8333333333333334e-05,
+      "loss": 0.1576,
+      "step": 290
+    },
+    {
+      "epoch": 0.33731797435340144,
+      "grad_norm": 0.4045778214931488,
+      "learning_rate": 4.8327586206896555e-05,
+      "loss": 0.1555,
+      "step": 291
+    },
+    {
+      "epoch": 0.3384771426501485,
+      "grad_norm": 0.3956005871295929,
+      "learning_rate": 4.832183908045978e-05,
+      "loss": 0.17,
+      "step": 292
+    },
+    {
+      "epoch": 0.3396363109468956,
+      "grad_norm": 0.2896331548690796,
+      "learning_rate": 4.831609195402299e-05,
+      "loss": 0.1436,
+      "step": 293
+    },
+    {
+      "epoch": 0.3407954792436427,
+      "grad_norm": 0.29543447494506836,
+      "learning_rate": 4.831034482758621e-05,
+      "loss": 0.1499,
+      "step": 294
+    },
+    {
+      "epoch": 0.3419546475403898,
+      "grad_norm": 0.32724907994270325,
+      "learning_rate": 4.830459770114943e-05,
+      "loss": 0.1568,
+      "step": 295
+    },
+    {
+      "epoch": 0.34311381583713685,
+      "grad_norm": 0.376208633184433,
+      "learning_rate": 4.829885057471264e-05,
+      "loss": 0.1449,
+      "step": 296
+    },
+    {
+      "epoch": 0.34427298413388396,
+      "grad_norm": 0.29767367243766785,
+      "learning_rate": 4.8293103448275865e-05,
+      "loss": 0.1444,
+      "step": 297
+    },
+    {
+      "epoch": 0.345432152430631,
+      "grad_norm": 0.30211395025253296,
+      "learning_rate": 4.828735632183909e-05,
+      "loss": 0.1375,
+      "step": 298
+    },
+    {
+      "epoch": 0.3465913207273781,
+      "grad_norm": 0.29970985651016235,
+      "learning_rate": 4.82816091954023e-05,
+      "loss": 0.1319,
+      "step": 299
+    },
+    {
+      "epoch": 0.3477504890241252,
+      "grad_norm": 0.30623510479927063,
+      "learning_rate": 4.827586206896552e-05,
+      "loss": 0.1489,
+      "step": 300
+    },
+    {
+      "epoch": 0.34890965732087226,
+      "grad_norm": 0.31533321738243103,
+      "learning_rate": 4.827011494252874e-05,
+      "loss": 0.1525,
+      "step": 301
+    },
+    {
+      "epoch": 0.3500688256176194,
+      "grad_norm": 0.35718074440956116,
+      "learning_rate": 4.826436781609195e-05,
+      "loss": 0.1465,
+      "step": 302
+    },
+    {
+      "epoch": 0.35122799391436643,
+      "grad_norm": 0.3315868675708771,
+      "learning_rate": 4.8258620689655175e-05,
+      "loss": 0.1438,
+      "step": 303
+    },
+    {
+      "epoch": 0.35238716221111355,
+      "grad_norm": 0.3789491355419159,
+      "learning_rate": 4.825287356321839e-05,
+      "loss": 0.1624,
+      "step": 304
+    },
+    {
+      "epoch": 0.3535463305078606,
+      "grad_norm": 0.3216198980808258,
+      "learning_rate": 4.824712643678161e-05,
+      "loss": 0.1495,
+      "step": 305
+    },
+    {
+      "epoch": 0.35470549880460767,
+      "grad_norm": 0.24692386388778687,
+      "learning_rate": 4.824137931034483e-05,
+      "loss": 0.1333,
+      "step": 306
+    },
+    {
+      "epoch": 0.3558646671013548,
+      "grad_norm": 0.26879554986953735,
+      "learning_rate": 4.823563218390805e-05,
+      "loss": 0.1328,
+      "step": 307
+    },
+    {
+      "epoch": 0.35702383539810184,
+      "grad_norm": 0.3616696894168854,
+      "learning_rate": 4.822988505747127e-05,
+      "loss": 0.1642,
+      "step": 308
+    },
+    {
+      "epoch": 0.35818300369484896,
+      "grad_norm": 0.4611773192882538,
+      "learning_rate": 4.8224137931034485e-05,
+      "loss": 0.1342,
+      "step": 309
+    },
+    {
+      "epoch": 0.359342171991596,
+      "grad_norm": 0.26914867758750916,
+      "learning_rate": 4.82183908045977e-05,
+      "loss": 0.1393,
+      "step": 310
+    },
+    {
+      "epoch": 0.36050134028834313,
+      "grad_norm": 0.3391886353492737,
+      "learning_rate": 4.821264367816092e-05,
+      "loss": 0.1502,
+      "step": 311
+    },
+    {
+      "epoch": 0.3616605085850902,
+      "grad_norm": 0.34171319007873535,
+      "learning_rate": 4.820689655172414e-05,
+      "loss": 0.1655,
+      "step": 312
+    },
+    {
+      "epoch": 0.3628196768818373,
+      "grad_norm": 0.2735805809497833,
+      "learning_rate": 4.820114942528736e-05,
+      "loss": 0.1211,
+      "step": 313
+    },
+    {
+      "epoch": 0.36397884517858436,
+      "grad_norm": 0.25337541103363037,
+      "learning_rate": 4.819540229885058e-05,
+      "loss": 0.1343,
+      "step": 314
+    },
+    {
+      "epoch": 0.3651380134753314,
+      "grad_norm": 0.3965268135070801,
+      "learning_rate": 4.8189655172413794e-05,
+      "loss": 0.1507,
+      "step": 315
+    },
+    {
+      "epoch": 0.36629718177207854,
+      "grad_norm": 0.26941707730293274,
+      "learning_rate": 4.8183908045977016e-05,
+      "loss": 0.1258,
+      "step": 316
+    },
+    {
+      "epoch": 0.3674563500688256,
+      "grad_norm": 0.31832900643348694,
+      "learning_rate": 4.817816091954023e-05,
+      "loss": 0.1693,
+      "step": 317
+    },
+    {
+      "epoch": 0.3686155183655727,
+      "grad_norm": 0.3194613456726074,
+      "learning_rate": 4.817241379310345e-05,
+      "loss": 0.1574,
+      "step": 318
+    },
+    {
+      "epoch": 0.3697746866623198,
+      "grad_norm": 0.3136231601238251,
+      "learning_rate": 4.8166666666666674e-05,
+      "loss": 0.1301,
+      "step": 319
+    },
+    {
+      "epoch": 0.3709338549590669,
+      "grad_norm": 0.302774578332901,
+      "learning_rate": 4.816091954022989e-05,
+      "loss": 0.1329,
+      "step": 320
+    },
+    {
+      "epoch": 0.37209302325581395,
+      "grad_norm": 0.3740023672580719,
+      "learning_rate": 4.8155172413793104e-05,
+      "loss": 0.1398,
+      "step": 321
+    },
+    {
+      "epoch": 0.37325219155256106,
+      "grad_norm": 0.35367128252983093,
+      "learning_rate": 4.8149425287356326e-05,
+      "loss": 0.1461,
+      "step": 322
+    },
+    {
+      "epoch": 0.3744113598493081,
+      "grad_norm": 0.35140734910964966,
+      "learning_rate": 4.814367816091954e-05,
+      "loss": 0.1545,
+      "step": 323
+    },
+    {
+      "epoch": 0.3755705281460552,
+      "grad_norm": 0.3153414726257324,
+      "learning_rate": 4.8137931034482755e-05,
+      "loss": 0.1283,
+      "step": 324
+    },
+    {
+      "epoch": 0.3767296964428023,
+      "grad_norm": 0.30042028427124023,
+      "learning_rate": 4.813218390804598e-05,
+      "loss": 0.151,
+      "step": 325
+    },
+    {
+      "epoch": 0.37788886473954936,
+      "grad_norm": 0.29036179184913635,
+      "learning_rate": 4.81264367816092e-05,
+      "loss": 0.1486,
+      "step": 326
+    },
+    {
+      "epoch": 0.3790480330362965,
+      "grad_norm": 0.2988058924674988,
+      "learning_rate": 4.812068965517242e-05,
+      "loss": 0.1545,
+      "step": 327
+    },
+    {
+      "epoch": 0.38020720133304353,
+      "grad_norm": 0.3292902410030365,
+      "learning_rate": 4.8114942528735635e-05,
+      "loss": 0.1629,
+      "step": 328
+    },
+    {
+      "epoch": 0.38136636962979065,
+      "grad_norm": 0.3768965005874634,
+      "learning_rate": 4.810919540229885e-05,
+      "loss": 0.1692,
+      "step": 329
+    },
+    {
+      "epoch": 0.3825255379265377,
+      "grad_norm": 0.28651896119117737,
+      "learning_rate": 4.810344827586207e-05,
+      "loss": 0.1534,
+      "step": 330
+    },
+    {
+      "epoch": 0.38368470622328477,
+      "grad_norm": 0.2976914942264557,
+      "learning_rate": 4.809770114942529e-05,
+      "loss": 0.1451,
+      "step": 331
+    },
+    {
+      "epoch": 0.3848438745200319,
+      "grad_norm": 0.3275349736213684,
+      "learning_rate": 4.809195402298851e-05,
+      "loss": 0.1632,
+      "step": 332
+    },
+    {
+      "epoch": 0.38600304281677894,
+      "grad_norm": 0.26491785049438477,
+      "learning_rate": 4.808620689655173e-05,
+      "loss": 0.1334,
+      "step": 333
+    },
+    {
+      "epoch": 0.38716221111352606,
+      "grad_norm": 0.28282400965690613,
+      "learning_rate": 4.8080459770114945e-05,
+      "loss": 0.1349,
+      "step": 334
+    },
+    {
+      "epoch": 0.3883213794102731,
+      "grad_norm": 0.31815335154533386,
+      "learning_rate": 4.807471264367817e-05,
+      "loss": 0.1514,
+      "step": 335
+    },
+    {
+      "epoch": 0.38948054770702023,
+      "grad_norm": 0.29894590377807617,
+      "learning_rate": 4.806896551724138e-05,
+      "loss": 0.1563,
+      "step": 336
+    },
+    {
+      "epoch": 0.3906397160037673,
+      "grad_norm": 0.31386902928352356,
+      "learning_rate": 4.80632183908046e-05,
+      "loss": 0.1553,
+      "step": 337
+    },
+    {
+      "epoch": 0.3917988843005144,
+      "grad_norm": 0.2859923541545868,
+      "learning_rate": 4.805747126436782e-05,
+      "loss": 0.1531,
+      "step": 338
+    },
+    {
+      "epoch": 0.39295805259726146,
+      "grad_norm": 0.3352300524711609,
+      "learning_rate": 4.805172413793104e-05,
+      "loss": 0.1412,
+      "step": 339
+    },
+    {
+      "epoch": 0.3941172208940085,
+      "grad_norm": 0.3148883283138275,
+      "learning_rate": 4.8045977011494255e-05,
+      "loss": 0.1297,
+      "step": 340
+    },
+    {
+      "epoch": 0.39527638919075564,
+      "grad_norm": 0.27482402324676514,
+      "learning_rate": 4.8040229885057477e-05,
+      "loss": 0.1333,
+      "step": 341
+    },
+    {
+      "epoch": 0.3964355574875027,
+      "grad_norm": 0.3114891052246094,
+      "learning_rate": 4.803448275862069e-05,
+      "loss": 0.1494,
+      "step": 342
+    },
+    {
+      "epoch": 0.3975947257842498,
+      "grad_norm": 0.3976518213748932,
+      "learning_rate": 4.8028735632183906e-05,
+      "loss": 0.1575,
+      "step": 343
+    },
+    {
+      "epoch": 0.3987538940809969,
+      "grad_norm": 0.4233146607875824,
+      "learning_rate": 4.802298850574713e-05,
+      "loss": 0.1459,
+      "step": 344
+    },
+    {
+      "epoch": 0.399913062377744,
+      "grad_norm": 0.30527806282043457,
+      "learning_rate": 4.801724137931034e-05,
+      "loss": 0.1369,
+      "step": 345
+    },
+    {
+      "epoch": 0.40107223067449105,
+      "grad_norm": 0.28466179966926575,
+      "learning_rate": 4.8011494252873565e-05,
+      "loss": 0.1283,
+      "step": 346
+    },
+    {
+      "epoch": 0.40223139897123816,
+      "grad_norm": 0.3994863033294678,
+      "learning_rate": 4.8005747126436786e-05,
+      "loss": 0.1623,
+      "step": 347
+    },
+    {
+      "epoch": 0.4033905672679852,
+      "grad_norm": 0.32415613532066345,
+      "learning_rate": 4.8e-05,
+      "loss": 0.155,
+      "step": 348
+    },
+    {
+      "epoch": 0.4045497355647323,
+      "grad_norm": 0.37542182207107544,
+      "learning_rate": 4.799425287356322e-05,
+      "loss": 0.1594,
+      "step": 349
+    },
+    {
+      "epoch": 0.4057089038614794,
+      "grad_norm": 0.33119359612464905,
+      "learning_rate": 4.798850574712644e-05,
+      "loss": 0.1555,
+      "step": 350
+    },
+    {
+      "epoch": 0.40686807215822646,
+      "grad_norm": 0.2560279369354248,
+      "learning_rate": 4.798275862068965e-05,
+      "loss": 0.1241,
+      "step": 351
+    },
+    {
+      "epoch": 0.40802724045497357,
+      "grad_norm": 0.2843601405620575,
+      "learning_rate": 4.7977011494252874e-05,
+      "loss": 0.1289,
+      "step": 352
+    },
+    {
+      "epoch": 0.40918640875172063,
+      "grad_norm": 0.29628679156303406,
+      "learning_rate": 4.7971264367816096e-05,
+      "loss": 0.1476,
+      "step": 353
+    },
+    {
+      "epoch": 0.41034557704846775,
+      "grad_norm": 0.36033913493156433,
+      "learning_rate": 4.796551724137932e-05,
+      "loss": 0.1579,
+      "step": 354
+    },
+    {
+      "epoch": 0.4115047453452148,
+      "grad_norm": 0.2819597125053406,
+      "learning_rate": 4.795977011494253e-05,
+      "loss": 0.1389,
+      "step": 355
+    },
+    {
+      "epoch": 0.41266391364196187,
+      "grad_norm": 0.2536991834640503,
+      "learning_rate": 4.795402298850575e-05,
+      "loss": 0.1342,
+      "step": 356
+    },
+    {
+      "epoch": 0.413823081938709,
+      "grad_norm": 0.339702844619751,
+      "learning_rate": 4.794827586206897e-05,
+      "loss": 0.1452,
+      "step": 357
+    },
+    {
+      "epoch": 0.41498225023545604,
+      "grad_norm": 0.29577532410621643,
+      "learning_rate": 4.7942528735632184e-05,
+      "loss": 0.1516,
+      "step": 358
+    },
+    {
+      "epoch": 0.41614141853220316,
+      "grad_norm": 0.30197975039482117,
+      "learning_rate": 4.7936781609195406e-05,
+      "loss": 0.1341,
+      "step": 359
+    },
+    {
+      "epoch": 0.4173005868289502,
+      "grad_norm": 0.2787219285964966,
+      "learning_rate": 4.793103448275863e-05,
+      "loss": 0.1426,
+      "step": 360
+    },
+    {
+      "epoch": 0.41845975512569733,
+      "grad_norm": 0.3281581997871399,
+      "learning_rate": 4.792528735632184e-05,
+      "loss": 0.1708,
+      "step": 361
+    },
+    {
+      "epoch": 0.4196189234224444,
+      "grad_norm": 0.2598758041858673,
+      "learning_rate": 4.791954022988506e-05,
+      "loss": 0.1323,
+      "step": 362
+    },
+    {
+      "epoch": 0.4207780917191915,
+      "grad_norm": 0.26790300011634827,
+      "learning_rate": 4.791379310344828e-05,
+      "loss": 0.1202,
+      "step": 363
+    },
+    {
+      "epoch": 0.42193726001593856,
+      "grad_norm": 0.2569654583930969,
+      "learning_rate": 4.7908045977011494e-05,
+      "loss": 0.1312,
+      "step": 364
+    },
+    {
+      "epoch": 0.4230964283126856,
+      "grad_norm": 0.3506622314453125,
+      "learning_rate": 4.7902298850574716e-05,
+      "loss": 0.1634,
+      "step": 365
+    },
+    {
+      "epoch": 0.42425559660943274,
+      "grad_norm": 0.3123525083065033,
+      "learning_rate": 4.789655172413793e-05,
+      "loss": 0.1588,
+      "step": 366
+    },
+    {
+      "epoch": 0.4254147649061798,
+      "grad_norm": 0.26198455691337585,
+      "learning_rate": 4.789080459770115e-05,
+      "loss": 0.1387,
+      "step": 367
+    },
+    {
+      "epoch": 0.4265739332029269,
+      "grad_norm": 0.3061137795448303,
+      "learning_rate": 4.7885057471264374e-05,
+      "loss": 0.1444,
+      "step": 368
+    },
+    {
+      "epoch": 0.427733101499674,
+      "grad_norm": 0.48860934376716614,
+      "learning_rate": 4.787931034482759e-05,
+      "loss": 0.151,
+      "step": 369
+    },
+    {
+      "epoch": 0.4288922697964211,
+      "grad_norm": 0.3341166377067566,
+      "learning_rate": 4.7873563218390804e-05,
+      "loss": 0.1377,
+      "step": 370
+    },
+    {
+      "epoch": 0.43005143809316815,
+      "grad_norm": 0.38944321870803833,
+      "learning_rate": 4.7867816091954025e-05,
+      "loss": 0.1635,
+      "step": 371
+    },
+    {
+      "epoch": 0.43121060638991526,
+      "grad_norm": 0.3289624750614166,
+      "learning_rate": 4.786206896551724e-05,
+      "loss": 0.1492,
+      "step": 372
+    },
+    {
+      "epoch": 0.4323697746866623,
+      "grad_norm": 0.2940187454223633,
+      "learning_rate": 4.785632183908046e-05,
+      "loss": 0.1386,
+      "step": 373
+    },
+    {
+      "epoch": 0.4335289429834094,
+      "grad_norm": 0.3036423921585083,
+      "learning_rate": 4.7850574712643684e-05,
+      "loss": 0.1383,
+      "step": 374
+    },
+    {
+      "epoch": 0.4346881112801565,
+      "grad_norm": 0.343669593334198,
+      "learning_rate": 4.78448275862069e-05,
+      "loss": 0.1605,
+      "step": 375
+    },
+    {
+      "epoch": 0.43584727957690356,
+      "grad_norm": 0.3061246871948242,
+      "learning_rate": 4.783908045977012e-05,
+      "loss": 0.1286,
+      "step": 376
+    },
+    {
+      "epoch": 0.43700644787365067,
+      "grad_norm": 0.33070918917655945,
+      "learning_rate": 4.7833333333333335e-05,
+      "loss": 0.1458,
+      "step": 377
+    },
+    {
+      "epoch": 0.43816561617039773,
+      "grad_norm": 0.9597157835960388,
+      "learning_rate": 4.782758620689655e-05,
+      "loss": 0.1461,
+      "step": 378
+    },
+    {
+      "epoch": 0.43932478446714485,
+      "grad_norm": 0.28434523940086365,
+      "learning_rate": 4.782183908045977e-05,
+      "loss": 0.1383,
+      "step": 379
+    },
+    {
+      "epoch": 0.4404839527638919,
+      "grad_norm": 0.2993975579738617,
+      "learning_rate": 4.781609195402299e-05,
+      "loss": 0.1398,
+      "step": 380
+    },
+    {
+      "epoch": 0.44164312106063897,
+      "grad_norm": 0.2816145420074463,
+      "learning_rate": 4.781034482758621e-05,
+      "loss": 0.1337,
+      "step": 381
+    },
+    {
+      "epoch": 0.4428022893573861,
+      "grad_norm": 0.906739354133606,
+      "learning_rate": 4.780459770114943e-05,
+      "loss": 0.1285,
+      "step": 382
+    },
+    {
+      "epoch": 0.44396145765413314,
+      "grad_norm": 0.25160372257232666,
+      "learning_rate": 4.7798850574712645e-05,
+      "loss": 0.1219,
+      "step": 383
+    },
+    {
+      "epoch": 0.44512062595088026,
+      "grad_norm": 0.3275229036808014,
+      "learning_rate": 4.7793103448275866e-05,
+      "loss": 0.1546,
+      "step": 384
+    },
+    {
+      "epoch": 0.4462797942476273,
+      "grad_norm": 0.2726188600063324,
+      "learning_rate": 4.778735632183908e-05,
+      "loss": 0.1479,
+      "step": 385
+    },
+    {
+      "epoch": 0.44743896254437443,
+      "grad_norm": 0.2804552912712097,
+      "learning_rate": 4.7781609195402296e-05,
+      "loss": 0.1386,
+      "step": 386
+    },
+    {
+      "epoch": 0.4485981308411215,
+      "grad_norm": 0.3268015682697296,
+      "learning_rate": 4.777586206896552e-05,
+      "loss": 0.1549,
+      "step": 387
+    },
+    {
+      "epoch": 0.4497572991378686,
+      "grad_norm": 0.263644814491272,
+      "learning_rate": 4.777011494252874e-05,
+      "loss": 0.1401,
+      "step": 388
+    },
+    {
+      "epoch": 0.45091646743461566,
+      "grad_norm": 0.28181982040405273,
+      "learning_rate": 4.7764367816091954e-05,
+      "loss": 0.1351,
+      "step": 389
+    },
+    {
+      "epoch": 0.4520756357313627,
+      "grad_norm": 0.2432672381401062,
+      "learning_rate": 4.7758620689655176e-05,
+      "loss": 0.1279,
+      "step": 390
+    },
+    {
+      "epoch": 0.45323480402810984,
+      "grad_norm": 0.24589940905570984,
+      "learning_rate": 4.775287356321839e-05,
+      "loss": 0.133,
+      "step": 391
+    },
+    {
+      "epoch": 0.4543939723248569,
+      "grad_norm": 0.29978325963020325,
+      "learning_rate": 4.774712643678161e-05,
+      "loss": 0.1392,
+      "step": 392
+    },
+    {
+      "epoch": 0.455553140621604,
+      "grad_norm": 0.32140663266181946,
+      "learning_rate": 4.774137931034483e-05,
+      "loss": 0.1376,
+      "step": 393
+    },
+    {
+      "epoch": 0.4567123089183511,
+      "grad_norm": 0.3055313527584076,
+      "learning_rate": 4.773563218390805e-05,
+      "loss": 0.148,
+      "step": 394
+    },
+    {
+      "epoch": 0.4578714772150982,
+      "grad_norm": 0.31229299306869507,
+      "learning_rate": 4.772988505747127e-05,
+      "loss": 0.1446,
+      "step": 395
+    },
+    {
+      "epoch": 0.45903064551184525,
+      "grad_norm": 0.3736218214035034,
+      "learning_rate": 4.7724137931034486e-05,
+      "loss": 0.1704,
+      "step": 396
+    },
+    {
+      "epoch": 0.4601898138085923,
+      "grad_norm": 0.2979601323604584,
+      "learning_rate": 4.77183908045977e-05,
+      "loss": 0.1347,
+      "step": 397
+    },
+    {
+      "epoch": 0.4613489821053394,
+      "grad_norm": 0.288699746131897,
+      "learning_rate": 4.771264367816092e-05,
+      "loss": 0.1335,
+      "step": 398
+    },
+    {
+      "epoch": 0.4625081504020865,
+      "grad_norm": 0.30045926570892334,
+      "learning_rate": 4.770689655172414e-05,
+      "loss": 0.1466,
+      "step": 399
+    },
+    {
+      "epoch": 0.4636673186988336,
+      "grad_norm": 0.41966867446899414,
+      "learning_rate": 4.770114942528736e-05,
+      "loss": 0.1628,
+      "step": 400
+    },
+    {
+      "epoch": 0.46482648699558066,
+      "grad_norm": 0.3311939537525177,
+      "learning_rate": 4.769540229885058e-05,
+      "loss": 0.1448,
+      "step": 401
+    },
+    {
+      "epoch": 0.46598565529232777,
+      "grad_norm": 0.2968727648258209,
+      "learning_rate": 4.7689655172413796e-05,
+      "loss": 0.1471,
+      "step": 402
+    },
+    {
+      "epoch": 0.46714482358907483,
+      "grad_norm": 0.3325969874858856,
+      "learning_rate": 4.768390804597702e-05,
+      "loss": 0.1527,
+      "step": 403
+    },
+    {
+      "epoch": 0.46830399188582195,
+      "grad_norm": 0.32422512769699097,
+      "learning_rate": 4.767816091954023e-05,
+      "loss": 0.1516,
+      "step": 404
+    },
+    {
+      "epoch": 0.469463160182569,
+      "grad_norm": 0.28346890211105347,
+      "learning_rate": 4.767241379310345e-05,
+      "loss": 0.1359,
+      "step": 405
+    },
+    {
+      "epoch": 0.47062232847931607,
+      "grad_norm": 0.27612075209617615,
+      "learning_rate": 4.766666666666667e-05,
+      "loss": 0.1373,
+      "step": 406
+    },
+    {
+      "epoch": 0.4717814967760632,
+      "grad_norm": 0.2808459401130676,
+      "learning_rate": 4.7660919540229884e-05,
+      "loss": 0.1408,
+      "step": 407
+    },
+    {
+      "epoch": 0.47294066507281024,
+      "grad_norm": 0.2832328677177429,
+      "learning_rate": 4.7655172413793105e-05,
+      "loss": 0.139,
+      "step": 408
+    },
+    {
+      "epoch": 0.47409983336955736,
+      "grad_norm": 0.27688664197921753,
+      "learning_rate": 4.764942528735633e-05,
+      "loss": 0.1392,
+      "step": 409
+    },
+    {
+      "epoch": 0.4752590016663044,
+      "grad_norm": 0.3172287344932556,
+      "learning_rate": 4.764367816091954e-05,
+      "loss": 0.1512,
+      "step": 410
+    },
+    {
+      "epoch": 0.47641816996305153,
+      "grad_norm": 0.2498215287923813,
+      "learning_rate": 4.7637931034482764e-05,
+      "loss": 0.12,
+      "step": 411
+    },
+    {
+      "epoch": 0.4775773382597986,
+      "grad_norm": 0.3091871440410614,
+      "learning_rate": 4.763218390804598e-05,
+      "loss": 0.1403,
+      "step": 412
+    },
+    {
+      "epoch": 0.4787365065565457,
+      "grad_norm": 0.29513412714004517,
+      "learning_rate": 4.7626436781609193e-05,
+      "loss": 0.1553,
+      "step": 413
+    },
+    {
+      "epoch": 0.47989567485329276,
+      "grad_norm": 0.31259381771087646,
+      "learning_rate": 4.7620689655172415e-05,
+      "loss": 0.138,
+      "step": 414
+    },
+    {
+      "epoch": 0.4810548431500398,
+      "grad_norm": 0.3411064147949219,
+      "learning_rate": 4.761494252873564e-05,
+      "loss": 0.1385,
+      "step": 415
+    },
+    {
+      "epoch": 0.48221401144678694,
+      "grad_norm": 0.5548298358917236,
+      "learning_rate": 4.760919540229885e-05,
+      "loss": 0.1357,
+      "step": 416
+    },
+    {
+      "epoch": 0.483373179743534,
+      "grad_norm": 0.3198058605194092,
+      "learning_rate": 4.7603448275862073e-05,
+      "loss": 0.1578,
+      "step": 417
+    },
+    {
+      "epoch": 0.4845323480402811,
+      "grad_norm": 0.2812342643737793,
+      "learning_rate": 4.759770114942529e-05,
+      "loss": 0.1559,
+      "step": 418
+    },
+    {
+      "epoch": 0.4856915163370282,
+      "grad_norm": 0.36544597148895264,
+      "learning_rate": 4.759195402298851e-05,
+      "loss": 0.1513,
+      "step": 419
+    },
+    {
+      "epoch": 0.4868506846337753,
+      "grad_norm": 0.29660576581954956,
+      "learning_rate": 4.7586206896551725e-05,
+      "loss": 0.1506,
+      "step": 420
+    },
+    {
+      "epoch": 0.48800985293052235,
+      "grad_norm": 0.2850414216518402,
+      "learning_rate": 4.7580459770114947e-05,
+      "loss": 0.1371,
+      "step": 421
+    },
+    {
+      "epoch": 0.4891690212272694,
+      "grad_norm": 0.3292869031429291,
+      "learning_rate": 4.757471264367817e-05,
+      "loss": 0.1758,
+      "step": 422
+    },
+    {
+      "epoch": 0.4903281895240165,
+      "grad_norm": 0.2677929997444153,
+      "learning_rate": 4.756896551724138e-05,
+      "loss": 0.1345,
+      "step": 423
+    },
+    {
+      "epoch": 0.4914873578207636,
+      "grad_norm": 0.32202059030532837,
+      "learning_rate": 4.75632183908046e-05,
+      "loss": 0.1465,
+      "step": 424
+    },
+    {
+      "epoch": 0.4926465261175107,
+      "grad_norm": 0.2961694300174713,
+      "learning_rate": 4.755747126436782e-05,
+      "loss": 0.1461,
+      "step": 425
+    },
+    {
+      "epoch": 0.49380569441425776,
+      "grad_norm": 0.27797091007232666,
+      "learning_rate": 4.7551724137931035e-05,
+      "loss": 0.1186,
+      "step": 426
+    },
+    {
+      "epoch": 0.49496486271100487,
+      "grad_norm": 0.42147669196128845,
+      "learning_rate": 4.754597701149425e-05,
+      "loss": 0.1351,
+      "step": 427
+    },
+    {
+      "epoch": 0.49612403100775193,
+      "grad_norm": 0.41197213530540466,
+      "learning_rate": 4.754022988505747e-05,
+      "loss": 0.1242,
+      "step": 428
+    },
+    {
+      "epoch": 0.49728319930449905,
+      "grad_norm": 0.352716326713562,
+      "learning_rate": 4.753448275862069e-05,
+      "loss": 0.1541,
+      "step": 429
+    },
+    {
+      "epoch": 0.4984423676012461,
+      "grad_norm": 0.29174548387527466,
+      "learning_rate": 4.7528735632183915e-05,
+      "loss": 0.1466,
+      "step": 430
+    },
+    {
+      "epoch": 0.49960153589799317,
+      "grad_norm": 0.32256200909614563,
+      "learning_rate": 4.752298850574713e-05,
+      "loss": 0.1365,
+      "step": 431
+    },
+    {
+      "epoch": 0.5007607041947403,
+      "grad_norm": 0.35533028841018677,
+      "learning_rate": 4.7517241379310344e-05,
+      "loss": 0.1489,
+      "step": 432
+    },
+    {
+      "epoch": 0.5019198724914874,
+      "grad_norm": 0.3250570297241211,
+      "learning_rate": 4.7511494252873566e-05,
+      "loss": 0.157,
+      "step": 433
+    },
+    {
+      "epoch": 0.5030790407882344,
+      "grad_norm": 0.36209189891815186,
+      "learning_rate": 4.750574712643678e-05,
+      "loss": 0.1579,
+      "step": 434
+    },
+    {
+      "epoch": 0.5042382090849815,
+      "grad_norm": 0.33881399035453796,
+      "learning_rate": 4.75e-05,
+      "loss": 0.1516,
+      "step": 435
+    },
+    {
+      "epoch": 0.5053973773817286,
+      "grad_norm": 0.2697497010231018,
+      "learning_rate": 4.7494252873563224e-05,
+      "loss": 0.1347,
+      "step": 436
+    },
+    {
+      "epoch": 0.5065565456784757,
+      "grad_norm": 0.3097374439239502,
+      "learning_rate": 4.748850574712644e-05,
+      "loss": 0.1506,
+      "step": 437
+    },
+    {
+      "epoch": 0.5077157139752227,
+      "grad_norm": 0.28055351972579956,
+      "learning_rate": 4.748275862068966e-05,
+      "loss": 0.1418,
+      "step": 438
+    },
+    {
+      "epoch": 0.5088748822719699,
+      "grad_norm": 0.31767186522483826,
+      "learning_rate": 4.7477011494252876e-05,
+      "loss": 0.1502,
+      "step": 439
+    },
+    {
+      "epoch": 0.510034050568717,
+      "grad_norm": 0.3027343451976776,
+      "learning_rate": 4.747126436781609e-05,
+      "loss": 0.1437,
+      "step": 440
+    },
+    {
+      "epoch": 0.511193218865464,
+      "grad_norm": 0.2773297131061554,
+      "learning_rate": 4.746551724137931e-05,
+      "loss": 0.139,
+      "step": 441
+    },
+    {
+      "epoch": 0.5123523871622111,
+      "grad_norm": 0.2976417541503906,
+      "learning_rate": 4.7459770114942534e-05,
+      "loss": 0.147,
+      "step": 442
+    },
+    {
+      "epoch": 0.5135115554589582,
+      "grad_norm": 0.30484509468078613,
+      "learning_rate": 4.745402298850575e-05,
+      "loss": 0.1371,
+      "step": 443
+    },
+    {
+      "epoch": 0.5146707237557053,
+      "grad_norm": 0.3369823694229126,
+      "learning_rate": 4.744827586206897e-05,
+      "loss": 0.1674,
+      "step": 444
+    },
+    {
+      "epoch": 0.5158298920524523,
+      "grad_norm": 0.27961060404777527,
+      "learning_rate": 4.7442528735632186e-05,
+      "loss": 0.141,
+      "step": 445
+    },
+    {
+      "epoch": 0.5169890603491994,
+      "grad_norm": 0.29113152623176575,
+      "learning_rate": 4.74367816091954e-05,
+      "loss": 0.1432,
+      "step": 446
+    },
+    {
+      "epoch": 0.5181482286459466,
+      "grad_norm": 0.363936185836792,
+      "learning_rate": 4.743103448275862e-05,
+      "loss": 0.1637,
+      "step": 447
+    },
+    {
+      "epoch": 0.5193073969426936,
+      "grad_norm": 0.3023361563682556,
+      "learning_rate": 4.742528735632184e-05,
+      "loss": 0.1368,
+      "step": 448
+    },
+    {
+      "epoch": 0.5204665652394407,
+      "grad_norm": 0.33404213190078735,
+      "learning_rate": 4.7419540229885065e-05,
+      "loss": 0.1555,
+      "step": 449
+    },
+    {
+      "epoch": 0.5216257335361878,
+      "grad_norm": 0.302898645401001,
+      "learning_rate": 4.741379310344828e-05,
+      "loss": 0.1354,
+      "step": 450
+    },
+    {
+      "epoch": 0.5227849018329349,
+      "grad_norm": 0.35367482900619507,
+      "learning_rate": 4.7408045977011495e-05,
+      "loss": 0.1583,
+      "step": 451
+    },
+    {
+      "epoch": 0.5239440701296819,
+      "grad_norm": 0.31517934799194336,
+      "learning_rate": 4.740229885057472e-05,
+      "loss": 0.1457,
+      "step": 452
+    },
+    {
+      "epoch": 0.525103238426429,
+      "grad_norm": 0.2723785936832428,
+      "learning_rate": 4.739655172413793e-05,
+      "loss": 0.1381,
+      "step": 453
+    },
+    {
+      "epoch": 0.5262624067231761,
+      "grad_norm": 0.2963480055332184,
+      "learning_rate": 4.739080459770115e-05,
+      "loss": 0.1408,
+      "step": 454
+    },
+    {
+      "epoch": 0.5274215750199232,
+      "grad_norm": 0.31834137439727783,
+      "learning_rate": 4.738505747126437e-05,
+      "loss": 0.1411,
+      "step": 455
+    },
+    {
+      "epoch": 0.5285807433166703,
+      "grad_norm": 0.31244561076164246,
+      "learning_rate": 4.737931034482759e-05,
+      "loss": 0.1368,
+      "step": 456
+    },
+    {
+      "epoch": 0.5297399116134174,
+      "grad_norm": 0.29169803857803345,
+      "learning_rate": 4.737356321839081e-05,
+      "loss": 0.1367,
+      "step": 457
+    },
+    {
+      "epoch": 0.5308990799101645,
+      "grad_norm": 0.2684137523174286,
+      "learning_rate": 4.736781609195403e-05,
+      "loss": 0.1482,
+      "step": 458
+    },
+    {
+      "epoch": 0.5320582482069115,
+      "grad_norm": 0.26147085428237915,
+      "learning_rate": 4.736206896551724e-05,
+      "loss": 0.1406,
+      "step": 459
+    },
+    {
+      "epoch": 0.5332174165036586,
+      "grad_norm": 0.2554711401462555,
+      "learning_rate": 4.735632183908046e-05,
+      "loss": 0.131,
+      "step": 460
+    },
+    {
+      "epoch": 0.5343765848004057,
+      "grad_norm": 0.2564930319786072,
+      "learning_rate": 4.735057471264368e-05,
+      "loss": 0.126,
+      "step": 461
+    },
+    {
+      "epoch": 0.5355357530971528,
+      "grad_norm": 0.2714848220348358,
+      "learning_rate": 4.73448275862069e-05,
+      "loss": 0.1308,
+      "step": 462
+    },
+    {
+      "epoch": 0.5366949213938998,
+      "grad_norm": 0.2883203327655792,
+      "learning_rate": 4.733908045977012e-05,
+      "loss": 0.1286,
+      "step": 463
+    },
+    {
+      "epoch": 0.537854089690647,
+      "grad_norm": 0.29676780104637146,
+      "learning_rate": 4.7333333333333336e-05,
+      "loss": 0.1499,
+      "step": 464
+    },
+    {
+      "epoch": 0.5390132579873941,
+      "grad_norm": 0.2832816541194916,
+      "learning_rate": 4.732758620689655e-05,
+      "loss": 0.1452,
+      "step": 465
+    },
+    {
+      "epoch": 0.5401724262841411,
+      "grad_norm": 0.28993991017341614,
+      "learning_rate": 4.732183908045977e-05,
+      "loss": 0.1501,
+      "step": 466
+    },
+    {
+      "epoch": 0.5413315945808882,
+      "grad_norm": 0.3065154254436493,
+      "learning_rate": 4.731609195402299e-05,
+      "loss": 0.1545,
+      "step": 467
+    },
+    {
+      "epoch": 0.5424907628776353,
+      "grad_norm": 0.28335148096084595,
+      "learning_rate": 4.731034482758621e-05,
+      "loss": 0.1305,
+      "step": 468
+    },
+    {
+      "epoch": 0.5436499311743824,
+      "grad_norm": 0.29181987047195435,
+      "learning_rate": 4.730459770114943e-05,
+      "loss": 0.1379,
+      "step": 469
+    },
+    {
+      "epoch": 0.5448090994711294,
+      "grad_norm": 0.2934599220752716,
+      "learning_rate": 4.7298850574712646e-05,
+      "loss": 0.1435,
+      "step": 470
+    },
+    {
+      "epoch": 0.5459682677678765,
+      "grad_norm": 0.2953728437423706,
+      "learning_rate": 4.729310344827587e-05,
+      "loss": 0.1453,
+      "step": 471
+    },
+    {
+      "epoch": 0.5471274360646237,
+      "grad_norm": 0.2823188006877899,
+      "learning_rate": 4.728735632183908e-05,
+      "loss": 0.1312,
+      "step": 472
+    },
+    {
+      "epoch": 0.5482866043613707,
+      "grad_norm": 0.2421528846025467,
+      "learning_rate": 4.72816091954023e-05,
+      "loss": 0.1207,
+      "step": 473
+    },
+    {
+      "epoch": 0.5494457726581178,
+      "grad_norm": 0.3098517954349518,
+      "learning_rate": 4.727586206896552e-05,
+      "loss": 0.1492,
+      "step": 474
+    },
+    {
+      "epoch": 0.5506049409548649,
+      "grad_norm": 0.2447943240404129,
+      "learning_rate": 4.7270114942528734e-05,
+      "loss": 0.128,
+      "step": 475
+    },
+    {
+      "epoch": 0.551764109251612,
+      "grad_norm": 0.37644991278648376,
+      "learning_rate": 4.7264367816091956e-05,
+      "loss": 0.1414,
+      "step": 476
+    },
+    {
+      "epoch": 0.552923277548359,
+      "grad_norm": 0.2898421883583069,
+      "learning_rate": 4.725862068965518e-05,
+      "loss": 0.1382,
+      "step": 477
+    },
+    {
+      "epoch": 0.5540824458451061,
+      "grad_norm": 0.3101235032081604,
+      "learning_rate": 4.725287356321839e-05,
+      "loss": 0.1507,
+      "step": 478
+    },
+    {
+      "epoch": 0.5552416141418532,
+      "grad_norm": 0.31535589694976807,
+      "learning_rate": 4.7247126436781614e-05,
+      "loss": 0.1479,
+      "step": 479
+    },
+    {
+      "epoch": 0.5564007824386002,
+      "grad_norm": 0.3017010986804962,
+      "learning_rate": 4.724137931034483e-05,
+      "loss": 0.1491,
+      "step": 480
+    },
+    {
+      "epoch": 0.5575599507353474,
+      "grad_norm": 0.27675431966781616,
+      "learning_rate": 4.7235632183908044e-05,
+      "loss": 0.1343,
+      "step": 481
+    },
+    {
+      "epoch": 0.5587191190320945,
+      "grad_norm": 0.28481778502464294,
+      "learning_rate": 4.7229885057471266e-05,
+      "loss": 0.1598,
+      "step": 482
+    },
+    {
+      "epoch": 0.5598782873288416,
+      "grad_norm": 0.2603989541530609,
+      "learning_rate": 4.722413793103449e-05,
+      "loss": 0.1365,
+      "step": 483
+    },
+    {
+      "epoch": 0.5610374556255886,
+      "grad_norm": 0.2759985625743866,
+      "learning_rate": 4.72183908045977e-05,
+      "loss": 0.1291,
+      "step": 484
+    },
+    {
+      "epoch": 0.5621966239223357,
+      "grad_norm": 0.27497798204421997,
+      "learning_rate": 4.7212643678160924e-05,
+      "loss": 0.1302,
+      "step": 485
+    },
+    {
+      "epoch": 0.5633557922190828,
+      "grad_norm": 0.4083801209926605,
+      "learning_rate": 4.720689655172414e-05,
+      "loss": 0.1512,
+      "step": 486
+    },
+    {
+      "epoch": 0.5645149605158299,
+      "grad_norm": 0.25597989559173584,
+      "learning_rate": 4.720114942528736e-05,
+      "loss": 0.125,
+      "step": 487
+    },
+    {
+      "epoch": 0.565674128812577,
+      "grad_norm": 0.34843572974205017,
+      "learning_rate": 4.7195402298850575e-05,
+      "loss": 0.1537,
+      "step": 488
+    },
+    {
+      "epoch": 0.5668332971093241,
+      "grad_norm": 0.34892401099205017,
+      "learning_rate": 4.718965517241379e-05,
+      "loss": 0.1342,
+      "step": 489
+    },
+    {
+      "epoch": 0.5679924654060712,
+      "grad_norm": 0.28030136227607727,
+      "learning_rate": 4.718390804597702e-05,
+      "loss": 0.1477,
+      "step": 490
+    },
+    {
+      "epoch": 0.5691516337028182,
+      "grad_norm": 0.27783283591270447,
+      "learning_rate": 4.7178160919540234e-05,
+      "loss": 0.1431,
+      "step": 491
+    },
+    {
+      "epoch": 0.5703108019995653,
+      "grad_norm": 0.23067143559455872,
+      "learning_rate": 4.717241379310345e-05,
+      "loss": 0.1242,
+      "step": 492
+    },
+    {
+      "epoch": 0.5714699702963124,
+      "grad_norm": 0.25984445214271545,
+      "learning_rate": 4.716666666666667e-05,
+      "loss": 0.1298,
+      "step": 493
+    },
+    {
+      "epoch": 0.5726291385930595,
+      "grad_norm": 0.255754292011261,
+      "learning_rate": 4.7160919540229885e-05,
+      "loss": 0.1398,
+      "step": 494
+    },
+    {
+      "epoch": 0.5737883068898065,
+      "grad_norm": 0.28669071197509766,
+      "learning_rate": 4.715517241379311e-05,
+      "loss": 0.1428,
+      "step": 495
+    },
+    {
+      "epoch": 0.5749474751865536,
+      "grad_norm": 0.29052338004112244,
+      "learning_rate": 4.714942528735632e-05,
+      "loss": 0.1399,
+      "step": 496
+    },
+    {
+      "epoch": 0.5761066434833008,
+      "grad_norm": 0.24293456971645355,
+      "learning_rate": 4.714367816091954e-05,
+      "loss": 0.122,
+      "step": 497
+    },
+    {
+      "epoch": 0.5772658117800478,
+      "grad_norm": 0.278358519077301,
+      "learning_rate": 4.7137931034482765e-05,
+      "loss": 0.146,
+      "step": 498
+    },
+    {
+      "epoch": 0.5784249800767949,
+      "grad_norm": 0.25874063372612,
+      "learning_rate": 4.713218390804598e-05,
+      "loss": 0.1298,
+      "step": 499
+    },
+    {
+      "epoch": 0.579584148373542,
+      "grad_norm": 0.4070255756378174,
+      "learning_rate": 4.7126436781609195e-05,
+      "loss": 0.1511,
+      "step": 500
+    },
+    {
+      "epoch": 0.5807433166702891,
+      "grad_norm": 0.4098628759384155,
+      "learning_rate": 4.7120689655172417e-05,
+      "loss": 0.1399,
+      "step": 501
+    },
+    {
+      "epoch": 0.5819024849670361,
+      "grad_norm": 0.33255958557128906,
+      "learning_rate": 4.711494252873563e-05,
+      "loss": 0.1489,
+      "step": 502
+    },
+    {
+      "epoch": 0.5830616532637832,
+      "grad_norm": 0.2659994959831238,
+      "learning_rate": 4.710919540229885e-05,
+      "loss": 0.1258,
+      "step": 503
+    },
+    {
+      "epoch": 0.5842208215605303,
+      "grad_norm": 0.30784937739372253,
+      "learning_rate": 4.7103448275862075e-05,
+      "loss": 0.1468,
+      "step": 504
+    },
+    {
+      "epoch": 0.5853799898572773,
+      "grad_norm": 0.30625712871551514,
+      "learning_rate": 4.709770114942529e-05,
+      "loss": 0.1355,
+      "step": 505
+    },
+    {
+      "epoch": 0.5865391581540245,
+      "grad_norm": 0.32840919494628906,
+      "learning_rate": 4.709195402298851e-05,
+      "loss": 0.1346,
+      "step": 506
+    },
+    {
+      "epoch": 0.5876983264507716,
+      "grad_norm": 0.2901023030281067,
+      "learning_rate": 4.7086206896551726e-05,
+      "loss": 0.1621,
+      "step": 507
+    },
+    {
+      "epoch": 0.5888574947475187,
+      "grad_norm": 0.29218408465385437,
+      "learning_rate": 4.708045977011494e-05,
+      "loss": 0.1415,
+      "step": 508
+    },
+    {
+      "epoch": 0.5900166630442657,
+      "grad_norm": 0.2910397946834564,
+      "learning_rate": 4.707471264367816e-05,
+      "loss": 0.1536,
+      "step": 509
+    },
+    {
+      "epoch": 0.5911758313410128,
+      "grad_norm": 0.23362290859222412,
+      "learning_rate": 4.7068965517241385e-05,
+      "loss": 0.1204,
+      "step": 510
+    },
+    {
+      "epoch": 0.5923349996377599,
+      "grad_norm": 0.29354870319366455,
+      "learning_rate": 4.70632183908046e-05,
+      "loss": 0.1392,
+      "step": 511
+    },
+    {
+      "epoch": 0.593494167934507,
+      "grad_norm": 0.26072168350219727,
+      "learning_rate": 4.705747126436782e-05,
+      "loss": 0.1376,
+      "step": 512
+    },
+    {
+      "epoch": 0.594653336231254,
+      "grad_norm": 0.32599276304244995,
+      "learning_rate": 4.7051724137931036e-05,
+      "loss": 0.1371,
+      "step": 513
+    },
+    {
+      "epoch": 0.5958125045280012,
+      "grad_norm": 0.2970302402973175,
+      "learning_rate": 4.704597701149426e-05,
+      "loss": 0.1376,
+      "step": 514
+    },
+    {
+      "epoch": 0.5969716728247483,
+      "grad_norm": 0.2292068749666214,
+      "learning_rate": 4.704022988505747e-05,
+      "loss": 0.1264,
+      "step": 515
+    },
+    {
+      "epoch": 0.5981308411214953,
+      "grad_norm": 0.2724842131137848,
+      "learning_rate": 4.703448275862069e-05,
+      "loss": 0.136,
+      "step": 516
+    },
+    {
+      "epoch": 0.5992900094182424,
+      "grad_norm": 0.2579404413700104,
+      "learning_rate": 4.702873563218391e-05,
+      "loss": 0.1378,
+      "step": 517
+    },
+    {
+      "epoch": 0.6004491777149895,
+      "grad_norm": 0.28821900486946106,
+      "learning_rate": 4.702298850574713e-05,
+      "loss": 0.1481,
+      "step": 518
+    },
+    {
+      "epoch": 0.6016083460117366,
+      "grad_norm": 0.31739768385887146,
+      "learning_rate": 4.7017241379310346e-05,
+      "loss": 0.1536,
+      "step": 519
+    },
+    {
+      "epoch": 0.6027675143084836,
+      "grad_norm": 0.2741948366165161,
+      "learning_rate": 4.701149425287357e-05,
+      "loss": 0.1293,
+      "step": 520
+    },
+    {
+      "epoch": 0.6039266826052307,
+      "grad_norm": 0.2746334969997406,
+      "learning_rate": 4.700574712643678e-05,
+      "loss": 0.1309,
+      "step": 521
+    },
+    {
+      "epoch": 0.6050858509019779,
+      "grad_norm": 0.2569972276687622,
+      "learning_rate": 4.7e-05,
+      "loss": 0.1192,
+      "step": 522
+    },
+    {
+      "epoch": 0.6062450191987249,
+      "grad_norm": 0.3422163128852844,
+      "learning_rate": 4.699425287356322e-05,
+      "loss": 0.1518,
+      "step": 523
+    },
+    {
+      "epoch": 0.607404187495472,
+      "grad_norm": 0.30587247014045715,
+      "learning_rate": 4.698850574712644e-05,
+      "loss": 0.1599,
+      "step": 524
+    },
+    {
+      "epoch": 0.6085633557922191,
+      "grad_norm": 0.27892470359802246,
+      "learning_rate": 4.698275862068966e-05,
+      "loss": 0.1373,
+      "step": 525
+    },
+    {
+      "epoch": 0.6097225240889662,
+      "grad_norm": 0.3484020233154297,
+      "learning_rate": 4.697701149425288e-05,
+      "loss": 0.1368,
+      "step": 526
+    },
+    {
+      "epoch": 0.6108816923857132,
+      "grad_norm": 0.2913890779018402,
+      "learning_rate": 4.697126436781609e-05,
+      "loss": 0.1351,
+      "step": 527
+    },
+    {
+      "epoch": 0.6120408606824603,
+      "grad_norm": 0.3555639982223511,
+      "learning_rate": 4.6965517241379314e-05,
+      "loss": 0.1608,
+      "step": 528
+    },
+    {
+      "epoch": 0.6132000289792074,
+      "grad_norm": 0.2778078317642212,
+      "learning_rate": 4.695977011494253e-05,
+      "loss": 0.1446,
+      "step": 529
+    },
+    {
+      "epoch": 0.6143591972759544,
+      "grad_norm": 0.24142858386039734,
+      "learning_rate": 4.695402298850575e-05,
+      "loss": 0.1161,
+      "step": 530
+    },
+    {
+      "epoch": 0.6155183655727016,
+      "grad_norm": 0.2555319368839264,
+      "learning_rate": 4.694827586206897e-05,
+      "loss": 0.1229,
+      "step": 531
+    },
+    {
+      "epoch": 0.6166775338694487,
+      "grad_norm": 0.2874903678894043,
+      "learning_rate": 4.694252873563219e-05,
+      "loss": 0.1487,
+      "step": 532
+    },
+    {
+      "epoch": 0.6178367021661958,
+      "grad_norm": 0.27659285068511963,
+      "learning_rate": 4.693678160919541e-05,
+      "loss": 0.1458,
+      "step": 533
+    },
+    {
+      "epoch": 0.6189958704629428,
+      "grad_norm": 0.3088036775588989,
+      "learning_rate": 4.6931034482758623e-05,
+      "loss": 0.1448,
+      "step": 534
+    },
+    {
+      "epoch": 0.6201550387596899,
+      "grad_norm": 0.338383287191391,
+      "learning_rate": 4.692528735632184e-05,
+      "loss": 0.124,
+      "step": 535
+    },
+    {
+      "epoch": 0.621314207056437,
+      "grad_norm": 0.27692559361457825,
+      "learning_rate": 4.691954022988506e-05,
+      "loss": 0.1417,
+      "step": 536
+    },
+    {
+      "epoch": 0.6224733753531841,
+      "grad_norm": 0.2864556610584259,
+      "learning_rate": 4.6913793103448275e-05,
+      "loss": 0.1406,
+      "step": 537
+    },
+    {
+      "epoch": 0.6236325436499311,
+      "grad_norm": 0.3573722243309021,
+      "learning_rate": 4.69080459770115e-05,
+      "loss": 0.1445,
+      "step": 538
+    },
+    {
+      "epoch": 0.6247917119466783,
+      "grad_norm": 0.27817103266716003,
+      "learning_rate": 4.690229885057472e-05,
+      "loss": 0.1359,
+      "step": 539
+    },
+    {
+      "epoch": 0.6259508802434254,
+      "grad_norm": 0.3294277787208557,
+      "learning_rate": 4.689655172413793e-05,
+      "loss": 0.1455,
+      "step": 540
+    },
+    {
+      "epoch": 0.6271100485401724,
+      "grad_norm": 0.2555405795574188,
+      "learning_rate": 4.689080459770115e-05,
+      "loss": 0.1352,
+      "step": 541
+    },
+    {
+      "epoch": 0.6282692168369195,
+      "grad_norm": 0.27222466468811035,
+      "learning_rate": 4.688505747126437e-05,
+      "loss": 0.1391,
+      "step": 542
+    },
+    {
+      "epoch": 0.6294283851336666,
+      "grad_norm": 0.26646101474761963,
+      "learning_rate": 4.6879310344827585e-05,
+      "loss": 0.1405,
+      "step": 543
+    },
+    {
+      "epoch": 0.6305875534304137,
+      "grad_norm": 0.295244961977005,
+      "learning_rate": 4.6873563218390806e-05,
+      "loss": 0.1401,
+      "step": 544
+    },
+    {
+      "epoch": 0.6317467217271607,
+      "grad_norm": 0.26528140902519226,
+      "learning_rate": 4.686781609195403e-05,
+      "loss": 0.1342,
+      "step": 545
+    },
+    {
+      "epoch": 0.6329058900239078,
+      "grad_norm": 0.2771555781364441,
+      "learning_rate": 4.686206896551724e-05,
+      "loss": 0.1525,
+      "step": 546
+    },
+    {
+      "epoch": 0.634065058320655,
+      "grad_norm": 0.3281809091567993,
+      "learning_rate": 4.6856321839080465e-05,
+      "loss": 0.1653,
+      "step": 547
+    },
+    {
+      "epoch": 0.635224226617402,
+      "grad_norm": 0.3082864284515381,
+      "learning_rate": 4.685057471264368e-05,
+      "loss": 0.125,
+      "step": 548
+    },
+    {
+      "epoch": 0.6363833949141491,
+      "grad_norm": 0.30822286009788513,
+      "learning_rate": 4.6844827586206894e-05,
+      "loss": 0.1455,
+      "step": 549
+    },
+    {
+      "epoch": 0.6375425632108962,
+      "grad_norm": 0.3183049261569977,
+      "learning_rate": 4.6839080459770116e-05,
+      "loss": 0.1363,
+      "step": 550
+    },
+    {
+      "epoch": 0.6387017315076433,
+      "grad_norm": 0.2250736802816391,
+      "learning_rate": 4.683333333333334e-05,
+      "loss": 0.1147,
+      "step": 551
+    },
+    {
+      "epoch": 0.6398608998043903,
+      "grad_norm": 0.2534612715244293,
+      "learning_rate": 4.682758620689656e-05,
+      "loss": 0.1447,
+      "step": 552
+    },
+    {
+      "epoch": 0.6410200681011374,
+      "grad_norm": 0.2627923786640167,
+      "learning_rate": 4.6821839080459774e-05,
+      "loss": 0.1404,
+      "step": 553
+    },
+    {
+      "epoch": 0.6421792363978845,
+      "grad_norm": 0.25790974497795105,
+      "learning_rate": 4.681609195402299e-05,
+      "loss": 0.1267,
+      "step": 554
+    },
+    {
+      "epoch": 0.6433384046946315,
+      "grad_norm": 0.2535199224948883,
+      "learning_rate": 4.681034482758621e-05,
+      "loss": 0.1377,
+      "step": 555
+    },
+    {
+      "epoch": 0.6444975729913787,
+      "grad_norm": 0.27798426151275635,
+      "learning_rate": 4.6804597701149426e-05,
+      "loss": 0.1422,
+      "step": 556
+    },
+    {
+      "epoch": 0.6456567412881258,
+      "grad_norm": 0.26984500885009766,
+      "learning_rate": 4.679885057471264e-05,
+      "loss": 0.1383,
+      "step": 557
+    },
+    {
+      "epoch": 0.6468159095848729,
+      "grad_norm": 0.28631749749183655,
+      "learning_rate": 4.679310344827586e-05,
+      "loss": 0.1402,
+      "step": 558
+    },
+    {
+      "epoch": 0.6479750778816199,
+      "grad_norm": 0.29483139514923096,
+      "learning_rate": 4.6787356321839084e-05,
+      "loss": 0.1384,
+      "step": 559
+    },
+    {
+      "epoch": 0.649134246178367,
+      "grad_norm": 0.23459696769714355,
+      "learning_rate": 4.67816091954023e-05,
+      "loss": 0.1242,
+      "step": 560
+    },
+    {
+      "epoch": 0.6502934144751141,
+      "grad_norm": 0.28944167494773865,
+      "learning_rate": 4.677586206896552e-05,
+      "loss": 0.1505,
+      "step": 561
+    },
+    {
+      "epoch": 0.6514525827718612,
+      "grad_norm": 0.3058931529521942,
+      "learning_rate": 4.6770114942528736e-05,
+      "loss": 0.1369,
+      "step": 562
+    },
+    {
+      "epoch": 0.6526117510686082,
+      "grad_norm": 0.28087151050567627,
+      "learning_rate": 4.676436781609196e-05,
+      "loss": 0.1372,
+      "step": 563
+    },
+    {
+      "epoch": 0.6537709193653554,
+      "grad_norm": 0.3356817364692688,
+      "learning_rate": 4.675862068965517e-05,
+      "loss": 0.1603,
+      "step": 564
+    },
+    {
+      "epoch": 0.6549300876621025,
+      "grad_norm": 0.24305835366249084,
+      "learning_rate": 4.6752873563218394e-05,
+      "loss": 0.1336,
+      "step": 565
+    },
+    {
+      "epoch": 0.6560892559588495,
+      "grad_norm": 0.26663926243782043,
+      "learning_rate": 4.6747126436781616e-05,
+      "loss": 0.1326,
+      "step": 566
+    },
+    {
+      "epoch": 0.6572484242555966,
+      "grad_norm": 0.3176979124546051,
+      "learning_rate": 4.674137931034483e-05,
+      "loss": 0.1586,
+      "step": 567
+    },
+    {
+      "epoch": 0.6584075925523437,
+      "grad_norm": 0.2597542405128479,
+      "learning_rate": 4.6735632183908045e-05,
+      "loss": 0.1325,
+      "step": 568
+    },
+    {
+      "epoch": 0.6595667608490908,
+      "grad_norm": 0.2565508782863617,
+      "learning_rate": 4.672988505747127e-05,
+      "loss": 0.1403,
+      "step": 569
+    },
+    {
+      "epoch": 0.6607259291458378,
+      "grad_norm": 0.22864797711372375,
+      "learning_rate": 4.672413793103448e-05,
+      "loss": 0.13,
+      "step": 570
+    },
+    {
+      "epoch": 0.661885097442585,
+      "grad_norm": 0.27961140871047974,
+      "learning_rate": 4.6718390804597704e-05,
+      "loss": 0.145,
+      "step": 571
+    },
+    {
+      "epoch": 0.6630442657393321,
+      "grad_norm": 0.23423443734645844,
+      "learning_rate": 4.6712643678160925e-05,
+      "loss": 0.1325,
+      "step": 572
+    },
+    {
+      "epoch": 0.6642034340360791,
+      "grad_norm": 0.26428887248039246,
+      "learning_rate": 4.670689655172414e-05,
+      "loss": 0.1299,
+      "step": 573
+    },
+    {
+      "epoch": 0.6653626023328262,
+      "grad_norm": 0.2903366684913635,
+      "learning_rate": 4.670114942528736e-05,
+      "loss": 0.1445,
+      "step": 574
+    },
+    {
+      "epoch": 0.6665217706295733,
+      "grad_norm": 0.3106566369533539,
+      "learning_rate": 4.669540229885058e-05,
+      "loss": 0.1396,
+      "step": 575
+    },
+    {
+      "epoch": 0.6676809389263204,
+      "grad_norm": 0.2907349169254303,
+      "learning_rate": 4.668965517241379e-05,
+      "loss": 0.1224,
+      "step": 576
+    },
+    {
+      "epoch": 0.6688401072230674,
+      "grad_norm": 0.2840725779533386,
+      "learning_rate": 4.668390804597701e-05,
+      "loss": 0.1356,
+      "step": 577
+    },
+    {
+      "epoch": 0.6699992755198145,
+      "grad_norm": 0.3185088634490967,
+      "learning_rate": 4.667816091954023e-05,
+      "loss": 0.1499,
+      "step": 578
+    },
+    {
+      "epoch": 0.6711584438165616,
+      "grad_norm": 0.2769455909729004,
+      "learning_rate": 4.667241379310345e-05,
+      "loss": 0.1354,
+      "step": 579
+    },
+    {
+      "epoch": 0.6723176121133086,
+      "grad_norm": 0.3882596492767334,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 0.1551,
+      "step": 580
+    },
+    {
+      "epoch": 0.6734767804100558,
+      "grad_norm": 0.29669293761253357,
+      "learning_rate": 4.6660919540229887e-05,
+      "loss": 0.1437,
+      "step": 581
+    },
+    {
+      "epoch": 0.6746359487068029,
+      "grad_norm": 0.24709312617778778,
+      "learning_rate": 4.665517241379311e-05,
+      "loss": 0.1349,
+      "step": 582
+    },
+    {
+      "epoch": 0.67579511700355,
+      "grad_norm": 0.29217076301574707,
+      "learning_rate": 4.664942528735632e-05,
+      "loss": 0.1327,
+      "step": 583
+    },
+    {
+      "epoch": 0.676954285300297,
+      "grad_norm": 0.2767103612422943,
+      "learning_rate": 4.664367816091954e-05,
+      "loss": 0.1341,
+      "step": 584
+    },
+    {
+      "epoch": 0.6781134535970441,
+      "grad_norm": 0.3487448990345001,
+      "learning_rate": 4.663793103448276e-05,
+      "loss": 0.1567,
+      "step": 585
+    },
+    {
+      "epoch": 0.6792726218937912,
+      "grad_norm": 0.30940744280815125,
+      "learning_rate": 4.663218390804598e-05,
+      "loss": 0.1486,
+      "step": 586
+    },
+    {
+      "epoch": 0.6804317901905383,
+      "grad_norm": 0.2427985668182373,
+      "learning_rate": 4.6626436781609196e-05,
+      "loss": 0.1298,
+      "step": 587
+    },
+    {
+      "epoch": 0.6815909584872853,
+      "grad_norm": 0.2587006092071533,
+      "learning_rate": 4.662068965517242e-05,
+      "loss": 0.1557,
+      "step": 588
+    },
+    {
+      "epoch": 0.6827501267840325,
+      "grad_norm": 0.2949361205101013,
+      "learning_rate": 4.661494252873563e-05,
+      "loss": 0.1513,
+      "step": 589
+    },
+    {
+      "epoch": 0.6839092950807796,
+      "grad_norm": 0.2784793972969055,
+      "learning_rate": 4.6609195402298855e-05,
+      "loss": 0.1407,
+      "step": 590
+    },
+    {
+      "epoch": 0.6850684633775266,
+      "grad_norm": 0.28113847970962524,
+      "learning_rate": 4.660344827586207e-05,
+      "loss": 0.1346,
+      "step": 591
+    },
+    {
+      "epoch": 0.6862276316742737,
+      "grad_norm": 0.30229294300079346,
+      "learning_rate": 4.659770114942529e-05,
+      "loss": 0.1533,
+      "step": 592
+    },
+    {
+      "epoch": 0.6873867999710208,
+      "grad_norm": 0.30034735798835754,
+      "learning_rate": 4.659195402298851e-05,
+      "loss": 0.1382,
+      "step": 593
+    },
+    {
+      "epoch": 0.6885459682677679,
+      "grad_norm": 0.23939953744411469,
+      "learning_rate": 4.658620689655173e-05,
+      "loss": 0.1248,
+      "step": 594
+    },
+    {
+      "epoch": 0.6897051365645149,
+      "grad_norm": 0.2364337295293808,
+      "learning_rate": 4.658045977011494e-05,
+      "loss": 0.1335,
+      "step": 595
+    },
+    {
+      "epoch": 0.690864304861262,
+      "grad_norm": 0.2641420364379883,
+      "learning_rate": 4.6574712643678164e-05,
+      "loss": 0.1492,
+      "step": 596
+    },
+    {
+      "epoch": 0.6920234731580092,
+      "grad_norm": 0.29211926460266113,
+      "learning_rate": 4.656896551724138e-05,
+      "loss": 0.1397,
+      "step": 597
+    },
+    {
+      "epoch": 0.6931826414547562,
+      "grad_norm": 0.2939442992210388,
+      "learning_rate": 4.65632183908046e-05,
+      "loss": 0.1345,
+      "step": 598
+    },
+    {
+      "epoch": 0.6943418097515033,
+      "grad_norm": 0.2861159145832062,
+      "learning_rate": 4.6557471264367816e-05,
+      "loss": 0.1359,
+      "step": 599
+    },
+    {
+      "epoch": 0.6955009780482504,
+      "grad_norm": 0.3505733907222748,
+      "learning_rate": 4.655172413793104e-05,
+      "loss": 0.1464,
+      "step": 600
+    },
+    {
+      "epoch": 0.6966601463449975,
+      "grad_norm": 0.26291075348854065,
+      "learning_rate": 4.654597701149426e-05,
+      "loss": 0.1354,
+      "step": 601
+    },
+    {
+      "epoch": 0.6978193146417445,
+      "grad_norm": 0.27308157086372375,
+      "learning_rate": 4.6540229885057474e-05,
+      "loss": 0.1374,
+      "step": 602
+    },
+    {
+      "epoch": 0.6989784829384916,
+      "grad_norm": 0.3093547224998474,
+      "learning_rate": 4.653448275862069e-05,
+      "loss": 0.1366,
+      "step": 603
+    },
+    {
+      "epoch": 0.7001376512352387,
+      "grad_norm": 0.29862943291664124,
+      "learning_rate": 4.652873563218391e-05,
+      "loss": 0.1417,
+      "step": 604
+    },
+    {
+      "epoch": 0.7012968195319857,
+      "grad_norm": 0.35645461082458496,
+      "learning_rate": 4.6522988505747125e-05,
+      "loss": 0.1348,
+      "step": 605
+    },
+    {
+      "epoch": 0.7024559878287329,
+      "grad_norm": 0.3341381847858429,
+      "learning_rate": 4.651724137931035e-05,
+      "loss": 0.1426,
+      "step": 606
+    },
+    {
+      "epoch": 0.70361515612548,
+      "grad_norm": 0.22590871155261993,
+      "learning_rate": 4.651149425287357e-05,
+      "loss": 0.1192,
+      "step": 607
+    },
+    {
+      "epoch": 0.7047743244222271,
+      "grad_norm": 0.22133195400238037,
+      "learning_rate": 4.6505747126436784e-05,
+      "loss": 0.1176,
+      "step": 608
+    },
+    {
+      "epoch": 0.7059334927189741,
+      "grad_norm": 0.2593124806880951,
+      "learning_rate": 4.6500000000000005e-05,
+      "loss": 0.1356,
+      "step": 609
+    },
+    {
+      "epoch": 0.7070926610157212,
+      "grad_norm": 0.28317561745643616,
+      "learning_rate": 4.649425287356322e-05,
+      "loss": 0.144,
+      "step": 610
+    },
+    {
+      "epoch": 0.7082518293124683,
+      "grad_norm": 0.23904190957546234,
+      "learning_rate": 4.6488505747126435e-05,
+      "loss": 0.1371,
+      "step": 611
+    },
+    {
+      "epoch": 0.7094109976092153,
+      "grad_norm": 0.23972614109516144,
+      "learning_rate": 4.648275862068966e-05,
+      "loss": 0.1268,
+      "step": 612
+    },
+    {
+      "epoch": 0.7105701659059624,
+      "grad_norm": 0.27289271354675293,
+      "learning_rate": 4.647701149425288e-05,
+      "loss": 0.1257,
+      "step": 613
+    },
+    {
+      "epoch": 0.7117293342027096,
+      "grad_norm": 0.2559848427772522,
+      "learning_rate": 4.6471264367816093e-05,
+      "loss": 0.139,
+      "step": 614
+    },
+    {
+      "epoch": 0.7128885024994567,
+      "grad_norm": 0.28095269203186035,
+      "learning_rate": 4.6465517241379315e-05,
+      "loss": 0.1508,
+      "step": 615
+    },
+    {
+      "epoch": 0.7140476707962037,
+      "grad_norm": 0.25798696279525757,
+      "learning_rate": 4.645977011494253e-05,
+      "loss": 0.1302,
+      "step": 616
+    },
+    {
+      "epoch": 0.7152068390929508,
+      "grad_norm": 0.28479090332984924,
+      "learning_rate": 4.645402298850575e-05,
+      "loss": 0.1462,
+      "step": 617
+    },
+    {
+      "epoch": 0.7163660073896979,
+      "grad_norm": 0.3032209575176239,
+      "learning_rate": 4.644827586206897e-05,
+      "loss": 0.1581,
+      "step": 618
+    },
+    {
+      "epoch": 0.717525175686445,
+      "grad_norm": 0.26038414239883423,
+      "learning_rate": 4.644252873563218e-05,
+      "loss": 0.133,
+      "step": 619
+    },
+    {
+      "epoch": 0.718684343983192,
+      "grad_norm": 0.2712029218673706,
+      "learning_rate": 4.643678160919541e-05,
+      "loss": 0.1456,
+      "step": 620
+    },
+    {
+      "epoch": 0.7198435122799391,
+      "grad_norm": 0.2711297571659088,
+      "learning_rate": 4.6431034482758625e-05,
+      "loss": 0.1485,
+      "step": 621
+    },
+    {
+      "epoch": 0.7210026805766863,
+      "grad_norm": 0.26714852452278137,
+      "learning_rate": 4.642528735632184e-05,
+      "loss": 0.1409,
+      "step": 622
+    },
+    {
+      "epoch": 0.7221618488734333,
+      "grad_norm": 0.2638694941997528,
+      "learning_rate": 4.641954022988506e-05,
+      "loss": 0.1415,
+      "step": 623
+    },
+    {
+      "epoch": 0.7233210171701804,
+      "grad_norm": 0.24792924523353577,
+      "learning_rate": 4.6413793103448276e-05,
+      "loss": 0.144,
+      "step": 624
+    },
+    {
+      "epoch": 0.7244801854669275,
+      "grad_norm": 0.33576005697250366,
+      "learning_rate": 4.640804597701149e-05,
+      "loss": 0.1677,
+      "step": 625
+    },
+    {
+      "epoch": 0.7256393537636746,
+      "grad_norm": 0.26434826850891113,
+      "learning_rate": 4.640229885057471e-05,
+      "loss": 0.1413,
+      "step": 626
+    },
+    {
+      "epoch": 0.7267985220604216,
+      "grad_norm": 0.22275975346565247,
+      "learning_rate": 4.6396551724137935e-05,
+      "loss": 0.1265,
+      "step": 627
+    },
+    {
+      "epoch": 0.7279576903571687,
+      "grad_norm": 0.25366446375846863,
+      "learning_rate": 4.6390804597701156e-05,
+      "loss": 0.1353,
+      "step": 628
+    },
+    {
+      "epoch": 0.7291168586539158,
+      "grad_norm": 0.24983662366867065,
+      "learning_rate": 4.638505747126437e-05,
+      "loss": 0.1338,
+      "step": 629
+    },
+    {
+      "epoch": 0.7302760269506628,
+      "grad_norm": 0.24769724905490875,
+      "learning_rate": 4.6379310344827586e-05,
+      "loss": 0.1242,
+      "step": 630
+    },
+    {
+      "epoch": 0.73143519524741,
+      "grad_norm": 0.2604047358036041,
+      "learning_rate": 4.637356321839081e-05,
+      "loss": 0.126,
+      "step": 631
+    },
+    {
+      "epoch": 0.7325943635441571,
+      "grad_norm": 0.2878481447696686,
+      "learning_rate": 4.636781609195402e-05,
+      "loss": 0.1573,
+      "step": 632
+    },
+    {
+      "epoch": 0.7337535318409042,
+      "grad_norm": 0.25279513001441956,
+      "learning_rate": 4.6362068965517244e-05,
+      "loss": 0.1361,
+      "step": 633
+    },
+    {
+      "epoch": 0.7349127001376512,
+      "grad_norm": 0.28761938214302063,
+      "learning_rate": 4.6356321839080466e-05,
+      "loss": 0.1437,
+      "step": 634
+    },
+    {
+      "epoch": 0.7360718684343983,
+      "grad_norm": 0.30495256185531616,
+      "learning_rate": 4.635057471264368e-05,
+      "loss": 0.1549,
+      "step": 635
+    },
+    {
+      "epoch": 0.7372310367311454,
+      "grad_norm": 0.2510074973106384,
+      "learning_rate": 4.63448275862069e-05,
+      "loss": 0.1286,
+      "step": 636
+    },
+    {
+      "epoch": 0.7383902050278924,
+      "grad_norm": 0.2822047472000122,
+      "learning_rate": 4.633908045977012e-05,
+      "loss": 0.1279,
+      "step": 637
+    },
+    {
+      "epoch": 0.7395493733246395,
+      "grad_norm": 0.23230963945388794,
+      "learning_rate": 4.633333333333333e-05,
+      "loss": 0.1406,
+      "step": 638
+    },
+    {
+      "epoch": 0.7407085416213867,
+      "grad_norm": 0.2472311556339264,
+      "learning_rate": 4.6327586206896554e-05,
+      "loss": 0.1244,
+      "step": 639
+    },
+    {
+      "epoch": 0.7418677099181338,
+      "grad_norm": 0.2912672460079193,
+      "learning_rate": 4.632183908045977e-05,
+      "loss": 0.136,
+      "step": 640
+    },
+    {
+      "epoch": 0.7430268782148808,
+      "grad_norm": 0.35469523072242737,
+      "learning_rate": 4.631609195402299e-05,
+      "loss": 0.1484,
+      "step": 641
+    },
+    {
+      "epoch": 0.7441860465116279,
+      "grad_norm": 0.26528200507164,
+      "learning_rate": 4.631034482758621e-05,
+      "loss": 0.1252,
+      "step": 642
+    },
+    {
+      "epoch": 0.745345214808375,
+      "grad_norm": 0.2616005539894104,
+      "learning_rate": 4.630459770114943e-05,
+      "loss": 0.1222,
+      "step": 643
+    },
+    {
+      "epoch": 0.7465043831051221,
+      "grad_norm": 0.26047012209892273,
+      "learning_rate": 4.629885057471264e-05,
+      "loss": 0.1401,
+      "step": 644
+    },
+    {
+      "epoch": 0.7476635514018691,
+      "grad_norm": 0.27605798840522766,
+      "learning_rate": 4.6293103448275864e-05,
+      "loss": 0.13,
+      "step": 645
+    },
+    {
+      "epoch": 0.7488227196986162,
+      "grad_norm": 0.30614152550697327,
+      "learning_rate": 4.628735632183908e-05,
+      "loss": 0.1486,
+      "step": 646
+    },
+    {
+      "epoch": 0.7499818879953634,
+      "grad_norm": 0.34578898549079895,
+      "learning_rate": 4.62816091954023e-05,
+      "loss": 0.1515,
+      "step": 647
+    },
+    {
+      "epoch": 0.7511410562921104,
+      "grad_norm": 0.24491111934185028,
+      "learning_rate": 4.627586206896552e-05,
+      "loss": 0.1234,
+      "step": 648
+    },
+    {
+      "epoch": 0.7523002245888575,
+      "grad_norm": 0.2561955153942108,
+      "learning_rate": 4.627011494252874e-05,
+      "loss": 0.1255,
+      "step": 649
+    },
+    {
+      "epoch": 0.7534593928856046,
+      "grad_norm": 0.24703799188137054,
+      "learning_rate": 4.626436781609196e-05,
+      "loss": 0.139,
+      "step": 650
+    },
+    {
+      "epoch": 0.7546185611823517,
+      "grad_norm": 0.26947158575057983,
+      "learning_rate": 4.6258620689655174e-05,
+      "loss": 0.1435,
+      "step": 651
+    },
+    {
+      "epoch": 0.7557777294790987,
+      "grad_norm": 0.2430969476699829,
+      "learning_rate": 4.625287356321839e-05,
+      "loss": 0.1316,
+      "step": 652
+    },
+    {
+      "epoch": 0.7569368977758458,
+      "grad_norm": 0.40103209018707275,
+      "learning_rate": 4.624712643678161e-05,
+      "loss": 0.1605,
+      "step": 653
+    },
+    {
+      "epoch": 0.758096066072593,
+      "grad_norm": 0.25342556834220886,
+      "learning_rate": 4.624137931034483e-05,
+      "loss": 0.1357,
+      "step": 654
+    },
+    {
+      "epoch": 0.75925523436934,
+      "grad_norm": 0.27793052792549133,
+      "learning_rate": 4.6235632183908054e-05,
+      "loss": 0.1392,
+      "step": 655
+    },
+    {
+      "epoch": 0.7604144026660871,
+      "grad_norm": 0.27625927329063416,
+      "learning_rate": 4.622988505747127e-05,
+      "loss": 0.1312,
+      "step": 656
+    },
+    {
+      "epoch": 0.7615735709628342,
+      "grad_norm": 0.2726586163043976,
+      "learning_rate": 4.622413793103448e-05,
+      "loss": 0.1439,
+      "step": 657
+    },
+    {
+      "epoch": 0.7627327392595813,
+      "grad_norm": 0.25028151273727417,
+      "learning_rate": 4.6218390804597705e-05,
+      "loss": 0.14,
+      "step": 658
+    },
+    {
+      "epoch": 0.7638919075563283,
+      "grad_norm": 0.2604377865791321,
+      "learning_rate": 4.621264367816092e-05,
+      "loss": 0.1318,
+      "step": 659
+    },
+    {
+      "epoch": 0.7650510758530754,
+      "grad_norm": 0.35959815979003906,
+      "learning_rate": 4.6206896551724135e-05,
+      "loss": 0.149,
+      "step": 660
+    },
+    {
+      "epoch": 0.7662102441498225,
+      "grad_norm": 0.21311073005199432,
+      "learning_rate": 4.620114942528736e-05,
+      "loss": 0.1123,
+      "step": 661
+    },
+    {
+      "epoch": 0.7673694124465695,
+      "grad_norm": 0.24613121151924133,
+      "learning_rate": 4.619540229885058e-05,
+      "loss": 0.1293,
+      "step": 662
+    },
+    {
+      "epoch": 0.7685285807433166,
+      "grad_norm": 0.27359437942504883,
+      "learning_rate": 4.618965517241379e-05,
+      "loss": 0.1359,
+      "step": 663
+    },
+    {
+      "epoch": 0.7696877490400638,
+      "grad_norm": 0.24835547804832458,
+      "learning_rate": 4.6183908045977015e-05,
+      "loss": 0.1335,
+      "step": 664
+    },
+    {
+      "epoch": 0.7708469173368109,
+      "grad_norm": 0.3183259665966034,
+      "learning_rate": 4.617816091954023e-05,
+      "loss": 0.1455,
+      "step": 665
+    },
+    {
+      "epoch": 0.7720060856335579,
+      "grad_norm": 0.23825454711914062,
+      "learning_rate": 4.617241379310345e-05,
+      "loss": 0.1327,
+      "step": 666
+    },
+    {
+      "epoch": 0.773165253930305,
+      "grad_norm": 0.24269577860832214,
+      "learning_rate": 4.6166666666666666e-05,
+      "loss": 0.1321,
+      "step": 667
+    },
+    {
+      "epoch": 0.7743244222270521,
+      "grad_norm": 0.26813822984695435,
+      "learning_rate": 4.616091954022989e-05,
+      "loss": 0.1385,
+      "step": 668
+    },
+    {
+      "epoch": 0.7754835905237992,
+      "grad_norm": 0.2735162377357483,
+      "learning_rate": 4.615517241379311e-05,
+      "loss": 0.1382,
+      "step": 669
+    },
+    {
+      "epoch": 0.7766427588205462,
+      "grad_norm": 0.26851484179496765,
+      "learning_rate": 4.6149425287356324e-05,
+      "loss": 0.1323,
+      "step": 670
+    },
+    {
+      "epoch": 0.7778019271172933,
+      "grad_norm": 0.3302837312221527,
+      "learning_rate": 4.614367816091954e-05,
+      "loss": 0.1644,
+      "step": 671
+    },
+    {
+      "epoch": 0.7789610954140405,
+      "grad_norm": 0.28512895107269287,
+      "learning_rate": 4.613793103448276e-05,
+      "loss": 0.1541,
+      "step": 672
+    },
+    {
+      "epoch": 0.7801202637107875,
+      "grad_norm": 0.26242977380752563,
+      "learning_rate": 4.6132183908045976e-05,
+      "loss": 0.1327,
+      "step": 673
+    },
+    {
+      "epoch": 0.7812794320075346,
+      "grad_norm": 0.286178320646286,
+      "learning_rate": 4.61264367816092e-05,
+      "loss": 0.1301,
+      "step": 674
+    },
+    {
+      "epoch": 0.7824386003042817,
+      "grad_norm": 0.22549080848693848,
+      "learning_rate": 4.612068965517242e-05,
+      "loss": 0.1237,
+      "step": 675
+    },
+    {
+      "epoch": 0.7835977686010288,
+      "grad_norm": 0.24352504312992096,
+      "learning_rate": 4.6114942528735634e-05,
+      "loss": 0.1298,
+      "step": 676
+    },
+    {
+      "epoch": 0.7847569368977758,
+      "grad_norm": 0.23466962575912476,
+      "learning_rate": 4.6109195402298856e-05,
+      "loss": 0.133,
+      "step": 677
+    },
+    {
+      "epoch": 0.7859161051945229,
+      "grad_norm": 0.2620813250541687,
+      "learning_rate": 4.610344827586207e-05,
+      "loss": 0.1311,
+      "step": 678
+    },
+    {
+      "epoch": 0.78707527349127,
+      "grad_norm": 0.2720955014228821,
+      "learning_rate": 4.6097701149425286e-05,
+      "loss": 0.1207,
+      "step": 679
+    },
+    {
+      "epoch": 0.788234441788017,
+      "grad_norm": 0.2539190649986267,
+      "learning_rate": 4.609195402298851e-05,
+      "loss": 0.1287,
+      "step": 680
+    },
+    {
+      "epoch": 0.7893936100847642,
+      "grad_norm": 0.2616521716117859,
+      "learning_rate": 4.608620689655173e-05,
+      "loss": 0.144,
+      "step": 681
+    },
+    {
+      "epoch": 0.7905527783815113,
+      "grad_norm": 0.2984738349914551,
+      "learning_rate": 4.6080459770114944e-05,
+      "loss": 0.1408,
+      "step": 682
+    },
+    {
+      "epoch": 0.7917119466782584,
+      "grad_norm": 0.25363418459892273,
+      "learning_rate": 4.6074712643678166e-05,
+      "loss": 0.121,
+      "step": 683
+    },
+    {
+      "epoch": 0.7928711149750054,
+      "grad_norm": 0.22894874215126038,
+      "learning_rate": 4.606896551724138e-05,
+      "loss": 0.1137,
+      "step": 684
+    },
+    {
+      "epoch": 0.7940302832717525,
+      "grad_norm": 0.25778335332870483,
+      "learning_rate": 4.60632183908046e-05,
+      "loss": 0.1338,
+      "step": 685
+    },
+    {
+      "epoch": 0.7951894515684996,
+      "grad_norm": 0.2576935887336731,
+      "learning_rate": 4.605747126436782e-05,
+      "loss": 0.1416,
+      "step": 686
+    },
+    {
+      "epoch": 0.7963486198652466,
+      "grad_norm": 0.3121855556964874,
+      "learning_rate": 4.605172413793103e-05,
+      "loss": 0.1284,
+      "step": 687
+    },
+    {
+      "epoch": 0.7975077881619937,
+      "grad_norm": 0.24522997438907623,
+      "learning_rate": 4.6045977011494254e-05,
+      "loss": 0.1367,
+      "step": 688
+    },
+    {
+      "epoch": 0.7986669564587409,
+      "grad_norm": 0.3264451324939728,
+      "learning_rate": 4.6040229885057475e-05,
+      "loss": 0.1574,
+      "step": 689
+    },
+    {
+      "epoch": 0.799826124755488,
+      "grad_norm": 0.29765743017196655,
+      "learning_rate": 4.603448275862069e-05,
+      "loss": 0.1336,
+      "step": 690
+    },
+    {
+      "epoch": 0.800985293052235,
+      "grad_norm": 0.302561491727829,
+      "learning_rate": 4.602873563218391e-05,
+      "loss": 0.1449,
+      "step": 691
+    },
+    {
+      "epoch": 0.8021444613489821,
+      "grad_norm": 0.2951429486274719,
+      "learning_rate": 4.602298850574713e-05,
+      "loss": 0.1373,
+      "step": 692
+    },
+    {
+      "epoch": 0.8033036296457292,
+      "grad_norm": 0.26344770193099976,
+      "learning_rate": 4.601724137931035e-05,
+      "loss": 0.1355,
+      "step": 693
+    },
+    {
+      "epoch": 0.8044627979424763,
+      "grad_norm": 0.23573601245880127,
+      "learning_rate": 4.6011494252873563e-05,
+      "loss": 0.1264,
+      "step": 694
+    },
+    {
+      "epoch": 0.8056219662392233,
+      "grad_norm": 0.2914588153362274,
+      "learning_rate": 4.6005747126436785e-05,
+      "loss": 0.1496,
+      "step": 695
+    },
+    {
+      "epoch": 0.8067811345359704,
+      "grad_norm": 0.23518706858158112,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.125,
+      "step": 696
+    },
+    {
+      "epoch": 0.8079403028327176,
+      "grad_norm": 0.2899051606655121,
+      "learning_rate": 4.599425287356322e-05,
+      "loss": 0.1379,
+      "step": 697
+    },
+    {
+      "epoch": 0.8090994711294646,
+      "grad_norm": 0.25654926896095276,
+      "learning_rate": 4.598850574712644e-05,
+      "loss": 0.1469,
+      "step": 698
+    },
+    {
+      "epoch": 0.8102586394262117,
+      "grad_norm": 0.36468809843063354,
+      "learning_rate": 4.598275862068966e-05,
+      "loss": 0.1424,
+      "step": 699
+    },
+    {
+      "epoch": 0.8114178077229588,
+      "grad_norm": 0.2551107108592987,
+      "learning_rate": 4.597701149425287e-05,
+      "loss": 0.1441,
+      "step": 700
+    },
+    {
+      "epoch": 0.8125769760197059,
+      "grad_norm": 0.21090874075889587,
+      "learning_rate": 4.597126436781609e-05,
+      "loss": 0.1253,
+      "step": 701
+    },
+    {
+      "epoch": 0.8137361443164529,
+      "grad_norm": 0.2588905096054077,
+      "learning_rate": 4.5965517241379317e-05,
+      "loss": 0.1327,
+      "step": 702
+    },
+    {
+      "epoch": 0.8148953126132,
+      "grad_norm": 0.24421721696853638,
+      "learning_rate": 4.595977011494253e-05,
+      "loss": 0.1294,
+      "step": 703
+    },
+    {
+      "epoch": 0.8160544809099471,
+      "grad_norm": 0.25245368480682373,
+      "learning_rate": 4.595402298850575e-05,
+      "loss": 0.1247,
+      "step": 704
+    },
+    {
+      "epoch": 0.8172136492066941,
+      "grad_norm": 0.3936619758605957,
+      "learning_rate": 4.594827586206897e-05,
+      "loss": 0.1321,
+      "step": 705
+    },
+    {
+      "epoch": 0.8183728175034413,
+      "grad_norm": 0.25811803340911865,
+      "learning_rate": 4.594252873563218e-05,
+      "loss": 0.1337,
+      "step": 706
+    },
+    {
+      "epoch": 0.8195319858001884,
+      "grad_norm": 0.23244990408420563,
+      "learning_rate": 4.5936781609195405e-05,
+      "loss": 0.1219,
+      "step": 707
+    },
+    {
+      "epoch": 0.8206911540969355,
+      "grad_norm": 0.2513156235218048,
+      "learning_rate": 4.593103448275862e-05,
+      "loss": 0.1422,
+      "step": 708
+    },
+    {
+      "epoch": 0.8218503223936825,
+      "grad_norm": 0.26262450218200684,
+      "learning_rate": 4.592528735632184e-05,
+      "loss": 0.1419,
+      "step": 709
+    },
+    {
+      "epoch": 0.8230094906904296,
+      "grad_norm": 0.268996924161911,
+      "learning_rate": 4.591954022988506e-05,
+      "loss": 0.1427,
+      "step": 710
+    },
+    {
+      "epoch": 0.8241686589871767,
+      "grad_norm": 0.2628903090953827,
+      "learning_rate": 4.591379310344828e-05,
+      "loss": 0.1406,
+      "step": 711
+    },
+    {
+      "epoch": 0.8253278272839237,
+      "grad_norm": 0.24836388230323792,
+      "learning_rate": 4.59080459770115e-05,
+      "loss": 0.1344,
+      "step": 712
+    },
+    {
+      "epoch": 0.8264869955806708,
+      "grad_norm": 0.29568374156951904,
+      "learning_rate": 4.5902298850574714e-05,
+      "loss": 0.1353,
+      "step": 713
+    },
+    {
+      "epoch": 0.827646163877418,
+      "grad_norm": 0.30942583084106445,
+      "learning_rate": 4.589655172413793e-05,
+      "loss": 0.1381,
+      "step": 714
+    },
+    {
+      "epoch": 0.8288053321741651,
+      "grad_norm": 0.3070472478866577,
+      "learning_rate": 4.589080459770115e-05,
+      "loss": 0.142,
+      "step": 715
+    },
+    {
+      "epoch": 0.8299645004709121,
+      "grad_norm": 0.27771466970443726,
+      "learning_rate": 4.588505747126437e-05,
+      "loss": 0.1397,
+      "step": 716
+    },
+    {
+      "epoch": 0.8311236687676592,
+      "grad_norm": 0.2712878882884979,
+      "learning_rate": 4.587931034482759e-05,
+      "loss": 0.1252,
+      "step": 717
+    },
+    {
+      "epoch": 0.8322828370644063,
+      "grad_norm": 0.24696029722690582,
+      "learning_rate": 4.587356321839081e-05,
+      "loss": 0.1327,
+      "step": 718
+    },
+    {
+      "epoch": 0.8334420053611534,
+      "grad_norm": 0.26035380363464355,
+      "learning_rate": 4.5867816091954024e-05,
+      "loss": 0.1276,
+      "step": 719
+    },
+    {
+      "epoch": 0.8346011736579004,
+      "grad_norm": 0.23014302551746368,
+      "learning_rate": 4.586206896551724e-05,
+      "loss": 0.1294,
+      "step": 720
+    },
+    {
+      "epoch": 0.8357603419546475,
+      "grad_norm": 0.22767721116542816,
+      "learning_rate": 4.585632183908046e-05,
+      "loss": 0.1204,
+      "step": 721
+    },
+    {
+      "epoch": 0.8369195102513947,
+      "grad_norm": 0.24614818394184113,
+      "learning_rate": 4.585057471264368e-05,
+      "loss": 0.1412,
+      "step": 722
+    },
+    {
+      "epoch": 0.8380786785481417,
+      "grad_norm": 0.2815050184726715,
+      "learning_rate": 4.5844827586206904e-05,
+      "loss": 0.142,
+      "step": 723
+    },
+    {
+      "epoch": 0.8392378468448888,
+      "grad_norm": 0.24825121462345123,
+      "learning_rate": 4.583908045977012e-05,
+      "loss": 0.1477,
+      "step": 724
+    },
+    {
+      "epoch": 0.8403970151416359,
+      "grad_norm": 0.2300599366426468,
+      "learning_rate": 4.5833333333333334e-05,
+      "loss": 0.1413,
+      "step": 725
+    },
+    {
+      "epoch": 0.841556183438383,
+      "grad_norm": 0.27150389552116394,
+      "learning_rate": 4.5827586206896556e-05,
+      "loss": 0.1553,
+      "step": 726
+    },
+    {
+      "epoch": 0.84271535173513,
+      "grad_norm": 0.33325132727622986,
+      "learning_rate": 4.582183908045977e-05,
+      "loss": 0.1699,
+      "step": 727
+    },
+    {
+      "epoch": 0.8438745200318771,
+      "grad_norm": 0.21544939279556274,
+      "learning_rate": 4.5816091954022985e-05,
+      "loss": 0.1279,
+      "step": 728
+    },
+    {
+      "epoch": 0.8450336883286242,
+      "grad_norm": 0.23038536310195923,
+      "learning_rate": 4.581034482758621e-05,
+      "loss": 0.1227,
+      "step": 729
+    },
+    {
+      "epoch": 0.8461928566253712,
+      "grad_norm": 0.25293296575546265,
+      "learning_rate": 4.580459770114943e-05,
+      "loss": 0.1497,
+      "step": 730
+    },
+    {
+      "epoch": 0.8473520249221184,
+      "grad_norm": 0.2448996752500534,
+      "learning_rate": 4.579885057471265e-05,
+      "loss": 0.1475,
+      "step": 731
+    },
+    {
+      "epoch": 0.8485111932188655,
+      "grad_norm": 0.2752504050731659,
+      "learning_rate": 4.5793103448275865e-05,
+      "loss": 0.1453,
+      "step": 732
+    },
+    {
+      "epoch": 0.8496703615156126,
+      "grad_norm": 0.2612292170524597,
+      "learning_rate": 4.578735632183908e-05,
+      "loss": 0.1306,
+      "step": 733
+    },
+    {
+      "epoch": 0.8508295298123596,
+      "grad_norm": 0.2557094395160675,
+      "learning_rate": 4.57816091954023e-05,
+      "loss": 0.1288,
+      "step": 734
+    },
+    {
+      "epoch": 0.8519886981091067,
+      "grad_norm": 0.24499356746673584,
+      "learning_rate": 4.577586206896552e-05,
+      "loss": 0.1391,
+      "step": 735
+    },
+    {
+      "epoch": 0.8531478664058538,
+      "grad_norm": 0.23510660231113434,
+      "learning_rate": 4.577011494252874e-05,
+      "loss": 0.1277,
+      "step": 736
+    },
+    {
+      "epoch": 0.8543070347026008,
+      "grad_norm": 0.25060582160949707,
+      "learning_rate": 4.576436781609196e-05,
+      "loss": 0.1301,
+      "step": 737
+    },
+    {
+      "epoch": 0.855466202999348,
+      "grad_norm": 0.2850019931793213,
+      "learning_rate": 4.5758620689655175e-05,
+      "loss": 0.1406,
+      "step": 738
+    },
+    {
+      "epoch": 0.8566253712960951,
+      "grad_norm": 0.30304649472236633,
+      "learning_rate": 4.575287356321839e-05,
+      "loss": 0.1455,
+      "step": 739
+    },
+    {
+      "epoch": 0.8577845395928422,
+      "grad_norm": 0.27735939621925354,
+      "learning_rate": 4.574712643678161e-05,
+      "loss": 0.133,
+      "step": 740
+    },
+    {
+      "epoch": 0.8589437078895892,
+      "grad_norm": 0.258037805557251,
+      "learning_rate": 4.5741379310344826e-05,
+      "loss": 0.1359,
+      "step": 741
+    },
+    {
+      "epoch": 0.8601028761863363,
+      "grad_norm": 0.2623947858810425,
+      "learning_rate": 4.573563218390805e-05,
+      "loss": 0.141,
+      "step": 742
+    },
+    {
+      "epoch": 0.8612620444830834,
+      "grad_norm": 0.2677944004535675,
+      "learning_rate": 4.572988505747127e-05,
+      "loss": 0.1398,
+      "step": 743
+    },
+    {
+      "epoch": 0.8624212127798305,
+      "grad_norm": 0.28370678424835205,
+      "learning_rate": 4.5724137931034485e-05,
+      "loss": 0.1551,
+      "step": 744
+    },
+    {
+      "epoch": 0.8635803810765775,
+      "grad_norm": 0.259971022605896,
+      "learning_rate": 4.5718390804597706e-05,
+      "loss": 0.1488,
+      "step": 745
+    },
+    {
+      "epoch": 0.8647395493733246,
+      "grad_norm": 0.22784096002578735,
+      "learning_rate": 4.571264367816092e-05,
+      "loss": 0.123,
+      "step": 746
+    },
+    {
+      "epoch": 0.8658987176700718,
+      "grad_norm": 0.25658029317855835,
+      "learning_rate": 4.5706896551724136e-05,
+      "loss": 0.1364,
+      "step": 747
+    },
+    {
+      "epoch": 0.8670578859668188,
+      "grad_norm": 0.24773739278316498,
+      "learning_rate": 4.570114942528736e-05,
+      "loss": 0.128,
+      "step": 748
+    },
+    {
+      "epoch": 0.8682170542635659,
+      "grad_norm": 0.2921466827392578,
+      "learning_rate": 4.569540229885057e-05,
+      "loss": 0.1432,
+      "step": 749
+    },
+    {
+      "epoch": 0.869376222560313,
+      "grad_norm": 0.26510924100875854,
+      "learning_rate": 4.5689655172413794e-05,
+      "loss": 0.1334,
+      "step": 750
+    },
+    {
+      "epoch": 0.8705353908570601,
+      "grad_norm": 0.2811342179775238,
+      "learning_rate": 4.5683908045977016e-05,
+      "loss": 0.1446,
+      "step": 751
+    },
+    {
+      "epoch": 0.8716945591538071,
+      "grad_norm": 0.24528606235980988,
+      "learning_rate": 4.567816091954023e-05,
+      "loss": 0.1254,
+      "step": 752
+    },
+    {
+      "epoch": 0.8728537274505542,
+      "grad_norm": 0.31388193368911743,
+      "learning_rate": 4.567241379310345e-05,
+      "loss": 0.148,
+      "step": 753
+    },
+    {
+      "epoch": 0.8740128957473013,
+      "grad_norm": 0.26188236474990845,
+      "learning_rate": 4.566666666666667e-05,
+      "loss": 0.1347,
+      "step": 754
+    },
+    {
+      "epoch": 0.8751720640440483,
+      "grad_norm": 0.2681477665901184,
+      "learning_rate": 4.566091954022988e-05,
+      "loss": 0.1251,
+      "step": 755
+    },
+    {
+      "epoch": 0.8763312323407955,
+      "grad_norm": 0.24694494903087616,
+      "learning_rate": 4.5655172413793104e-05,
+      "loss": 0.1341,
+      "step": 756
+    },
+    {
+      "epoch": 0.8774904006375426,
+      "grad_norm": 0.27526146173477173,
+      "learning_rate": 4.5649425287356326e-05,
+      "loss": 0.1425,
+      "step": 757
+    },
+    {
+      "epoch": 0.8786495689342897,
+      "grad_norm": 0.28800222277641296,
+      "learning_rate": 4.564367816091955e-05,
+      "loss": 0.1396,
+      "step": 758
+    },
+    {
+      "epoch": 0.8798087372310367,
+      "grad_norm": 0.29053691029548645,
+      "learning_rate": 4.563793103448276e-05,
+      "loss": 0.1427,
+      "step": 759
+    },
+    {
+      "epoch": 0.8809679055277838,
+      "grad_norm": 0.29743799567222595,
+      "learning_rate": 4.563218390804598e-05,
+      "loss": 0.1513,
+      "step": 760
+    },
+    {
+      "epoch": 0.8821270738245309,
+      "grad_norm": 0.27560529112815857,
+      "learning_rate": 4.56264367816092e-05,
+      "loss": 0.1356,
+      "step": 761
+    },
+    {
+      "epoch": 0.8832862421212779,
+      "grad_norm": 0.23586055636405945,
+      "learning_rate": 4.5620689655172414e-05,
+      "loss": 0.132,
+      "step": 762
+    },
+    {
+      "epoch": 0.884445410418025,
+      "grad_norm": 0.26681259274482727,
+      "learning_rate": 4.5614942528735636e-05,
+      "loss": 0.1584,
+      "step": 763
+    },
+    {
+      "epoch": 0.8856045787147722,
+      "grad_norm": 0.3042534589767456,
+      "learning_rate": 4.560919540229886e-05,
+      "loss": 0.149,
+      "step": 764
+    },
+    {
+      "epoch": 0.8867637470115193,
+      "grad_norm": 0.24351336061954498,
+      "learning_rate": 4.560344827586207e-05,
+      "loss": 0.1311,
+      "step": 765
+    },
+    {
+      "epoch": 0.8879229153082663,
+      "grad_norm": 0.2620246708393097,
+      "learning_rate": 4.559770114942529e-05,
+      "loss": 0.1448,
+      "step": 766
+    },
+    {
+      "epoch": 0.8890820836050134,
+      "grad_norm": 0.2437165081501007,
+      "learning_rate": 4.559195402298851e-05,
+      "loss": 0.1435,
+      "step": 767
+    },
+    {
+      "epoch": 0.8902412519017605,
+      "grad_norm": 0.231397345662117,
+      "learning_rate": 4.5586206896551724e-05,
+      "loss": 0.1378,
+      "step": 768
+    },
+    {
+      "epoch": 0.8914004201985076,
+      "grad_norm": 0.22732099890708923,
+      "learning_rate": 4.5580459770114945e-05,
+      "loss": 0.1399,
+      "step": 769
+    },
+    {
+      "epoch": 0.8925595884952546,
+      "grad_norm": 0.2291109710931778,
+      "learning_rate": 4.557471264367816e-05,
+      "loss": 0.1207,
+      "step": 770
+    },
+    {
+      "epoch": 0.8937187567920017,
+      "grad_norm": 0.23525553941726685,
+      "learning_rate": 4.556896551724138e-05,
+      "loss": 0.1269,
+      "step": 771
+    },
+    {
+      "epoch": 0.8948779250887489,
+      "grad_norm": 0.2587391436100006,
+      "learning_rate": 4.5563218390804604e-05,
+      "loss": 0.1208,
+      "step": 772
+    },
+    {
+      "epoch": 0.8960370933854959,
+      "grad_norm": 0.24662849307060242,
+      "learning_rate": 4.555747126436782e-05,
+      "loss": 0.1413,
+      "step": 773
+    },
+    {
+      "epoch": 0.897196261682243,
+      "grad_norm": 0.2599044144153595,
+      "learning_rate": 4.5551724137931033e-05,
+      "loss": 0.1322,
+      "step": 774
+    },
+    {
+      "epoch": 0.8983554299789901,
+      "grad_norm": 0.2424292415380478,
+      "learning_rate": 4.5545977011494255e-05,
+      "loss": 0.1359,
+      "step": 775
+    },
+    {
+      "epoch": 0.8995145982757372,
+      "grad_norm": 0.25603169202804565,
+      "learning_rate": 4.554022988505747e-05,
+      "loss": 0.1324,
+      "step": 776
+    },
+    {
+      "epoch": 0.9006737665724842,
+      "grad_norm": 0.2570304870605469,
+      "learning_rate": 4.553448275862069e-05,
+      "loss": 0.122,
+      "step": 777
+    },
+    {
+      "epoch": 0.9018329348692313,
+      "grad_norm": 0.2565818130970001,
+      "learning_rate": 4.552873563218391e-05,
+      "loss": 0.1246,
+      "step": 778
+    },
+    {
+      "epoch": 0.9029921031659784,
+      "grad_norm": 0.2947520613670349,
+      "learning_rate": 4.552298850574713e-05,
+      "loss": 0.1376,
+      "step": 779
+    },
+    {
+      "epoch": 0.9041512714627254,
+      "grad_norm": 0.31323108077049255,
+      "learning_rate": 4.551724137931035e-05,
+      "loss": 0.148,
+      "step": 780
+    },
+    {
+      "epoch": 0.9053104397594726,
+      "grad_norm": 0.3410814702510834,
+      "learning_rate": 4.5511494252873565e-05,
+      "loss": 0.1493,
+      "step": 781
+    },
+    {
+      "epoch": 0.9064696080562197,
+      "grad_norm": 0.2659667134284973,
+      "learning_rate": 4.550574712643678e-05,
+      "loss": 0.1303,
+      "step": 782
+    },
+    {
+      "epoch": 0.9076287763529668,
+      "grad_norm": 0.2651742696762085,
+      "learning_rate": 4.55e-05,
+      "loss": 0.1327,
+      "step": 783
+    },
+    {
+      "epoch": 0.9087879446497138,
+      "grad_norm": 0.24942578375339508,
+      "learning_rate": 4.549425287356322e-05,
+      "loss": 0.1342,
+      "step": 784
+    },
+    {
+      "epoch": 0.9099471129464609,
+      "grad_norm": 0.2682301104068756,
+      "learning_rate": 4.548850574712644e-05,
+      "loss": 0.1449,
+      "step": 785
+    },
+    {
+      "epoch": 0.911106281243208,
+      "grad_norm": 0.2612859606742859,
+      "learning_rate": 4.548275862068966e-05,
+      "loss": 0.1371,
+      "step": 786
+    },
+    {
+      "epoch": 0.912265449539955,
+      "grad_norm": 0.30771300196647644,
+      "learning_rate": 4.5477011494252875e-05,
+      "loss": 0.1446,
+      "step": 787
+    },
+    {
+      "epoch": 0.9134246178367021,
+      "grad_norm": 0.33343127369880676,
+      "learning_rate": 4.5471264367816096e-05,
+      "loss": 0.1375,
+      "step": 788
+    },
+    {
+      "epoch": 0.9145837861334493,
+      "grad_norm": 0.2678926885128021,
+      "learning_rate": 4.546551724137931e-05,
+      "loss": 0.1347,
+      "step": 789
+    },
+    {
+      "epoch": 0.9157429544301964,
+      "grad_norm": 0.2823614478111267,
+      "learning_rate": 4.5459770114942526e-05,
+      "loss": 0.1461,
+      "step": 790
+    },
+    {
+      "epoch": 0.9169021227269434,
+      "grad_norm": 0.27924710512161255,
+      "learning_rate": 4.545402298850575e-05,
+      "loss": 0.1343,
+      "step": 791
+    },
+    {
+      "epoch": 0.9180612910236905,
+      "grad_norm": 0.27781474590301514,
+      "learning_rate": 4.544827586206897e-05,
+      "loss": 0.1433,
+      "step": 792
+    },
+    {
+      "epoch": 0.9192204593204376,
+      "grad_norm": 0.23757725954055786,
+      "learning_rate": 4.5442528735632184e-05,
+      "loss": 0.1214,
+      "step": 793
+    },
+    {
+      "epoch": 0.9203796276171846,
+      "grad_norm": 0.305899053812027,
+      "learning_rate": 4.5436781609195406e-05,
+      "loss": 0.136,
+      "step": 794
+    },
+    {
+      "epoch": 0.9215387959139317,
+      "grad_norm": 0.29127955436706543,
+      "learning_rate": 4.543103448275862e-05,
+      "loss": 0.1551,
+      "step": 795
+    },
+    {
+      "epoch": 0.9226979642106788,
+      "grad_norm": 0.29604771733283997,
+      "learning_rate": 4.542528735632184e-05,
+      "loss": 0.1378,
+      "step": 796
+    },
+    {
+      "epoch": 0.923857132507426,
+      "grad_norm": 0.23887404799461365,
+      "learning_rate": 4.541954022988506e-05,
+      "loss": 0.1268,
+      "step": 797
+    },
+    {
+      "epoch": 0.925016300804173,
+      "grad_norm": 0.2829357981681824,
+      "learning_rate": 4.541379310344828e-05,
+      "loss": 0.1432,
+      "step": 798
+    },
+    {
+      "epoch": 0.9261754691009201,
+      "grad_norm": 0.2251252681016922,
+      "learning_rate": 4.54080459770115e-05,
+      "loss": 0.1298,
+      "step": 799
+    },
+    {
+      "epoch": 0.9273346373976672,
+      "grad_norm": 0.26304125785827637,
+      "learning_rate": 4.5402298850574716e-05,
+      "loss": 0.1376,
+      "step": 800
+    },
+    {
+      "epoch": 0.9284938056944143,
+      "grad_norm": 0.22513070702552795,
+      "learning_rate": 4.539655172413793e-05,
+      "loss": 0.127,
+      "step": 801
+    },
+    {
+      "epoch": 0.9296529739911613,
+      "grad_norm": 0.25344815850257874,
+      "learning_rate": 4.539080459770115e-05,
+      "loss": 0.1395,
+      "step": 802
+    },
+    {
+      "epoch": 0.9308121422879084,
+      "grad_norm": 0.23576590418815613,
+      "learning_rate": 4.538505747126437e-05,
+      "loss": 0.1269,
+      "step": 803
+    },
+    {
+      "epoch": 0.9319713105846555,
+      "grad_norm": 0.24553732573986053,
+      "learning_rate": 4.537931034482759e-05,
+      "loss": 0.1408,
+      "step": 804
+    },
+    {
+      "epoch": 0.9331304788814025,
+      "grad_norm": 0.24802154302597046,
+      "learning_rate": 4.537356321839081e-05,
+      "loss": 0.1307,
+      "step": 805
+    },
+    {
+      "epoch": 0.9342896471781497,
+      "grad_norm": 0.255938321352005,
+      "learning_rate": 4.5367816091954025e-05,
+      "loss": 0.1446,
+      "step": 806
+    },
+    {
+      "epoch": 0.9354488154748968,
+      "grad_norm": 0.23458512127399445,
+      "learning_rate": 4.536206896551725e-05,
+      "loss": 0.1371,
+      "step": 807
+    },
+    {
+      "epoch": 0.9366079837716439,
+      "grad_norm": 0.2571949064731598,
+      "learning_rate": 4.535632183908046e-05,
+      "loss": 0.1319,
+      "step": 808
+    },
+    {
+      "epoch": 0.9377671520683909,
+      "grad_norm": 0.2464578002691269,
+      "learning_rate": 4.535057471264368e-05,
+      "loss": 0.1425,
+      "step": 809
+    },
+    {
+      "epoch": 0.938926320365138,
+      "grad_norm": 0.23102332651615143,
+      "learning_rate": 4.53448275862069e-05,
+      "loss": 0.1351,
+      "step": 810
+    },
+    {
+      "epoch": 0.9400854886618851,
+      "grad_norm": 0.27257415652275085,
+      "learning_rate": 4.5339080459770114e-05,
+      "loss": 0.1359,
+      "step": 811
+    },
+    {
+      "epoch": 0.9412446569586321,
+      "grad_norm": 0.24599824845790863,
+      "learning_rate": 4.5333333333333335e-05,
+      "loss": 0.1356,
+      "step": 812
+    },
+    {
+      "epoch": 0.9424038252553792,
+      "grad_norm": 0.23127569258213043,
+      "learning_rate": 4.532758620689656e-05,
+      "loss": 0.1412,
+      "step": 813
+    },
+    {
+      "epoch": 0.9435629935521264,
+      "grad_norm": 0.2365388125181198,
+      "learning_rate": 4.532183908045977e-05,
+      "loss": 0.1118,
+      "step": 814
+    },
+    {
+      "epoch": 0.9447221618488735,
+      "grad_norm": 0.2809307873249054,
+      "learning_rate": 4.5316091954022993e-05,
+      "loss": 0.1363,
+      "step": 815
+    },
+    {
+      "epoch": 0.9458813301456205,
+      "grad_norm": 0.258364737033844,
+      "learning_rate": 4.531034482758621e-05,
+      "loss": 0.1378,
+      "step": 816
+    },
+    {
+      "epoch": 0.9470404984423676,
+      "grad_norm": 0.27401670813560486,
+      "learning_rate": 4.530459770114942e-05,
+      "loss": 0.1465,
+      "step": 817
+    },
+    {
+      "epoch": 0.9481996667391147,
+      "grad_norm": 0.22026842832565308,
+      "learning_rate": 4.5298850574712645e-05,
+      "loss": 0.1283,
+      "step": 818
+    },
+    {
+      "epoch": 0.9493588350358617,
+      "grad_norm": 0.2709653675556183,
+      "learning_rate": 4.529310344827587e-05,
+      "loss": 0.1284,
+      "step": 819
+    },
+    {
+      "epoch": 0.9505180033326088,
+      "grad_norm": 0.2417244166135788,
+      "learning_rate": 4.528735632183908e-05,
+      "loss": 0.1363,
+      "step": 820
+    },
+    {
+      "epoch": 0.951677171629356,
+      "grad_norm": 0.2279566526412964,
+      "learning_rate": 4.52816091954023e-05,
+      "loss": 0.1285,
+      "step": 821
+    },
+    {
+      "epoch": 0.9528363399261031,
+      "grad_norm": 0.32052162289619446,
+      "learning_rate": 4.527586206896552e-05,
+      "loss": 0.1433,
+      "step": 822
+    },
+    {
+      "epoch": 0.9539955082228501,
+      "grad_norm": 0.3256911635398865,
+      "learning_rate": 4.527011494252873e-05,
+      "loss": 0.145,
+      "step": 823
+    },
+    {
+      "epoch": 0.9551546765195972,
+      "grad_norm": 0.30633530020713806,
+      "learning_rate": 4.5264367816091955e-05,
+      "loss": 0.1458,
+      "step": 824
+    },
+    {
+      "epoch": 0.9563138448163443,
+      "grad_norm": 0.24086791276931763,
+      "learning_rate": 4.5258620689655176e-05,
+      "loss": 0.1213,
+      "step": 825
+    },
+    {
+      "epoch": 0.9574730131130914,
+      "grad_norm": 0.3224896788597107,
+      "learning_rate": 4.52528735632184e-05,
+      "loss": 0.1543,
+      "step": 826
+    },
+    {
+      "epoch": 0.9586321814098384,
+      "grad_norm": 0.24555163085460663,
+      "learning_rate": 4.524712643678161e-05,
+      "loss": 0.1295,
+      "step": 827
+    },
+    {
+      "epoch": 0.9597913497065855,
+      "grad_norm": 0.2528969645500183,
+      "learning_rate": 4.524137931034483e-05,
+      "loss": 0.1389,
+      "step": 828
+    },
+    {
+      "epoch": 0.9609505180033326,
+      "grad_norm": 0.2498130351305008,
+      "learning_rate": 4.523563218390805e-05,
+      "loss": 0.1529,
+      "step": 829
+    },
+    {
+      "epoch": 0.9621096863000796,
+      "grad_norm": 0.26962774991989136,
+      "learning_rate": 4.5229885057471264e-05,
+      "loss": 0.1346,
+      "step": 830
+    },
+    {
+      "epoch": 0.9632688545968268,
+      "grad_norm": 0.2615809142589569,
+      "learning_rate": 4.522413793103448e-05,
+      "loss": 0.1549,
+      "step": 831
+    },
+    {
+      "epoch": 0.9644280228935739,
+      "grad_norm": 0.25109943747520447,
+      "learning_rate": 4.521839080459771e-05,
+      "loss": 0.1402,
+      "step": 832
+    },
+    {
+      "epoch": 0.965587191190321,
+      "grad_norm": 0.2414674311876297,
+      "learning_rate": 4.521264367816092e-05,
+      "loss": 0.1329,
+      "step": 833
+    },
+    {
+      "epoch": 0.966746359487068,
+      "grad_norm": 0.2999972701072693,
+      "learning_rate": 4.5206896551724144e-05,
+      "loss": 0.1524,
+      "step": 834
+    },
+    {
+      "epoch": 0.9679055277838151,
+      "grad_norm": 0.22573649883270264,
+      "learning_rate": 4.520114942528736e-05,
+      "loss": 0.1248,
+      "step": 835
+    },
+    {
+      "epoch": 0.9690646960805622,
+      "grad_norm": 0.2691646218299866,
+      "learning_rate": 4.5195402298850574e-05,
+      "loss": 0.1259,
+      "step": 836
+    },
+    {
+      "epoch": 0.9702238643773092,
+      "grad_norm": 0.2566002607345581,
+      "learning_rate": 4.5189655172413796e-05,
+      "loss": 0.1338,
+      "step": 837
+    },
+    {
+      "epoch": 0.9713830326740563,
+      "grad_norm": 0.30500656366348267,
+      "learning_rate": 4.518390804597701e-05,
+      "loss": 0.1393,
+      "step": 838
+    },
+    {
+      "epoch": 0.9725422009708035,
+      "grad_norm": 0.2786789834499359,
+      "learning_rate": 4.517816091954023e-05,
+      "loss": 0.1418,
+      "step": 839
+    },
+    {
+      "epoch": 0.9737013692675506,
+      "grad_norm": 0.2511579096317291,
+      "learning_rate": 4.5172413793103454e-05,
+      "loss": 0.1261,
+      "step": 840
+    },
+    {
+      "epoch": 0.9748605375642976,
+      "grad_norm": 0.2628697454929352,
+      "learning_rate": 4.516666666666667e-05,
+      "loss": 0.1255,
+      "step": 841
+    },
+    {
+      "epoch": 0.9760197058610447,
+      "grad_norm": 0.2691211700439453,
+      "learning_rate": 4.5160919540229884e-05,
+      "loss": 0.1469,
+      "step": 842
+    },
+    {
+      "epoch": 0.9771788741577918,
+      "grad_norm": 0.4240913689136505,
+      "learning_rate": 4.5155172413793106e-05,
+      "loss": 0.1345,
+      "step": 843
+    },
+    {
+      "epoch": 0.9783380424545388,
+      "grad_norm": 0.26439380645751953,
+      "learning_rate": 4.514942528735632e-05,
+      "loss": 0.1389,
+      "step": 844
+    },
+    {
+      "epoch": 0.9794972107512859,
+      "grad_norm": 0.2187066376209259,
+      "learning_rate": 4.514367816091954e-05,
+      "loss": 0.1191,
+      "step": 845
+    },
+    {
+      "epoch": 0.980656379048033,
+      "grad_norm": 0.2630019187927246,
+      "learning_rate": 4.5137931034482764e-05,
+      "loss": 0.1255,
+      "step": 846
+    },
+    {
+      "epoch": 0.9818155473447802,
+      "grad_norm": 0.2619337737560272,
+      "learning_rate": 4.513218390804598e-05,
+      "loss": 0.146,
+      "step": 847
+    },
+    {
+      "epoch": 0.9829747156415272,
+      "grad_norm": 0.24650199711322784,
+      "learning_rate": 4.51264367816092e-05,
+      "loss": 0.1246,
+      "step": 848
+    },
+    {
+      "epoch": 0.9841338839382743,
+      "grad_norm": 0.33000877499580383,
+      "learning_rate": 4.5120689655172415e-05,
+      "loss": 0.1448,
+      "step": 849
+    },
+    {
+      "epoch": 0.9852930522350214,
+      "grad_norm": 0.39497750997543335,
+      "learning_rate": 4.511494252873563e-05,
+      "loss": 0.1332,
+      "step": 850
+    },
+    {
+      "epoch": 0.9864522205317685,
+      "grad_norm": 0.27251842617988586,
+      "learning_rate": 4.510919540229885e-05,
+      "loss": 0.127,
+      "step": 851
+    },
+    {
+      "epoch": 0.9876113888285155,
+      "grad_norm": 0.4220561981201172,
+      "learning_rate": 4.510344827586207e-05,
+      "loss": 0.1353,
+      "step": 852
+    },
+    {
+      "epoch": 0.9887705571252626,
+      "grad_norm": 0.24371755123138428,
+      "learning_rate": 4.5097701149425295e-05,
+      "loss": 0.1436,
+      "step": 853
+    },
+    {
+      "epoch": 0.9899297254220097,
+      "grad_norm": 0.2443200647830963,
+      "learning_rate": 4.509195402298851e-05,
+      "loss": 0.1405,
+      "step": 854
+    },
+    {
+      "epoch": 0.9910888937187567,
+      "grad_norm": 0.22042927145957947,
+      "learning_rate": 4.5086206896551725e-05,
+      "loss": 0.1294,
+      "step": 855
+    },
+    {
+      "epoch": 0.9922480620155039,
+      "grad_norm": 0.26495447754859924,
+      "learning_rate": 4.508045977011495e-05,
+      "loss": 0.1442,
+      "step": 856
+    },
+    {
+      "epoch": 0.993407230312251,
+      "grad_norm": 0.28576844930648804,
+      "learning_rate": 4.507471264367816e-05,
+      "loss": 0.1422,
+      "step": 857
+    },
+    {
+      "epoch": 0.9945663986089981,
+      "grad_norm": 0.24880261719226837,
+      "learning_rate": 4.5068965517241377e-05,
+      "loss": 0.1404,
+      "step": 858
+    },
+    {
+      "epoch": 0.9957255669057451,
+      "grad_norm": 0.2431262731552124,
+      "learning_rate": 4.50632183908046e-05,
+      "loss": 0.1376,
+      "step": 859
+    },
+    {
+      "epoch": 0.9968847352024922,
+      "grad_norm": 0.2703196704387665,
+      "learning_rate": 4.505747126436782e-05,
+      "loss": 0.1407,
+      "step": 860
+    },
+    {
+      "epoch": 0.9980439034992393,
+      "grad_norm": 0.3191419839859009,
+      "learning_rate": 4.5051724137931035e-05,
+      "loss": 0.1558,
+      "step": 861
+    },
+    {
+      "epoch": 0.9992030717959863,
+      "grad_norm": 0.25417354702949524,
+      "learning_rate": 4.5045977011494257e-05,
+      "loss": 0.1386,
+      "step": 862
+    },
+    {
+      "epoch": 0.9992030717959863,
+      "eval_loss": 0.13573063910007477,
+      "eval_runtime": 265.6894,
+      "eval_samples_per_second": 5.774,
+      "eval_steps_per_second": 5.774,
+      "step": 862
+    },
+    {
+      "epoch": 1.0003622400927334,
+      "grad_norm": 0.21144159138202667,
+      "learning_rate": 4.504022988505747e-05,
+      "loss": 0.119,
+      "step": 863
+    },
+    {
+      "epoch": 1.0015214083894806,
+      "grad_norm": 0.27048224210739136,
+      "learning_rate": 4.503448275862069e-05,
+      "loss": 0.1282,
+      "step": 864
+    },
+    {
+      "epoch": 1.0026805766862277,
+      "grad_norm": 0.2059887945652008,
+      "learning_rate": 4.502873563218391e-05,
+      "loss": 0.1389,
+      "step": 865
+    },
+    {
+      "epoch": 1.0038397449829748,
+      "grad_norm": 0.25673580169677734,
+      "learning_rate": 4.502298850574713e-05,
+      "loss": 0.134,
+      "step": 866
+    },
+    {
+      "epoch": 1.004998913279722,
+      "grad_norm": 0.2381593883037567,
+      "learning_rate": 4.501724137931035e-05,
+      "loss": 0.1271,
+      "step": 867
+    },
+    {
+      "epoch": 1.0061580815764688,
+      "grad_norm": 0.2609866261482239,
+      "learning_rate": 4.5011494252873566e-05,
+      "loss": 0.1349,
+      "step": 868
+    },
+    {
+      "epoch": 1.007317249873216,
+      "grad_norm": 0.24942117929458618,
+      "learning_rate": 4.500574712643678e-05,
+      "loss": 0.1251,
+      "step": 869
+    },
+    {
+      "epoch": 1.008476418169963,
+      "grad_norm": 0.2552180886268616,
+      "learning_rate": 4.5e-05,
+      "loss": 0.1333,
+      "step": 870
+    },
+    {
+      "epoch": 1.0096355864667101,
+      "grad_norm": 0.2939591407775879,
+      "learning_rate": 4.499425287356322e-05,
+      "loss": 0.1433,
+      "step": 871
+    },
+    {
+      "epoch": 1.0107947547634573,
+      "grad_norm": 0.2345464825630188,
+      "learning_rate": 4.498850574712644e-05,
+      "loss": 0.1273,
+      "step": 872
+    },
+    {
+      "epoch": 1.0119539230602044,
+      "grad_norm": 0.3161166310310364,
+      "learning_rate": 4.498275862068966e-05,
+      "loss": 0.1301,
+      "step": 873
+    },
+    {
+      "epoch": 1.0131130913569515,
+      "grad_norm": 0.2409006655216217,
+      "learning_rate": 4.4977011494252876e-05,
+      "loss": 0.1145,
+      "step": 874
+    },
+    {
+      "epoch": 1.0142722596536984,
+      "grad_norm": 0.26523053646087646,
+      "learning_rate": 4.49712643678161e-05,
+      "loss": 0.1386,
+      "step": 875
+    },
+    {
+      "epoch": 1.0154314279504455,
+      "grad_norm": 0.2150416225194931,
+      "learning_rate": 4.496551724137931e-05,
+      "loss": 0.1285,
+      "step": 876
+    },
+    {
+      "epoch": 1.0165905962471926,
+      "grad_norm": 0.23828035593032837,
+      "learning_rate": 4.495977011494253e-05,
+      "loss": 0.1161,
+      "step": 877
+    },
+    {
+      "epoch": 1.0177497645439397,
+      "grad_norm": 0.319975882768631,
+      "learning_rate": 4.495402298850575e-05,
+      "loss": 0.1444,
+      "step": 878
+    },
+    {
+      "epoch": 1.0189089328406868,
+      "grad_norm": 0.22034992277622223,
+      "learning_rate": 4.4948275862068964e-05,
+      "loss": 0.1195,
+      "step": 879
+    },
+    {
+      "epoch": 1.020068101137434,
+      "grad_norm": 0.2223641574382782,
+      "learning_rate": 4.4942528735632186e-05,
+      "loss": 0.1181,
+      "step": 880
+    },
+    {
+      "epoch": 1.021227269434181,
+      "grad_norm": 0.21697430312633514,
+      "learning_rate": 4.493678160919541e-05,
+      "loss": 0.1131,
+      "step": 881
+    },
+    {
+      "epoch": 1.022386437730928,
+      "grad_norm": 0.24257703125476837,
+      "learning_rate": 4.493103448275862e-05,
+      "loss": 0.123,
+      "step": 882
+    },
+    {
+      "epoch": 1.023545606027675,
+      "grad_norm": 0.2622239291667938,
+      "learning_rate": 4.4925287356321844e-05,
+      "loss": 0.1325,
+      "step": 883
+    },
+    {
+      "epoch": 1.0247047743244222,
+      "grad_norm": 0.2764127254486084,
+      "learning_rate": 4.491954022988506e-05,
+      "loss": 0.1228,
+      "step": 884
+    },
+    {
+      "epoch": 1.0258639426211693,
+      "grad_norm": 0.25700658559799194,
+      "learning_rate": 4.4913793103448274e-05,
+      "loss": 0.1212,
+      "step": 885
+    },
+    {
+      "epoch": 1.0270231109179164,
+      "grad_norm": 0.27795660495758057,
+      "learning_rate": 4.4908045977011495e-05,
+      "loss": 0.1328,
+      "step": 886
+    },
+    {
+      "epoch": 1.0281822792146635,
+      "grad_norm": 0.28292378783226013,
+      "learning_rate": 4.490229885057472e-05,
+      "loss": 0.1197,
+      "step": 887
+    },
+    {
+      "epoch": 1.0293414475114107,
+      "grad_norm": 0.27904993295669556,
+      "learning_rate": 4.489655172413793e-05,
+      "loss": 0.1208,
+      "step": 888
+    },
+    {
+      "epoch": 1.0305006158081575,
+      "grad_norm": 0.23822936415672302,
+      "learning_rate": 4.4890804597701154e-05,
+      "loss": 0.1187,
+      "step": 889
+    },
+    {
+      "epoch": 1.0316597841049047,
+      "grad_norm": 0.23615434765815735,
+      "learning_rate": 4.488505747126437e-05,
+      "loss": 0.119,
+      "step": 890
+    },
+    {
+      "epoch": 1.0328189524016518,
+      "grad_norm": 0.2193615883588791,
+      "learning_rate": 4.487931034482759e-05,
+      "loss": 0.1251,
+      "step": 891
+    },
+    {
+      "epoch": 1.033978120698399,
+      "grad_norm": 0.2385019063949585,
+      "learning_rate": 4.4873563218390805e-05,
+      "loss": 0.1261,
+      "step": 892
+    },
+    {
+      "epoch": 1.035137288995146,
+      "grad_norm": 0.262351393699646,
+      "learning_rate": 4.486781609195403e-05,
+      "loss": 0.1285,
+      "step": 893
+    },
+    {
+      "epoch": 1.0362964572918931,
+      "grad_norm": 0.2790246307849884,
+      "learning_rate": 4.486206896551725e-05,
+      "loss": 0.13,
+      "step": 894
+    },
+    {
+      "epoch": 1.0374556255886402,
+      "grad_norm": 0.28902125358581543,
+      "learning_rate": 4.4856321839080463e-05,
+      "loss": 0.1266,
+      "step": 895
+    },
+    {
+      "epoch": 1.0386147938853871,
+      "grad_norm": 0.2775196433067322,
+      "learning_rate": 4.485057471264368e-05,
+      "loss": 0.1453,
+      "step": 896
+    },
+    {
+      "epoch": 1.0397739621821342,
+      "grad_norm": 0.3169780671596527,
+      "learning_rate": 4.48448275862069e-05,
+      "loss": 0.1372,
+      "step": 897
+    },
+    {
+      "epoch": 1.0409331304788814,
+      "grad_norm": 0.24293041229248047,
+      "learning_rate": 4.4839080459770115e-05,
+      "loss": 0.1267,
+      "step": 898
+    },
+    {
+      "epoch": 1.0420922987756285,
+      "grad_norm": 0.3456263244152069,
+      "learning_rate": 4.483333333333333e-05,
+      "loss": 0.1566,
+      "step": 899
+    },
+    {
+      "epoch": 1.0432514670723756,
+      "grad_norm": 0.2542770504951477,
+      "learning_rate": 4.482758620689655e-05,
+      "loss": 0.1292,
+      "step": 900
+    },
+    {
+      "epoch": 1.0444106353691227,
+      "grad_norm": 0.308881551027298,
+      "learning_rate": 4.482183908045977e-05,
+      "loss": 0.1492,
+      "step": 901
+    },
+    {
+      "epoch": 1.0455698036658698,
+      "grad_norm": 0.26184189319610596,
+      "learning_rate": 4.4816091954022995e-05,
+      "loss": 0.1225,
+      "step": 902
+    },
+    {
+      "epoch": 1.0467289719626167,
+      "grad_norm": 0.22493098676204681,
+      "learning_rate": 4.481034482758621e-05,
+      "loss": 0.1239,
+      "step": 903
+    },
+    {
+      "epoch": 1.0478881402593638,
+      "grad_norm": 0.23027804493904114,
+      "learning_rate": 4.4804597701149425e-05,
+      "loss": 0.1253,
+      "step": 904
+    },
+    {
+      "epoch": 1.049047308556111,
+      "grad_norm": 0.2515313923358917,
+      "learning_rate": 4.4798850574712646e-05,
+      "loss": 0.1268,
+      "step": 905
+    },
+    {
+      "epoch": 1.050206476852858,
+      "grad_norm": 0.30108925700187683,
+      "learning_rate": 4.479310344827586e-05,
+      "loss": 0.1306,
+      "step": 906
+    },
+    {
+      "epoch": 1.0513656451496052,
+      "grad_norm": 0.25616785883903503,
+      "learning_rate": 4.478735632183908e-05,
+      "loss": 0.1292,
+      "step": 907
+    },
+    {
+      "epoch": 1.0525248134463523,
+      "grad_norm": 0.27651333808898926,
+      "learning_rate": 4.4781609195402305e-05,
+      "loss": 0.1293,
+      "step": 908
+    },
+    {
+      "epoch": 1.0536839817430994,
+      "grad_norm": 0.26399362087249756,
+      "learning_rate": 4.477586206896552e-05,
+      "loss": 0.1379,
+      "step": 909
+    },
+    {
+      "epoch": 1.0548431500398463,
+      "grad_norm": 0.27588364481925964,
+      "learning_rate": 4.477011494252874e-05,
+      "loss": 0.1369,
+      "step": 910
+    },
+    {
+      "epoch": 1.0560023183365934,
+      "grad_norm": 0.22311675548553467,
+      "learning_rate": 4.4764367816091956e-05,
+      "loss": 0.119,
+      "step": 911
+    },
+    {
+      "epoch": 1.0571614866333405,
+      "grad_norm": 0.25451788306236267,
+      "learning_rate": 4.475862068965517e-05,
+      "loss": 0.1204,
+      "step": 912
+    },
+    {
+      "epoch": 1.0583206549300876,
+      "grad_norm": 0.25941815972328186,
+      "learning_rate": 4.475287356321839e-05,
+      "loss": 0.1291,
+      "step": 913
+    },
+    {
+      "epoch": 1.0594798232268348,
+      "grad_norm": 0.2863259017467499,
+      "learning_rate": 4.4747126436781614e-05,
+      "loss": 0.1269,
+      "step": 914
+    },
+    {
+      "epoch": 1.0606389915235819,
+      "grad_norm": 0.26459577679634094,
+      "learning_rate": 4.474137931034483e-05,
+      "loss": 0.1291,
+      "step": 915
+    },
+    {
+      "epoch": 1.061798159820329,
+      "grad_norm": 0.2957836985588074,
+      "learning_rate": 4.473563218390805e-05,
+      "loss": 0.1313,
+      "step": 916
+    },
+    {
+      "epoch": 1.062957328117076,
+      "grad_norm": 0.33022555708885193,
+      "learning_rate": 4.4729885057471266e-05,
+      "loss": 0.1444,
+      "step": 917
+    },
+    {
+      "epoch": 1.064116496413823,
+      "grad_norm": 0.26044270396232605,
+      "learning_rate": 4.472413793103448e-05,
+      "loss": 0.1297,
+      "step": 918
+    },
+    {
+      "epoch": 1.0652756647105701,
+      "grad_norm": 0.25342807173728943,
+      "learning_rate": 4.47183908045977e-05,
+      "loss": 0.1134,
+      "step": 919
+    },
+    {
+      "epoch": 1.0664348330073172,
+      "grad_norm": 0.27540135383605957,
+      "learning_rate": 4.471264367816092e-05,
+      "loss": 0.1218,
+      "step": 920
+    },
+    {
+      "epoch": 1.0675940013040643,
+      "grad_norm": 0.24183742702007294,
+      "learning_rate": 4.470689655172414e-05,
+      "loss": 0.1227,
+      "step": 921
+    },
+    {
+      "epoch": 1.0687531696008115,
+      "grad_norm": 0.493832528591156,
+      "learning_rate": 4.470114942528736e-05,
+      "loss": 0.132,
+      "step": 922
+    },
+    {
+      "epoch": 1.0699123378975586,
+      "grad_norm": 0.31524115800857544,
+      "learning_rate": 4.4695402298850576e-05,
+      "loss": 0.135,
+      "step": 923
+    },
+    {
+      "epoch": 1.0710715061943057,
+      "grad_norm": 0.272127240896225,
+      "learning_rate": 4.46896551724138e-05,
+      "loss": 0.13,
+      "step": 924
+    },
+    {
+      "epoch": 1.0722306744910526,
+      "grad_norm": 0.24013520777225494,
+      "learning_rate": 4.468390804597701e-05,
+      "loss": 0.1231,
+      "step": 925
+    },
+    {
+      "epoch": 1.0733898427877997,
+      "grad_norm": 0.44068050384521484,
+      "learning_rate": 4.467816091954023e-05,
+      "loss": 0.1219,
+      "step": 926
+    },
+    {
+      "epoch": 1.0745490110845468,
+      "grad_norm": 0.2556682825088501,
+      "learning_rate": 4.467241379310345e-05,
+      "loss": 0.132,
+      "step": 927
+    },
+    {
+      "epoch": 1.075708179381294,
+      "grad_norm": 0.22146081924438477,
+      "learning_rate": 4.466666666666667e-05,
+      "loss": 0.1174,
+      "step": 928
+    },
+    {
+      "epoch": 1.076867347678041,
+      "grad_norm": 0.2589811086654663,
+      "learning_rate": 4.466091954022989e-05,
+      "loss": 0.1235,
+      "step": 929
+    },
+    {
+      "epoch": 1.0780265159747882,
+      "grad_norm": 0.26327693462371826,
+      "learning_rate": 4.465517241379311e-05,
+      "loss": 0.1303,
+      "step": 930
+    },
+    {
+      "epoch": 1.0791856842715353,
+      "grad_norm": 0.25532066822052,
+      "learning_rate": 4.464942528735632e-05,
+      "loss": 0.1442,
+      "step": 931
+    },
+    {
+      "epoch": 1.0803448525682822,
+      "grad_norm": 0.31189385056495667,
+      "learning_rate": 4.4643678160919544e-05,
+      "loss": 0.1448,
+      "step": 932
+    },
+    {
+      "epoch": 1.0815040208650293,
+      "grad_norm": 0.26370447874069214,
+      "learning_rate": 4.463793103448276e-05,
+      "loss": 0.1257,
+      "step": 933
+    },
+    {
+      "epoch": 1.0826631891617764,
+      "grad_norm": 0.2603430151939392,
+      "learning_rate": 4.463218390804598e-05,
+      "loss": 0.1248,
+      "step": 934
+    },
+    {
+      "epoch": 1.0838223574585235,
+      "grad_norm": 0.2369643896818161,
+      "learning_rate": 4.46264367816092e-05,
+      "loss": 0.1211,
+      "step": 935
+    },
+    {
+      "epoch": 1.0849815257552706,
+      "grad_norm": 0.2757248878479004,
+      "learning_rate": 4.462068965517242e-05,
+      "loss": 0.1319,
+      "step": 936
+    },
+    {
+      "epoch": 1.0861406940520177,
+      "grad_norm": 0.249775692820549,
+      "learning_rate": 4.461494252873564e-05,
+      "loss": 0.1349,
+      "step": 937
+    },
+    {
+      "epoch": 1.0872998623487649,
+      "grad_norm": 0.25326481461524963,
+      "learning_rate": 4.460919540229885e-05,
+      "loss": 0.1313,
+      "step": 938
+    },
+    {
+      "epoch": 1.0884590306455117,
+      "grad_norm": 0.20008519291877747,
+      "learning_rate": 4.460344827586207e-05,
+      "loss": 0.1142,
+      "step": 939
+    },
+    {
+      "epoch": 1.0896181989422589,
+      "grad_norm": 0.29447588324546814,
+      "learning_rate": 4.459770114942529e-05,
+      "loss": 0.1376,
+      "step": 940
+    },
+    {
+      "epoch": 1.090777367239006,
+      "grad_norm": 0.22719112038612366,
+      "learning_rate": 4.4591954022988505e-05,
+      "loss": 0.1122,
+      "step": 941
+    },
+    {
+      "epoch": 1.091936535535753,
+      "grad_norm": 0.2523985207080841,
+      "learning_rate": 4.4586206896551726e-05,
+      "loss": 0.1289,
+      "step": 942
+    },
+    {
+      "epoch": 1.0930957038325002,
+      "grad_norm": 0.2906104028224945,
+      "learning_rate": 4.458045977011495e-05,
+      "loss": 0.1285,
+      "step": 943
+    },
+    {
+      "epoch": 1.0942548721292473,
+      "grad_norm": 0.344566285610199,
+      "learning_rate": 4.457471264367816e-05,
+      "loss": 0.1237,
+      "step": 944
+    },
+    {
+      "epoch": 1.0954140404259944,
+      "grad_norm": 0.2776789963245392,
+      "learning_rate": 4.456896551724138e-05,
+      "loss": 0.1273,
+      "step": 945
+    },
+    {
+      "epoch": 1.0965732087227413,
+      "grad_norm": 0.2559017241001129,
+      "learning_rate": 4.45632183908046e-05,
+      "loss": 0.1239,
+      "step": 946
+    },
+    {
+      "epoch": 1.0977323770194884,
+      "grad_norm": 0.2877829670906067,
+      "learning_rate": 4.4557471264367815e-05,
+      "loss": 0.1337,
+      "step": 947
+    },
+    {
+      "epoch": 1.0988915453162356,
+      "grad_norm": 0.3006168007850647,
+      "learning_rate": 4.4551724137931036e-05,
+      "loss": 0.1387,
+      "step": 948
+    },
+    {
+      "epoch": 1.1000507136129827,
+      "grad_norm": 0.2508884370326996,
+      "learning_rate": 4.454597701149426e-05,
+      "loss": 0.1225,
+      "step": 949
+    },
+    {
+      "epoch": 1.1012098819097298,
+      "grad_norm": 0.2920970618724823,
+      "learning_rate": 4.454022988505747e-05,
+      "loss": 0.1384,
+      "step": 950
+    },
+    {
+      "epoch": 1.102369050206477,
+      "grad_norm": 0.28268909454345703,
+      "learning_rate": 4.4534482758620694e-05,
+      "loss": 0.1344,
+      "step": 951
+    },
+    {
+      "epoch": 1.103528218503224,
+      "grad_norm": 0.41611185669898987,
+      "learning_rate": 4.452873563218391e-05,
+      "loss": 0.1416,
+      "step": 952
+    },
+    {
+      "epoch": 1.104687386799971,
+      "grad_norm": 0.27131763100624084,
+      "learning_rate": 4.4522988505747124e-05,
+      "loss": 0.1279,
+      "step": 953
+    },
+    {
+      "epoch": 1.105846555096718,
+      "grad_norm": 0.250305712223053,
+      "learning_rate": 4.4517241379310346e-05,
+      "loss": 0.1136,
+      "step": 954
+    },
+    {
+      "epoch": 1.1070057233934651,
+      "grad_norm": 0.35008758306503296,
+      "learning_rate": 4.451149425287357e-05,
+      "loss": 0.1449,
+      "step": 955
+    },
+    {
+      "epoch": 1.1081648916902123,
+      "grad_norm": 0.23740458488464355,
+      "learning_rate": 4.450574712643679e-05,
+      "loss": 0.1094,
+      "step": 956
+    },
+    {
+      "epoch": 1.1093240599869594,
+      "grad_norm": 0.25829577445983887,
+      "learning_rate": 4.4500000000000004e-05,
+      "loss": 0.1097,
+      "step": 957
+    },
+    {
+      "epoch": 1.1104832282837065,
+      "grad_norm": 0.24298064410686493,
+      "learning_rate": 4.449425287356322e-05,
+      "loss": 0.1226,
+      "step": 958
+    },
+    {
+      "epoch": 1.1116423965804536,
+      "grad_norm": 0.27012696862220764,
+      "learning_rate": 4.448850574712644e-05,
+      "loss": 0.1289,
+      "step": 959
+    },
+    {
+      "epoch": 1.1128015648772007,
+      "grad_norm": 0.289583683013916,
+      "learning_rate": 4.4482758620689656e-05,
+      "loss": 0.138,
+      "step": 960
+    },
+    {
+      "epoch": 1.1139607331739476,
+      "grad_norm": 0.23818224668502808,
+      "learning_rate": 4.447701149425287e-05,
+      "loss": 0.1302,
+      "step": 961
+    },
+    {
+      "epoch": 1.1151199014706947,
+      "grad_norm": 0.2659807801246643,
+      "learning_rate": 4.447126436781609e-05,
+      "loss": 0.1248,
+      "step": 962
+    },
+    {
+      "epoch": 1.1162790697674418,
+      "grad_norm": 0.27049756050109863,
+      "learning_rate": 4.4465517241379314e-05,
+      "loss": 0.1355,
+      "step": 963
+    },
+    {
+      "epoch": 1.117438238064189,
+      "grad_norm": 0.2728050947189331,
+      "learning_rate": 4.445977011494253e-05,
+      "loss": 0.1264,
+      "step": 964
+    },
+    {
+      "epoch": 1.118597406360936,
+      "grad_norm": 0.2654384672641754,
+      "learning_rate": 4.445402298850575e-05,
+      "loss": 0.1303,
+      "step": 965
+    },
+    {
+      "epoch": 1.1197565746576832,
+      "grad_norm": 0.25535136461257935,
+      "learning_rate": 4.4448275862068965e-05,
+      "loss": 0.1189,
+      "step": 966
+    },
+    {
+      "epoch": 1.12091574295443,
+      "grad_norm": 0.2649828791618347,
+      "learning_rate": 4.444252873563219e-05,
+      "loss": 0.1289,
+      "step": 967
+    },
+    {
+      "epoch": 1.1220749112511772,
+      "grad_norm": 0.26751139760017395,
+      "learning_rate": 4.44367816091954e-05,
+      "loss": 0.1306,
+      "step": 968
+    },
+    {
+      "epoch": 1.1232340795479243,
+      "grad_norm": 0.2819879353046417,
+      "learning_rate": 4.4431034482758624e-05,
+      "loss": 0.132,
+      "step": 969
+    },
+    {
+      "epoch": 1.1243932478446714,
+      "grad_norm": 0.24419303238391876,
+      "learning_rate": 4.4425287356321845e-05,
+      "loss": 0.1302,
+      "step": 970
+    },
+    {
+      "epoch": 1.1255524161414185,
+      "grad_norm": 0.23565685749053955,
+      "learning_rate": 4.441954022988506e-05,
+      "loss": 0.1215,
+      "step": 971
+    },
+    {
+      "epoch": 1.1267115844381657,
+      "grad_norm": 0.24753518402576447,
+      "learning_rate": 4.4413793103448275e-05,
+      "loss": 0.1195,
+      "step": 972
+    },
+    {
+      "epoch": 1.1278707527349128,
+      "grad_norm": 0.26664602756500244,
+      "learning_rate": 4.44080459770115e-05,
+      "loss": 0.1289,
+      "step": 973
+    },
+    {
+      "epoch": 1.1290299210316599,
+      "grad_norm": 0.2846687436103821,
+      "learning_rate": 4.440229885057471e-05,
+      "loss": 0.1374,
+      "step": 974
+    },
+    {
+      "epoch": 1.1301890893284068,
+      "grad_norm": 0.2790849506855011,
+      "learning_rate": 4.4396551724137933e-05,
+      "loss": 0.1427,
+      "step": 975
+    },
+    {
+      "epoch": 1.131348257625154,
+      "grad_norm": 0.27603012323379517,
+      "learning_rate": 4.4390804597701155e-05,
+      "loss": 0.1335,
+      "step": 976
+    },
+    {
+      "epoch": 1.132507425921901,
+      "grad_norm": 0.2694251835346222,
+      "learning_rate": 4.438505747126437e-05,
+      "loss": 0.1278,
+      "step": 977
+    },
+    {
+      "epoch": 1.1336665942186481,
+      "grad_norm": 0.25316059589385986,
+      "learning_rate": 4.437931034482759e-05,
+      "loss": 0.1387,
+      "step": 978
+    },
+    {
+      "epoch": 1.1348257625153952,
+      "grad_norm": 0.2663499116897583,
+      "learning_rate": 4.437356321839081e-05,
+      "loss": 0.1206,
+      "step": 979
+    },
+    {
+      "epoch": 1.1359849308121424,
+      "grad_norm": 0.2719017267227173,
+      "learning_rate": 4.436781609195402e-05,
+      "loss": 0.1401,
+      "step": 980
+    },
+    {
+      "epoch": 1.1371440991088893,
+      "grad_norm": 0.4000224769115448,
+      "learning_rate": 4.436206896551724e-05,
+      "loss": 0.1259,
+      "step": 981
+    },
+    {
+      "epoch": 1.1383032674056364,
+      "grad_norm": 0.2817740738391876,
+      "learning_rate": 4.435632183908046e-05,
+      "loss": 0.1307,
+      "step": 982
+    },
+    {
+      "epoch": 1.1394624357023835,
+      "grad_norm": 0.28263017535209656,
+      "learning_rate": 4.435057471264368e-05,
+      "loss": 0.1328,
+      "step": 983
+    },
+    {
+      "epoch": 1.1406216039991306,
+      "grad_norm": 0.2730136513710022,
+      "learning_rate": 4.43448275862069e-05,
+      "loss": 0.125,
+      "step": 984
+    },
+    {
+      "epoch": 1.1417807722958777,
+      "grad_norm": 0.24145883321762085,
+      "learning_rate": 4.4339080459770116e-05,
+      "loss": 0.1206,
+      "step": 985
+    },
+    {
+      "epoch": 1.1429399405926248,
+      "grad_norm": 0.29111185669898987,
+      "learning_rate": 4.433333333333334e-05,
+      "loss": 0.1469,
+      "step": 986
+    },
+    {
+      "epoch": 1.144099108889372,
+      "grad_norm": 0.28760775923728943,
+      "learning_rate": 4.432758620689655e-05,
+      "loss": 0.1338,
+      "step": 987
+    },
+    {
+      "epoch": 1.145258277186119,
+      "grad_norm": 0.2725701630115509,
+      "learning_rate": 4.432183908045977e-05,
+      "loss": 0.1314,
+      "step": 988
+    },
+    {
+      "epoch": 1.146417445482866,
+      "grad_norm": 0.2519955635070801,
+      "learning_rate": 4.431609195402299e-05,
+      "loss": 0.1277,
+      "step": 989
+    },
+    {
+      "epoch": 1.147576613779613,
+      "grad_norm": 0.28241047263145447,
+      "learning_rate": 4.431034482758621e-05,
+      "loss": 0.1309,
+      "step": 990
+    },
+    {
+      "epoch": 1.1487357820763602,
+      "grad_norm": 0.284697562456131,
+      "learning_rate": 4.4304597701149426e-05,
+      "loss": 0.1396,
+      "step": 991
+    },
+    {
+      "epoch": 1.1498949503731073,
+      "grad_norm": 0.2153967171907425,
+      "learning_rate": 4.429885057471265e-05,
+      "loss": 0.1074,
+      "step": 992
+    },
+    {
+      "epoch": 1.1510541186698544,
+      "grad_norm": 0.28190892934799194,
+      "learning_rate": 4.429310344827586e-05,
+      "loss": 0.1408,
+      "step": 993
+    },
+    {
+      "epoch": 1.1522132869666015,
+      "grad_norm": 0.26553207635879517,
+      "learning_rate": 4.4287356321839084e-05,
+      "loss": 0.1191,
+      "step": 994
+    },
+    {
+      "epoch": 1.1533724552633486,
+      "grad_norm": 0.3092179596424103,
+      "learning_rate": 4.42816091954023e-05,
+      "loss": 0.1408,
+      "step": 995
+    },
+    {
+      "epoch": 1.1545316235600955,
+      "grad_norm": 0.2503480315208435,
+      "learning_rate": 4.427586206896552e-05,
+      "loss": 0.1152,
+      "step": 996
+    },
+    {
+      "epoch": 1.1556907918568426,
+      "grad_norm": 0.279032826423645,
+      "learning_rate": 4.427011494252874e-05,
+      "loss": 0.118,
+      "step": 997
+    },
+    {
+      "epoch": 1.1568499601535898,
+      "grad_norm": 0.283666729927063,
+      "learning_rate": 4.426436781609196e-05,
+      "loss": 0.143,
+      "step": 998
+    },
+    {
+      "epoch": 1.1580091284503369,
+      "grad_norm": 0.3145494759082794,
+      "learning_rate": 4.425862068965517e-05,
+      "loss": 0.1467,
+      "step": 999
+    },
+    {
+      "epoch": 1.159168296747084,
+      "grad_norm": 0.24866163730621338,
+      "learning_rate": 4.4252873563218394e-05,
+      "loss": 0.1217,
+      "step": 1000
+    },
+    {
+      "epoch": 1.160327465043831,
+      "grad_norm": 0.29811036586761475,
+      "learning_rate": 4.424712643678161e-05,
+      "loss": 0.1343,
+      "step": 1001
+    },
+    {
+      "epoch": 1.1614866333405782,
+      "grad_norm": 0.2376304566860199,
+      "learning_rate": 4.4241379310344824e-05,
+      "loss": 0.1111,
+      "step": 1002
+    },
+    {
+      "epoch": 1.1626458016373253,
+      "grad_norm": 0.2960997223854065,
+      "learning_rate": 4.423563218390805e-05,
+      "loss": 0.1537,
+      "step": 1003
+    },
+    {
+      "epoch": 1.1638049699340722,
+      "grad_norm": 0.2618663012981415,
+      "learning_rate": 4.422988505747127e-05,
+      "loss": 0.1244,
+      "step": 1004
+    },
+    {
+      "epoch": 1.1649641382308193,
+      "grad_norm": 0.2316623479127884,
+      "learning_rate": 4.422413793103449e-05,
+      "loss": 0.1276,
+      "step": 1005
+    },
+    {
+      "epoch": 1.1661233065275665,
+      "grad_norm": 0.225045844912529,
+      "learning_rate": 4.4218390804597704e-05,
+      "loss": 0.12,
+      "step": 1006
+    },
+    {
+      "epoch": 1.1672824748243136,
+      "grad_norm": 0.2224287986755371,
+      "learning_rate": 4.421264367816092e-05,
+      "loss": 0.1281,
+      "step": 1007
+    },
+    {
+      "epoch": 1.1684416431210607,
+      "grad_norm": 0.23176699876785278,
+      "learning_rate": 4.420689655172414e-05,
+      "loss": 0.1115,
+      "step": 1008
+    },
+    {
+      "epoch": 1.1696008114178078,
+      "grad_norm": 0.2502862215042114,
+      "learning_rate": 4.4201149425287355e-05,
+      "loss": 0.1183,
+      "step": 1009
+    },
+    {
+      "epoch": 1.1707599797145547,
+      "grad_norm": 0.24579674005508423,
+      "learning_rate": 4.419540229885058e-05,
+      "loss": 0.1302,
+      "step": 1010
+    },
+    {
+      "epoch": 1.1719191480113018,
+      "grad_norm": 0.26011398434638977,
+      "learning_rate": 4.41896551724138e-05,
+      "loss": 0.1386,
+      "step": 1011
+    },
+    {
+      "epoch": 1.173078316308049,
+      "grad_norm": 0.26081839203834534,
+      "learning_rate": 4.4183908045977014e-05,
+      "loss": 0.1511,
+      "step": 1012
+    },
+    {
+      "epoch": 1.174237484604796,
+      "grad_norm": 0.22634555399417877,
+      "learning_rate": 4.4178160919540235e-05,
+      "loss": 0.1304,
+      "step": 1013
+    },
+    {
+      "epoch": 1.1753966529015432,
+      "grad_norm": 0.23882032930850983,
+      "learning_rate": 4.417241379310345e-05,
+      "loss": 0.1289,
+      "step": 1014
+    },
+    {
+      "epoch": 1.1765558211982903,
+      "grad_norm": 0.26218292117118835,
+      "learning_rate": 4.4166666666666665e-05,
+      "loss": 0.1373,
+      "step": 1015
+    },
+    {
+      "epoch": 1.1777149894950374,
+      "grad_norm": 0.3141017556190491,
+      "learning_rate": 4.416091954022989e-05,
+      "loss": 0.1389,
+      "step": 1016
+    },
+    {
+      "epoch": 1.1788741577917845,
+      "grad_norm": 0.2978722155094147,
+      "learning_rate": 4.415517241379311e-05,
+      "loss": 0.14,
+      "step": 1017
+    },
+    {
+      "epoch": 1.1800333260885314,
+      "grad_norm": 0.30704426765441895,
+      "learning_rate": 4.414942528735632e-05,
+      "loss": 0.1286,
+      "step": 1018
+    },
+    {
+      "epoch": 1.1811924943852785,
+      "grad_norm": 0.3598279058933258,
+      "learning_rate": 4.4143678160919545e-05,
+      "loss": 0.1354,
+      "step": 1019
+    },
+    {
+      "epoch": 1.1823516626820256,
+      "grad_norm": 0.2837865650653839,
+      "learning_rate": 4.413793103448276e-05,
+      "loss": 0.1311,
+      "step": 1020
+    },
+    {
+      "epoch": 1.1835108309787727,
+      "grad_norm": 0.2859646677970886,
+      "learning_rate": 4.4132183908045975e-05,
+      "loss": 0.1153,
+      "step": 1021
+    },
+    {
+      "epoch": 1.1846699992755199,
+      "grad_norm": 0.2848678529262543,
+      "learning_rate": 4.4126436781609196e-05,
+      "loss": 0.1344,
+      "step": 1022
+    },
+    {
+      "epoch": 1.185829167572267,
+      "grad_norm": 0.27502962946891785,
+      "learning_rate": 4.412068965517241e-05,
+      "loss": 0.1336,
+      "step": 1023
+    },
+    {
+      "epoch": 1.1869883358690139,
+      "grad_norm": 0.23905035853385925,
+      "learning_rate": 4.411494252873564e-05,
+      "loss": 0.1415,
+      "step": 1024
+    },
+    {
+      "epoch": 1.188147504165761,
+      "grad_norm": 0.3138931691646576,
+      "learning_rate": 4.4109195402298855e-05,
+      "loss": 0.1388,
+      "step": 1025
+    },
+    {
+      "epoch": 1.189306672462508,
+      "grad_norm": 0.2796839773654938,
+      "learning_rate": 4.410344827586207e-05,
+      "loss": 0.126,
+      "step": 1026
+    },
+    {
+      "epoch": 1.1904658407592552,
+      "grad_norm": 0.24468503892421722,
+      "learning_rate": 4.409770114942529e-05,
+      "loss": 0.1356,
+      "step": 1027
+    },
+    {
+      "epoch": 1.1916250090560023,
+      "grad_norm": 0.2509666383266449,
+      "learning_rate": 4.4091954022988506e-05,
+      "loss": 0.1223,
+      "step": 1028
+    },
+    {
+      "epoch": 1.1927841773527494,
+      "grad_norm": 0.2291174829006195,
+      "learning_rate": 4.408620689655172e-05,
+      "loss": 0.1114,
+      "step": 1029
+    },
+    {
+      "epoch": 1.1939433456494966,
+      "grad_norm": 0.21360966563224792,
+      "learning_rate": 4.408045977011494e-05,
+      "loss": 0.1149,
+      "step": 1030
+    },
+    {
+      "epoch": 1.1951025139462437,
+      "grad_norm": 0.273369163274765,
+      "learning_rate": 4.4074712643678164e-05,
+      "loss": 0.14,
+      "step": 1031
+    },
+    {
+      "epoch": 1.1962616822429906,
+      "grad_norm": 0.24241317808628082,
+      "learning_rate": 4.4068965517241386e-05,
+      "loss": 0.1242,
+      "step": 1032
+    },
+    {
+      "epoch": 1.1974208505397377,
+      "grad_norm": 0.29952454566955566,
+      "learning_rate": 4.40632183908046e-05,
+      "loss": 0.1363,
+      "step": 1033
+    },
+    {
+      "epoch": 1.1985800188364848,
+      "grad_norm": 0.2337886095046997,
+      "learning_rate": 4.4057471264367816e-05,
+      "loss": 0.1229,
+      "step": 1034
+    },
+    {
+      "epoch": 1.199739187133232,
+      "grad_norm": 0.21567285060882568,
+      "learning_rate": 4.405172413793104e-05,
+      "loss": 0.1208,
+      "step": 1035
+    },
+    {
+      "epoch": 1.200898355429979,
+      "grad_norm": 0.2876308262348175,
+      "learning_rate": 4.404597701149425e-05,
+      "loss": 0.1376,
+      "step": 1036
+    },
+    {
+      "epoch": 1.2020575237267261,
+      "grad_norm": 0.30270469188690186,
+      "learning_rate": 4.4040229885057474e-05,
+      "loss": 0.1444,
+      "step": 1037
+    },
+    {
+      "epoch": 1.203216692023473,
+      "grad_norm": 0.2885245680809021,
+      "learning_rate": 4.4034482758620696e-05,
+      "loss": 0.1327,
+      "step": 1038
+    },
+    {
+      "epoch": 1.2043758603202201,
+      "grad_norm": 0.3696737289428711,
+      "learning_rate": 4.402873563218391e-05,
+      "loss": 0.1444,
+      "step": 1039
+    },
+    {
+      "epoch": 1.2055350286169673,
+      "grad_norm": 0.26480332016944885,
+      "learning_rate": 4.4022988505747126e-05,
+      "loss": 0.1332,
+      "step": 1040
+    },
+    {
+      "epoch": 1.2066941969137144,
+      "grad_norm": 0.25360003113746643,
+      "learning_rate": 4.401724137931035e-05,
+      "loss": 0.1315,
+      "step": 1041
+    },
+    {
+      "epoch": 1.2078533652104615,
+      "grad_norm": 0.22617582976818085,
+      "learning_rate": 4.401149425287356e-05,
+      "loss": 0.1289,
+      "step": 1042
+    },
+    {
+      "epoch": 1.2090125335072086,
+      "grad_norm": 0.22125768661499023,
+      "learning_rate": 4.4005747126436784e-05,
+      "loss": 0.125,
+      "step": 1043
+    },
+    {
+      "epoch": 1.2101717018039557,
+      "grad_norm": 0.23760303854942322,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.1201,
+      "step": 1044
+    },
+    {
+      "epoch": 1.2113308701007028,
+      "grad_norm": 0.28465020656585693,
+      "learning_rate": 4.399425287356322e-05,
+      "loss": 0.1306,
+      "step": 1045
+    },
+    {
+      "epoch": 1.21249003839745,
+      "grad_norm": 0.2427520602941513,
+      "learning_rate": 4.398850574712644e-05,
+      "loss": 0.1252,
+      "step": 1046
+    },
+    {
+      "epoch": 1.2136492066941968,
+      "grad_norm": 0.29493942856788635,
+      "learning_rate": 4.398275862068966e-05,
+      "loss": 0.1131,
+      "step": 1047
+    },
+    {
+      "epoch": 1.214808374990944,
+      "grad_norm": 0.24606949090957642,
+      "learning_rate": 4.397701149425287e-05,
+      "loss": 0.1254,
+      "step": 1048
+    },
+    {
+      "epoch": 1.215967543287691,
+      "grad_norm": 0.26658526062965393,
+      "learning_rate": 4.3971264367816094e-05,
+      "loss": 0.1299,
+      "step": 1049
+    },
+    {
+      "epoch": 1.2171267115844382,
+      "grad_norm": 0.22725453972816467,
+      "learning_rate": 4.396551724137931e-05,
+      "loss": 0.1209,
+      "step": 1050
+    },
+    {
+      "epoch": 1.2182858798811853,
+      "grad_norm": 0.26789578795433044,
+      "learning_rate": 4.395977011494253e-05,
+      "loss": 0.131,
+      "step": 1051
+    },
+    {
+      "epoch": 1.2194450481779324,
+      "grad_norm": 0.2898029685020447,
+      "learning_rate": 4.395402298850575e-05,
+      "loss": 0.1321,
+      "step": 1052
+    },
+    {
+      "epoch": 1.2206042164746793,
+      "grad_norm": 0.2893812656402588,
+      "learning_rate": 4.394827586206897e-05,
+      "loss": 0.151,
+      "step": 1053
+    },
+    {
+      "epoch": 1.2217633847714264,
+      "grad_norm": 0.2428007870912552,
+      "learning_rate": 4.394252873563219e-05,
+      "loss": 0.1221,
+      "step": 1054
+    },
+    {
+      "epoch": 1.2229225530681735,
+      "grad_norm": 0.31453728675842285,
+      "learning_rate": 4.3936781609195403e-05,
+      "loss": 0.127,
+      "step": 1055
+    },
+    {
+      "epoch": 1.2240817213649207,
+      "grad_norm": 0.22855356335639954,
+      "learning_rate": 4.393103448275862e-05,
+      "loss": 0.1232,
+      "step": 1056
+    },
+    {
+      "epoch": 1.2252408896616678,
+      "grad_norm": 0.21666891872882843,
+      "learning_rate": 4.392528735632184e-05,
+      "loss": 0.1174,
+      "step": 1057
+    },
+    {
+      "epoch": 1.226400057958415,
+      "grad_norm": 0.2279151976108551,
+      "learning_rate": 4.391954022988506e-05,
+      "loss": 0.1307,
+      "step": 1058
+    },
+    {
+      "epoch": 1.227559226255162,
+      "grad_norm": 0.2704702615737915,
+      "learning_rate": 4.3913793103448277e-05,
+      "loss": 0.1328,
+      "step": 1059
+    },
+    {
+      "epoch": 1.2287183945519091,
+      "grad_norm": 0.2585502564907074,
+      "learning_rate": 4.39080459770115e-05,
+      "loss": 0.1359,
+      "step": 1060
+    },
+    {
+      "epoch": 1.229877562848656,
+      "grad_norm": 0.29662445187568665,
+      "learning_rate": 4.390229885057471e-05,
+      "loss": 0.1302,
+      "step": 1061
+    },
+    {
+      "epoch": 1.2310367311454031,
+      "grad_norm": 0.24643655121326447,
+      "learning_rate": 4.3896551724137935e-05,
+      "loss": 0.1233,
+      "step": 1062
+    },
+    {
+      "epoch": 1.2321958994421502,
+      "grad_norm": 0.238943949341774,
+      "learning_rate": 4.389080459770115e-05,
+      "loss": 0.1271,
+      "step": 1063
+    },
+    {
+      "epoch": 1.2333550677388974,
+      "grad_norm": 0.27087101340293884,
+      "learning_rate": 4.388505747126437e-05,
+      "loss": 0.125,
+      "step": 1064
+    },
+    {
+      "epoch": 1.2345142360356445,
+      "grad_norm": 0.29983842372894287,
+      "learning_rate": 4.387931034482759e-05,
+      "loss": 0.135,
+      "step": 1065
+    },
+    {
+      "epoch": 1.2356734043323916,
+      "grad_norm": 0.24061062932014465,
+      "learning_rate": 4.387356321839081e-05,
+      "loss": 0.1365,
+      "step": 1066
+    },
+    {
+      "epoch": 1.2368325726291385,
+      "grad_norm": 0.24052466452121735,
+      "learning_rate": 4.386781609195402e-05,
+      "loss": 0.1175,
+      "step": 1067
+    },
+    {
+      "epoch": 1.2379917409258856,
+      "grad_norm": 0.2357567995786667,
+      "learning_rate": 4.3862068965517245e-05,
+      "loss": 0.1337,
+      "step": 1068
+    },
+    {
+      "epoch": 1.2391509092226327,
+      "grad_norm": 0.3319186270236969,
+      "learning_rate": 4.385632183908046e-05,
+      "loss": 0.1434,
+      "step": 1069
+    },
+    {
+      "epoch": 1.2403100775193798,
+      "grad_norm": 0.34508636593818665,
+      "learning_rate": 4.385057471264368e-05,
+      "loss": 0.1511,
+      "step": 1070
+    },
+    {
+      "epoch": 1.241469245816127,
+      "grad_norm": 0.22875401377677917,
+      "learning_rate": 4.3844827586206896e-05,
+      "loss": 0.1278,
+      "step": 1071
+    },
+    {
+      "epoch": 1.242628414112874,
+      "grad_norm": 0.2499420940876007,
+      "learning_rate": 4.383908045977012e-05,
+      "loss": 0.1204,
+      "step": 1072
+    },
+    {
+      "epoch": 1.2437875824096212,
+      "grad_norm": 0.24523814022541046,
+      "learning_rate": 4.383333333333334e-05,
+      "loss": 0.1351,
+      "step": 1073
+    },
+    {
+      "epoch": 1.2449467507063683,
+      "grad_norm": 0.28577345609664917,
+      "learning_rate": 4.3827586206896554e-05,
+      "loss": 0.121,
+      "step": 1074
+    },
+    {
+      "epoch": 1.2461059190031152,
+      "grad_norm": 0.21815964579582214,
+      "learning_rate": 4.382183908045977e-05,
+      "loss": 0.113,
+      "step": 1075
+    },
+    {
+      "epoch": 1.2472650872998623,
+      "grad_norm": 0.24391821026802063,
+      "learning_rate": 4.381609195402299e-05,
+      "loss": 0.1196,
+      "step": 1076
+    },
+    {
+      "epoch": 1.2484242555966094,
+      "grad_norm": 0.2738696336746216,
+      "learning_rate": 4.3810344827586206e-05,
+      "loss": 0.1436,
+      "step": 1077
+    },
+    {
+      "epoch": 1.2495834238933565,
+      "grad_norm": 0.24671784043312073,
+      "learning_rate": 4.380459770114943e-05,
+      "loss": 0.1221,
+      "step": 1078
+    },
+    {
+      "epoch": 1.2507425921901036,
+      "grad_norm": 0.249884694814682,
+      "learning_rate": 4.379885057471265e-05,
+      "loss": 0.1199,
+      "step": 1079
+    },
+    {
+      "epoch": 1.2519017604868508,
+      "grad_norm": 0.25286155939102173,
+      "learning_rate": 4.3793103448275864e-05,
+      "loss": 0.1381,
+      "step": 1080
+    },
+    {
+      "epoch": 1.2530609287835976,
+      "grad_norm": 0.36500653624534607,
+      "learning_rate": 4.3787356321839086e-05,
+      "loss": 0.13,
+      "step": 1081
+    },
+    {
+      "epoch": 1.2542200970803448,
+      "grad_norm": 0.2865230441093445,
+      "learning_rate": 4.37816091954023e-05,
+      "loss": 0.1223,
+      "step": 1082
+    },
+    {
+      "epoch": 1.2553792653770919,
+      "grad_norm": 0.2717812657356262,
+      "learning_rate": 4.3775862068965516e-05,
+      "loss": 0.118,
+      "step": 1083
+    },
+    {
+      "epoch": 1.256538433673839,
+      "grad_norm": 0.23932139575481415,
+      "learning_rate": 4.377011494252874e-05,
+      "loss": 0.1171,
+      "step": 1084
+    },
+    {
+      "epoch": 1.257697601970586,
+      "grad_norm": 0.25386103987693787,
+      "learning_rate": 4.376436781609196e-05,
+      "loss": 0.127,
+      "step": 1085
+    },
+    {
+      "epoch": 1.2588567702673332,
+      "grad_norm": 0.25412800908088684,
+      "learning_rate": 4.3758620689655174e-05,
+      "loss": 0.1394,
+      "step": 1086
+    },
+    {
+      "epoch": 1.2600159385640803,
+      "grad_norm": 0.30397337675094604,
+      "learning_rate": 4.3752873563218395e-05,
+      "loss": 0.1087,
+      "step": 1087
+    },
+    {
+      "epoch": 1.2611751068608275,
+      "grad_norm": 0.2501704692840576,
+      "learning_rate": 4.374712643678161e-05,
+      "loss": 0.1342,
+      "step": 1088
+    },
+    {
+      "epoch": 1.2623342751575746,
+      "grad_norm": 0.2684697210788727,
+      "learning_rate": 4.374137931034483e-05,
+      "loss": 0.1225,
+      "step": 1089
+    },
+    {
+      "epoch": 1.2634934434543215,
+      "grad_norm": 0.2351657599210739,
+      "learning_rate": 4.373563218390805e-05,
+      "loss": 0.1106,
+      "step": 1090
+    },
+    {
+      "epoch": 1.2646526117510686,
+      "grad_norm": 0.3316408097743988,
+      "learning_rate": 4.372988505747126e-05,
+      "loss": 0.1367,
+      "step": 1091
+    },
+    {
+      "epoch": 1.2658117800478157,
+      "grad_norm": 0.2677469849586487,
+      "learning_rate": 4.3724137931034484e-05,
+      "loss": 0.1221,
+      "step": 1092
+    },
+    {
+      "epoch": 1.2669709483445628,
+      "grad_norm": 0.5877587199211121,
+      "learning_rate": 4.3718390804597705e-05,
+      "loss": 0.1538,
+      "step": 1093
+    },
+    {
+      "epoch": 1.26813011664131,
+      "grad_norm": 0.2598472833633423,
+      "learning_rate": 4.371264367816092e-05,
+      "loss": 0.1325,
+      "step": 1094
+    },
+    {
+      "epoch": 1.2692892849380568,
+      "grad_norm": 0.3260670602321625,
+      "learning_rate": 4.370689655172414e-05,
+      "loss": 0.1372,
+      "step": 1095
+    },
+    {
+      "epoch": 1.270448453234804,
+      "grad_norm": 0.24232131242752075,
+      "learning_rate": 4.370114942528736e-05,
+      "loss": 0.1255,
+      "step": 1096
+    },
+    {
+      "epoch": 1.271607621531551,
+      "grad_norm": 0.3213041126728058,
+      "learning_rate": 4.369540229885058e-05,
+      "loss": 0.1325,
+      "step": 1097
+    },
+    {
+      "epoch": 1.2727667898282982,
+      "grad_norm": 0.2717669904232025,
+      "learning_rate": 4.368965517241379e-05,
+      "loss": 0.1447,
+      "step": 1098
+    },
+    {
+      "epoch": 1.2739259581250453,
+      "grad_norm": 0.2611459791660309,
+      "learning_rate": 4.3683908045977015e-05,
+      "loss": 0.1349,
+      "step": 1099
+    },
+    {
+      "epoch": 1.2750851264217924,
+      "grad_norm": 0.21271906793117523,
+      "learning_rate": 4.367816091954024e-05,
+      "loss": 0.1062,
+      "step": 1100
+    },
+    {
+      "epoch": 1.2762442947185395,
+      "grad_norm": 0.28656983375549316,
+      "learning_rate": 4.367241379310345e-05,
+      "loss": 0.1535,
+      "step": 1101
+    },
+    {
+      "epoch": 1.2774034630152866,
+      "grad_norm": 0.24881134927272797,
+      "learning_rate": 4.3666666666666666e-05,
+      "loss": 0.132,
+      "step": 1102
+    },
+    {
+      "epoch": 1.2785626313120337,
+      "grad_norm": 0.3379434645175934,
+      "learning_rate": 4.366091954022989e-05,
+      "loss": 0.1276,
+      "step": 1103
+    },
+    {
+      "epoch": 1.2797217996087806,
+      "grad_norm": 0.32010799646377563,
+      "learning_rate": 4.36551724137931e-05,
+      "loss": 0.133,
+      "step": 1104
+    },
+    {
+      "epoch": 1.2808809679055277,
+      "grad_norm": 0.22083823382854462,
+      "learning_rate": 4.3649425287356325e-05,
+      "loss": 0.1258,
+      "step": 1105
+    },
+    {
+      "epoch": 1.2820401362022749,
+      "grad_norm": 0.27177348732948303,
+      "learning_rate": 4.3643678160919546e-05,
+      "loss": 0.1299,
+      "step": 1106
+    },
+    {
+      "epoch": 1.283199304499022,
+      "grad_norm": 0.2737819254398346,
+      "learning_rate": 4.363793103448276e-05,
+      "loss": 0.1322,
+      "step": 1107
+    },
+    {
+      "epoch": 1.284358472795769,
+      "grad_norm": 0.2882055640220642,
+      "learning_rate": 4.363218390804598e-05,
+      "loss": 0.1317,
+      "step": 1108
+    },
+    {
+      "epoch": 1.285517641092516,
+      "grad_norm": 0.24766948819160461,
+      "learning_rate": 4.36264367816092e-05,
+      "loss": 0.1247,
+      "step": 1109
+    },
+    {
+      "epoch": 1.286676809389263,
+      "grad_norm": 0.27174562215805054,
+      "learning_rate": 4.362068965517241e-05,
+      "loss": 0.1404,
+      "step": 1110
+    },
+    {
+      "epoch": 1.2878359776860102,
+      "grad_norm": 0.25926515460014343,
+      "learning_rate": 4.3614942528735634e-05,
+      "loss": 0.1275,
+      "step": 1111
+    },
+    {
+      "epoch": 1.2889951459827573,
+      "grad_norm": 0.2558022439479828,
+      "learning_rate": 4.360919540229885e-05,
+      "loss": 0.1224,
+      "step": 1112
+    },
+    {
+      "epoch": 1.2901543142795044,
+      "grad_norm": 0.24418483674526215,
+      "learning_rate": 4.360344827586207e-05,
+      "loss": 0.1274,
+      "step": 1113
+    },
+    {
+      "epoch": 1.2913134825762516,
+      "grad_norm": 0.2943531274795532,
+      "learning_rate": 4.359770114942529e-05,
+      "loss": 0.1242,
+      "step": 1114
+    },
+    {
+      "epoch": 1.2924726508729987,
+      "grad_norm": 0.24752426147460938,
+      "learning_rate": 4.359195402298851e-05,
+      "loss": 0.1336,
+      "step": 1115
+    },
+    {
+      "epoch": 1.2936318191697458,
+      "grad_norm": 0.24908795952796936,
+      "learning_rate": 4.358620689655173e-05,
+      "loss": 0.1176,
+      "step": 1116
+    },
+    {
+      "epoch": 1.294790987466493,
+      "grad_norm": 0.290752649307251,
+      "learning_rate": 4.3580459770114944e-05,
+      "loss": 0.1532,
+      "step": 1117
+    },
+    {
+      "epoch": 1.2959501557632398,
+      "grad_norm": 0.2473299503326416,
+      "learning_rate": 4.357471264367816e-05,
+      "loss": 0.1166,
+      "step": 1118
+    },
+    {
+      "epoch": 1.297109324059987,
+      "grad_norm": 0.24173453450202942,
+      "learning_rate": 4.356896551724138e-05,
+      "loss": 0.1069,
+      "step": 1119
+    },
+    {
+      "epoch": 1.298268492356734,
+      "grad_norm": 0.2775953412055969,
+      "learning_rate": 4.35632183908046e-05,
+      "loss": 0.1163,
+      "step": 1120
+    },
+    {
+      "epoch": 1.2994276606534811,
+      "grad_norm": 0.2543400824069977,
+      "learning_rate": 4.355747126436782e-05,
+      "loss": 0.1359,
+      "step": 1121
+    },
+    {
+      "epoch": 1.3005868289502283,
+      "grad_norm": 0.2600906193256378,
+      "learning_rate": 4.355172413793104e-05,
+      "loss": 0.1297,
+      "step": 1122
+    },
+    {
+      "epoch": 1.3017459972469754,
+      "grad_norm": 0.21949701011180878,
+      "learning_rate": 4.3545977011494254e-05,
+      "loss": 0.1223,
+      "step": 1123
+    },
+    {
+      "epoch": 1.3029051655437223,
+      "grad_norm": 0.2765035629272461,
+      "learning_rate": 4.354022988505747e-05,
+      "loss": 0.1302,
+      "step": 1124
+    },
+    {
+      "epoch": 1.3040643338404694,
+      "grad_norm": 0.2629549205303192,
+      "learning_rate": 4.353448275862069e-05,
+      "loss": 0.1321,
+      "step": 1125
+    },
+    {
+      "epoch": 1.3052235021372165,
+      "grad_norm": 0.278767853975296,
+      "learning_rate": 4.352873563218391e-05,
+      "loss": 0.1252,
+      "step": 1126
+    },
+    {
+      "epoch": 1.3063826704339636,
+      "grad_norm": 0.3192477226257324,
+      "learning_rate": 4.3522988505747134e-05,
+      "loss": 0.1559,
+      "step": 1127
+    },
+    {
+      "epoch": 1.3075418387307107,
+      "grad_norm": 0.24097472429275513,
+      "learning_rate": 4.351724137931035e-05,
+      "loss": 0.1239,
+      "step": 1128
+    },
+    {
+      "epoch": 1.3087010070274578,
+      "grad_norm": 0.27867624163627625,
+      "learning_rate": 4.3511494252873564e-05,
+      "loss": 0.142,
+      "step": 1129
+    },
+    {
+      "epoch": 1.309860175324205,
+      "grad_norm": 0.2419125735759735,
+      "learning_rate": 4.3505747126436785e-05,
+      "loss": 0.1126,
+      "step": 1130
+    },
+    {
+      "epoch": 1.311019343620952,
+      "grad_norm": 0.24018104374408722,
+      "learning_rate": 4.35e-05,
+      "loss": 0.1324,
+      "step": 1131
+    },
+    {
+      "epoch": 1.3121785119176992,
+      "grad_norm": 0.2754385769367218,
+      "learning_rate": 4.3494252873563215e-05,
+      "loss": 0.1422,
+      "step": 1132
+    },
+    {
+      "epoch": 1.313337680214446,
+      "grad_norm": 0.24360622465610504,
+      "learning_rate": 4.348850574712644e-05,
+      "loss": 0.1409,
+      "step": 1133
+    },
+    {
+      "epoch": 1.3144968485111932,
+      "grad_norm": 0.2462378591299057,
+      "learning_rate": 4.348275862068966e-05,
+      "loss": 0.1299,
+      "step": 1134
+    },
+    {
+      "epoch": 1.3156560168079403,
+      "grad_norm": 0.22275815904140472,
+      "learning_rate": 4.347701149425288e-05,
+      "loss": 0.1283,
+      "step": 1135
+    },
+    {
+      "epoch": 1.3168151851046874,
+      "grad_norm": 0.388450026512146,
+      "learning_rate": 4.3471264367816095e-05,
+      "loss": 0.152,
+      "step": 1136
+    },
+    {
+      "epoch": 1.3179743534014345,
+      "grad_norm": 0.24839439988136292,
+      "learning_rate": 4.346551724137931e-05,
+      "loss": 0.1333,
+      "step": 1137
+    },
+    {
+      "epoch": 1.3191335216981814,
+      "grad_norm": 0.24341842532157898,
+      "learning_rate": 4.345977011494253e-05,
+      "loss": 0.1176,
+      "step": 1138
+    },
+    {
+      "epoch": 1.3202926899949285,
+      "grad_norm": 0.22959062457084656,
+      "learning_rate": 4.3454022988505747e-05,
+      "loss": 0.1337,
+      "step": 1139
+    },
+    {
+      "epoch": 1.3214518582916757,
+      "grad_norm": 0.2183791995048523,
+      "learning_rate": 4.344827586206897e-05,
+      "loss": 0.1157,
+      "step": 1140
+    },
+    {
+      "epoch": 1.3226110265884228,
+      "grad_norm": 0.3063965141773224,
+      "learning_rate": 4.344252873563219e-05,
+      "loss": 0.1365,
+      "step": 1141
+    },
+    {
+      "epoch": 1.32377019488517,
+      "grad_norm": 0.24136781692504883,
+      "learning_rate": 4.3436781609195405e-05,
+      "loss": 0.1255,
+      "step": 1142
+    },
+    {
+      "epoch": 1.324929363181917,
+      "grad_norm": 0.21908266842365265,
+      "learning_rate": 4.343103448275862e-05,
+      "loss": 0.1144,
+      "step": 1143
+    },
+    {
+      "epoch": 1.3260885314786641,
+      "grad_norm": 0.2468622326850891,
+      "learning_rate": 4.342528735632184e-05,
+      "loss": 0.1298,
+      "step": 1144
+    },
+    {
+      "epoch": 1.3272476997754112,
+      "grad_norm": 0.28959596157073975,
+      "learning_rate": 4.3419540229885056e-05,
+      "loss": 0.1318,
+      "step": 1145
+    },
+    {
+      "epoch": 1.3284068680721584,
+      "grad_norm": 0.2605699300765991,
+      "learning_rate": 4.341379310344828e-05,
+      "loss": 0.137,
+      "step": 1146
+    },
+    {
+      "epoch": 1.3295660363689052,
+      "grad_norm": 0.3156377971172333,
+      "learning_rate": 4.34080459770115e-05,
+      "loss": 0.1417,
+      "step": 1147
+    },
+    {
+      "epoch": 1.3307252046656524,
+      "grad_norm": 0.24471645057201385,
+      "learning_rate": 4.3402298850574715e-05,
+      "loss": 0.1268,
+      "step": 1148
+    },
+    {
+      "epoch": 1.3318843729623995,
+      "grad_norm": 0.24388937652111053,
+      "learning_rate": 4.3396551724137936e-05,
+      "loss": 0.13,
+      "step": 1149
+    },
+    {
+      "epoch": 1.3330435412591466,
+      "grad_norm": 0.21283157169818878,
+      "learning_rate": 4.339080459770115e-05,
+      "loss": 0.1213,
+      "step": 1150
+    },
+    {
+      "epoch": 1.3342027095558937,
+      "grad_norm": 0.232645183801651,
+      "learning_rate": 4.3385057471264366e-05,
+      "loss": 0.1245,
+      "step": 1151
+    },
+    {
+      "epoch": 1.3353618778526406,
+      "grad_norm": 0.25331243872642517,
+      "learning_rate": 4.337931034482759e-05,
+      "loss": 0.1232,
+      "step": 1152
+    },
+    {
+      "epoch": 1.3365210461493877,
+      "grad_norm": 0.24029718339443207,
+      "learning_rate": 4.33735632183908e-05,
+      "loss": 0.1161,
+      "step": 1153
+    },
+    {
+      "epoch": 1.3376802144461348,
+      "grad_norm": 0.2674920856952667,
+      "learning_rate": 4.336781609195403e-05,
+      "loss": 0.1361,
+      "step": 1154
+    },
+    {
+      "epoch": 1.338839382742882,
+      "grad_norm": 0.19390049576759338,
+      "learning_rate": 4.3362068965517246e-05,
+      "loss": 0.1138,
+      "step": 1155
+    },
+    {
+      "epoch": 1.339998551039629,
+      "grad_norm": 0.2703855633735657,
+      "learning_rate": 4.335632183908046e-05,
+      "loss": 0.1334,
+      "step": 1156
+    },
+    {
+      "epoch": 1.3411577193363762,
+      "grad_norm": 0.23119378089904785,
+      "learning_rate": 4.335057471264368e-05,
+      "loss": 0.1278,
+      "step": 1157
+    },
+    {
+      "epoch": 1.3423168876331233,
+      "grad_norm": 0.22249457240104675,
+      "learning_rate": 4.33448275862069e-05,
+      "loss": 0.1122,
+      "step": 1158
+    },
+    {
+      "epoch": 1.3434760559298704,
+      "grad_norm": 0.22288426756858826,
+      "learning_rate": 4.333908045977011e-05,
+      "loss": 0.1251,
+      "step": 1159
+    },
+    {
+      "epoch": 1.3446352242266175,
+      "grad_norm": 0.22571322321891785,
+      "learning_rate": 4.3333333333333334e-05,
+      "loss": 0.1167,
+      "step": 1160
+    },
+    {
+      "epoch": 1.3457943925233644,
+      "grad_norm": 0.25278669595718384,
+      "learning_rate": 4.3327586206896556e-05,
+      "loss": 0.1247,
+      "step": 1161
+    },
+    {
+      "epoch": 1.3469535608201115,
+      "grad_norm": 0.24313043057918549,
+      "learning_rate": 4.332183908045977e-05,
+      "loss": 0.132,
+      "step": 1162
+    },
+    {
+      "epoch": 1.3481127291168586,
+      "grad_norm": 0.26298508048057556,
+      "learning_rate": 4.331609195402299e-05,
+      "loss": 0.1262,
+      "step": 1163
+    },
+    {
+      "epoch": 1.3492718974136058,
+      "grad_norm": 0.24116156995296478,
+      "learning_rate": 4.331034482758621e-05,
+      "loss": 0.1313,
+      "step": 1164
+    },
+    {
+      "epoch": 1.3504310657103529,
+      "grad_norm": 0.24901200830936432,
+      "learning_rate": 4.330459770114943e-05,
+      "loss": 0.1097,
+      "step": 1165
+    },
+    {
+      "epoch": 1.3515902340071,
+      "grad_norm": 0.30485498905181885,
+      "learning_rate": 4.3298850574712644e-05,
+      "loss": 0.1178,
+      "step": 1166
+    },
+    {
+      "epoch": 1.3527494023038469,
+      "grad_norm": 0.3070286810398102,
+      "learning_rate": 4.3293103448275865e-05,
+      "loss": 0.1386,
+      "step": 1167
+    },
+    {
+      "epoch": 1.353908570600594,
+      "grad_norm": 0.27904579043388367,
+      "learning_rate": 4.328735632183909e-05,
+      "loss": 0.1504,
+      "step": 1168
+    },
+    {
+      "epoch": 1.3550677388973411,
+      "grad_norm": 0.28574293851852417,
+      "learning_rate": 4.32816091954023e-05,
+      "loss": 0.1273,
+      "step": 1169
+    },
+    {
+      "epoch": 1.3562269071940882,
+      "grad_norm": 0.27783069014549255,
+      "learning_rate": 4.327586206896552e-05,
+      "loss": 0.13,
+      "step": 1170
+    },
+    {
+      "epoch": 1.3573860754908353,
+      "grad_norm": 0.21755783259868622,
+      "learning_rate": 4.327011494252874e-05,
+      "loss": 0.1189,
+      "step": 1171
+    },
+    {
+      "epoch": 1.3585452437875825,
+      "grad_norm": 0.27805066108703613,
+      "learning_rate": 4.3264367816091954e-05,
+      "loss": 0.1295,
+      "step": 1172
+    },
+    {
+      "epoch": 1.3597044120843296,
+      "grad_norm": 0.26228439807891846,
+      "learning_rate": 4.3258620689655175e-05,
+      "loss": 0.1312,
+      "step": 1173
+    },
+    {
+      "epoch": 1.3608635803810767,
+      "grad_norm": 0.31947600841522217,
+      "learning_rate": 4.325287356321839e-05,
+      "loss": 0.1407,
+      "step": 1174
+    },
+    {
+      "epoch": 1.3620227486778236,
+      "grad_norm": 0.2713332176208496,
+      "learning_rate": 4.324712643678161e-05,
+      "loss": 0.1229,
+      "step": 1175
+    },
+    {
+      "epoch": 1.3631819169745707,
+      "grad_norm": 0.3652999699115753,
+      "learning_rate": 4.3241379310344833e-05,
+      "loss": 0.133,
+      "step": 1176
+    },
+    {
+      "epoch": 1.3643410852713178,
+      "grad_norm": 0.26640525460243225,
+      "learning_rate": 4.323563218390805e-05,
+      "loss": 0.1254,
+      "step": 1177
+    },
+    {
+      "epoch": 1.365500253568065,
+      "grad_norm": 0.24308764934539795,
+      "learning_rate": 4.322988505747126e-05,
+      "loss": 0.1311,
+      "step": 1178
+    },
+    {
+      "epoch": 1.366659421864812,
+      "grad_norm": 0.2529830038547516,
+      "learning_rate": 4.3224137931034485e-05,
+      "loss": 0.1257,
+      "step": 1179
+    },
+    {
+      "epoch": 1.3678185901615592,
+      "grad_norm": 0.2467876523733139,
+      "learning_rate": 4.32183908045977e-05,
+      "loss": 0.1387,
+      "step": 1180
+    },
+    {
+      "epoch": 1.368977758458306,
+      "grad_norm": 0.27905914187431335,
+      "learning_rate": 4.321264367816092e-05,
+      "loss": 0.119,
+      "step": 1181
+    },
+    {
+      "epoch": 1.3701369267550532,
+      "grad_norm": 0.2613096535205841,
+      "learning_rate": 4.320689655172414e-05,
+      "loss": 0.1392,
+      "step": 1182
+    },
+    {
+      "epoch": 1.3712960950518003,
+      "grad_norm": 0.3503858745098114,
+      "learning_rate": 4.320114942528736e-05,
+      "loss": 0.1457,
+      "step": 1183
+    },
+    {
+      "epoch": 1.3724552633485474,
+      "grad_norm": 0.26688286662101746,
+      "learning_rate": 4.319540229885058e-05,
+      "loss": 0.141,
+      "step": 1184
+    },
+    {
+      "epoch": 1.3736144316452945,
+      "grad_norm": 0.2871832251548767,
+      "learning_rate": 4.3189655172413795e-05,
+      "loss": 0.1323,
+      "step": 1185
+    },
+    {
+      "epoch": 1.3747735999420416,
+      "grad_norm": 0.27119460701942444,
+      "learning_rate": 4.318390804597701e-05,
+      "loss": 0.1207,
+      "step": 1186
+    },
+    {
+      "epoch": 1.3759327682387887,
+      "grad_norm": 0.23951269686222076,
+      "learning_rate": 4.317816091954023e-05,
+      "loss": 0.1135,
+      "step": 1187
+    },
+    {
+      "epoch": 1.3770919365355359,
+      "grad_norm": 0.221263587474823,
+      "learning_rate": 4.317241379310345e-05,
+      "loss": 0.1169,
+      "step": 1188
+    },
+    {
+      "epoch": 1.378251104832283,
+      "grad_norm": 0.2704179883003235,
+      "learning_rate": 4.316666666666667e-05,
+      "loss": 0.1341,
+      "step": 1189
+    },
+    {
+      "epoch": 1.3794102731290299,
+      "grad_norm": 0.2637876272201538,
+      "learning_rate": 4.316091954022989e-05,
+      "loss": 0.13,
+      "step": 1190
+    },
+    {
+      "epoch": 1.380569441425777,
+      "grad_norm": 0.2561126947402954,
+      "learning_rate": 4.3155172413793104e-05,
+      "loss": 0.128,
+      "step": 1191
+    },
+    {
+      "epoch": 1.381728609722524,
+      "grad_norm": 0.2545618712902069,
+      "learning_rate": 4.3149425287356326e-05,
+      "loss": 0.1245,
+      "step": 1192
+    },
+    {
+      "epoch": 1.3828877780192712,
+      "grad_norm": 0.20275923609733582,
+      "learning_rate": 4.314367816091954e-05,
+      "loss": 0.1168,
+      "step": 1193
+    },
+    {
+      "epoch": 1.3840469463160183,
+      "grad_norm": 0.2271161824464798,
+      "learning_rate": 4.3137931034482756e-05,
+      "loss": 0.1326,
+      "step": 1194
+    },
+    {
+      "epoch": 1.3852061146127652,
+      "grad_norm": 0.21496017277240753,
+      "learning_rate": 4.3132183908045984e-05,
+      "loss": 0.1248,
+      "step": 1195
+    },
+    {
+      "epoch": 1.3863652829095123,
+      "grad_norm": 0.2341107428073883,
+      "learning_rate": 4.31264367816092e-05,
+      "loss": 0.1268,
+      "step": 1196
+    },
+    {
+      "epoch": 1.3875244512062594,
+      "grad_norm": 0.2202032208442688,
+      "learning_rate": 4.3120689655172414e-05,
+      "loss": 0.1246,
+      "step": 1197
+    },
+    {
+      "epoch": 1.3886836195030066,
+      "grad_norm": 0.2526317834854126,
+      "learning_rate": 4.3114942528735636e-05,
+      "loss": 0.1276,
+      "step": 1198
+    },
+    {
+      "epoch": 1.3898427877997537,
+      "grad_norm": 0.29777175188064575,
+      "learning_rate": 4.310919540229885e-05,
+      "loss": 0.119,
+      "step": 1199
+    },
+    {
+      "epoch": 1.3910019560965008,
+      "grad_norm": 0.28763195872306824,
+      "learning_rate": 4.3103448275862066e-05,
+      "loss": 0.1363,
+      "step": 1200
+    },
+    {
+      "epoch": 1.392161124393248,
+      "grad_norm": 0.22191934287548065,
+      "learning_rate": 4.309770114942529e-05,
+      "loss": 0.1113,
+      "step": 1201
+    },
+    {
+      "epoch": 1.393320292689995,
+      "grad_norm": 0.36686575412750244,
+      "learning_rate": 4.309195402298851e-05,
+      "loss": 0.1451,
+      "step": 1202
+    },
+    {
+      "epoch": 1.3944794609867421,
+      "grad_norm": 0.2628101110458374,
+      "learning_rate": 4.308620689655173e-05,
+      "loss": 0.1255,
+      "step": 1203
+    },
+    {
+      "epoch": 1.395638629283489,
+      "grad_norm": 0.2633892893791199,
+      "learning_rate": 4.3080459770114946e-05,
+      "loss": 0.1391,
+      "step": 1204
+    },
+    {
+      "epoch": 1.3967977975802361,
+      "grad_norm": 0.28212958574295044,
+      "learning_rate": 4.307471264367816e-05,
+      "loss": 0.1383,
+      "step": 1205
+    },
+    {
+      "epoch": 1.3979569658769833,
+      "grad_norm": 0.24369271099567413,
+      "learning_rate": 4.306896551724138e-05,
+      "loss": 0.1227,
+      "step": 1206
+    },
+    {
+      "epoch": 1.3991161341737304,
+      "grad_norm": 0.23191668093204498,
+      "learning_rate": 4.30632183908046e-05,
+      "loss": 0.1255,
+      "step": 1207
+    },
+    {
+      "epoch": 1.4002753024704775,
+      "grad_norm": 0.2150292694568634,
+      "learning_rate": 4.305747126436782e-05,
+      "loss": 0.1172,
+      "step": 1208
+    },
+    {
+      "epoch": 1.4014344707672244,
+      "grad_norm": 0.2306421399116516,
+      "learning_rate": 4.305172413793104e-05,
+      "loss": 0.1213,
+      "step": 1209
+    },
+    {
+      "epoch": 1.4025936390639715,
+      "grad_norm": 0.22703979909420013,
+      "learning_rate": 4.3045977011494255e-05,
+      "loss": 0.1059,
+      "step": 1210
+    },
+    {
+      "epoch": 1.4037528073607186,
+      "grad_norm": 0.2609024941921234,
+      "learning_rate": 4.304022988505748e-05,
+      "loss": 0.1393,
+      "step": 1211
+    },
+    {
+      "epoch": 1.4049119756574657,
+      "grad_norm": 0.23271793127059937,
+      "learning_rate": 4.303448275862069e-05,
+      "loss": 0.1306,
+      "step": 1212
+    },
+    {
+      "epoch": 1.4060711439542128,
+      "grad_norm": 0.25955161452293396,
+      "learning_rate": 4.302873563218391e-05,
+      "loss": 0.1345,
+      "step": 1213
+    },
+    {
+      "epoch": 1.40723031225096,
+      "grad_norm": 0.20886264741420746,
+      "learning_rate": 4.302298850574713e-05,
+      "loss": 0.1255,
+      "step": 1214
+    },
+    {
+      "epoch": 1.408389480547707,
+      "grad_norm": 0.2347596287727356,
+      "learning_rate": 4.301724137931035e-05,
+      "loss": 0.1219,
+      "step": 1215
+    },
+    {
+      "epoch": 1.4095486488444542,
+      "grad_norm": 0.25307637453079224,
+      "learning_rate": 4.3011494252873565e-05,
+      "loss": 0.1122,
+      "step": 1216
+    },
+    {
+      "epoch": 1.4107078171412013,
+      "grad_norm": 0.28534573316574097,
+      "learning_rate": 4.300574712643679e-05,
+      "loss": 0.1356,
+      "step": 1217
+    },
+    {
+      "epoch": 1.4118669854379482,
+      "grad_norm": 0.2502347528934479,
+      "learning_rate": 4.3e-05,
+      "loss": 0.1256,
+      "step": 1218
+    },
+    {
+      "epoch": 1.4130261537346953,
+      "grad_norm": 0.31428301334381104,
+      "learning_rate": 4.2994252873563217e-05,
+      "loss": 0.1375,
+      "step": 1219
+    },
+    {
+      "epoch": 1.4141853220314424,
+      "grad_norm": 0.23887260258197784,
+      "learning_rate": 4.298850574712644e-05,
+      "loss": 0.127,
+      "step": 1220
+    },
+    {
+      "epoch": 1.4153444903281895,
+      "grad_norm": 0.2826530933380127,
+      "learning_rate": 4.298275862068965e-05,
+      "loss": 0.1431,
+      "step": 1221
+    },
+    {
+      "epoch": 1.4165036586249367,
+      "grad_norm": 0.22156329452991486,
+      "learning_rate": 4.2977011494252875e-05,
+      "loss": 0.1132,
+      "step": 1222
+    },
+    {
+      "epoch": 1.4176628269216838,
+      "grad_norm": 0.23711679875850677,
+      "learning_rate": 4.2971264367816096e-05,
+      "loss": 0.1234,
+      "step": 1223
+    },
+    {
+      "epoch": 1.4188219952184307,
+      "grad_norm": 0.21596437692642212,
+      "learning_rate": 4.296551724137931e-05,
+      "loss": 0.1195,
+      "step": 1224
+    },
+    {
+      "epoch": 1.4199811635151778,
+      "grad_norm": 0.26402637362480164,
+      "learning_rate": 4.295977011494253e-05,
+      "loss": 0.1268,
+      "step": 1225
+    },
+    {
+      "epoch": 1.421140331811925,
+      "grad_norm": 0.20864778757095337,
+      "learning_rate": 4.295402298850575e-05,
+      "loss": 0.1287,
+      "step": 1226
+    },
+    {
+      "epoch": 1.422299500108672,
+      "grad_norm": 0.23605617880821228,
+      "learning_rate": 4.294827586206896e-05,
+      "loss": 0.1432,
+      "step": 1227
+    },
+    {
+      "epoch": 1.4234586684054191,
+      "grad_norm": 0.2601798176765442,
+      "learning_rate": 4.2942528735632185e-05,
+      "loss": 0.1317,
+      "step": 1228
+    },
+    {
+      "epoch": 1.4246178367021662,
+      "grad_norm": 0.2347661405801773,
+      "learning_rate": 4.2936781609195406e-05,
+      "loss": 0.1233,
+      "step": 1229
+    },
+    {
+      "epoch": 1.4257770049989134,
+      "grad_norm": 0.22487115859985352,
+      "learning_rate": 4.293103448275863e-05,
+      "loss": 0.1314,
+      "step": 1230
+    },
+    {
+      "epoch": 1.4269361732956605,
+      "grad_norm": 0.2523593306541443,
+      "learning_rate": 4.292528735632184e-05,
+      "loss": 0.127,
+      "step": 1231
+    },
+    {
+      "epoch": 1.4280953415924076,
+      "grad_norm": 0.25972914695739746,
+      "learning_rate": 4.291954022988506e-05,
+      "loss": 0.1404,
+      "step": 1232
+    },
+    {
+      "epoch": 1.4292545098891545,
+      "grad_norm": 0.2787970006465912,
+      "learning_rate": 4.291379310344828e-05,
+      "loss": 0.1309,
+      "step": 1233
+    },
+    {
+      "epoch": 1.4304136781859016,
+      "grad_norm": 0.3272540271282196,
+      "learning_rate": 4.2908045977011494e-05,
+      "loss": 0.1271,
+      "step": 1234
+    },
+    {
+      "epoch": 1.4315728464826487,
+      "grad_norm": 0.302356094121933,
+      "learning_rate": 4.290229885057471e-05,
+      "loss": 0.1341,
+      "step": 1235
+    },
+    {
+      "epoch": 1.4327320147793958,
+      "grad_norm": 0.19032296538352966,
+      "learning_rate": 4.289655172413794e-05,
+      "loss": 0.1179,
+      "step": 1236
+    },
+    {
+      "epoch": 1.433891183076143,
+      "grad_norm": 0.28379547595977783,
+      "learning_rate": 4.289080459770115e-05,
+      "loss": 0.127,
+      "step": 1237
+    },
+    {
+      "epoch": 1.4350503513728898,
+      "grad_norm": 0.25689971446990967,
+      "learning_rate": 4.288505747126437e-05,
+      "loss": 0.1253,
+      "step": 1238
+    },
+    {
+      "epoch": 1.436209519669637,
+      "grad_norm": 0.235792875289917,
+      "learning_rate": 4.287931034482759e-05,
+      "loss": 0.1287,
+      "step": 1239
+    },
+    {
+      "epoch": 1.437368687966384,
+      "grad_norm": 0.24741043150424957,
+      "learning_rate": 4.2873563218390804e-05,
+      "loss": 0.1317,
+      "step": 1240
+    },
+    {
+      "epoch": 1.4385278562631312,
+      "grad_norm": 0.25098156929016113,
+      "learning_rate": 4.2867816091954026e-05,
+      "loss": 0.1291,
+      "step": 1241
+    },
+    {
+      "epoch": 1.4396870245598783,
+      "grad_norm": 0.26148924231529236,
+      "learning_rate": 4.286206896551724e-05,
+      "loss": 0.1356,
+      "step": 1242
+    },
+    {
+      "epoch": 1.4408461928566254,
+      "grad_norm": 0.27373388409614563,
+      "learning_rate": 4.285632183908046e-05,
+      "loss": 0.1457,
+      "step": 1243
+    },
+    {
+      "epoch": 1.4420053611533725,
+      "grad_norm": 0.23837333917617798,
+      "learning_rate": 4.2850574712643684e-05,
+      "loss": 0.132,
+      "step": 1244
+    },
+    {
+      "epoch": 1.4431645294501196,
+      "grad_norm": 0.2650034725666046,
+      "learning_rate": 4.28448275862069e-05,
+      "loss": 0.1317,
+      "step": 1245
+    },
+    {
+      "epoch": 1.4443236977468668,
+      "grad_norm": 0.2445426732301712,
+      "learning_rate": 4.2839080459770114e-05,
+      "loss": 0.1183,
+      "step": 1246
+    },
+    {
+      "epoch": 1.4454828660436136,
+      "grad_norm": 0.2151249349117279,
+      "learning_rate": 4.2833333333333335e-05,
+      "loss": 0.1185,
+      "step": 1247
+    },
+    {
+      "epoch": 1.4466420343403608,
+      "grad_norm": 0.2612224817276001,
+      "learning_rate": 4.282758620689655e-05,
+      "loss": 0.1225,
+      "step": 1248
+    },
+    {
+      "epoch": 1.4478012026371079,
+      "grad_norm": 0.2132655531167984,
+      "learning_rate": 4.282183908045977e-05,
+      "loss": 0.1083,
+      "step": 1249
+    },
+    {
+      "epoch": 1.448960370933855,
+      "grad_norm": 0.28365272283554077,
+      "learning_rate": 4.2816091954022994e-05,
+      "loss": 0.1269,
+      "step": 1250
+    },
+    {
+      "epoch": 1.450119539230602,
+      "grad_norm": 0.24323928356170654,
+      "learning_rate": 4.281034482758621e-05,
+      "loss": 0.138,
+      "step": 1251
+    },
+    {
+      "epoch": 1.451278707527349,
+      "grad_norm": 0.2997300624847412,
+      "learning_rate": 4.280459770114943e-05,
+      "loss": 0.1181,
+      "step": 1252
+    },
+    {
+      "epoch": 1.4524378758240961,
+      "grad_norm": 0.28692933917045593,
+      "learning_rate": 4.2798850574712645e-05,
+      "loss": 0.136,
+      "step": 1253
+    },
+    {
+      "epoch": 1.4535970441208432,
+      "grad_norm": 0.3969144821166992,
+      "learning_rate": 4.279310344827586e-05,
+      "loss": 0.1519,
+      "step": 1254
+    },
+    {
+      "epoch": 1.4547562124175903,
+      "grad_norm": 0.29358184337615967,
+      "learning_rate": 4.278735632183908e-05,
+      "loss": 0.1243,
+      "step": 1255
+    },
+    {
+      "epoch": 1.4559153807143375,
+      "grad_norm": 0.2625267803668976,
+      "learning_rate": 4.2781609195402303e-05,
+      "loss": 0.1228,
+      "step": 1256
+    },
+    {
+      "epoch": 1.4570745490110846,
+      "grad_norm": 0.24338027834892273,
+      "learning_rate": 4.2775862068965525e-05,
+      "loss": 0.1182,
+      "step": 1257
+    },
+    {
+      "epoch": 1.4582337173078317,
+      "grad_norm": 0.25080496072769165,
+      "learning_rate": 4.277011494252874e-05,
+      "loss": 0.1171,
+      "step": 1258
+    },
+    {
+      "epoch": 1.4593928856045788,
+      "grad_norm": 0.2200382798910141,
+      "learning_rate": 4.2764367816091955e-05,
+      "loss": 0.1089,
+      "step": 1259
+    },
+    {
+      "epoch": 1.460552053901326,
+      "grad_norm": 0.5557222962379456,
+      "learning_rate": 4.275862068965518e-05,
+      "loss": 0.1408,
+      "step": 1260
+    },
+    {
+      "epoch": 1.4617112221980728,
+      "grad_norm": 0.21948528289794922,
+      "learning_rate": 4.275287356321839e-05,
+      "loss": 0.1188,
+      "step": 1261
+    },
+    {
+      "epoch": 1.46287039049482,
+      "grad_norm": 0.2826273441314697,
+      "learning_rate": 4.2747126436781606e-05,
+      "loss": 0.1208,
+      "step": 1262
+    },
+    {
+      "epoch": 1.464029558791567,
+      "grad_norm": 0.3197879195213318,
+      "learning_rate": 4.274137931034483e-05,
+      "loss": 0.1428,
+      "step": 1263
+    },
+    {
+      "epoch": 1.4651887270883142,
+      "grad_norm": 0.26854708790779114,
+      "learning_rate": 4.273563218390805e-05,
+      "loss": 0.1354,
+      "step": 1264
+    },
+    {
+      "epoch": 1.4663478953850613,
+      "grad_norm": 0.20866204798221588,
+      "learning_rate": 4.2729885057471265e-05,
+      "loss": 0.1205,
+      "step": 1265
+    },
+    {
+      "epoch": 1.4675070636818082,
+      "grad_norm": 0.22703410685062408,
+      "learning_rate": 4.2724137931034486e-05,
+      "loss": 0.1165,
+      "step": 1266
+    },
+    {
+      "epoch": 1.4686662319785553,
+      "grad_norm": 0.2805153727531433,
+      "learning_rate": 4.27183908045977e-05,
+      "loss": 0.129,
+      "step": 1267
+    },
+    {
+      "epoch": 1.4698254002753024,
+      "grad_norm": 0.25746089220046997,
+      "learning_rate": 4.271264367816092e-05,
+      "loss": 0.113,
+      "step": 1268
+    },
+    {
+      "epoch": 1.4709845685720495,
+      "grad_norm": 0.24136294424533844,
+      "learning_rate": 4.270689655172414e-05,
+      "loss": 0.1215,
+      "step": 1269
+    },
+    {
+      "epoch": 1.4721437368687966,
+      "grad_norm": 0.21415215730667114,
+      "learning_rate": 4.270114942528736e-05,
+      "loss": 0.113,
+      "step": 1270
+    },
+    {
+      "epoch": 1.4733029051655437,
+      "grad_norm": 0.23764324188232422,
+      "learning_rate": 4.269540229885058e-05,
+      "loss": 0.1283,
+      "step": 1271
+    },
+    {
+      "epoch": 1.4744620734622909,
+      "grad_norm": 0.3138484060764313,
+      "learning_rate": 4.2689655172413796e-05,
+      "loss": 0.1294,
+      "step": 1272
+    },
+    {
+      "epoch": 1.475621241759038,
+      "grad_norm": 0.29593080282211304,
+      "learning_rate": 4.268390804597701e-05,
+      "loss": 0.1109,
+      "step": 1273
+    },
+    {
+      "epoch": 1.476780410055785,
+      "grad_norm": 0.19137899577617645,
+      "learning_rate": 4.267816091954023e-05,
+      "loss": 0.1062,
+      "step": 1274
+    },
+    {
+      "epoch": 1.477939578352532,
+      "grad_norm": 0.2660485506057739,
+      "learning_rate": 4.267241379310345e-05,
+      "loss": 0.1331,
+      "step": 1275
+    },
+    {
+      "epoch": 1.479098746649279,
+      "grad_norm": 0.2462538778781891,
+      "learning_rate": 4.266666666666667e-05,
+      "loss": 0.1241,
+      "step": 1276
+    },
+    {
+      "epoch": 1.4802579149460262,
+      "grad_norm": 0.28121456503868103,
+      "learning_rate": 4.266091954022989e-05,
+      "loss": 0.1301,
+      "step": 1277
+    },
+    {
+      "epoch": 1.4814170832427733,
+      "grad_norm": 0.2737043797969818,
+      "learning_rate": 4.2655172413793106e-05,
+      "loss": 0.139,
+      "step": 1278
+    },
+    {
+      "epoch": 1.4825762515395204,
+      "grad_norm": 0.271465539932251,
+      "learning_rate": 4.264942528735633e-05,
+      "loss": 0.1272,
+      "step": 1279
+    },
+    {
+      "epoch": 1.4837354198362676,
+      "grad_norm": 0.35305821895599365,
+      "learning_rate": 4.264367816091954e-05,
+      "loss": 0.13,
+      "step": 1280
+    },
+    {
+      "epoch": 1.4848945881330144,
+      "grad_norm": 0.26508283615112305,
+      "learning_rate": 4.263793103448276e-05,
+      "loss": 0.1277,
+      "step": 1281
+    },
+    {
+      "epoch": 1.4860537564297616,
+      "grad_norm": 0.2650332450866699,
+      "learning_rate": 4.263218390804598e-05,
+      "loss": 0.1276,
+      "step": 1282
+    },
+    {
+      "epoch": 1.4872129247265087,
+      "grad_norm": 0.234884575009346,
+      "learning_rate": 4.2626436781609194e-05,
+      "loss": 0.1385,
+      "step": 1283
+    },
+    {
+      "epoch": 1.4883720930232558,
+      "grad_norm": 0.26508384943008423,
+      "learning_rate": 4.2620689655172416e-05,
+      "loss": 0.1308,
+      "step": 1284
+    },
+    {
+      "epoch": 1.489531261320003,
+      "grad_norm": 0.2862217426300049,
+      "learning_rate": 4.261494252873564e-05,
+      "loss": 0.1252,
+      "step": 1285
+    },
+    {
+      "epoch": 1.49069042961675,
+      "grad_norm": 0.2019922137260437,
+      "learning_rate": 4.260919540229885e-05,
+      "loss": 0.1138,
+      "step": 1286
+    },
+    {
+      "epoch": 1.4918495979134971,
+      "grad_norm": 0.278028279542923,
+      "learning_rate": 4.2603448275862074e-05,
+      "loss": 0.1363,
+      "step": 1287
+    },
+    {
+      "epoch": 1.4930087662102443,
+      "grad_norm": 0.29411014914512634,
+      "learning_rate": 4.259770114942529e-05,
+      "loss": 0.1316,
+      "step": 1288
+    },
+    {
+      "epoch": 1.4941679345069914,
+      "grad_norm": 0.22040338814258575,
+      "learning_rate": 4.2591954022988504e-05,
+      "loss": 0.1376,
+      "step": 1289
+    },
+    {
+      "epoch": 1.4953271028037383,
+      "grad_norm": 0.2261519730091095,
+      "learning_rate": 4.2586206896551725e-05,
+      "loss": 0.1237,
+      "step": 1290
+    },
+    {
+      "epoch": 1.4964862711004854,
+      "grad_norm": 0.3036960959434509,
+      "learning_rate": 4.258045977011495e-05,
+      "loss": 0.126,
+      "step": 1291
+    },
+    {
+      "epoch": 1.4976454393972325,
+      "grad_norm": 0.2828342020511627,
+      "learning_rate": 4.257471264367816e-05,
+      "loss": 0.1441,
+      "step": 1292
+    },
+    {
+      "epoch": 1.4988046076939796,
+      "grad_norm": 0.2660335302352905,
+      "learning_rate": 4.2568965517241384e-05,
+      "loss": 0.1235,
+      "step": 1293
+    },
+    {
+      "epoch": 1.4999637759907267,
+      "grad_norm": 0.22510606050491333,
+      "learning_rate": 4.25632183908046e-05,
+      "loss": 0.1087,
+      "step": 1294
+    },
+    {
+      "epoch": 1.5011229442874736,
+      "grad_norm": 0.2649793028831482,
+      "learning_rate": 4.255747126436782e-05,
+      "loss": 0.1258,
+      "step": 1295
+    },
+    {
+      "epoch": 1.5022821125842207,
+      "grad_norm": 0.2566182315349579,
+      "learning_rate": 4.2551724137931035e-05,
+      "loss": 0.1272,
+      "step": 1296
+    },
+    {
+      "epoch": 1.5034412808809678,
+      "grad_norm": 0.2644808888435364,
+      "learning_rate": 4.254597701149426e-05,
+      "loss": 0.1268,
+      "step": 1297
+    },
+    {
+      "epoch": 1.504600449177715,
+      "grad_norm": 0.32291561365127563,
+      "learning_rate": 4.254022988505748e-05,
+      "loss": 0.1329,
+      "step": 1298
+    },
+    {
+      "epoch": 1.505759617474462,
+      "grad_norm": 0.34708988666534424,
+      "learning_rate": 4.253448275862069e-05,
+      "loss": 0.1569,
+      "step": 1299
+    },
+    {
+      "epoch": 1.5069187857712092,
+      "grad_norm": 0.23418490588665009,
+      "learning_rate": 4.252873563218391e-05,
+      "loss": 0.1303,
+      "step": 1300
+    },
+    {
+      "epoch": 1.5080779540679563,
+      "grad_norm": 0.35759875178337097,
+      "learning_rate": 4.252298850574713e-05,
+      "loss": 0.1226,
+      "step": 1301
+    },
+    {
+      "epoch": 1.5092371223647034,
+      "grad_norm": 0.255636066198349,
+      "learning_rate": 4.2517241379310345e-05,
+      "loss": 0.12,
+      "step": 1302
+    },
+    {
+      "epoch": 1.5103962906614505,
+      "grad_norm": 0.2545788586139679,
+      "learning_rate": 4.251149425287356e-05,
+      "loss": 0.1326,
+      "step": 1303
+    },
+    {
+      "epoch": 1.5115554589581977,
+      "grad_norm": 0.2811221480369568,
+      "learning_rate": 4.250574712643678e-05,
+      "loss": 0.1441,
+      "step": 1304
+    },
+    {
+      "epoch": 1.5127146272549445,
+      "grad_norm": 0.2685137987136841,
+      "learning_rate": 4.25e-05,
+      "loss": 0.1367,
+      "step": 1305
+    },
+    {
+      "epoch": 1.5138737955516917,
+      "grad_norm": 0.26587024331092834,
+      "learning_rate": 4.2494252873563225e-05,
+      "loss": 0.1196,
+      "step": 1306
+    },
+    {
+      "epoch": 1.5150329638484388,
+      "grad_norm": 0.296543151140213,
+      "learning_rate": 4.248850574712644e-05,
+      "loss": 0.1502,
+      "step": 1307
+    },
+    {
+      "epoch": 1.5161921321451857,
+      "grad_norm": 0.2668646275997162,
+      "learning_rate": 4.2482758620689655e-05,
+      "loss": 0.1351,
+      "step": 1308
+    },
+    {
+      "epoch": 1.5173513004419328,
+      "grad_norm": 0.2467673420906067,
+      "learning_rate": 4.2477011494252876e-05,
+      "loss": 0.1281,
+      "step": 1309
+    },
+    {
+      "epoch": 1.51851046873868,
+      "grad_norm": 0.25315842032432556,
+      "learning_rate": 4.247126436781609e-05,
+      "loss": 0.1286,
+      "step": 1310
+    },
+    {
+      "epoch": 1.519669637035427,
+      "grad_norm": 0.25049182772636414,
+      "learning_rate": 4.246551724137931e-05,
+      "loss": 0.1356,
+      "step": 1311
+    },
+    {
+      "epoch": 1.5208288053321741,
+      "grad_norm": 0.23967556655406952,
+      "learning_rate": 4.2459770114942534e-05,
+      "loss": 0.1135,
+      "step": 1312
+    },
+    {
+      "epoch": 1.5219879736289212,
+      "grad_norm": 0.2868068218231201,
+      "learning_rate": 4.245402298850575e-05,
+      "loss": 0.1309,
+      "step": 1313
+    },
+    {
+      "epoch": 1.5231471419256684,
+      "grad_norm": 0.21902504563331604,
+      "learning_rate": 4.244827586206897e-05,
+      "loss": 0.1156,
+      "step": 1314
+    },
+    {
+      "epoch": 1.5243063102224155,
+      "grad_norm": 0.24290820956230164,
+      "learning_rate": 4.2442528735632186e-05,
+      "loss": 0.1296,
+      "step": 1315
+    },
+    {
+      "epoch": 1.5254654785191626,
+      "grad_norm": 0.21558572351932526,
+      "learning_rate": 4.24367816091954e-05,
+      "loss": 0.1213,
+      "step": 1316
+    },
+    {
+      "epoch": 1.5266246468159097,
+      "grad_norm": 0.28430676460266113,
+      "learning_rate": 4.243103448275862e-05,
+      "loss": 0.1275,
+      "step": 1317
+    },
+    {
+      "epoch": 1.5277838151126568,
+      "grad_norm": 0.28674018383026123,
+      "learning_rate": 4.2425287356321844e-05,
+      "loss": 0.1315,
+      "step": 1318
+    },
+    {
+      "epoch": 1.5289429834094037,
+      "grad_norm": 0.31648436188697815,
+      "learning_rate": 4.241954022988506e-05,
+      "loss": 0.1457,
+      "step": 1319
+    },
+    {
+      "epoch": 1.5301021517061508,
+      "grad_norm": 0.2692152261734009,
+      "learning_rate": 4.241379310344828e-05,
+      "loss": 0.1261,
+      "step": 1320
+    },
+    {
+      "epoch": 1.531261320002898,
+      "grad_norm": 0.21987690031528473,
+      "learning_rate": 4.2408045977011496e-05,
+      "loss": 0.1224,
+      "step": 1321
+    },
+    {
+      "epoch": 1.532420488299645,
+      "grad_norm": 0.32818254828453064,
+      "learning_rate": 4.240229885057471e-05,
+      "loss": 0.128,
+      "step": 1322
+    },
+    {
+      "epoch": 1.533579656596392,
+      "grad_norm": 0.2926555275917053,
+      "learning_rate": 4.239655172413793e-05,
+      "loss": 0.1425,
+      "step": 1323
+    },
+    {
+      "epoch": 1.534738824893139,
+      "grad_norm": 0.2183428853750229,
+      "learning_rate": 4.239080459770115e-05,
+      "loss": 0.121,
+      "step": 1324
+    },
+    {
+      "epoch": 1.5358979931898862,
+      "grad_norm": 0.26924946904182434,
+      "learning_rate": 4.238505747126437e-05,
+      "loss": 0.1474,
+      "step": 1325
+    },
+    {
+      "epoch": 1.5370571614866333,
+      "grad_norm": 0.1868240386247635,
+      "learning_rate": 4.237931034482759e-05,
+      "loss": 0.1293,
+      "step": 1326
+    },
+    {
+      "epoch": 1.5382163297833804,
+      "grad_norm": 0.20834980905056,
+      "learning_rate": 4.2373563218390805e-05,
+      "loss": 0.1159,
+      "step": 1327
+    },
+    {
+      "epoch": 1.5393754980801275,
+      "grad_norm": 0.2959633767604828,
+      "learning_rate": 4.236781609195403e-05,
+      "loss": 0.1421,
+      "step": 1328
+    },
+    {
+      "epoch": 1.5405346663768746,
+      "grad_norm": 0.222950279712677,
+      "learning_rate": 4.236206896551724e-05,
+      "loss": 0.1279,
+      "step": 1329
+    },
+    {
+      "epoch": 1.5416938346736218,
+      "grad_norm": 0.2185351401567459,
+      "learning_rate": 4.235632183908046e-05,
+      "loss": 0.1247,
+      "step": 1330
+    },
+    {
+      "epoch": 1.5428530029703689,
+      "grad_norm": 0.16755683720111847,
+      "learning_rate": 4.235057471264368e-05,
+      "loss": 0.101,
+      "step": 1331
+    },
+    {
+      "epoch": 1.544012171267116,
+      "grad_norm": 0.2677669823169708,
+      "learning_rate": 4.23448275862069e-05,
+      "loss": 0.1339,
+      "step": 1332
+    },
+    {
+      "epoch": 1.5451713395638629,
+      "grad_norm": 0.252986341714859,
+      "learning_rate": 4.233908045977012e-05,
+      "loss": 0.1202,
+      "step": 1333
+    },
+    {
+      "epoch": 1.54633050786061,
+      "grad_norm": 0.2726978659629822,
+      "learning_rate": 4.233333333333334e-05,
+      "loss": 0.1274,
+      "step": 1334
+    },
+    {
+      "epoch": 1.547489676157357,
+      "grad_norm": 0.24727889895439148,
+      "learning_rate": 4.232758620689655e-05,
+      "loss": 0.1292,
+      "step": 1335
+    },
+    {
+      "epoch": 1.5486488444541042,
+      "grad_norm": 0.24011863768100739,
+      "learning_rate": 4.2321839080459773e-05,
+      "loss": 0.1145,
+      "step": 1336
+    },
+    {
+      "epoch": 1.5498080127508511,
+      "grad_norm": 0.26071444153785706,
+      "learning_rate": 4.231609195402299e-05,
+      "loss": 0.1208,
+      "step": 1337
+    },
+    {
+      "epoch": 1.5509671810475982,
+      "grad_norm": 0.2095252424478531,
+      "learning_rate": 4.231034482758621e-05,
+      "loss": 0.1164,
+      "step": 1338
+    },
+    {
+      "epoch": 1.5521263493443453,
+      "grad_norm": 0.3264429569244385,
+      "learning_rate": 4.230459770114943e-05,
+      "loss": 0.1344,
+      "step": 1339
+    },
+    {
+      "epoch": 1.5532855176410925,
+      "grad_norm": 0.24655984342098236,
+      "learning_rate": 4.2298850574712647e-05,
+      "loss": 0.124,
+      "step": 1340
+    },
+    {
+      "epoch": 1.5544446859378396,
+      "grad_norm": 0.28563857078552246,
+      "learning_rate": 4.229310344827586e-05,
+      "loss": 0.1312,
+      "step": 1341
+    },
+    {
+      "epoch": 1.5556038542345867,
+      "grad_norm": 0.2881647050380707,
+      "learning_rate": 4.228735632183908e-05,
+      "loss": 0.1418,
+      "step": 1342
+    },
+    {
+      "epoch": 1.5567630225313338,
+      "grad_norm": 0.2571274936199188,
+      "learning_rate": 4.22816091954023e-05,
+      "loss": 0.1213,
+      "step": 1343
+    },
+    {
+      "epoch": 1.557922190828081,
+      "grad_norm": 0.26905322074890137,
+      "learning_rate": 4.227586206896552e-05,
+      "loss": 0.1294,
+      "step": 1344
+    },
+    {
+      "epoch": 1.559081359124828,
+      "grad_norm": 0.3038524389266968,
+      "learning_rate": 4.2270114942528735e-05,
+      "loss": 0.1398,
+      "step": 1345
+    },
+    {
+      "epoch": 1.5602405274215752,
+      "grad_norm": 0.31705963611602783,
+      "learning_rate": 4.2264367816091956e-05,
+      "loss": 0.1284,
+      "step": 1346
+    },
+    {
+      "epoch": 1.5613996957183223,
+      "grad_norm": 0.24810083210468292,
+      "learning_rate": 4.225862068965518e-05,
+      "loss": 0.126,
+      "step": 1347
+    },
+    {
+      "epoch": 1.5625588640150692,
+      "grad_norm": 0.2836195230484009,
+      "learning_rate": 4.225287356321839e-05,
+      "loss": 0.1318,
+      "step": 1348
+    },
+    {
+      "epoch": 1.5637180323118163,
+      "grad_norm": 0.26509571075439453,
+      "learning_rate": 4.224712643678161e-05,
+      "loss": 0.1279,
+      "step": 1349
+    },
+    {
+      "epoch": 1.5648772006085634,
+      "grad_norm": 0.24432741105556488,
+      "learning_rate": 4.224137931034483e-05,
+      "loss": 0.1238,
+      "step": 1350
+    },
+    {
+      "epoch": 1.5660363689053103,
+      "grad_norm": 0.25199276208877563,
+      "learning_rate": 4.2235632183908044e-05,
+      "loss": 0.1153,
+      "step": 1351
+    },
+    {
+      "epoch": 1.5671955372020574,
+      "grad_norm": 0.2422485649585724,
+      "learning_rate": 4.2229885057471266e-05,
+      "loss": 0.1222,
+      "step": 1352
+    },
+    {
+      "epoch": 1.5683547054988045,
+      "grad_norm": 0.28373682498931885,
+      "learning_rate": 4.222413793103449e-05,
+      "loss": 0.143,
+      "step": 1353
+    },
+    {
+      "epoch": 1.5695138737955516,
+      "grad_norm": 0.323743999004364,
+      "learning_rate": 4.22183908045977e-05,
+      "loss": 0.1185,
+      "step": 1354
+    },
+    {
+      "epoch": 1.5706730420922987,
+      "grad_norm": 0.27437567710876465,
+      "learning_rate": 4.2212643678160924e-05,
+      "loss": 0.1209,
+      "step": 1355
+    },
+    {
+      "epoch": 1.5718322103890459,
+      "grad_norm": 0.20446433126926422,
+      "learning_rate": 4.220689655172414e-05,
+      "loss": 0.1164,
+      "step": 1356
+    },
+    {
+      "epoch": 1.572991378685793,
+      "grad_norm": 0.21873438358306885,
+      "learning_rate": 4.2201149425287354e-05,
+      "loss": 0.1209,
+      "step": 1357
+    },
+    {
+      "epoch": 1.57415054698254,
+      "grad_norm": 0.2323230654001236,
+      "learning_rate": 4.2195402298850576e-05,
+      "loss": 0.1233,
+      "step": 1358
+    },
+    {
+      "epoch": 1.5753097152792872,
+      "grad_norm": 0.2679866552352905,
+      "learning_rate": 4.21896551724138e-05,
+      "loss": 0.1219,
+      "step": 1359
+    },
+    {
+      "epoch": 1.5764688835760343,
+      "grad_norm": 0.29342472553253174,
+      "learning_rate": 4.218390804597701e-05,
+      "loss": 0.1449,
+      "step": 1360
+    },
+    {
+      "epoch": 1.5776280518727814,
+      "grad_norm": 0.29322659969329834,
+      "learning_rate": 4.2178160919540234e-05,
+      "loss": 0.1494,
+      "step": 1361
+    },
+    {
+      "epoch": 1.5787872201695283,
+      "grad_norm": 0.23010887205600739,
+      "learning_rate": 4.217241379310345e-05,
+      "loss": 0.1304,
+      "step": 1362
+    },
+    {
+      "epoch": 1.5799463884662754,
+      "grad_norm": 0.20732039213180542,
+      "learning_rate": 4.216666666666667e-05,
+      "loss": 0.113,
+      "step": 1363
+    },
+    {
+      "epoch": 1.5811055567630226,
+      "grad_norm": 0.21723432838916779,
+      "learning_rate": 4.2160919540229886e-05,
+      "loss": 0.1193,
+      "step": 1364
+    },
+    {
+      "epoch": 1.5822647250597695,
+      "grad_norm": 0.268611878156662,
+      "learning_rate": 4.21551724137931e-05,
+      "loss": 0.1447,
+      "step": 1365
+    },
+    {
+      "epoch": 1.5834238933565166,
+      "grad_norm": 0.2262781858444214,
+      "learning_rate": 4.214942528735633e-05,
+      "loss": 0.1188,
+      "step": 1366
+    },
+    {
+      "epoch": 1.5845830616532637,
+      "grad_norm": 0.21195755898952484,
+      "learning_rate": 4.2143678160919544e-05,
+      "loss": 0.1156,
+      "step": 1367
+    },
+    {
+      "epoch": 1.5857422299500108,
+      "grad_norm": 0.26163250207901,
+      "learning_rate": 4.213793103448276e-05,
+      "loss": 0.1348,
+      "step": 1368
+    },
+    {
+      "epoch": 1.586901398246758,
+      "grad_norm": 0.3384718894958496,
+      "learning_rate": 4.213218390804598e-05,
+      "loss": 0.1473,
+      "step": 1369
+    },
+    {
+      "epoch": 1.588060566543505,
+      "grad_norm": 0.26248493790626526,
+      "learning_rate": 4.2126436781609195e-05,
+      "loss": 0.1212,
+      "step": 1370
+    },
+    {
+      "epoch": 1.5892197348402521,
+      "grad_norm": 0.21204310655593872,
+      "learning_rate": 4.212068965517242e-05,
+      "loss": 0.1131,
+      "step": 1371
+    },
+    {
+      "epoch": 1.5903789031369993,
+      "grad_norm": 0.2920737862586975,
+      "learning_rate": 4.211494252873563e-05,
+      "loss": 0.134,
+      "step": 1372
+    },
+    {
+      "epoch": 1.5915380714337464,
+      "grad_norm": 0.2683486044406891,
+      "learning_rate": 4.2109195402298854e-05,
+      "loss": 0.1297,
+      "step": 1373
+    },
+    {
+      "epoch": 1.5926972397304935,
+      "grad_norm": 0.25629928708076477,
+      "learning_rate": 4.2103448275862075e-05,
+      "loss": 0.1364,
+      "step": 1374
+    },
+    {
+      "epoch": 1.5938564080272406,
+      "grad_norm": 0.22611750662326813,
+      "learning_rate": 4.209770114942529e-05,
+      "loss": 0.1189,
+      "step": 1375
+    },
+    {
+      "epoch": 1.5950155763239875,
+      "grad_norm": 0.2171134501695633,
+      "learning_rate": 4.2091954022988505e-05,
+      "loss": 0.1224,
+      "step": 1376
+    },
+    {
+      "epoch": 1.5961747446207346,
+      "grad_norm": 0.3188816010951996,
+      "learning_rate": 4.208620689655173e-05,
+      "loss": 0.1325,
+      "step": 1377
+    },
+    {
+      "epoch": 1.5973339129174817,
+      "grad_norm": 0.25183406472206116,
+      "learning_rate": 4.208045977011494e-05,
+      "loss": 0.116,
+      "step": 1378
+    },
+    {
+      "epoch": 1.5984930812142288,
+      "grad_norm": 0.274368017911911,
+      "learning_rate": 4.207471264367816e-05,
+      "loss": 0.1388,
+      "step": 1379
+    },
+    {
+      "epoch": 1.5996522495109757,
+      "grad_norm": 0.2406664341688156,
+      "learning_rate": 4.2068965517241385e-05,
+      "loss": 0.1225,
+      "step": 1380
+    },
+    {
+      "epoch": 1.6008114178077228,
+      "grad_norm": 0.22044144570827484,
+      "learning_rate": 4.20632183908046e-05,
+      "loss": 0.1212,
+      "step": 1381
+    },
+    {
+      "epoch": 1.60197058610447,
+      "grad_norm": 0.2523006200790405,
+      "learning_rate": 4.205747126436782e-05,
+      "loss": 0.1142,
+      "step": 1382
+    },
+    {
+      "epoch": 1.603129754401217,
+      "grad_norm": 0.26407095789909363,
+      "learning_rate": 4.2051724137931036e-05,
+      "loss": 0.1252,
+      "step": 1383
+    },
+    {
+      "epoch": 1.6042889226979642,
+      "grad_norm": 0.22272075712680817,
+      "learning_rate": 4.204597701149425e-05,
+      "loss": 0.1177,
+      "step": 1384
+    },
+    {
+      "epoch": 1.6054480909947113,
+      "grad_norm": 0.2437649518251419,
+      "learning_rate": 4.204022988505747e-05,
+      "loss": 0.1238,
+      "step": 1385
+    },
+    {
+      "epoch": 1.6066072592914584,
+      "grad_norm": 0.29245513677597046,
+      "learning_rate": 4.203448275862069e-05,
+      "loss": 0.1389,
+      "step": 1386
+    },
+    {
+      "epoch": 1.6077664275882055,
+      "grad_norm": 0.25167253613471985,
+      "learning_rate": 4.202873563218391e-05,
+      "loss": 0.131,
+      "step": 1387
+    },
+    {
+      "epoch": 1.6089255958849527,
+      "grad_norm": 0.2610798478126526,
+      "learning_rate": 4.202298850574713e-05,
+      "loss": 0.1415,
+      "step": 1388
+    },
+    {
+      "epoch": 1.6100847641816998,
+      "grad_norm": 0.21162807941436768,
+      "learning_rate": 4.2017241379310346e-05,
+      "loss": 0.1296,
+      "step": 1389
+    },
+    {
+      "epoch": 1.6112439324784467,
+      "grad_norm": 0.26718345284461975,
+      "learning_rate": 4.201149425287357e-05,
+      "loss": 0.1346,
+      "step": 1390
+    },
+    {
+      "epoch": 1.6124031007751938,
+      "grad_norm": 0.2872393727302551,
+      "learning_rate": 4.200574712643678e-05,
+      "loss": 0.1452,
+      "step": 1391
+    },
+    {
+      "epoch": 1.613562269071941,
+      "grad_norm": 0.2846381366252899,
+      "learning_rate": 4.2e-05,
+      "loss": 0.1281,
+      "step": 1392
+    },
+    {
+      "epoch": 1.614721437368688,
+      "grad_norm": 0.2739090919494629,
+      "learning_rate": 4.199425287356322e-05,
+      "loss": 0.1248,
+      "step": 1393
+    },
+    {
+      "epoch": 1.615880605665435,
+      "grad_norm": 0.24870949983596802,
+      "learning_rate": 4.198850574712644e-05,
+      "loss": 0.1377,
+      "step": 1394
+    },
+    {
+      "epoch": 1.617039773962182,
+      "grad_norm": 0.26134321093559265,
+      "learning_rate": 4.1982758620689656e-05,
+      "loss": 0.1348,
+      "step": 1395
+    },
+    {
+      "epoch": 1.6181989422589291,
+      "grad_norm": 0.2896818220615387,
+      "learning_rate": 4.197701149425288e-05,
+      "loss": 0.1376,
+      "step": 1396
+    },
+    {
+      "epoch": 1.6193581105556762,
+      "grad_norm": 0.24523970484733582,
+      "learning_rate": 4.197126436781609e-05,
+      "loss": 0.1289,
+      "step": 1397
+    },
+    {
+      "epoch": 1.6205172788524234,
+      "grad_norm": 0.2446785867214203,
+      "learning_rate": 4.196551724137931e-05,
+      "loss": 0.1219,
+      "step": 1398
+    },
+    {
+      "epoch": 1.6216764471491705,
+      "grad_norm": 0.2920502722263336,
+      "learning_rate": 4.195977011494253e-05,
+      "loss": 0.1609,
+      "step": 1399
+    },
+    {
+      "epoch": 1.6228356154459176,
+      "grad_norm": 0.24321280419826508,
+      "learning_rate": 4.195402298850575e-05,
+      "loss": 0.1217,
+      "step": 1400
+    },
+    {
+      "epoch": 1.6239947837426647,
+      "grad_norm": 0.20927608013153076,
+      "learning_rate": 4.194827586206897e-05,
+      "loss": 0.1071,
+      "step": 1401
+    },
+    {
+      "epoch": 1.6251539520394118,
+      "grad_norm": 0.2517341077327728,
+      "learning_rate": 4.194252873563219e-05,
+      "loss": 0.1359,
+      "step": 1402
+    },
+    {
+      "epoch": 1.626313120336159,
+      "grad_norm": 0.2402840554714203,
+      "learning_rate": 4.19367816091954e-05,
+      "loss": 0.1352,
+      "step": 1403
+    },
+    {
+      "epoch": 1.627472288632906,
+      "grad_norm": 0.22628480195999146,
+      "learning_rate": 4.1931034482758624e-05,
+      "loss": 0.1271,
+      "step": 1404
+    },
+    {
+      "epoch": 1.628631456929653,
+      "grad_norm": 0.24545376002788544,
+      "learning_rate": 4.192528735632184e-05,
+      "loss": 0.1338,
+      "step": 1405
+    },
+    {
+      "epoch": 1.6297906252264,
+      "grad_norm": 0.2462320774793625,
+      "learning_rate": 4.1919540229885054e-05,
+      "loss": 0.1375,
+      "step": 1406
+    },
+    {
+      "epoch": 1.6309497935231472,
+      "grad_norm": 0.2645837068557739,
+      "learning_rate": 4.191379310344828e-05,
+      "loss": 0.1327,
+      "step": 1407
+    },
+    {
+      "epoch": 1.632108961819894,
+      "grad_norm": 0.2443280965089798,
+      "learning_rate": 4.19080459770115e-05,
+      "loss": 0.1211,
+      "step": 1408
+    },
+    {
+      "epoch": 1.6332681301166412,
+      "grad_norm": 0.260450154542923,
+      "learning_rate": 4.190229885057472e-05,
+      "loss": 0.1318,
+      "step": 1409
+    },
+    {
+      "epoch": 1.6344272984133883,
+      "grad_norm": 0.22310788929462433,
+      "learning_rate": 4.1896551724137934e-05,
+      "loss": 0.123,
+      "step": 1410
+    },
+    {
+      "epoch": 1.6355864667101354,
+      "grad_norm": 0.24180607497692108,
+      "learning_rate": 4.189080459770115e-05,
+      "loss": 0.1214,
+      "step": 1411
+    },
+    {
+      "epoch": 1.6367456350068825,
+      "grad_norm": 0.33246803283691406,
+      "learning_rate": 4.188505747126437e-05,
+      "loss": 0.159,
+      "step": 1412
+    },
+    {
+      "epoch": 1.6379048033036296,
+      "grad_norm": 0.2154608517885208,
+      "learning_rate": 4.1879310344827585e-05,
+      "loss": 0.1242,
+      "step": 1413
+    },
+    {
+      "epoch": 1.6390639716003768,
+      "grad_norm": 0.24036885797977448,
+      "learning_rate": 4.187356321839081e-05,
+      "loss": 0.1097,
+      "step": 1414
+    },
+    {
+      "epoch": 1.6402231398971239,
+      "grad_norm": 0.21203725039958954,
+      "learning_rate": 4.186781609195403e-05,
+      "loss": 0.1232,
+      "step": 1415
+    },
+    {
+      "epoch": 1.641382308193871,
+      "grad_norm": 0.24248747527599335,
+      "learning_rate": 4.1862068965517243e-05,
+      "loss": 0.1223,
+      "step": 1416
+    },
+    {
+      "epoch": 1.642541476490618,
+      "grad_norm": 0.2965581715106964,
+      "learning_rate": 4.185632183908046e-05,
+      "loss": 0.1277,
+      "step": 1417
+    },
+    {
+      "epoch": 1.6437006447873652,
+      "grad_norm": 0.24066859483718872,
+      "learning_rate": 4.185057471264368e-05,
+      "loss": 0.1388,
+      "step": 1418
+    },
+    {
+      "epoch": 1.644859813084112,
+      "grad_norm": 0.23044906556606293,
+      "learning_rate": 4.1844827586206895e-05,
+      "loss": 0.1173,
+      "step": 1419
+    },
+    {
+      "epoch": 1.6460189813808592,
+      "grad_norm": 0.2829865515232086,
+      "learning_rate": 4.1839080459770117e-05,
+      "loss": 0.1264,
+      "step": 1420
+    },
+    {
+      "epoch": 1.6471781496776063,
+      "grad_norm": 0.37790974974632263,
+      "learning_rate": 4.183333333333334e-05,
+      "loss": 0.1158,
+      "step": 1421
+    },
+    {
+      "epoch": 1.6483373179743535,
+      "grad_norm": 0.23219698667526245,
+      "learning_rate": 4.182758620689655e-05,
+      "loss": 0.1293,
+      "step": 1422
+    },
+    {
+      "epoch": 1.6494964862711003,
+      "grad_norm": 0.26675254106521606,
+      "learning_rate": 4.1821839080459775e-05,
+      "loss": 0.1245,
+      "step": 1423
+    },
+    {
+      "epoch": 1.6506556545678475,
+      "grad_norm": 0.2584593594074249,
+      "learning_rate": 4.181609195402299e-05,
+      "loss": 0.1132,
+      "step": 1424
+    },
+    {
+      "epoch": 1.6518148228645946,
+      "grad_norm": 0.2532110810279846,
+      "learning_rate": 4.1810344827586205e-05,
+      "loss": 0.1265,
+      "step": 1425
+    },
+    {
+      "epoch": 1.6529739911613417,
+      "grad_norm": 0.27470529079437256,
+      "learning_rate": 4.1804597701149426e-05,
+      "loss": 0.1422,
+      "step": 1426
+    },
+    {
+      "epoch": 1.6541331594580888,
+      "grad_norm": 0.24903489649295807,
+      "learning_rate": 4.179885057471265e-05,
+      "loss": 0.1065,
+      "step": 1427
+    },
+    {
+      "epoch": 1.655292327754836,
+      "grad_norm": 0.29945898056030273,
+      "learning_rate": 4.179310344827587e-05,
+      "loss": 0.1418,
+      "step": 1428
+    },
+    {
+      "epoch": 1.656451496051583,
+      "grad_norm": 0.19737482070922852,
+      "learning_rate": 4.1787356321839085e-05,
+      "loss": 0.1162,
+      "step": 1429
+    },
+    {
+      "epoch": 1.6576106643483302,
+      "grad_norm": 0.31681379675865173,
+      "learning_rate": 4.17816091954023e-05,
+      "loss": 0.1521,
+      "step": 1430
+    },
+    {
+      "epoch": 1.6587698326450773,
+      "grad_norm": 0.23246780037879944,
+      "learning_rate": 4.177586206896552e-05,
+      "loss": 0.1156,
+      "step": 1431
+    },
+    {
+      "epoch": 1.6599290009418244,
+      "grad_norm": 0.2456778883934021,
+      "learning_rate": 4.1770114942528736e-05,
+      "loss": 0.1194,
+      "step": 1432
+    },
+    {
+      "epoch": 1.6610881692385713,
+      "grad_norm": 0.23620393872261047,
+      "learning_rate": 4.176436781609195e-05,
+      "loss": 0.1231,
+      "step": 1433
+    },
+    {
+      "epoch": 1.6622473375353184,
+      "grad_norm": 0.20471753180027008,
+      "learning_rate": 4.175862068965517e-05,
+      "loss": 0.109,
+      "step": 1434
+    },
+    {
+      "epoch": 1.6634065058320655,
+      "grad_norm": 0.2827279269695282,
+      "learning_rate": 4.1752873563218394e-05,
+      "loss": 0.1467,
+      "step": 1435
+    },
+    {
+      "epoch": 1.6645656741288126,
+      "grad_norm": 0.21254539489746094,
+      "learning_rate": 4.1747126436781616e-05,
+      "loss": 0.1182,
+      "step": 1436
+    },
+    {
+      "epoch": 1.6657248424255595,
+      "grad_norm": 0.24518734216690063,
+      "learning_rate": 4.174137931034483e-05,
+      "loss": 0.1326,
+      "step": 1437
+    },
+    {
+      "epoch": 1.6668840107223066,
+      "grad_norm": 0.2632474899291992,
+      "learning_rate": 4.1735632183908046e-05,
+      "loss": 0.1357,
+      "step": 1438
+    },
+    {
+      "epoch": 1.6680431790190537,
+      "grad_norm": 0.23681853711605072,
+      "learning_rate": 4.172988505747127e-05,
+      "loss": 0.1272,
+      "step": 1439
+    },
+    {
+      "epoch": 1.6692023473158009,
+      "grad_norm": 0.25653204321861267,
+      "learning_rate": 4.172413793103448e-05,
+      "loss": 0.1305,
+      "step": 1440
+    },
+    {
+      "epoch": 1.670361515612548,
+      "grad_norm": 0.22474606335163116,
+      "learning_rate": 4.1718390804597704e-05,
+      "loss": 0.1192,
+      "step": 1441
+    },
+    {
+      "epoch": 1.671520683909295,
+      "grad_norm": 0.2875736951828003,
+      "learning_rate": 4.1712643678160926e-05,
+      "loss": 0.1271,
+      "step": 1442
+    },
+    {
+      "epoch": 1.6726798522060422,
+      "grad_norm": 0.27800050377845764,
+      "learning_rate": 4.170689655172414e-05,
+      "loss": 0.1348,
+      "step": 1443
+    },
+    {
+      "epoch": 1.6738390205027893,
+      "grad_norm": 0.2516680955886841,
+      "learning_rate": 4.1701149425287356e-05,
+      "loss": 0.1128,
+      "step": 1444
+    },
+    {
+      "epoch": 1.6749981887995364,
+      "grad_norm": 0.3367115259170532,
+      "learning_rate": 4.169540229885058e-05,
+      "loss": 0.1362,
+      "step": 1445
+    },
+    {
+      "epoch": 1.6761573570962836,
+      "grad_norm": 0.3319508731365204,
+      "learning_rate": 4.168965517241379e-05,
+      "loss": 0.1358,
+      "step": 1446
+    },
+    {
+      "epoch": 1.6773165253930307,
+      "grad_norm": 0.2614979147911072,
+      "learning_rate": 4.1683908045977014e-05,
+      "loss": 0.1244,
+      "step": 1447
+    },
+    {
+      "epoch": 1.6784756936897776,
+      "grad_norm": 0.24296019971370697,
+      "learning_rate": 4.1678160919540235e-05,
+      "loss": 0.1223,
+      "step": 1448
+    },
+    {
+      "epoch": 1.6796348619865247,
+      "grad_norm": 0.26165443658828735,
+      "learning_rate": 4.167241379310345e-05,
+      "loss": 0.1326,
+      "step": 1449
+    },
+    {
+      "epoch": 1.6807940302832718,
+      "grad_norm": 0.24626079201698303,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 0.1189,
+      "step": 1450
+    },
+    {
+      "epoch": 1.6819531985800187,
+      "grad_norm": 0.24544042348861694,
+      "learning_rate": 4.166091954022989e-05,
+      "loss": 0.1343,
+      "step": 1451
+    },
+    {
+      "epoch": 1.6831123668767658,
+      "grad_norm": 0.23547019064426422,
+      "learning_rate": 4.16551724137931e-05,
+      "loss": 0.1176,
+      "step": 1452
+    },
+    {
+      "epoch": 1.684271535173513,
+      "grad_norm": 0.2702592611312866,
+      "learning_rate": 4.1649425287356324e-05,
+      "loss": 0.1488,
+      "step": 1453
+    },
+    {
+      "epoch": 1.68543070347026,
+      "grad_norm": 0.24035340547561646,
+      "learning_rate": 4.164367816091954e-05,
+      "loss": 0.1348,
+      "step": 1454
+    },
+    {
+      "epoch": 1.6865898717670071,
+      "grad_norm": 0.22424210608005524,
+      "learning_rate": 4.163793103448276e-05,
+      "loss": 0.13,
+      "step": 1455
+    },
+    {
+      "epoch": 1.6877490400637543,
+      "grad_norm": 0.2859560549259186,
+      "learning_rate": 4.163218390804598e-05,
+      "loss": 0.1486,
+      "step": 1456
+    },
+    {
+      "epoch": 1.6889082083605014,
+      "grad_norm": 0.21824337542057037,
+      "learning_rate": 4.16264367816092e-05,
+      "loss": 0.1174,
+      "step": 1457
+    },
+    {
+      "epoch": 1.6900673766572485,
+      "grad_norm": 0.24200040102005005,
+      "learning_rate": 4.162068965517242e-05,
+      "loss": 0.1324,
+      "step": 1458
+    },
+    {
+      "epoch": 1.6912265449539956,
+      "grad_norm": 0.2552363872528076,
+      "learning_rate": 4.161494252873563e-05,
+      "loss": 0.1313,
+      "step": 1459
+    },
+    {
+      "epoch": 1.6923857132507427,
+      "grad_norm": 0.19345875084400177,
+      "learning_rate": 4.160919540229885e-05,
+      "loss": 0.1174,
+      "step": 1460
+    },
+    {
+      "epoch": 1.6935448815474898,
+      "grad_norm": 0.25094732642173767,
+      "learning_rate": 4.160344827586207e-05,
+      "loss": 0.1326,
+      "step": 1461
+    },
+    {
+      "epoch": 1.6947040498442367,
+      "grad_norm": 0.24605903029441833,
+      "learning_rate": 4.159770114942529e-05,
+      "loss": 0.1155,
+      "step": 1462
+    },
+    {
+      "epoch": 1.6958632181409838,
+      "grad_norm": 0.3044376075267792,
+      "learning_rate": 4.1591954022988506e-05,
+      "loss": 0.1273,
+      "step": 1463
+    },
+    {
+      "epoch": 1.697022386437731,
+      "grad_norm": 0.25969693064689636,
+      "learning_rate": 4.158620689655173e-05,
+      "loss": 0.1261,
+      "step": 1464
+    },
+    {
+      "epoch": 1.6981815547344778,
+      "grad_norm": 0.2134069949388504,
+      "learning_rate": 4.158045977011494e-05,
+      "loss": 0.129,
+      "step": 1465
+    },
+    {
+      "epoch": 1.699340723031225,
+      "grad_norm": 0.24624426662921906,
+      "learning_rate": 4.1574712643678165e-05,
+      "loss": 0.1254,
+      "step": 1466
+    },
+    {
+      "epoch": 1.700499891327972,
+      "grad_norm": 0.27227187156677246,
+      "learning_rate": 4.156896551724138e-05,
+      "loss": 0.1241,
+      "step": 1467
+    },
+    {
+      "epoch": 1.7016590596247192,
+      "grad_norm": 0.2617661952972412,
+      "learning_rate": 4.15632183908046e-05,
+      "loss": 0.1241,
+      "step": 1468
+    },
+    {
+      "epoch": 1.7028182279214663,
+      "grad_norm": 0.2572365701198578,
+      "learning_rate": 4.155747126436782e-05,
+      "loss": 0.1308,
+      "step": 1469
+    },
+    {
+      "epoch": 1.7039773962182134,
+      "grad_norm": 0.24519048631191254,
+      "learning_rate": 4.155172413793104e-05,
+      "loss": 0.1333,
+      "step": 1470
+    },
+    {
+      "epoch": 1.7051365645149605,
+      "grad_norm": 0.23390959203243256,
+      "learning_rate": 4.154597701149425e-05,
+      "loss": 0.1335,
+      "step": 1471
+    },
+    {
+      "epoch": 1.7062957328117077,
+      "grad_norm": 0.19565661251544952,
+      "learning_rate": 4.1540229885057474e-05,
+      "loss": 0.1161,
+      "step": 1472
+    },
+    {
+      "epoch": 1.7074549011084548,
+      "grad_norm": 0.19742251932621002,
+      "learning_rate": 4.153448275862069e-05,
+      "loss": 0.1153,
+      "step": 1473
+    },
+    {
+      "epoch": 1.7086140694052019,
+      "grad_norm": 0.2252473384141922,
+      "learning_rate": 4.152873563218391e-05,
+      "loss": 0.1223,
+      "step": 1474
+    },
+    {
+      "epoch": 1.709773237701949,
+      "grad_norm": 0.25856614112854004,
+      "learning_rate": 4.1522988505747126e-05,
+      "loss": 0.1225,
+      "step": 1475
+    },
+    {
+      "epoch": 1.710932405998696,
+      "grad_norm": 0.20960158109664917,
+      "learning_rate": 4.151724137931035e-05,
+      "loss": 0.1257,
+      "step": 1476
+    },
+    {
+      "epoch": 1.712091574295443,
+      "grad_norm": 0.26653629541397095,
+      "learning_rate": 4.151149425287357e-05,
+      "loss": 0.123,
+      "step": 1477
+    },
+    {
+      "epoch": 1.7132507425921901,
+      "grad_norm": 0.19926515221595764,
+      "learning_rate": 4.1505747126436784e-05,
+      "loss": 0.1199,
+      "step": 1478
+    },
+    {
+      "epoch": 1.7144099108889372,
+      "grad_norm": 0.2077583223581314,
+      "learning_rate": 4.15e-05,
+      "loss": 0.1212,
+      "step": 1479
+    },
+    {
+      "epoch": 1.7155690791856841,
+      "grad_norm": 0.22840292751789093,
+      "learning_rate": 4.149425287356322e-05,
+      "loss": 0.1244,
+      "step": 1480
+    },
+    {
+      "epoch": 1.7167282474824312,
+      "grad_norm": 0.26933014392852783,
+      "learning_rate": 4.1488505747126436e-05,
+      "loss": 0.1388,
+      "step": 1481
+    },
+    {
+      "epoch": 1.7178874157791784,
+      "grad_norm": 0.2531909942626953,
+      "learning_rate": 4.148275862068966e-05,
+      "loss": 0.128,
+      "step": 1482
+    },
+    {
+      "epoch": 1.7190465840759255,
+      "grad_norm": 0.3011757731437683,
+      "learning_rate": 4.147701149425288e-05,
+      "loss": 0.1574,
+      "step": 1483
+    },
+    {
+      "epoch": 1.7202057523726726,
+      "grad_norm": 0.27856916189193726,
+      "learning_rate": 4.1471264367816094e-05,
+      "loss": 0.1275,
+      "step": 1484
+    },
+    {
+      "epoch": 1.7213649206694197,
+      "grad_norm": 0.23599520325660706,
+      "learning_rate": 4.1465517241379316e-05,
+      "loss": 0.1206,
+      "step": 1485
+    },
+    {
+      "epoch": 1.7225240889661668,
+      "grad_norm": 0.2702353000640869,
+      "learning_rate": 4.145977011494253e-05,
+      "loss": 0.1296,
+      "step": 1486
+    },
+    {
+      "epoch": 1.723683257262914,
+      "grad_norm": 0.21505098044872284,
+      "learning_rate": 4.1454022988505745e-05,
+      "loss": 0.1131,
+      "step": 1487
+    },
+    {
+      "epoch": 1.724842425559661,
+      "grad_norm": 0.27001845836639404,
+      "learning_rate": 4.144827586206897e-05,
+      "loss": 0.1265,
+      "step": 1488
+    },
+    {
+      "epoch": 1.7260015938564082,
+      "grad_norm": 0.2392754852771759,
+      "learning_rate": 4.144252873563219e-05,
+      "loss": 0.1159,
+      "step": 1489
+    },
+    {
+      "epoch": 1.727160762153155,
+      "grad_norm": 0.24627234041690826,
+      "learning_rate": 4.1436781609195404e-05,
+      "loss": 0.1284,
+      "step": 1490
+    },
+    {
+      "epoch": 1.7283199304499022,
+      "grad_norm": 0.23222601413726807,
+      "learning_rate": 4.1431034482758625e-05,
+      "loss": 0.1288,
+      "step": 1491
+    },
+    {
+      "epoch": 1.7294790987466493,
+      "grad_norm": 0.2943344712257385,
+      "learning_rate": 4.142528735632184e-05,
+      "loss": 0.1299,
+      "step": 1492
+    },
+    {
+      "epoch": 1.7306382670433964,
+      "grad_norm": 0.2248132824897766,
+      "learning_rate": 4.141954022988506e-05,
+      "loss": 0.1191,
+      "step": 1493
+    },
+    {
+      "epoch": 1.7317974353401433,
+      "grad_norm": 0.24535109102725983,
+      "learning_rate": 4.141379310344828e-05,
+      "loss": 0.129,
+      "step": 1494
+    },
+    {
+      "epoch": 1.7329566036368904,
+      "grad_norm": 0.2560446262359619,
+      "learning_rate": 4.140804597701149e-05,
+      "loss": 0.1343,
+      "step": 1495
+    },
+    {
+      "epoch": 1.7341157719336375,
+      "grad_norm": 0.23877164721488953,
+      "learning_rate": 4.1402298850574713e-05,
+      "loss": 0.1026,
+      "step": 1496
+    },
+    {
+      "epoch": 1.7352749402303846,
+      "grad_norm": 0.2307940572500229,
+      "learning_rate": 4.1396551724137935e-05,
+      "loss": 0.1158,
+      "step": 1497
+    },
+    {
+      "epoch": 1.7364341085271318,
+      "grad_norm": 0.23847486078739166,
+      "learning_rate": 4.139080459770115e-05,
+      "loss": 0.1177,
+      "step": 1498
+    },
+    {
+      "epoch": 1.7375932768238789,
+      "grad_norm": 0.2667318880558014,
+      "learning_rate": 4.138505747126437e-05,
+      "loss": 0.1483,
+      "step": 1499
+    },
+    {
+      "epoch": 1.738752445120626,
+      "grad_norm": 0.2531675100326538,
+      "learning_rate": 4.1379310344827587e-05,
+      "loss": 0.1209,
+      "step": 1500
+    },
+    {
+      "epoch": 1.739911613417373,
+      "grad_norm": 0.4089076817035675,
+      "learning_rate": 4.13735632183908e-05,
+      "loss": 0.1485,
+      "step": 1501
+    },
+    {
+      "epoch": 1.7410707817141202,
+      "grad_norm": 0.2458110749721527,
+      "learning_rate": 4.136781609195402e-05,
+      "loss": 0.1252,
+      "step": 1502
+    },
+    {
+      "epoch": 1.7422299500108673,
+      "grad_norm": 0.23186376690864563,
+      "learning_rate": 4.1362068965517245e-05,
+      "loss": 0.1292,
+      "step": 1503
+    },
+    {
+      "epoch": 1.7433891183076144,
+      "grad_norm": 0.24878862500190735,
+      "learning_rate": 4.1356321839080466e-05,
+      "loss": 0.1307,
+      "step": 1504
+    },
+    {
+      "epoch": 1.7445482866043613,
+      "grad_norm": 0.2361961454153061,
+      "learning_rate": 4.135057471264368e-05,
+      "loss": 0.118,
+      "step": 1505
+    },
+    {
+      "epoch": 1.7457074549011085,
+      "grad_norm": 0.24874532222747803,
+      "learning_rate": 4.1344827586206896e-05,
+      "loss": 0.1122,
+      "step": 1506
+    },
+    {
+      "epoch": 1.7468666231978556,
+      "grad_norm": 0.2600504159927368,
+      "learning_rate": 4.133908045977012e-05,
+      "loss": 0.1229,
+      "step": 1507
+    },
+    {
+      "epoch": 1.7480257914946025,
+      "grad_norm": 0.28750351071357727,
+      "learning_rate": 4.133333333333333e-05,
+      "loss": 0.1385,
+      "step": 1508
+    },
+    {
+      "epoch": 1.7491849597913496,
+      "grad_norm": 0.26855355501174927,
+      "learning_rate": 4.1327586206896555e-05,
+      "loss": 0.1394,
+      "step": 1509
+    },
+    {
+      "epoch": 1.7503441280880967,
+      "grad_norm": 0.28046318888664246,
+      "learning_rate": 4.1321839080459776e-05,
+      "loss": 0.1362,
+      "step": 1510
+    },
+    {
+      "epoch": 1.7515032963848438,
+      "grad_norm": 0.27256345748901367,
+      "learning_rate": 4.131609195402299e-05,
+      "loss": 0.1283,
+      "step": 1511
+    },
+    {
+      "epoch": 1.752662464681591,
+      "grad_norm": 0.22228793799877167,
+      "learning_rate": 4.131034482758621e-05,
+      "loss": 0.1104,
+      "step": 1512
+    },
+    {
+      "epoch": 1.753821632978338,
+      "grad_norm": 0.24911850690841675,
+      "learning_rate": 4.130459770114943e-05,
+      "loss": 0.1212,
+      "step": 1513
+    },
+    {
+      "epoch": 1.7549808012750852,
+      "grad_norm": 0.21466509997844696,
+      "learning_rate": 4.129885057471264e-05,
+      "loss": 0.1141,
+      "step": 1514
+    },
+    {
+      "epoch": 1.7561399695718323,
+      "grad_norm": 0.2404317706823349,
+      "learning_rate": 4.1293103448275864e-05,
+      "loss": 0.119,
+      "step": 1515
+    },
+    {
+      "epoch": 1.7572991378685794,
+      "grad_norm": 0.25505301356315613,
+      "learning_rate": 4.128735632183908e-05,
+      "loss": 0.115,
+      "step": 1516
+    },
+    {
+      "epoch": 1.7584583061653265,
+      "grad_norm": 0.2898204028606415,
+      "learning_rate": 4.12816091954023e-05,
+      "loss": 0.1508,
+      "step": 1517
+    },
+    {
+      "epoch": 1.7596174744620736,
+      "grad_norm": 0.24544641375541687,
+      "learning_rate": 4.127586206896552e-05,
+      "loss": 0.1278,
+      "step": 1518
+    },
+    {
+      "epoch": 1.7607766427588205,
+      "grad_norm": 0.21617454290390015,
+      "learning_rate": 4.127011494252874e-05,
+      "loss": 0.1206,
+      "step": 1519
+    },
+    {
+      "epoch": 1.7619358110555676,
+      "grad_norm": 0.22762687504291534,
+      "learning_rate": 4.126436781609195e-05,
+      "loss": 0.1354,
+      "step": 1520
+    },
+    {
+      "epoch": 1.7630949793523147,
+      "grad_norm": 0.22020873427391052,
+      "learning_rate": 4.1258620689655174e-05,
+      "loss": 0.1361,
+      "step": 1521
+    },
+    {
+      "epoch": 1.7642541476490616,
+      "grad_norm": 0.21312019228935242,
+      "learning_rate": 4.125287356321839e-05,
+      "loss": 0.1258,
+      "step": 1522
+    },
+    {
+      "epoch": 1.7654133159458087,
+      "grad_norm": 0.22142820060253143,
+      "learning_rate": 4.124712643678161e-05,
+      "loss": 0.1167,
+      "step": 1523
+    },
+    {
+      "epoch": 1.7665724842425559,
+      "grad_norm": 0.23192018270492554,
+      "learning_rate": 4.124137931034483e-05,
+      "loss": 0.1149,
+      "step": 1524
+    },
+    {
+      "epoch": 1.767731652539303,
+      "grad_norm": 0.23699016869068146,
+      "learning_rate": 4.123563218390805e-05,
+      "loss": 0.1356,
+      "step": 1525
+    },
+    {
+      "epoch": 1.76889082083605,
+      "grad_norm": 0.25063955783843994,
+      "learning_rate": 4.122988505747127e-05,
+      "loss": 0.1232,
+      "step": 1526
+    },
+    {
+      "epoch": 1.7700499891327972,
+      "grad_norm": 0.2514595687389374,
+      "learning_rate": 4.1224137931034484e-05,
+      "loss": 0.1332,
+      "step": 1527
+    },
+    {
+      "epoch": 1.7712091574295443,
+      "grad_norm": 0.24346187710762024,
+      "learning_rate": 4.12183908045977e-05,
+      "loss": 0.1239,
+      "step": 1528
+    },
+    {
+      "epoch": 1.7723683257262914,
+      "grad_norm": 0.22228626906871796,
+      "learning_rate": 4.121264367816092e-05,
+      "loss": 0.1061,
+      "step": 1529
+    },
+    {
+      "epoch": 1.7735274940230386,
+      "grad_norm": 0.2860787510871887,
+      "learning_rate": 4.120689655172414e-05,
+      "loss": 0.1254,
+      "step": 1530
+    },
+    {
+      "epoch": 1.7746866623197857,
+      "grad_norm": 0.2642548680305481,
+      "learning_rate": 4.1201149425287364e-05,
+      "loss": 0.126,
+      "step": 1531
+    },
+    {
+      "epoch": 1.7758458306165328,
+      "grad_norm": 0.2488432377576828,
+      "learning_rate": 4.119540229885058e-05,
+      "loss": 0.1175,
+      "step": 1532
+    },
+    {
+      "epoch": 1.7770049989132797,
+      "grad_norm": 0.23602858185768127,
+      "learning_rate": 4.1189655172413794e-05,
+      "loss": 0.1244,
+      "step": 1533
+    },
+    {
+      "epoch": 1.7781641672100268,
+      "grad_norm": 0.23683659732341766,
+      "learning_rate": 4.1183908045977015e-05,
+      "loss": 0.1186,
+      "step": 1534
+    },
+    {
+      "epoch": 1.779323335506774,
+      "grad_norm": 0.2419913113117218,
+      "learning_rate": 4.117816091954023e-05,
+      "loss": 0.1364,
+      "step": 1535
+    },
+    {
+      "epoch": 1.780482503803521,
+      "grad_norm": 0.26852262020111084,
+      "learning_rate": 4.1172413793103445e-05,
+      "loss": 0.1287,
+      "step": 1536
+    },
+    {
+      "epoch": 1.781641672100268,
+      "grad_norm": 0.2520958185195923,
+      "learning_rate": 4.116666666666667e-05,
+      "loss": 0.1381,
+      "step": 1537
+    },
+    {
+      "epoch": 1.782800840397015,
+      "grad_norm": 0.2294151484966278,
+      "learning_rate": 4.116091954022989e-05,
+      "loss": 0.1458,
+      "step": 1538
+    },
+    {
+      "epoch": 1.7839600086937621,
+      "grad_norm": 0.20725025236606598,
+      "learning_rate": 4.11551724137931e-05,
+      "loss": 0.1358,
+      "step": 1539
+    },
+    {
+      "epoch": 1.7851191769905093,
+      "grad_norm": 0.22536805272102356,
+      "learning_rate": 4.1149425287356325e-05,
+      "loss": 0.1273,
+      "step": 1540
+    },
+    {
+      "epoch": 1.7862783452872564,
+      "grad_norm": 0.2271406203508377,
+      "learning_rate": 4.114367816091954e-05,
+      "loss": 0.1254,
+      "step": 1541
+    },
+    {
+      "epoch": 1.7874375135840035,
+      "grad_norm": 0.25207868218421936,
+      "learning_rate": 4.113793103448276e-05,
+      "loss": 0.1325,
+      "step": 1542
+    },
+    {
+      "epoch": 1.7885966818807506,
+      "grad_norm": 0.22057871520519257,
+      "learning_rate": 4.1132183908045976e-05,
+      "loss": 0.1287,
+      "step": 1543
+    },
+    {
+      "epoch": 1.7897558501774977,
+      "grad_norm": 0.21857698261737823,
+      "learning_rate": 4.11264367816092e-05,
+      "loss": 0.118,
+      "step": 1544
+    },
+    {
+      "epoch": 1.7909150184742448,
+      "grad_norm": 0.26904162764549255,
+      "learning_rate": 4.112068965517242e-05,
+      "loss": 0.137,
+      "step": 1545
+    },
+    {
+      "epoch": 1.792074186770992,
+      "grad_norm": 0.20811136066913605,
+      "learning_rate": 4.1114942528735635e-05,
+      "loss": 0.1148,
+      "step": 1546
+    },
+    {
+      "epoch": 1.7932333550677388,
+      "grad_norm": 0.2898993194103241,
+      "learning_rate": 4.110919540229885e-05,
+      "loss": 0.1396,
+      "step": 1547
+    },
+    {
+      "epoch": 1.794392523364486,
+      "grad_norm": 0.2406226396560669,
+      "learning_rate": 4.110344827586207e-05,
+      "loss": 0.1177,
+      "step": 1548
+    },
+    {
+      "epoch": 1.795551691661233,
+      "grad_norm": 0.21909652650356293,
+      "learning_rate": 4.1097701149425286e-05,
+      "loss": 0.1221,
+      "step": 1549
+    },
+    {
+      "epoch": 1.7967108599579802,
+      "grad_norm": 0.24359643459320068,
+      "learning_rate": 4.109195402298851e-05,
+      "loss": 0.1249,
+      "step": 1550
+    },
+    {
+      "epoch": 1.797870028254727,
+      "grad_norm": 0.20829685032367706,
+      "learning_rate": 4.108620689655173e-05,
+      "loss": 0.1155,
+      "step": 1551
+    },
+    {
+      "epoch": 1.7990291965514742,
+      "grad_norm": 0.34604066610336304,
+      "learning_rate": 4.1080459770114944e-05,
+      "loss": 0.1366,
+      "step": 1552
+    },
+    {
+      "epoch": 1.8001883648482213,
+      "grad_norm": 0.2357710599899292,
+      "learning_rate": 4.1074712643678166e-05,
+      "loss": 0.1238,
+      "step": 1553
+    },
+    {
+      "epoch": 1.8013475331449684,
+      "grad_norm": 0.2505914866924286,
+      "learning_rate": 4.106896551724138e-05,
+      "loss": 0.1381,
+      "step": 1554
+    },
+    {
+      "epoch": 1.8025067014417155,
+      "grad_norm": 0.3045206665992737,
+      "learning_rate": 4.1063218390804596e-05,
+      "loss": 0.125,
+      "step": 1555
+    },
+    {
+      "epoch": 1.8036658697384627,
+      "grad_norm": 0.22790023684501648,
+      "learning_rate": 4.105747126436782e-05,
+      "loss": 0.1189,
+      "step": 1556
+    },
+    {
+      "epoch": 1.8048250380352098,
+      "grad_norm": 0.267311155796051,
+      "learning_rate": 4.105172413793103e-05,
+      "loss": 0.133,
+      "step": 1557
+    },
+    {
+      "epoch": 1.8059842063319569,
+      "grad_norm": 0.2867993712425232,
+      "learning_rate": 4.1045977011494254e-05,
+      "loss": 0.1197,
+      "step": 1558
+    },
+    {
+      "epoch": 1.807143374628704,
+      "grad_norm": 0.2740671634674072,
+      "learning_rate": 4.1040229885057476e-05,
+      "loss": 0.1247,
+      "step": 1559
+    },
+    {
+      "epoch": 1.8083025429254511,
+      "grad_norm": 0.23246349394321442,
+      "learning_rate": 4.103448275862069e-05,
+      "loss": 0.1204,
+      "step": 1560
+    },
+    {
+      "epoch": 1.8094617112221982,
+      "grad_norm": 0.2556561231613159,
+      "learning_rate": 4.102873563218391e-05,
+      "loss": 0.1381,
+      "step": 1561
+    },
+    {
+      "epoch": 1.8106208795189451,
+      "grad_norm": 0.22063712775707245,
+      "learning_rate": 4.102298850574713e-05,
+      "loss": 0.1288,
+      "step": 1562
+    },
+    {
+      "epoch": 1.8117800478156922,
+      "grad_norm": 0.26971542835235596,
+      "learning_rate": 4.101724137931034e-05,
+      "loss": 0.1348,
+      "step": 1563
+    },
+    {
+      "epoch": 1.8129392161124394,
+      "grad_norm": 0.28952643275260925,
+      "learning_rate": 4.1011494252873564e-05,
+      "loss": 0.1408,
+      "step": 1564
+    },
+    {
+      "epoch": 1.8140983844091862,
+      "grad_norm": 0.24396105110645294,
+      "learning_rate": 4.1005747126436786e-05,
+      "loss": 0.1284,
+      "step": 1565
+    },
+    {
+      "epoch": 1.8152575527059334,
+      "grad_norm": 0.2277979552745819,
+      "learning_rate": 4.1e-05,
+      "loss": 0.1234,
+      "step": 1566
+    },
+    {
+      "epoch": 1.8164167210026805,
+      "grad_norm": 0.1920943409204483,
+      "learning_rate": 4.099425287356322e-05,
+      "loss": 0.1183,
+      "step": 1567
+    },
+    {
+      "epoch": 1.8175758892994276,
+      "grad_norm": 0.25110727548599243,
+      "learning_rate": 4.098850574712644e-05,
+      "loss": 0.1338,
+      "step": 1568
+    },
+    {
+      "epoch": 1.8187350575961747,
+      "grad_norm": 0.2476564645767212,
+      "learning_rate": 4.098275862068966e-05,
+      "loss": 0.1311,
+      "step": 1569
+    },
+    {
+      "epoch": 1.8198942258929218,
+      "grad_norm": 0.2378225028514862,
+      "learning_rate": 4.0977011494252874e-05,
+      "loss": 0.131,
+      "step": 1570
+    },
+    {
+      "epoch": 1.821053394189669,
+      "grad_norm": 0.2154211848974228,
+      "learning_rate": 4.0971264367816095e-05,
+      "loss": 0.1192,
+      "step": 1571
+    },
+    {
+      "epoch": 1.822212562486416,
+      "grad_norm": 0.23698244988918304,
+      "learning_rate": 4.096551724137932e-05,
+      "loss": 0.1302,
+      "step": 1572
+    },
+    {
+      "epoch": 1.8233717307831632,
+      "grad_norm": 0.22462879121303558,
+      "learning_rate": 4.095977011494253e-05,
+      "loss": 0.1225,
+      "step": 1573
+    },
+    {
+      "epoch": 1.8245308990799103,
+      "grad_norm": 0.21854135394096375,
+      "learning_rate": 4.095402298850575e-05,
+      "loss": 0.1185,
+      "step": 1574
+    },
+    {
+      "epoch": 1.8256900673766574,
+      "grad_norm": 0.24488233029842377,
+      "learning_rate": 4.094827586206897e-05,
+      "loss": 0.1434,
+      "step": 1575
+    },
+    {
+      "epoch": 1.8268492356734043,
+      "grad_norm": 0.22526147961616516,
+      "learning_rate": 4.094252873563218e-05,
+      "loss": 0.1272,
+      "step": 1576
+    },
+    {
+      "epoch": 1.8280084039701514,
+      "grad_norm": 0.206687793135643,
+      "learning_rate": 4.09367816091954e-05,
+      "loss": 0.1109,
+      "step": 1577
+    },
+    {
+      "epoch": 1.8291675722668985,
+      "grad_norm": 0.26105666160583496,
+      "learning_rate": 4.093103448275863e-05,
+      "loss": 0.1389,
+      "step": 1578
+    },
+    {
+      "epoch": 1.8303267405636456,
+      "grad_norm": 0.3510555326938629,
+      "learning_rate": 4.092528735632184e-05,
+      "loss": 0.1294,
+      "step": 1579
+    },
+    {
+      "epoch": 1.8314859088603925,
+      "grad_norm": 0.238253653049469,
+      "learning_rate": 4.091954022988506e-05,
+      "loss": 0.123,
+      "step": 1580
+    },
+    {
+      "epoch": 1.8326450771571396,
+      "grad_norm": 0.2679709792137146,
+      "learning_rate": 4.091379310344828e-05,
+      "loss": 0.1326,
+      "step": 1581
+    },
+    {
+      "epoch": 1.8338042454538868,
+      "grad_norm": 0.22641777992248535,
+      "learning_rate": 4.090804597701149e-05,
+      "loss": 0.1299,
+      "step": 1582
+    },
+    {
+      "epoch": 1.8349634137506339,
+      "grad_norm": 0.25320807099342346,
+      "learning_rate": 4.0902298850574715e-05,
+      "loss": 0.1223,
+      "step": 1583
+    },
+    {
+      "epoch": 1.836122582047381,
+      "grad_norm": 0.24081723392009735,
+      "learning_rate": 4.089655172413793e-05,
+      "loss": 0.1327,
+      "step": 1584
+    },
+    {
+      "epoch": 1.837281750344128,
+      "grad_norm": 0.2431461215019226,
+      "learning_rate": 4.089080459770115e-05,
+      "loss": 0.1347,
+      "step": 1585
+    },
+    {
+      "epoch": 1.8384409186408752,
+      "grad_norm": 0.2519475817680359,
+      "learning_rate": 4.088505747126437e-05,
+      "loss": 0.1438,
+      "step": 1586
+    },
+    {
+      "epoch": 1.8396000869376223,
+      "grad_norm": 0.252816766500473,
+      "learning_rate": 4.087931034482759e-05,
+      "loss": 0.138,
+      "step": 1587
+    },
+    {
+      "epoch": 1.8407592552343695,
+      "grad_norm": 0.22462011873722076,
+      "learning_rate": 4.087356321839081e-05,
+      "loss": 0.1192,
+      "step": 1588
+    },
+    {
+      "epoch": 1.8419184235311166,
+      "grad_norm": 0.28114578127861023,
+      "learning_rate": 4.0867816091954025e-05,
+      "loss": 0.1384,
+      "step": 1589
+    },
+    {
+      "epoch": 1.8430775918278635,
+      "grad_norm": 0.24173279106616974,
+      "learning_rate": 4.086206896551724e-05,
+      "loss": 0.1389,
+      "step": 1590
+    },
+    {
+      "epoch": 1.8442367601246106,
+      "grad_norm": 0.19824741780757904,
+      "learning_rate": 4.085632183908046e-05,
+      "loss": 0.1117,
+      "step": 1591
+    },
+    {
+      "epoch": 1.8453959284213577,
+      "grad_norm": 0.2025434970855713,
+      "learning_rate": 4.085057471264368e-05,
+      "loss": 0.1103,
+      "step": 1592
+    },
+    {
+      "epoch": 1.8465550967181048,
+      "grad_norm": 0.2460830956697464,
+      "learning_rate": 4.08448275862069e-05,
+      "loss": 0.1327,
+      "step": 1593
+    },
+    {
+      "epoch": 1.8477142650148517,
+      "grad_norm": 0.2609751224517822,
+      "learning_rate": 4.083908045977012e-05,
+      "loss": 0.1315,
+      "step": 1594
+    },
+    {
+      "epoch": 1.8488734333115988,
+      "grad_norm": 0.20882025361061096,
+      "learning_rate": 4.0833333333333334e-05,
+      "loss": 0.1076,
+      "step": 1595
+    },
+    {
+      "epoch": 1.850032601608346,
+      "grad_norm": 0.20100267231464386,
+      "learning_rate": 4.0827586206896556e-05,
+      "loss": 0.1116,
+      "step": 1596
+    },
+    {
+      "epoch": 1.851191769905093,
+      "grad_norm": 0.2182759940624237,
+      "learning_rate": 4.082183908045977e-05,
+      "loss": 0.135,
+      "step": 1597
+    },
+    {
+      "epoch": 1.8523509382018402,
+      "grad_norm": 0.21987977623939514,
+      "learning_rate": 4.0816091954022986e-05,
+      "loss": 0.1287,
+      "step": 1598
+    },
+    {
+      "epoch": 1.8535101064985873,
+      "grad_norm": 0.2634388506412506,
+      "learning_rate": 4.0810344827586214e-05,
+      "loss": 0.1267,
+      "step": 1599
+    },
+    {
+      "epoch": 1.8546692747953344,
+      "grad_norm": 0.2175815850496292,
+      "learning_rate": 4.080459770114943e-05,
+      "loss": 0.1175,
+      "step": 1600
+    },
+    {
+      "epoch": 1.8558284430920815,
+      "grad_norm": 0.24021421372890472,
+      "learning_rate": 4.0798850574712644e-05,
+      "loss": 0.1298,
+      "step": 1601
+    },
+    {
+      "epoch": 1.8569876113888286,
+      "grad_norm": 0.2455061972141266,
+      "learning_rate": 4.0793103448275866e-05,
+      "loss": 0.1318,
+      "step": 1602
+    },
+    {
+      "epoch": 1.8581467796855757,
+      "grad_norm": 0.5049688220024109,
+      "learning_rate": 4.078735632183908e-05,
+      "loss": 0.1285,
+      "step": 1603
+    },
+    {
+      "epoch": 1.8593059479823228,
+      "grad_norm": 0.3722595274448395,
+      "learning_rate": 4.0781609195402295e-05,
+      "loss": 0.1359,
+      "step": 1604
+    },
+    {
+      "epoch": 1.8604651162790697,
+      "grad_norm": 0.3047929108142853,
+      "learning_rate": 4.077586206896552e-05,
+      "loss": 0.1284,
+      "step": 1605
+    },
+    {
+      "epoch": 1.8616242845758169,
+      "grad_norm": 0.48975053429603577,
+      "learning_rate": 4.077011494252874e-05,
+      "loss": 0.1326,
+      "step": 1606
+    },
+    {
+      "epoch": 1.862783452872564,
+      "grad_norm": 0.27146080136299133,
+      "learning_rate": 4.076436781609196e-05,
+      "loss": 0.1237,
+      "step": 1607
+    },
+    {
+      "epoch": 1.8639426211693109,
+      "grad_norm": 0.26589420437812805,
+      "learning_rate": 4.0758620689655175e-05,
+      "loss": 0.1225,
+      "step": 1608
+    },
+    {
+      "epoch": 1.865101789466058,
+      "grad_norm": 0.2550943195819855,
+      "learning_rate": 4.075287356321839e-05,
+      "loss": 0.1348,
+      "step": 1609
+    },
+    {
+      "epoch": 1.866260957762805,
+      "grad_norm": 0.22643890976905823,
+      "learning_rate": 4.074712643678161e-05,
+      "loss": 0.1199,
+      "step": 1610
+    },
+    {
+      "epoch": 1.8674201260595522,
+      "grad_norm": 0.28529438376426697,
+      "learning_rate": 4.074137931034483e-05,
+      "loss": 0.1186,
+      "step": 1611
+    },
+    {
+      "epoch": 1.8685792943562993,
+      "grad_norm": 0.24935007095336914,
+      "learning_rate": 4.073563218390805e-05,
+      "loss": 0.1218,
+      "step": 1612
+    },
+    {
+      "epoch": 1.8697384626530464,
+      "grad_norm": 0.23699697852134705,
+      "learning_rate": 4.072988505747127e-05,
+      "loss": 0.136,
+      "step": 1613
+    },
+    {
+      "epoch": 1.8708976309497936,
+      "grad_norm": 0.4210551381111145,
+      "learning_rate": 4.0724137931034485e-05,
+      "loss": 0.1095,
+      "step": 1614
+    },
+    {
+      "epoch": 1.8720567992465407,
+      "grad_norm": 0.2276671975851059,
+      "learning_rate": 4.071839080459771e-05,
+      "loss": 0.1346,
+      "step": 1615
+    },
+    {
+      "epoch": 1.8732159675432878,
+      "grad_norm": 0.22973059117794037,
+      "learning_rate": 4.071264367816092e-05,
+      "loss": 0.122,
+      "step": 1616
+    },
+    {
+      "epoch": 1.874375135840035,
+      "grad_norm": 0.37576723098754883,
+      "learning_rate": 4.070689655172414e-05,
+      "loss": 0.1173,
+      "step": 1617
+    },
+    {
+      "epoch": 1.875534304136782,
+      "grad_norm": 0.23602044582366943,
+      "learning_rate": 4.070114942528736e-05,
+      "loss": 0.1426,
+      "step": 1618
+    },
+    {
+      "epoch": 1.876693472433529,
+      "grad_norm": 0.22846823930740356,
+      "learning_rate": 4.069540229885058e-05,
+      "loss": 0.1252,
+      "step": 1619
+    },
+    {
+      "epoch": 1.877852640730276,
+      "grad_norm": 0.21576084196567535,
+      "learning_rate": 4.0689655172413795e-05,
+      "loss": 0.1176,
+      "step": 1620
+    },
+    {
+      "epoch": 1.8790118090270231,
+      "grad_norm": 0.21635417640209198,
+      "learning_rate": 4.0683908045977017e-05,
+      "loss": 0.1256,
+      "step": 1621
+    },
+    {
+      "epoch": 1.88017097732377,
+      "grad_norm": 0.20854774117469788,
+      "learning_rate": 4.067816091954023e-05,
+      "loss": 0.1167,
+      "step": 1622
+    },
+    {
+      "epoch": 1.8813301456205171,
+      "grad_norm": 0.25999489426612854,
+      "learning_rate": 4.0672413793103446e-05,
+      "loss": 0.1225,
+      "step": 1623
+    },
+    {
+      "epoch": 1.8824893139172643,
+      "grad_norm": 0.2581290602684021,
+      "learning_rate": 4.066666666666667e-05,
+      "loss": 0.1254,
+      "step": 1624
+    },
+    {
+      "epoch": 1.8836484822140114,
+      "grad_norm": 0.2655247449874878,
+      "learning_rate": 4.066091954022988e-05,
+      "loss": 0.1326,
+      "step": 1625
+    },
+    {
+      "epoch": 1.8848076505107585,
+      "grad_norm": 0.2115846574306488,
+      "learning_rate": 4.0655172413793105e-05,
+      "loss": 0.1275,
+      "step": 1626
+    },
+    {
+      "epoch": 1.8859668188075056,
+      "grad_norm": 0.23177579045295715,
+      "learning_rate": 4.0649425287356326e-05,
+      "loss": 0.1169,
+      "step": 1627
+    },
+    {
+      "epoch": 1.8871259871042527,
+      "grad_norm": 0.45042136311531067,
+      "learning_rate": 4.064367816091954e-05,
+      "loss": 0.1309,
+      "step": 1628
+    },
+    {
+      "epoch": 1.8882851554009998,
+      "grad_norm": 0.26189303398132324,
+      "learning_rate": 4.063793103448276e-05,
+      "loss": 0.1309,
+      "step": 1629
+    },
+    {
+      "epoch": 1.889444323697747,
+      "grad_norm": 0.2205989956855774,
+      "learning_rate": 4.063218390804598e-05,
+      "loss": 0.1242,
+      "step": 1630
+    },
+    {
+      "epoch": 1.890603491994494,
+      "grad_norm": 0.3108113706111908,
+      "learning_rate": 4.062643678160919e-05,
+      "loss": 0.1392,
+      "step": 1631
+    },
+    {
+      "epoch": 1.8917626602912412,
+      "grad_norm": 0.2217811793088913,
+      "learning_rate": 4.0620689655172414e-05,
+      "loss": 0.1243,
+      "step": 1632
+    },
+    {
+      "epoch": 1.892921828587988,
+      "grad_norm": 0.21977943181991577,
+      "learning_rate": 4.0614942528735636e-05,
+      "loss": 0.1202,
+      "step": 1633
+    },
+    {
+      "epoch": 1.8940809968847352,
+      "grad_norm": 0.20839521288871765,
+      "learning_rate": 4.060919540229886e-05,
+      "loss": 0.1205,
+      "step": 1634
+    },
+    {
+      "epoch": 1.8952401651814823,
+      "grad_norm": 0.2656031548976898,
+      "learning_rate": 4.060344827586207e-05,
+      "loss": 0.1355,
+      "step": 1635
+    },
+    {
+      "epoch": 1.8963993334782294,
+      "grad_norm": 0.23771442472934723,
+      "learning_rate": 4.059770114942529e-05,
+      "loss": 0.1294,
+      "step": 1636
+    },
+    {
+      "epoch": 1.8975585017749763,
+      "grad_norm": 0.27091866731643677,
+      "learning_rate": 4.059195402298851e-05,
+      "loss": 0.1364,
+      "step": 1637
+    },
+    {
+      "epoch": 1.8987176700717234,
+      "grad_norm": 0.250888854265213,
+      "learning_rate": 4.0586206896551724e-05,
+      "loss": 0.1362,
+      "step": 1638
+    },
+    {
+      "epoch": 1.8998768383684705,
+      "grad_norm": 0.26287785172462463,
+      "learning_rate": 4.0580459770114946e-05,
+      "loss": 0.1268,
+      "step": 1639
+    },
+    {
+      "epoch": 1.9010360066652177,
+      "grad_norm": 0.2068604677915573,
+      "learning_rate": 4.057471264367817e-05,
+      "loss": 0.1186,
+      "step": 1640
+    },
+    {
+      "epoch": 1.9021951749619648,
+      "grad_norm": 0.27759426832199097,
+      "learning_rate": 4.056896551724138e-05,
+      "loss": 0.1424,
+      "step": 1641
+    },
+    {
+      "epoch": 1.903354343258712,
+      "grad_norm": 0.2885616719722748,
+      "learning_rate": 4.05632183908046e-05,
+      "loss": 0.1431,
+      "step": 1642
+    },
+    {
+      "epoch": 1.904513511555459,
+      "grad_norm": 0.2596352994441986,
+      "learning_rate": 4.055747126436782e-05,
+      "loss": 0.1301,
+      "step": 1643
+    },
+    {
+      "epoch": 1.9056726798522061,
+      "grad_norm": 0.22041890025138855,
+      "learning_rate": 4.0551724137931034e-05,
+      "loss": 0.1252,
+      "step": 1644
+    },
+    {
+      "epoch": 1.9068318481489532,
+      "grad_norm": 0.23935045301914215,
+      "learning_rate": 4.0545977011494256e-05,
+      "loss": 0.1234,
+      "step": 1645
+    },
+    {
+      "epoch": 1.9079910164457003,
+      "grad_norm": 0.23104766011238098,
+      "learning_rate": 4.054022988505747e-05,
+      "loss": 0.1247,
+      "step": 1646
+    },
+    {
+      "epoch": 1.9091501847424472,
+      "grad_norm": 0.25777825713157654,
+      "learning_rate": 4.053448275862069e-05,
+      "loss": 0.1221,
+      "step": 1647
+    },
+    {
+      "epoch": 1.9103093530391944,
+      "grad_norm": 0.21101577579975128,
+      "learning_rate": 4.0528735632183914e-05,
+      "loss": 0.1213,
+      "step": 1648
+    },
+    {
+      "epoch": 1.9114685213359415,
+      "grad_norm": 0.27736735343933105,
+      "learning_rate": 4.052298850574713e-05,
+      "loss": 0.1315,
+      "step": 1649
+    },
+    {
+      "epoch": 1.9126276896326886,
+      "grad_norm": 0.2874712646007538,
+      "learning_rate": 4.0517241379310344e-05,
+      "loss": 0.1267,
+      "step": 1650
+    },
+    {
+      "epoch": 1.9137868579294355,
+      "grad_norm": 0.27713993191719055,
+      "learning_rate": 4.0511494252873565e-05,
+      "loss": 0.1236,
+      "step": 1651
+    },
+    {
+      "epoch": 1.9149460262261826,
+      "grad_norm": 0.22386078536510468,
+      "learning_rate": 4.050574712643678e-05,
+      "loss": 0.1147,
+      "step": 1652
+    },
+    {
+      "epoch": 1.9161051945229297,
+      "grad_norm": 0.21959929168224335,
+      "learning_rate": 4.05e-05,
+      "loss": 0.1079,
+      "step": 1653
+    },
+    {
+      "epoch": 1.9172643628196768,
+      "grad_norm": 0.2387046068906784,
+      "learning_rate": 4.0494252873563224e-05,
+      "loss": 0.1267,
+      "step": 1654
+    },
+    {
+      "epoch": 1.918423531116424,
+      "grad_norm": 0.22587932646274567,
+      "learning_rate": 4.048850574712644e-05,
+      "loss": 0.124,
+      "step": 1655
+    },
+    {
+      "epoch": 1.919582699413171,
+      "grad_norm": 0.2760310471057892,
+      "learning_rate": 4.048275862068966e-05,
+      "loss": 0.1384,
+      "step": 1656
+    },
+    {
+      "epoch": 1.9207418677099182,
+      "grad_norm": 0.24281680583953857,
+      "learning_rate": 4.0477011494252875e-05,
+      "loss": 0.1264,
+      "step": 1657
+    },
+    {
+      "epoch": 1.9219010360066653,
+      "grad_norm": 0.23182785511016846,
+      "learning_rate": 4.047126436781609e-05,
+      "loss": 0.1184,
+      "step": 1658
+    },
+    {
+      "epoch": 1.9230602043034124,
+      "grad_norm": 0.2864990532398224,
+      "learning_rate": 4.046551724137931e-05,
+      "loss": 0.1398,
+      "step": 1659
+    },
+    {
+      "epoch": 1.9242193726001595,
+      "grad_norm": 0.267501562833786,
+      "learning_rate": 4.045977011494253e-05,
+      "loss": 0.135,
+      "step": 1660
+    },
+    {
+      "epoch": 1.9253785408969066,
+      "grad_norm": 0.6067526936531067,
+      "learning_rate": 4.045402298850575e-05,
+      "loss": 0.1321,
+      "step": 1661
+    },
+    {
+      "epoch": 1.9265377091936535,
+      "grad_norm": 0.2144288271665573,
+      "learning_rate": 4.044827586206897e-05,
+      "loss": 0.1195,
+      "step": 1662
+    },
+    {
+      "epoch": 1.9276968774904006,
+      "grad_norm": 0.2767769396305084,
+      "learning_rate": 4.0442528735632185e-05,
+      "loss": 0.121,
+      "step": 1663
+    },
+    {
+      "epoch": 1.9288560457871478,
+      "grad_norm": 0.258683443069458,
+      "learning_rate": 4.0436781609195406e-05,
+      "loss": 0.1358,
+      "step": 1664
+    },
+    {
+      "epoch": 1.9300152140838946,
+      "grad_norm": 0.3497917056083679,
+      "learning_rate": 4.043103448275862e-05,
+      "loss": 0.1271,
+      "step": 1665
+    },
+    {
+      "epoch": 1.9311743823806418,
+      "grad_norm": 0.2050793319940567,
+      "learning_rate": 4.0425287356321836e-05,
+      "loss": 0.0987,
+      "step": 1666
+    },
+    {
+      "epoch": 1.9323335506773889,
+      "grad_norm": 0.25068920850753784,
+      "learning_rate": 4.041954022988506e-05,
+      "loss": 0.1228,
+      "step": 1667
+    },
+    {
+      "epoch": 1.933492718974136,
+      "grad_norm": 0.31791725754737854,
+      "learning_rate": 4.041379310344828e-05,
+      "loss": 0.1245,
+      "step": 1668
+    },
+    {
+      "epoch": 1.934651887270883,
+      "grad_norm": 0.38801631331443787,
+      "learning_rate": 4.0408045977011495e-05,
+      "loss": 0.1264,
+      "step": 1669
+    },
+    {
+      "epoch": 1.9358110555676302,
+      "grad_norm": 0.26199209690093994,
+      "learning_rate": 4.0402298850574716e-05,
+      "loss": 0.1332,
+      "step": 1670
+    },
+    {
+      "epoch": 1.9369702238643773,
+      "grad_norm": 0.2558247148990631,
+      "learning_rate": 4.039655172413793e-05,
+      "loss": 0.126,
+      "step": 1671
+    },
+    {
+      "epoch": 1.9381293921611245,
+      "grad_norm": 0.2440565675497055,
+      "learning_rate": 4.039080459770115e-05,
+      "loss": 0.1265,
+      "step": 1672
+    },
+    {
+      "epoch": 1.9392885604578716,
+      "grad_norm": 0.35634905099868774,
+      "learning_rate": 4.038505747126437e-05,
+      "loss": 0.1492,
+      "step": 1673
+    },
+    {
+      "epoch": 1.9404477287546187,
+      "grad_norm": 0.2178521454334259,
+      "learning_rate": 4.037931034482759e-05,
+      "loss": 0.1166,
+      "step": 1674
+    },
+    {
+      "epoch": 1.9416068970513658,
+      "grad_norm": 0.25191420316696167,
+      "learning_rate": 4.037356321839081e-05,
+      "loss": 0.1329,
+      "step": 1675
+    },
+    {
+      "epoch": 1.9427660653481127,
+      "grad_norm": 0.25486209988594055,
+      "learning_rate": 4.0367816091954026e-05,
+      "loss": 0.1294,
+      "step": 1676
+    },
+    {
+      "epoch": 1.9439252336448598,
+      "grad_norm": 0.21692296862602234,
+      "learning_rate": 4.036206896551724e-05,
+      "loss": 0.1204,
+      "step": 1677
+    },
+    {
+      "epoch": 1.945084401941607,
+      "grad_norm": 0.24165906012058258,
+      "learning_rate": 4.035632183908046e-05,
+      "loss": 0.1244,
+      "step": 1678
+    },
+    {
+      "epoch": 1.9462435702383538,
+      "grad_norm": 0.2571072578430176,
+      "learning_rate": 4.035057471264368e-05,
+      "loss": 0.1387,
+      "step": 1679
+    },
+    {
+      "epoch": 1.947402738535101,
+      "grad_norm": 0.22820697724819183,
+      "learning_rate": 4.03448275862069e-05,
+      "loss": 0.1331,
+      "step": 1680
+    },
+    {
+      "epoch": 1.948561906831848,
+      "grad_norm": 0.2674961984157562,
+      "learning_rate": 4.033908045977012e-05,
+      "loss": 0.128,
+      "step": 1681
+    },
+    {
+      "epoch": 1.9497210751285952,
+      "grad_norm": 0.23374570906162262,
+      "learning_rate": 4.0333333333333336e-05,
+      "loss": 0.1256,
+      "step": 1682
+    },
+    {
+      "epoch": 1.9508802434253423,
+      "grad_norm": 0.23953098058700562,
+      "learning_rate": 4.032758620689656e-05,
+      "loss": 0.1354,
+      "step": 1683
+    },
+    {
+      "epoch": 1.9520394117220894,
+      "grad_norm": 0.21969419717788696,
+      "learning_rate": 4.032183908045977e-05,
+      "loss": 0.1273,
+      "step": 1684
+    },
+    {
+      "epoch": 1.9531985800188365,
+      "grad_norm": 0.2675417959690094,
+      "learning_rate": 4.031609195402299e-05,
+      "loss": 0.145,
+      "step": 1685
+    },
+    {
+      "epoch": 1.9543577483155836,
+      "grad_norm": 0.22775276005268097,
+      "learning_rate": 4.031034482758621e-05,
+      "loss": 0.1368,
+      "step": 1686
+    },
+    {
+      "epoch": 1.9555169166123307,
+      "grad_norm": 0.25551965832710266,
+      "learning_rate": 4.0304597701149424e-05,
+      "loss": 0.1395,
+      "step": 1687
+    },
+    {
+      "epoch": 1.9566760849090779,
+      "grad_norm": 0.20460672676563263,
+      "learning_rate": 4.0298850574712645e-05,
+      "loss": 0.1111,
+      "step": 1688
+    },
+    {
+      "epoch": 1.957835253205825,
+      "grad_norm": 0.42244213819503784,
+      "learning_rate": 4.029310344827587e-05,
+      "loss": 0.1504,
+      "step": 1689
+    },
+    {
+      "epoch": 1.9589944215025719,
+      "grad_norm": 0.2375163733959198,
+      "learning_rate": 4.028735632183908e-05,
+      "loss": 0.1243,
+      "step": 1690
+    },
+    {
+      "epoch": 1.960153589799319,
+      "grad_norm": 0.23860155045986176,
+      "learning_rate": 4.0281609195402304e-05,
+      "loss": 0.1317,
+      "step": 1691
+    },
+    {
+      "epoch": 1.961312758096066,
+      "grad_norm": 0.20756982266902924,
+      "learning_rate": 4.027586206896552e-05,
+      "loss": 0.1252,
+      "step": 1692
+    },
+    {
+      "epoch": 1.9624719263928132,
+      "grad_norm": 0.25197523832321167,
+      "learning_rate": 4.0270114942528733e-05,
+      "loss": 0.1406,
+      "step": 1693
+    },
+    {
+      "epoch": 1.96363109468956,
+      "grad_norm": 0.2613564133644104,
+      "learning_rate": 4.0264367816091955e-05,
+      "loss": 0.1342,
+      "step": 1694
+    },
+    {
+      "epoch": 1.9647902629863072,
+      "grad_norm": 0.21359771490097046,
+      "learning_rate": 4.025862068965518e-05,
+      "loss": 0.1171,
+      "step": 1695
+    },
+    {
+      "epoch": 1.9659494312830543,
+      "grad_norm": 0.2883491516113281,
+      "learning_rate": 4.025287356321839e-05,
+      "loss": 0.1262,
+      "step": 1696
+    },
+    {
+      "epoch": 1.9671085995798014,
+      "grad_norm": 0.22171908617019653,
+      "learning_rate": 4.0247126436781613e-05,
+      "loss": 0.1259,
+      "step": 1697
+    },
+    {
+      "epoch": 1.9682677678765486,
+      "grad_norm": 0.2951849699020386,
+      "learning_rate": 4.024137931034483e-05,
+      "loss": 0.1447,
+      "step": 1698
+    },
+    {
+      "epoch": 1.9694269361732957,
+      "grad_norm": 0.2592881917953491,
+      "learning_rate": 4.023563218390804e-05,
+      "loss": 0.1294,
+      "step": 1699
+    },
+    {
+      "epoch": 1.9705861044700428,
+      "grad_norm": 0.22371990978717804,
+      "learning_rate": 4.0229885057471265e-05,
+      "loss": 0.1238,
+      "step": 1700
+    },
+    {
+      "epoch": 1.97174527276679,
+      "grad_norm": 0.23679739236831665,
+      "learning_rate": 4.0224137931034487e-05,
+      "loss": 0.128,
+      "step": 1701
+    },
+    {
+      "epoch": 1.972904441063537,
+      "grad_norm": 0.208289235830307,
+      "learning_rate": 4.021839080459771e-05,
+      "loss": 0.1221,
+      "step": 1702
+    },
+    {
+      "epoch": 1.9740636093602841,
+      "grad_norm": 0.3069906532764435,
+      "learning_rate": 4.021264367816092e-05,
+      "loss": 0.1264,
+      "step": 1703
+    },
+    {
+      "epoch": 1.975222777657031,
+      "grad_norm": 0.23526400327682495,
+      "learning_rate": 4.020689655172414e-05,
+      "loss": 0.1314,
+      "step": 1704
+    },
+    {
+      "epoch": 1.9763819459537781,
+      "grad_norm": 0.36365655064582825,
+      "learning_rate": 4.020114942528736e-05,
+      "loss": 0.135,
+      "step": 1705
+    },
+    {
+      "epoch": 1.9775411142505253,
+      "grad_norm": 0.21487149596214294,
+      "learning_rate": 4.0195402298850575e-05,
+      "loss": 0.1166,
+      "step": 1706
+    },
+    {
+      "epoch": 1.9787002825472724,
+      "grad_norm": 0.1872834861278534,
+      "learning_rate": 4.018965517241379e-05,
+      "loss": 0.114,
+      "step": 1707
+    },
+    {
+      "epoch": 1.9798594508440193,
+      "grad_norm": 0.2666516602039337,
+      "learning_rate": 4.018390804597701e-05,
+      "loss": 0.1396,
+      "step": 1708
+    },
+    {
+      "epoch": 1.9810186191407664,
+      "grad_norm": 0.23788093030452728,
+      "learning_rate": 4.017816091954023e-05,
+      "loss": 0.1303,
+      "step": 1709
+    },
+    {
+      "epoch": 1.9821777874375135,
+      "grad_norm": 0.23278290033340454,
+      "learning_rate": 4.0172413793103455e-05,
+      "loss": 0.127,
+      "step": 1710
+    },
+    {
+      "epoch": 1.9833369557342606,
+      "grad_norm": 0.2589450180530548,
+      "learning_rate": 4.016666666666667e-05,
+      "loss": 0.1311,
+      "step": 1711
+    },
+    {
+      "epoch": 1.9844961240310077,
+      "grad_norm": 0.20115303993225098,
+      "learning_rate": 4.0160919540229884e-05,
+      "loss": 0.1255,
+      "step": 1712
+    },
+    {
+      "epoch": 1.9856552923277548,
+      "grad_norm": 0.24653758108615875,
+      "learning_rate": 4.0155172413793106e-05,
+      "loss": 0.1346,
+      "step": 1713
+    },
+    {
+      "epoch": 1.986814460624502,
+      "grad_norm": 0.25631844997406006,
+      "learning_rate": 4.014942528735632e-05,
+      "loss": 0.1385,
+      "step": 1714
+    },
+    {
+      "epoch": 1.987973628921249,
+      "grad_norm": 0.21562731266021729,
+      "learning_rate": 4.014367816091954e-05,
+      "loss": 0.118,
+      "step": 1715
+    },
+    {
+      "epoch": 1.9891327972179962,
+      "grad_norm": 0.20288534462451935,
+      "learning_rate": 4.0137931034482764e-05,
+      "loss": 0.1183,
+      "step": 1716
+    },
+    {
+      "epoch": 1.9902919655147433,
+      "grad_norm": 0.2522091865539551,
+      "learning_rate": 4.013218390804598e-05,
+      "loss": 0.1337,
+      "step": 1717
+    },
+    {
+      "epoch": 1.9914511338114904,
+      "grad_norm": 0.2482483983039856,
+      "learning_rate": 4.0126436781609194e-05,
+      "loss": 0.1347,
+      "step": 1718
+    },
+    {
+      "epoch": 1.9926103021082373,
+      "grad_norm": 0.26897165179252625,
+      "learning_rate": 4.0120689655172416e-05,
+      "loss": 0.1463,
+      "step": 1719
+    },
+    {
+      "epoch": 1.9937694704049844,
+      "grad_norm": 0.2053130716085434,
+      "learning_rate": 4.011494252873563e-05,
+      "loss": 0.1138,
+      "step": 1720
+    },
+    {
+      "epoch": 1.9949286387017315,
+      "grad_norm": 0.22863611578941345,
+      "learning_rate": 4.010919540229885e-05,
+      "loss": 0.1255,
+      "step": 1721
+    },
+    {
+      "epoch": 1.9960878069984784,
+      "grad_norm": 0.25083255767822266,
+      "learning_rate": 4.0103448275862074e-05,
+      "loss": 0.137,
+      "step": 1722
+    },
+    {
+      "epoch": 1.9972469752952255,
+      "grad_norm": 0.2726476788520813,
+      "learning_rate": 4.009770114942529e-05,
+      "loss": 0.1459,
+      "step": 1723
+    },
+    {
+      "epoch": 1.9984061435919727,
+      "grad_norm": 0.18482446670532227,
+      "learning_rate": 4.009195402298851e-05,
+      "loss": 0.1116,
+      "step": 1724
+    },
+    {
+      "epoch": 1.9995653118887198,
+      "grad_norm": 0.22042682766914368,
+      "learning_rate": 4.0086206896551726e-05,
+      "loss": 0.1282,
+      "step": 1725
+    },
+    {
+      "epoch": 1.9995653118887198,
+      "eval_loss": 0.13071221113204956,
+      "eval_runtime": 265.8343,
+      "eval_samples_per_second": 5.771,
+      "eval_steps_per_second": 5.771,
+      "step": 1725
+    },
+    {
+      "epoch": 2.000724480185467,
+      "grad_norm": 0.20493850111961365,
+      "learning_rate": 4.008045977011494e-05,
+      "loss": 0.115,
+      "step": 1726
+    },
+    {
+      "epoch": 2.001883648482214,
+      "grad_norm": 0.21877332031726837,
+      "learning_rate": 4.007471264367816e-05,
+      "loss": 0.1288,
+      "step": 1727
+    },
+    {
+      "epoch": 2.003042816778961,
+      "grad_norm": 0.17066803574562073,
+      "learning_rate": 4.006896551724138e-05,
+      "loss": 0.103,
+      "step": 1728
+    },
+    {
+      "epoch": 2.0042019850757082,
+      "grad_norm": 0.2311452031135559,
+      "learning_rate": 4.0063218390804605e-05,
+      "loss": 0.123,
+      "step": 1729
+    },
+    {
+      "epoch": 2.0053611533724554,
+      "grad_norm": 0.18617844581604004,
+      "learning_rate": 4.005747126436782e-05,
+      "loss": 0.1075,
+      "step": 1730
+    },
+    {
+      "epoch": 2.0065203216692025,
+      "grad_norm": 0.17932945489883423,
+      "learning_rate": 4.0051724137931035e-05,
+      "loss": 0.1124,
+      "step": 1731
+    },
+    {
+      "epoch": 2.0076794899659496,
+      "grad_norm": 0.2430291473865509,
+      "learning_rate": 4.004597701149426e-05,
+      "loss": 0.112,
+      "step": 1732
+    },
+    {
+      "epoch": 2.0088386582626967,
+      "grad_norm": 0.1980556845664978,
+      "learning_rate": 4.004022988505747e-05,
+      "loss": 0.1257,
+      "step": 1733
+    },
+    {
+      "epoch": 2.009997826559444,
+      "grad_norm": 0.2553955614566803,
+      "learning_rate": 4.003448275862069e-05,
+      "loss": 0.1168,
+      "step": 1734
+    },
+    {
+      "epoch": 2.0111569948561905,
+      "grad_norm": 0.5730363130569458,
+      "learning_rate": 4.002873563218391e-05,
+      "loss": 0.1191,
+      "step": 1735
+    },
+    {
+      "epoch": 2.0123161631529376,
+      "grad_norm": 0.27074134349823,
+      "learning_rate": 4.002298850574713e-05,
+      "loss": 0.1228,
+      "step": 1736
+    },
+    {
+      "epoch": 2.0134753314496847,
+      "grad_norm": 0.2566532492637634,
+      "learning_rate": 4.0017241379310345e-05,
+      "loss": 0.1185,
+      "step": 1737
+    },
+    {
+      "epoch": 2.014634499746432,
+      "grad_norm": 0.2553810179233551,
+      "learning_rate": 4.001149425287357e-05,
+      "loss": 0.1207,
+      "step": 1738
+    },
+    {
+      "epoch": 2.015793668043179,
+      "grad_norm": 0.314544677734375,
+      "learning_rate": 4.000574712643678e-05,
+      "loss": 0.1338,
+      "step": 1739
+    },
+    {
+      "epoch": 2.016952836339926,
+      "grad_norm": 0.26933231949806213,
+      "learning_rate": 4e-05,
+      "loss": 0.1175,
+      "step": 1740
+    },
+    {
+      "epoch": 2.018112004636673,
+      "grad_norm": 0.4147557020187378,
+      "learning_rate": 3.999425287356322e-05,
+      "loss": 0.1191,
+      "step": 1741
+    },
+    {
+      "epoch": 2.0192711729334203,
+      "grad_norm": 0.4365951716899872,
+      "learning_rate": 3.998850574712644e-05,
+      "loss": 0.1246,
+      "step": 1742
+    },
+    {
+      "epoch": 2.0204303412301674,
+      "grad_norm": 0.22776885330677032,
+      "learning_rate": 3.998275862068966e-05,
+      "loss": 0.1277,
+      "step": 1743
+    },
+    {
+      "epoch": 2.0215895095269145,
+      "grad_norm": 0.5201489925384521,
+      "learning_rate": 3.9977011494252876e-05,
+      "loss": 0.108,
+      "step": 1744
+    },
+    {
+      "epoch": 2.0227486778236616,
+      "grad_norm": 0.2805781960487366,
+      "learning_rate": 3.997126436781609e-05,
+      "loss": 0.1168,
+      "step": 1745
+    },
+    {
+      "epoch": 2.0239078461204087,
+      "grad_norm": 0.25759008526802063,
+      "learning_rate": 3.996551724137931e-05,
+      "loss": 0.1312,
+      "step": 1746
+    },
+    {
+      "epoch": 2.025067014417156,
+      "grad_norm": 0.2724314332008362,
+      "learning_rate": 3.995977011494253e-05,
+      "loss": 0.1102,
+      "step": 1747
+    },
+    {
+      "epoch": 2.026226182713903,
+      "grad_norm": 0.2714131772518158,
+      "learning_rate": 3.995402298850575e-05,
+      "loss": 0.1136,
+      "step": 1748
+    },
+    {
+      "epoch": 2.0273853510106497,
+      "grad_norm": 0.5604748725891113,
+      "learning_rate": 3.994827586206897e-05,
+      "loss": 0.1387,
+      "step": 1749
+    },
+    {
+      "epoch": 2.0285445193073968,
+      "grad_norm": 0.20117241144180298,
+      "learning_rate": 3.9942528735632186e-05,
+      "loss": 0.1045,
+      "step": 1750
+    },
+    {
+      "epoch": 2.029703687604144,
+      "grad_norm": 0.2628777027130127,
+      "learning_rate": 3.993678160919541e-05,
+      "loss": 0.1219,
+      "step": 1751
+    },
+    {
+      "epoch": 2.030862855900891,
+      "grad_norm": 0.29843881726264954,
+      "learning_rate": 3.993103448275862e-05,
+      "loss": 0.1282,
+      "step": 1752
+    },
+    {
+      "epoch": 2.032022024197638,
+      "grad_norm": 0.4298798441886902,
+      "learning_rate": 3.992528735632184e-05,
+      "loss": 0.1108,
+      "step": 1753
+    },
+    {
+      "epoch": 2.0331811924943852,
+      "grad_norm": 0.24070972204208374,
+      "learning_rate": 3.991954022988506e-05,
+      "loss": 0.1171,
+      "step": 1754
+    },
+    {
+      "epoch": 2.0343403607911323,
+      "grad_norm": 0.2516588568687439,
+      "learning_rate": 3.9913793103448274e-05,
+      "loss": 0.1178,
+      "step": 1755
+    },
+    {
+      "epoch": 2.0354995290878795,
+      "grad_norm": 0.3218182623386383,
+      "learning_rate": 3.9908045977011496e-05,
+      "loss": 0.1193,
+      "step": 1756
+    },
+    {
+      "epoch": 2.0366586973846266,
+      "grad_norm": 0.2040579468011856,
+      "learning_rate": 3.990229885057472e-05,
+      "loss": 0.1125,
+      "step": 1757
+    },
+    {
+      "epoch": 2.0378178656813737,
+      "grad_norm": 0.2935560643672943,
+      "learning_rate": 3.989655172413793e-05,
+      "loss": 0.1269,
+      "step": 1758
+    },
+    {
+      "epoch": 2.038977033978121,
+      "grad_norm": 0.22673268616199493,
+      "learning_rate": 3.9890804597701154e-05,
+      "loss": 0.1145,
+      "step": 1759
+    },
+    {
+      "epoch": 2.040136202274868,
+      "grad_norm": 0.37315088510513306,
+      "learning_rate": 3.988505747126437e-05,
+      "loss": 0.1226,
+      "step": 1760
+    },
+    {
+      "epoch": 2.041295370571615,
+      "grad_norm": 0.21398569643497467,
+      "learning_rate": 3.9879310344827584e-05,
+      "loss": 0.1196,
+      "step": 1761
+    },
+    {
+      "epoch": 2.042454538868362,
+      "grad_norm": 0.3148459494113922,
+      "learning_rate": 3.9873563218390806e-05,
+      "loss": 0.1321,
+      "step": 1762
+    },
+    {
+      "epoch": 2.043613707165109,
+      "grad_norm": 0.5063501596450806,
+      "learning_rate": 3.986781609195403e-05,
+      "loss": 0.1237,
+      "step": 1763
+    },
+    {
+      "epoch": 2.044772875461856,
+      "grad_norm": 0.25269943475723267,
+      "learning_rate": 3.986206896551724e-05,
+      "loss": 0.1149,
+      "step": 1764
+    },
+    {
+      "epoch": 2.045932043758603,
+      "grad_norm": 0.32295992970466614,
+      "learning_rate": 3.9856321839080464e-05,
+      "loss": 0.1087,
+      "step": 1765
+    },
+    {
+      "epoch": 2.04709121205535,
+      "grad_norm": 0.23823142051696777,
+      "learning_rate": 3.985057471264368e-05,
+      "loss": 0.1246,
+      "step": 1766
+    },
+    {
+      "epoch": 2.0482503803520973,
+      "grad_norm": 0.3159748315811157,
+      "learning_rate": 3.98448275862069e-05,
+      "loss": 0.1114,
+      "step": 1767
+    },
+    {
+      "epoch": 2.0494095486488444,
+      "grad_norm": 0.21690689027309418,
+      "learning_rate": 3.9839080459770115e-05,
+      "loss": 0.1105,
+      "step": 1768
+    },
+    {
+      "epoch": 2.0505687169455915,
+      "grad_norm": 0.26003125309944153,
+      "learning_rate": 3.983333333333333e-05,
+      "loss": 0.1257,
+      "step": 1769
+    },
+    {
+      "epoch": 2.0517278852423386,
+      "grad_norm": 0.27152520418167114,
+      "learning_rate": 3.982758620689656e-05,
+      "loss": 0.1207,
+      "step": 1770
+    },
+    {
+      "epoch": 2.0528870535390857,
+      "grad_norm": 0.2450665384531021,
+      "learning_rate": 3.9821839080459774e-05,
+      "loss": 0.1198,
+      "step": 1771
+    },
+    {
+      "epoch": 2.054046221835833,
+      "grad_norm": 0.20077472925186157,
+      "learning_rate": 3.981609195402299e-05,
+      "loss": 0.107,
+      "step": 1772
+    },
+    {
+      "epoch": 2.05520539013258,
+      "grad_norm": 0.4394693374633789,
+      "learning_rate": 3.981034482758621e-05,
+      "loss": 0.1103,
+      "step": 1773
+    },
+    {
+      "epoch": 2.056364558429327,
+      "grad_norm": 0.26243650913238525,
+      "learning_rate": 3.9804597701149425e-05,
+      "loss": 0.1128,
+      "step": 1774
+    },
+    {
+      "epoch": 2.057523726726074,
+      "grad_norm": 0.2425994873046875,
+      "learning_rate": 3.979885057471265e-05,
+      "loss": 0.123,
+      "step": 1775
+    },
+    {
+      "epoch": 2.0586828950228213,
+      "grad_norm": 0.23687340319156647,
+      "learning_rate": 3.979310344827586e-05,
+      "loss": 0.122,
+      "step": 1776
+    },
+    {
+      "epoch": 2.0598420633195684,
+      "grad_norm": 0.2466685026884079,
+      "learning_rate": 3.9787356321839083e-05,
+      "loss": 0.1201,
+      "step": 1777
+    },
+    {
+      "epoch": 2.061001231616315,
+      "grad_norm": 0.26380082964897156,
+      "learning_rate": 3.9781609195402305e-05,
+      "loss": 0.1216,
+      "step": 1778
+    },
+    {
+      "epoch": 2.062160399913062,
+      "grad_norm": 0.2581118643283844,
+      "learning_rate": 3.977586206896552e-05,
+      "loss": 0.1261,
+      "step": 1779
+    },
+    {
+      "epoch": 2.0633195682098093,
+      "grad_norm": 0.24851976335048676,
+      "learning_rate": 3.9770114942528735e-05,
+      "loss": 0.1212,
+      "step": 1780
+    },
+    {
+      "epoch": 2.0644787365065564,
+      "grad_norm": 0.3059890568256378,
+      "learning_rate": 3.9764367816091957e-05,
+      "loss": 0.1233,
+      "step": 1781
+    },
+    {
+      "epoch": 2.0656379048033036,
+      "grad_norm": 0.2686820924282074,
+      "learning_rate": 3.975862068965517e-05,
+      "loss": 0.1291,
+      "step": 1782
+    },
+    {
+      "epoch": 2.0667970731000507,
+      "grad_norm": 0.24418821930885315,
+      "learning_rate": 3.975287356321839e-05,
+      "loss": 0.115,
+      "step": 1783
+    },
+    {
+      "epoch": 2.067956241396798,
+      "grad_norm": 0.27120018005371094,
+      "learning_rate": 3.9747126436781615e-05,
+      "loss": 0.1192,
+      "step": 1784
+    },
+    {
+      "epoch": 2.069115409693545,
+      "grad_norm": 0.22046048939228058,
+      "learning_rate": 3.974137931034483e-05,
+      "loss": 0.1033,
+      "step": 1785
+    },
+    {
+      "epoch": 2.070274577990292,
+      "grad_norm": 0.24162054061889648,
+      "learning_rate": 3.973563218390805e-05,
+      "loss": 0.116,
+      "step": 1786
+    },
+    {
+      "epoch": 2.071433746287039,
+      "grad_norm": 0.26987624168395996,
+      "learning_rate": 3.9729885057471266e-05,
+      "loss": 0.127,
+      "step": 1787
+    },
+    {
+      "epoch": 2.0725929145837862,
+      "grad_norm": 0.2681949734687805,
+      "learning_rate": 3.972413793103448e-05,
+      "loss": 0.1266,
+      "step": 1788
+    },
+    {
+      "epoch": 2.0737520828805334,
+      "grad_norm": 0.26860347390174866,
+      "learning_rate": 3.97183908045977e-05,
+      "loss": 0.1186,
+      "step": 1789
+    },
+    {
+      "epoch": 2.0749112511772805,
+      "grad_norm": 0.2559060752391815,
+      "learning_rate": 3.9712643678160925e-05,
+      "loss": 0.108,
+      "step": 1790
+    },
+    {
+      "epoch": 2.0760704194740276,
+      "grad_norm": 0.1907990276813507,
+      "learning_rate": 3.970689655172414e-05,
+      "loss": 0.0989,
+      "step": 1791
+    },
+    {
+      "epoch": 2.0772295877707743,
+      "grad_norm": 0.2304263859987259,
+      "learning_rate": 3.970114942528736e-05,
+      "loss": 0.1186,
+      "step": 1792
+    },
+    {
+      "epoch": 2.0783887560675214,
+      "grad_norm": 0.21806973218917847,
+      "learning_rate": 3.9695402298850576e-05,
+      "loss": 0.1055,
+      "step": 1793
+    },
+    {
+      "epoch": 2.0795479243642685,
+      "grad_norm": 0.27678120136260986,
+      "learning_rate": 3.96896551724138e-05,
+      "loss": 0.1309,
+      "step": 1794
+    },
+    {
+      "epoch": 2.0807070926610156,
+      "grad_norm": 0.25782880187034607,
+      "learning_rate": 3.968390804597701e-05,
+      "loss": 0.1066,
+      "step": 1795
+    },
+    {
+      "epoch": 2.0818662609577627,
+      "grad_norm": 0.2833830416202545,
+      "learning_rate": 3.967816091954023e-05,
+      "loss": 0.1258,
+      "step": 1796
+    },
+    {
+      "epoch": 2.08302542925451,
+      "grad_norm": 0.231010302901268,
+      "learning_rate": 3.967241379310345e-05,
+      "loss": 0.1035,
+      "step": 1797
+    },
+    {
+      "epoch": 2.084184597551257,
+      "grad_norm": 0.3302299678325653,
+      "learning_rate": 3.966666666666667e-05,
+      "loss": 0.1137,
+      "step": 1798
+    },
+    {
+      "epoch": 2.085343765848004,
+      "grad_norm": 0.270266056060791,
+      "learning_rate": 3.9660919540229886e-05,
+      "loss": 0.1117,
+      "step": 1799
+    },
+    {
+      "epoch": 2.086502934144751,
+      "grad_norm": 0.23533837497234344,
+      "learning_rate": 3.965517241379311e-05,
+      "loss": 0.1112,
+      "step": 1800
+    },
+    {
+      "epoch": 2.0876621024414983,
+      "grad_norm": 0.316643089056015,
+      "learning_rate": 3.964942528735632e-05,
+      "loss": 0.1243,
+      "step": 1801
+    },
+    {
+      "epoch": 2.0888212707382454,
+      "grad_norm": 0.2741736173629761,
+      "learning_rate": 3.964367816091954e-05,
+      "loss": 0.1095,
+      "step": 1802
+    },
+    {
+      "epoch": 2.0899804390349925,
+      "grad_norm": 0.2951594889163971,
+      "learning_rate": 3.963793103448276e-05,
+      "loss": 0.1246,
+      "step": 1803
+    },
+    {
+      "epoch": 2.0911396073317396,
+      "grad_norm": 0.26919370889663696,
+      "learning_rate": 3.963218390804598e-05,
+      "loss": 0.1181,
+      "step": 1804
+    },
+    {
+      "epoch": 2.0922987756284868,
+      "grad_norm": 0.31926748156547546,
+      "learning_rate": 3.96264367816092e-05,
+      "loss": 0.1216,
+      "step": 1805
+    },
+    {
+      "epoch": 2.0934579439252334,
+      "grad_norm": 0.30833154916763306,
+      "learning_rate": 3.962068965517242e-05,
+      "loss": 0.1245,
+      "step": 1806
+    },
+    {
+      "epoch": 2.0946171122219805,
+      "grad_norm": 0.2637370824813843,
+      "learning_rate": 3.961494252873563e-05,
+      "loss": 0.1168,
+      "step": 1807
+    },
+    {
+      "epoch": 2.0957762805187277,
+      "grad_norm": 0.2496730089187622,
+      "learning_rate": 3.9609195402298854e-05,
+      "loss": 0.1152,
+      "step": 1808
+    },
+    {
+      "epoch": 2.096935448815475,
+      "grad_norm": 0.241145521402359,
+      "learning_rate": 3.960344827586207e-05,
+      "loss": 0.1257,
+      "step": 1809
+    },
+    {
+      "epoch": 2.098094617112222,
+      "grad_norm": 0.2335578054189682,
+      "learning_rate": 3.959770114942529e-05,
+      "loss": 0.1138,
+      "step": 1810
+    },
+    {
+      "epoch": 2.099253785408969,
+      "grad_norm": 0.23772041499614716,
+      "learning_rate": 3.959195402298851e-05,
+      "loss": 0.1115,
+      "step": 1811
+    },
+    {
+      "epoch": 2.100412953705716,
+      "grad_norm": 0.25079551339149475,
+      "learning_rate": 3.958620689655173e-05,
+      "loss": 0.1185,
+      "step": 1812
+    },
+    {
+      "epoch": 2.1015721220024632,
+      "grad_norm": 0.26366278529167175,
+      "learning_rate": 3.958045977011495e-05,
+      "loss": 0.1177,
+      "step": 1813
+    },
+    {
+      "epoch": 2.1027312902992104,
+      "grad_norm": 0.2486356645822525,
+      "learning_rate": 3.9574712643678164e-05,
+      "loss": 0.1204,
+      "step": 1814
+    },
+    {
+      "epoch": 2.1038904585959575,
+      "grad_norm": 0.24018944799900055,
+      "learning_rate": 3.956896551724138e-05,
+      "loss": 0.1333,
+      "step": 1815
+    },
+    {
+      "epoch": 2.1050496268927046,
+      "grad_norm": 0.25310537219047546,
+      "learning_rate": 3.95632183908046e-05,
+      "loss": 0.1243,
+      "step": 1816
+    },
+    {
+      "epoch": 2.1062087951894517,
+      "grad_norm": 0.26761263608932495,
+      "learning_rate": 3.9557471264367815e-05,
+      "loss": 0.1184,
+      "step": 1817
+    },
+    {
+      "epoch": 2.107367963486199,
+      "grad_norm": 0.32471829652786255,
+      "learning_rate": 3.955172413793104e-05,
+      "loss": 0.1234,
+      "step": 1818
+    },
+    {
+      "epoch": 2.108527131782946,
+      "grad_norm": 0.25897079706192017,
+      "learning_rate": 3.954597701149426e-05,
+      "loss": 0.1142,
+      "step": 1819
+    },
+    {
+      "epoch": 2.1096863000796926,
+      "grad_norm": 0.2632877826690674,
+      "learning_rate": 3.954022988505747e-05,
+      "loss": 0.1189,
+      "step": 1820
+    },
+    {
+      "epoch": 2.1108454683764397,
+      "grad_norm": 0.3021185100078583,
+      "learning_rate": 3.953448275862069e-05,
+      "loss": 0.1283,
+      "step": 1821
+    },
+    {
+      "epoch": 2.112004636673187,
+      "grad_norm": 0.2785482704639435,
+      "learning_rate": 3.952873563218391e-05,
+      "loss": 0.1302,
+      "step": 1822
+    },
+    {
+      "epoch": 2.113163804969934,
+      "grad_norm": 0.33356723189353943,
+      "learning_rate": 3.9522988505747125e-05,
+      "loss": 0.1194,
+      "step": 1823
+    },
+    {
+      "epoch": 2.114322973266681,
+      "grad_norm": 0.29633092880249023,
+      "learning_rate": 3.9517241379310346e-05,
+      "loss": 0.1243,
+      "step": 1824
+    },
+    {
+      "epoch": 2.115482141563428,
+      "grad_norm": 0.2905955910682678,
+      "learning_rate": 3.951149425287357e-05,
+      "loss": 0.1128,
+      "step": 1825
+    },
+    {
+      "epoch": 2.1166413098601753,
+      "grad_norm": 0.24889424443244934,
+      "learning_rate": 3.950574712643678e-05,
+      "loss": 0.1029,
+      "step": 1826
+    },
+    {
+      "epoch": 2.1178004781569224,
+      "grad_norm": 0.24264870584011078,
+      "learning_rate": 3.9500000000000005e-05,
+      "loss": 0.1125,
+      "step": 1827
+    },
+    {
+      "epoch": 2.1189596464536695,
+      "grad_norm": 0.2387196123600006,
+      "learning_rate": 3.949425287356322e-05,
+      "loss": 0.1123,
+      "step": 1828
+    },
+    {
+      "epoch": 2.1201188147504166,
+      "grad_norm": 0.23533257842063904,
+      "learning_rate": 3.9488505747126434e-05,
+      "loss": 0.118,
+      "step": 1829
+    },
+    {
+      "epoch": 2.1212779830471638,
+      "grad_norm": 0.24829024076461792,
+      "learning_rate": 3.9482758620689656e-05,
+      "loss": 0.1194,
+      "step": 1830
+    },
+    {
+      "epoch": 2.122437151343911,
+      "grad_norm": 0.26718825101852417,
+      "learning_rate": 3.947701149425288e-05,
+      "loss": 0.1139,
+      "step": 1831
+    },
+    {
+      "epoch": 2.123596319640658,
+      "grad_norm": 0.24464312195777893,
+      "learning_rate": 3.94712643678161e-05,
+      "loss": 0.1101,
+      "step": 1832
+    },
+    {
+      "epoch": 2.124755487937405,
+      "grad_norm": 0.22388160228729248,
+      "learning_rate": 3.9465517241379314e-05,
+      "loss": 0.1088,
+      "step": 1833
+    },
+    {
+      "epoch": 2.125914656234152,
+      "grad_norm": 0.25908219814300537,
+      "learning_rate": 3.945977011494253e-05,
+      "loss": 0.1202,
+      "step": 1834
+    },
+    {
+      "epoch": 2.127073824530899,
+      "grad_norm": 0.2020428478717804,
+      "learning_rate": 3.945402298850575e-05,
+      "loss": 0.108,
+      "step": 1835
+    },
+    {
+      "epoch": 2.128232992827646,
+      "grad_norm": 0.26430872082710266,
+      "learning_rate": 3.9448275862068966e-05,
+      "loss": 0.1235,
+      "step": 1836
+    },
+    {
+      "epoch": 2.129392161124393,
+      "grad_norm": 0.3125210702419281,
+      "learning_rate": 3.944252873563218e-05,
+      "loss": 0.1134,
+      "step": 1837
+    },
+    {
+      "epoch": 2.1305513294211402,
+      "grad_norm": 0.24071168899536133,
+      "learning_rate": 3.94367816091954e-05,
+      "loss": 0.1153,
+      "step": 1838
+    },
+    {
+      "epoch": 2.1317104977178873,
+      "grad_norm": 0.22722084820270538,
+      "learning_rate": 3.9431034482758624e-05,
+      "loss": 0.1135,
+      "step": 1839
+    },
+    {
+      "epoch": 2.1328696660146345,
+      "grad_norm": 0.2722453474998474,
+      "learning_rate": 3.942528735632184e-05,
+      "loss": 0.1159,
+      "step": 1840
+    },
+    {
+      "epoch": 2.1340288343113816,
+      "grad_norm": 0.26020506024360657,
+      "learning_rate": 3.941954022988506e-05,
+      "loss": 0.1083,
+      "step": 1841
+    },
+    {
+      "epoch": 2.1351880026081287,
+      "grad_norm": 0.38387203216552734,
+      "learning_rate": 3.9413793103448276e-05,
+      "loss": 0.1304,
+      "step": 1842
+    },
+    {
+      "epoch": 2.136347170904876,
+      "grad_norm": 0.28882285952568054,
+      "learning_rate": 3.94080459770115e-05,
+      "loss": 0.1188,
+      "step": 1843
+    },
+    {
+      "epoch": 2.137506339201623,
+      "grad_norm": 0.2755765914916992,
+      "learning_rate": 3.940229885057471e-05,
+      "loss": 0.1346,
+      "step": 1844
+    },
+    {
+      "epoch": 2.13866550749837,
+      "grad_norm": 0.2223798632621765,
+      "learning_rate": 3.9396551724137934e-05,
+      "loss": 0.1112,
+      "step": 1845
+    },
+    {
+      "epoch": 2.139824675795117,
+      "grad_norm": 0.30250582098960876,
+      "learning_rate": 3.9390804597701156e-05,
+      "loss": 0.1267,
+      "step": 1846
+    },
+    {
+      "epoch": 2.1409838440918643,
+      "grad_norm": 0.2759857177734375,
+      "learning_rate": 3.938505747126437e-05,
+      "loss": 0.1198,
+      "step": 1847
+    },
+    {
+      "epoch": 2.1421430123886114,
+      "grad_norm": 0.2635420560836792,
+      "learning_rate": 3.9379310344827585e-05,
+      "loss": 0.116,
+      "step": 1848
+    },
+    {
+      "epoch": 2.143302180685358,
+      "grad_norm": 0.22998470067977905,
+      "learning_rate": 3.937356321839081e-05,
+      "loss": 0.1124,
+      "step": 1849
+    },
+    {
+      "epoch": 2.144461348982105,
+      "grad_norm": 0.22204086184501648,
+      "learning_rate": 3.936781609195402e-05,
+      "loss": 0.1087,
+      "step": 1850
+    },
+    {
+      "epoch": 2.1456205172788523,
+      "grad_norm": 0.2902623116970062,
+      "learning_rate": 3.9362068965517244e-05,
+      "loss": 0.1307,
+      "step": 1851
+    },
+    {
+      "epoch": 2.1467796855755994,
+      "grad_norm": 0.2636187970638275,
+      "learning_rate": 3.9356321839080465e-05,
+      "loss": 0.1168,
+      "step": 1852
+    },
+    {
+      "epoch": 2.1479388538723465,
+      "grad_norm": 0.29677683115005493,
+      "learning_rate": 3.935057471264368e-05,
+      "loss": 0.1328,
+      "step": 1853
+    },
+    {
+      "epoch": 2.1490980221690936,
+      "grad_norm": 0.2324220985174179,
+      "learning_rate": 3.93448275862069e-05,
+      "loss": 0.1198,
+      "step": 1854
+    },
+    {
+      "epoch": 2.1502571904658407,
+      "grad_norm": 0.2689639627933502,
+      "learning_rate": 3.933908045977012e-05,
+      "loss": 0.1285,
+      "step": 1855
+    },
+    {
+      "epoch": 2.151416358762588,
+      "grad_norm": 0.264710932970047,
+      "learning_rate": 3.933333333333333e-05,
+      "loss": 0.1235,
+      "step": 1856
+    },
+    {
+      "epoch": 2.152575527059335,
+      "grad_norm": 0.26649346947669983,
+      "learning_rate": 3.932758620689655e-05,
+      "loss": 0.1212,
+      "step": 1857
+    },
+    {
+      "epoch": 2.153734695356082,
+      "grad_norm": 0.27043336629867554,
+      "learning_rate": 3.932183908045977e-05,
+      "loss": 0.1123,
+      "step": 1858
+    },
+    {
+      "epoch": 2.154893863652829,
+      "grad_norm": 0.24677076935768127,
+      "learning_rate": 3.931609195402299e-05,
+      "loss": 0.1058,
+      "step": 1859
+    },
+    {
+      "epoch": 2.1560530319495763,
+      "grad_norm": 0.3722219169139862,
+      "learning_rate": 3.931034482758621e-05,
+      "loss": 0.13,
+      "step": 1860
+    },
+    {
+      "epoch": 2.1572122002463234,
+      "grad_norm": 0.3465796411037445,
+      "learning_rate": 3.9304597701149427e-05,
+      "loss": 0.1323,
+      "step": 1861
+    },
+    {
+      "epoch": 2.1583713685430705,
+      "grad_norm": 0.24927546083927155,
+      "learning_rate": 3.929885057471265e-05,
+      "loss": 0.1133,
+      "step": 1862
+    },
+    {
+      "epoch": 2.159530536839817,
+      "grad_norm": 0.2714444696903229,
+      "learning_rate": 3.929310344827586e-05,
+      "loss": 0.114,
+      "step": 1863
+    },
+    {
+      "epoch": 2.1606897051365643,
+      "grad_norm": 0.2666366994380951,
+      "learning_rate": 3.928735632183908e-05,
+      "loss": 0.1182,
+      "step": 1864
+    },
+    {
+      "epoch": 2.1618488734333114,
+      "grad_norm": 0.24803896248340607,
+      "learning_rate": 3.92816091954023e-05,
+      "loss": 0.1162,
+      "step": 1865
+    },
+    {
+      "epoch": 2.1630080417300586,
+      "grad_norm": 0.23639126121997833,
+      "learning_rate": 3.927586206896552e-05,
+      "loss": 0.1154,
+      "step": 1866
+    },
+    {
+      "epoch": 2.1641672100268057,
+      "grad_norm": 0.2096773087978363,
+      "learning_rate": 3.9270114942528736e-05,
+      "loss": 0.1079,
+      "step": 1867
+    },
+    {
+      "epoch": 2.165326378323553,
+      "grad_norm": 0.3063262104988098,
+      "learning_rate": 3.926436781609196e-05,
+      "loss": 0.1197,
+      "step": 1868
+    },
+    {
+      "epoch": 2.1664855466203,
+      "grad_norm": 0.24331098794937134,
+      "learning_rate": 3.925862068965517e-05,
+      "loss": 0.1114,
+      "step": 1869
+    },
+    {
+      "epoch": 2.167644714917047,
+      "grad_norm": 0.25354528427124023,
+      "learning_rate": 3.9252873563218395e-05,
+      "loss": 0.1231,
+      "step": 1870
+    },
+    {
+      "epoch": 2.168803883213794,
+      "grad_norm": 0.2932867109775543,
+      "learning_rate": 3.924712643678161e-05,
+      "loss": 0.131,
+      "step": 1871
+    },
+    {
+      "epoch": 2.1699630515105413,
+      "grad_norm": 0.22387802600860596,
+      "learning_rate": 3.924137931034483e-05,
+      "loss": 0.1101,
+      "step": 1872
+    },
+    {
+      "epoch": 2.1711222198072884,
+      "grad_norm": 0.250288724899292,
+      "learning_rate": 3.923563218390805e-05,
+      "loss": 0.1229,
+      "step": 1873
+    },
+    {
+      "epoch": 2.1722813881040355,
+      "grad_norm": 0.39069998264312744,
+      "learning_rate": 3.922988505747127e-05,
+      "loss": 0.1259,
+      "step": 1874
+    },
+    {
+      "epoch": 2.1734405564007826,
+      "grad_norm": 0.20509329438209534,
+      "learning_rate": 3.922413793103448e-05,
+      "loss": 0.1273,
+      "step": 1875
+    },
+    {
+      "epoch": 2.1745997246975297,
+      "grad_norm": 0.27866530418395996,
+      "learning_rate": 3.9218390804597704e-05,
+      "loss": 0.1229,
+      "step": 1876
+    },
+    {
+      "epoch": 2.175758892994277,
+      "grad_norm": 0.2512810528278351,
+      "learning_rate": 3.921264367816092e-05,
+      "loss": 0.1142,
+      "step": 1877
+    },
+    {
+      "epoch": 2.1769180612910235,
+      "grad_norm": 0.2705281674861908,
+      "learning_rate": 3.9206896551724134e-05,
+      "loss": 0.1132,
+      "step": 1878
+    },
+    {
+      "epoch": 2.1780772295877706,
+      "grad_norm": 0.22448644042015076,
+      "learning_rate": 3.9201149425287356e-05,
+      "loss": 0.1077,
+      "step": 1879
+    },
+    {
+      "epoch": 2.1792363978845177,
+      "grad_norm": 0.30466312170028687,
+      "learning_rate": 3.919540229885058e-05,
+      "loss": 0.1246,
+      "step": 1880
+    },
+    {
+      "epoch": 2.180395566181265,
+      "grad_norm": 0.2996283769607544,
+      "learning_rate": 3.91896551724138e-05,
+      "loss": 0.1262,
+      "step": 1881
+    },
+    {
+      "epoch": 2.181554734478012,
+      "grad_norm": 0.3339924216270447,
+      "learning_rate": 3.9183908045977014e-05,
+      "loss": 0.1344,
+      "step": 1882
+    },
+    {
+      "epoch": 2.182713902774759,
+      "grad_norm": 0.24632121622562408,
+      "learning_rate": 3.917816091954023e-05,
+      "loss": 0.1099,
+      "step": 1883
+    },
+    {
+      "epoch": 2.183873071071506,
+      "grad_norm": 0.24614660441875458,
+      "learning_rate": 3.917241379310345e-05,
+      "loss": 0.1136,
+      "step": 1884
+    },
+    {
+      "epoch": 2.1850322393682533,
+      "grad_norm": 0.31324732303619385,
+      "learning_rate": 3.9166666666666665e-05,
+      "loss": 0.1242,
+      "step": 1885
+    },
+    {
+      "epoch": 2.1861914076650004,
+      "grad_norm": 0.3063158392906189,
+      "learning_rate": 3.916091954022989e-05,
+      "loss": 0.1313,
+      "step": 1886
+    },
+    {
+      "epoch": 2.1873505759617475,
+      "grad_norm": 0.3112230598926544,
+      "learning_rate": 3.915517241379311e-05,
+      "loss": 0.1244,
+      "step": 1887
+    },
+    {
+      "epoch": 2.1885097442584946,
+      "grad_norm": 0.2197721004486084,
+      "learning_rate": 3.9149425287356324e-05,
+      "loss": 0.1093,
+      "step": 1888
+    },
+    {
+      "epoch": 2.1896689125552418,
+      "grad_norm": 0.2518065273761749,
+      "learning_rate": 3.9143678160919545e-05,
+      "loss": 0.1126,
+      "step": 1889
+    },
+    {
+      "epoch": 2.190828080851989,
+      "grad_norm": 0.24468325078487396,
+      "learning_rate": 3.913793103448276e-05,
+      "loss": 0.1264,
+      "step": 1890
+    },
+    {
+      "epoch": 2.1919872491487356,
+      "grad_norm": 0.2638467252254486,
+      "learning_rate": 3.9132183908045975e-05,
+      "loss": 0.1299,
+      "step": 1891
+    },
+    {
+      "epoch": 2.1931464174454827,
+      "grad_norm": 0.29071396589279175,
+      "learning_rate": 3.91264367816092e-05,
+      "loss": 0.1406,
+      "step": 1892
+    },
+    {
+      "epoch": 2.19430558574223,
+      "grad_norm": 0.26825428009033203,
+      "learning_rate": 3.912068965517242e-05,
+      "loss": 0.133,
+      "step": 1893
+    },
+    {
+      "epoch": 2.195464754038977,
+      "grad_norm": 0.28922030329704285,
+      "learning_rate": 3.9114942528735633e-05,
+      "loss": 0.1303,
+      "step": 1894
+    },
+    {
+      "epoch": 2.196623922335724,
+      "grad_norm": 0.22582325339317322,
+      "learning_rate": 3.9109195402298855e-05,
+      "loss": 0.1118,
+      "step": 1895
+    },
+    {
+      "epoch": 2.197783090632471,
+      "grad_norm": 0.31042325496673584,
+      "learning_rate": 3.910344827586207e-05,
+      "loss": 0.1292,
+      "step": 1896
+    },
+    {
+      "epoch": 2.1989422589292182,
+      "grad_norm": 0.207808256149292,
+      "learning_rate": 3.9097701149425285e-05,
+      "loss": 0.1151,
+      "step": 1897
+    },
+    {
+      "epoch": 2.2001014272259654,
+      "grad_norm": 0.2200191169977188,
+      "learning_rate": 3.909195402298851e-05,
+      "loss": 0.1085,
+      "step": 1898
+    },
+    {
+      "epoch": 2.2012605955227125,
+      "grad_norm": 0.2516978979110718,
+      "learning_rate": 3.908620689655172e-05,
+      "loss": 0.122,
+      "step": 1899
+    },
+    {
+      "epoch": 2.2024197638194596,
+      "grad_norm": 0.2803818881511688,
+      "learning_rate": 3.908045977011495e-05,
+      "loss": 0.1214,
+      "step": 1900
+    },
+    {
+      "epoch": 2.2035789321162067,
+      "grad_norm": 0.24263282120227814,
+      "learning_rate": 3.9074712643678165e-05,
+      "loss": 0.1072,
+      "step": 1901
+    },
+    {
+      "epoch": 2.204738100412954,
+      "grad_norm": 0.24859733879566193,
+      "learning_rate": 3.906896551724138e-05,
+      "loss": 0.1286,
+      "step": 1902
+    },
+    {
+      "epoch": 2.205897268709701,
+      "grad_norm": 0.2725147306919098,
+      "learning_rate": 3.90632183908046e-05,
+      "loss": 0.1282,
+      "step": 1903
+    },
+    {
+      "epoch": 2.207056437006448,
+      "grad_norm": 0.42582276463508606,
+      "learning_rate": 3.9057471264367816e-05,
+      "loss": 0.1239,
+      "step": 1904
+    },
+    {
+      "epoch": 2.208215605303195,
+      "grad_norm": 0.2799397110939026,
+      "learning_rate": 3.905172413793103e-05,
+      "loss": 0.126,
+      "step": 1905
+    },
+    {
+      "epoch": 2.209374773599942,
+      "grad_norm": 0.23212167620658875,
+      "learning_rate": 3.904597701149425e-05,
+      "loss": 0.1143,
+      "step": 1906
+    },
+    {
+      "epoch": 2.210533941896689,
+      "grad_norm": 0.26310837268829346,
+      "learning_rate": 3.9040229885057475e-05,
+      "loss": 0.1218,
+      "step": 1907
+    },
+    {
+      "epoch": 2.211693110193436,
+      "grad_norm": 0.289028137922287,
+      "learning_rate": 3.9034482758620696e-05,
+      "loss": 0.1142,
+      "step": 1908
+    },
+    {
+      "epoch": 2.212852278490183,
+      "grad_norm": 0.2884521484375,
+      "learning_rate": 3.902873563218391e-05,
+      "loss": 0.1215,
+      "step": 1909
+    },
+    {
+      "epoch": 2.2140114467869303,
+      "grad_norm": 0.2507173717021942,
+      "learning_rate": 3.9022988505747126e-05,
+      "loss": 0.1184,
+      "step": 1910
+    },
+    {
+      "epoch": 2.2151706150836774,
+      "grad_norm": 0.23451244831085205,
+      "learning_rate": 3.901724137931035e-05,
+      "loss": 0.1123,
+      "step": 1911
+    },
+    {
+      "epoch": 2.2163297833804245,
+      "grad_norm": 0.2503047287464142,
+      "learning_rate": 3.901149425287356e-05,
+      "loss": 0.1204,
+      "step": 1912
+    },
+    {
+      "epoch": 2.2174889516771716,
+      "grad_norm": 0.2952342927455902,
+      "learning_rate": 3.9005747126436784e-05,
+      "loss": 0.1154,
+      "step": 1913
+    },
+    {
+      "epoch": 2.2186481199739188,
+      "grad_norm": 0.28569087386131287,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 0.1348,
+      "step": 1914
+    },
+    {
+      "epoch": 2.219807288270666,
+      "grad_norm": 0.3290340304374695,
+      "learning_rate": 3.899425287356322e-05,
+      "loss": 0.1197,
+      "step": 1915
+    },
+    {
+      "epoch": 2.220966456567413,
+      "grad_norm": 0.20089855790138245,
+      "learning_rate": 3.8988505747126436e-05,
+      "loss": 0.1161,
+      "step": 1916
+    },
+    {
+      "epoch": 2.22212562486416,
+      "grad_norm": 0.2501339912414551,
+      "learning_rate": 3.898275862068966e-05,
+      "loss": 0.1122,
+      "step": 1917
+    },
+    {
+      "epoch": 2.223284793160907,
+      "grad_norm": 0.27215123176574707,
+      "learning_rate": 3.897701149425287e-05,
+      "loss": 0.1152,
+      "step": 1918
+    },
+    {
+      "epoch": 2.2244439614576543,
+      "grad_norm": 0.2323666214942932,
+      "learning_rate": 3.8971264367816094e-05,
+      "loss": 0.1029,
+      "step": 1919
+    },
+    {
+      "epoch": 2.2256031297544014,
+      "grad_norm": 0.3195997476577759,
+      "learning_rate": 3.896551724137931e-05,
+      "loss": 0.1307,
+      "step": 1920
+    },
+    {
+      "epoch": 2.226762298051148,
+      "grad_norm": 0.34427744150161743,
+      "learning_rate": 3.895977011494253e-05,
+      "loss": 0.1279,
+      "step": 1921
+    },
+    {
+      "epoch": 2.2279214663478952,
+      "grad_norm": 0.3101141154766083,
+      "learning_rate": 3.895402298850575e-05,
+      "loss": 0.1152,
+      "step": 1922
+    },
+    {
+      "epoch": 2.2290806346446423,
+      "grad_norm": 0.3297821283340454,
+      "learning_rate": 3.894827586206897e-05,
+      "loss": 0.1249,
+      "step": 1923
+    },
+    {
+      "epoch": 2.2302398029413895,
+      "grad_norm": 0.27412423491477966,
+      "learning_rate": 3.894252873563218e-05,
+      "loss": 0.1201,
+      "step": 1924
+    },
+    {
+      "epoch": 2.2313989712381366,
+      "grad_norm": 0.2107117623090744,
+      "learning_rate": 3.8936781609195404e-05,
+      "loss": 0.1083,
+      "step": 1925
+    },
+    {
+      "epoch": 2.2325581395348837,
+      "grad_norm": 0.3204614222049713,
+      "learning_rate": 3.893103448275862e-05,
+      "loss": 0.1319,
+      "step": 1926
+    },
+    {
+      "epoch": 2.233717307831631,
+      "grad_norm": 0.32438015937805176,
+      "learning_rate": 3.892528735632184e-05,
+      "loss": 0.1343,
+      "step": 1927
+    },
+    {
+      "epoch": 2.234876476128378,
+      "grad_norm": 0.22636090219020844,
+      "learning_rate": 3.891954022988506e-05,
+      "loss": 0.1092,
+      "step": 1928
+    },
+    {
+      "epoch": 2.236035644425125,
+      "grad_norm": 0.27245032787323,
+      "learning_rate": 3.891379310344828e-05,
+      "loss": 0.1141,
+      "step": 1929
+    },
+    {
+      "epoch": 2.237194812721872,
+      "grad_norm": 0.3505677580833435,
+      "learning_rate": 3.89080459770115e-05,
+      "loss": 0.1366,
+      "step": 1930
+    },
+    {
+      "epoch": 2.2383539810186193,
+      "grad_norm": 0.2798290252685547,
+      "learning_rate": 3.8902298850574714e-05,
+      "loss": 0.1153,
+      "step": 1931
+    },
+    {
+      "epoch": 2.2395131493153664,
+      "grad_norm": 0.20595116913318634,
+      "learning_rate": 3.889655172413793e-05,
+      "loss": 0.1109,
+      "step": 1932
+    },
+    {
+      "epoch": 2.2406723176121135,
+      "grad_norm": 0.26562798023223877,
+      "learning_rate": 3.889080459770115e-05,
+      "loss": 0.1309,
+      "step": 1933
+    },
+    {
+      "epoch": 2.24183148590886,
+      "grad_norm": 0.28645437955856323,
+      "learning_rate": 3.888505747126437e-05,
+      "loss": 0.124,
+      "step": 1934
+    },
+    {
+      "epoch": 2.2429906542056073,
+      "grad_norm": 0.26411932706832886,
+      "learning_rate": 3.8879310344827594e-05,
+      "loss": 0.1231,
+      "step": 1935
+    },
+    {
+      "epoch": 2.2441498225023544,
+      "grad_norm": 0.2324962168931961,
+      "learning_rate": 3.887356321839081e-05,
+      "loss": 0.1213,
+      "step": 1936
+    },
+    {
+      "epoch": 2.2453089907991015,
+      "grad_norm": 0.2609705924987793,
+      "learning_rate": 3.886781609195402e-05,
+      "loss": 0.1246,
+      "step": 1937
+    },
+    {
+      "epoch": 2.2464681590958486,
+      "grad_norm": 0.24605713784694672,
+      "learning_rate": 3.8862068965517245e-05,
+      "loss": 0.1193,
+      "step": 1938
+    },
+    {
+      "epoch": 2.2476273273925957,
+      "grad_norm": 0.3854371905326843,
+      "learning_rate": 3.885632183908046e-05,
+      "loss": 0.1324,
+      "step": 1939
+    },
+    {
+      "epoch": 2.248786495689343,
+      "grad_norm": 0.2347278594970703,
+      "learning_rate": 3.8850574712643675e-05,
+      "loss": 0.1185,
+      "step": 1940
+    },
+    {
+      "epoch": 2.24994566398609,
+      "grad_norm": 0.25814783573150635,
+      "learning_rate": 3.88448275862069e-05,
+      "loss": 0.1163,
+      "step": 1941
+    },
+    {
+      "epoch": 2.251104832282837,
+      "grad_norm": 0.2590005695819855,
+      "learning_rate": 3.883908045977012e-05,
+      "loss": 0.1159,
+      "step": 1942
+    },
+    {
+      "epoch": 2.252264000579584,
+      "grad_norm": 0.2866111099720001,
+      "learning_rate": 3.883333333333333e-05,
+      "loss": 0.119,
+      "step": 1943
+    },
+    {
+      "epoch": 2.2534231688763313,
+      "grad_norm": 0.20185330510139465,
+      "learning_rate": 3.8827586206896555e-05,
+      "loss": 0.1127,
+      "step": 1944
+    },
+    {
+      "epoch": 2.2545823371730784,
+      "grad_norm": 0.2572873830795288,
+      "learning_rate": 3.882183908045977e-05,
+      "loss": 0.1235,
+      "step": 1945
+    },
+    {
+      "epoch": 2.2557415054698255,
+      "grad_norm": 0.29143568873405457,
+      "learning_rate": 3.881609195402299e-05,
+      "loss": 0.1161,
+      "step": 1946
+    },
+    {
+      "epoch": 2.2569006737665727,
+      "grad_norm": 0.25425946712493896,
+      "learning_rate": 3.8810344827586206e-05,
+      "loss": 0.1022,
+      "step": 1947
+    },
+    {
+      "epoch": 2.2580598420633198,
+      "grad_norm": 0.25637203454971313,
+      "learning_rate": 3.880459770114943e-05,
+      "loss": 0.1139,
+      "step": 1948
+    },
+    {
+      "epoch": 2.2592190103600664,
+      "grad_norm": 0.30507326126098633,
+      "learning_rate": 3.879885057471265e-05,
+      "loss": 0.1264,
+      "step": 1949
+    },
+    {
+      "epoch": 2.2603781786568136,
+      "grad_norm": 0.28318384289741516,
+      "learning_rate": 3.8793103448275865e-05,
+      "loss": 0.1243,
+      "step": 1950
+    },
+    {
+      "epoch": 2.2615373469535607,
+      "grad_norm": 0.3025624752044678,
+      "learning_rate": 3.878735632183908e-05,
+      "loss": 0.1223,
+      "step": 1951
+    },
+    {
+      "epoch": 2.262696515250308,
+      "grad_norm": 0.3020472824573517,
+      "learning_rate": 3.87816091954023e-05,
+      "loss": 0.1205,
+      "step": 1952
+    },
+    {
+      "epoch": 2.263855683547055,
+      "grad_norm": 0.3075084388256073,
+      "learning_rate": 3.8775862068965516e-05,
+      "loss": 0.1151,
+      "step": 1953
+    },
+    {
+      "epoch": 2.265014851843802,
+      "grad_norm": 0.3189045786857605,
+      "learning_rate": 3.877011494252874e-05,
+      "loss": 0.1102,
+      "step": 1954
+    },
+    {
+      "epoch": 2.266174020140549,
+      "grad_norm": 0.347697913646698,
+      "learning_rate": 3.876436781609196e-05,
+      "loss": 0.1333,
+      "step": 1955
+    },
+    {
+      "epoch": 2.2673331884372963,
+      "grad_norm": 0.27193987369537354,
+      "learning_rate": 3.8758620689655174e-05,
+      "loss": 0.1237,
+      "step": 1956
+    },
+    {
+      "epoch": 2.2684923567340434,
+      "grad_norm": 0.22569780051708221,
+      "learning_rate": 3.8752873563218396e-05,
+      "loss": 0.1018,
+      "step": 1957
+    },
+    {
+      "epoch": 2.2696515250307905,
+      "grad_norm": 0.2848230302333832,
+      "learning_rate": 3.874712643678161e-05,
+      "loss": 0.1088,
+      "step": 1958
+    },
+    {
+      "epoch": 2.2708106933275376,
+      "grad_norm": 0.20593014359474182,
+      "learning_rate": 3.8741379310344826e-05,
+      "loss": 0.1036,
+      "step": 1959
+    },
+    {
+      "epoch": 2.2719698616242847,
+      "grad_norm": 0.2758769989013672,
+      "learning_rate": 3.873563218390805e-05,
+      "loss": 0.1227,
+      "step": 1960
+    },
+    {
+      "epoch": 2.273129029921032,
+      "grad_norm": 0.2546898126602173,
+      "learning_rate": 3.872988505747127e-05,
+      "loss": 0.1075,
+      "step": 1961
+    },
+    {
+      "epoch": 2.2742881982177785,
+      "grad_norm": 0.3119022250175476,
+      "learning_rate": 3.8724137931034484e-05,
+      "loss": 0.1343,
+      "step": 1962
+    },
+    {
+      "epoch": 2.275447366514526,
+      "grad_norm": 0.2919405400753021,
+      "learning_rate": 3.8718390804597706e-05,
+      "loss": 0.1392,
+      "step": 1963
+    },
+    {
+      "epoch": 2.2766065348112727,
+      "grad_norm": 0.29379087686538696,
+      "learning_rate": 3.871264367816092e-05,
+      "loss": 0.137,
+      "step": 1964
+    },
+    {
+      "epoch": 2.27776570310802,
+      "grad_norm": 0.32056736946105957,
+      "learning_rate": 3.870689655172414e-05,
+      "loss": 0.1112,
+      "step": 1965
+    },
+    {
+      "epoch": 2.278924871404767,
+      "grad_norm": 0.25349029898643494,
+      "learning_rate": 3.870114942528736e-05,
+      "loss": 0.1178,
+      "step": 1966
+    },
+    {
+      "epoch": 2.280084039701514,
+      "grad_norm": 0.24394166469573975,
+      "learning_rate": 3.869540229885057e-05,
+      "loss": 0.1072,
+      "step": 1967
+    },
+    {
+      "epoch": 2.281243207998261,
+      "grad_norm": 0.23791316151618958,
+      "learning_rate": 3.8689655172413794e-05,
+      "loss": 0.1076,
+      "step": 1968
+    },
+    {
+      "epoch": 2.2824023762950083,
+      "grad_norm": 0.24641607701778412,
+      "learning_rate": 3.8683908045977015e-05,
+      "loss": 0.1218,
+      "step": 1969
+    },
+    {
+      "epoch": 2.2835615445917554,
+      "grad_norm": 0.23730486631393433,
+      "learning_rate": 3.867816091954023e-05,
+      "loss": 0.1135,
+      "step": 1970
+    },
+    {
+      "epoch": 2.2847207128885025,
+      "grad_norm": 0.2812163531780243,
+      "learning_rate": 3.867241379310345e-05,
+      "loss": 0.1196,
+      "step": 1971
+    },
+    {
+      "epoch": 2.2858798811852497,
+      "grad_norm": 0.25188305974006653,
+      "learning_rate": 3.866666666666667e-05,
+      "loss": 0.1174,
+      "step": 1972
+    },
+    {
+      "epoch": 2.2870390494819968,
+      "grad_norm": 0.2505231499671936,
+      "learning_rate": 3.866091954022989e-05,
+      "loss": 0.1175,
+      "step": 1973
+    },
+    {
+      "epoch": 2.288198217778744,
+      "grad_norm": 0.36217159032821655,
+      "learning_rate": 3.8655172413793103e-05,
+      "loss": 0.1261,
+      "step": 1974
+    },
+    {
+      "epoch": 2.289357386075491,
+      "grad_norm": 0.26633283495903015,
+      "learning_rate": 3.8649425287356325e-05,
+      "loss": 0.126,
+      "step": 1975
+    },
+    {
+      "epoch": 2.290516554372238,
+      "grad_norm": 0.286703884601593,
+      "learning_rate": 3.864367816091955e-05,
+      "loss": 0.1083,
+      "step": 1976
+    },
+    {
+      "epoch": 2.291675722668985,
+      "grad_norm": 0.2949945032596588,
+      "learning_rate": 3.863793103448276e-05,
+      "loss": 0.1201,
+      "step": 1977
+    },
+    {
+      "epoch": 2.292834890965732,
+      "grad_norm": 0.22798071801662445,
+      "learning_rate": 3.863218390804598e-05,
+      "loss": 0.1158,
+      "step": 1978
+    },
+    {
+      "epoch": 2.293994059262479,
+      "grad_norm": 0.2760827839374542,
+      "learning_rate": 3.86264367816092e-05,
+      "loss": 0.1322,
+      "step": 1979
+    },
+    {
+      "epoch": 2.295153227559226,
+      "grad_norm": 0.25793665647506714,
+      "learning_rate": 3.862068965517241e-05,
+      "loss": 0.1113,
+      "step": 1980
+    },
+    {
+      "epoch": 2.2963123958559732,
+      "grad_norm": 0.28689324855804443,
+      "learning_rate": 3.861494252873563e-05,
+      "loss": 0.1064,
+      "step": 1981
+    },
+    {
+      "epoch": 2.2974715641527204,
+      "grad_norm": 0.2645778954029083,
+      "learning_rate": 3.8609195402298857e-05,
+      "loss": 0.1239,
+      "step": 1982
+    },
+    {
+      "epoch": 2.2986307324494675,
+      "grad_norm": 0.26537734270095825,
+      "learning_rate": 3.860344827586207e-05,
+      "loss": 0.1201,
+      "step": 1983
+    },
+    {
+      "epoch": 2.2997899007462146,
+      "grad_norm": 0.281444251537323,
+      "learning_rate": 3.859770114942529e-05,
+      "loss": 0.1199,
+      "step": 1984
+    },
+    {
+      "epoch": 2.3009490690429617,
+      "grad_norm": 0.25172534584999084,
+      "learning_rate": 3.859195402298851e-05,
+      "loss": 0.1133,
+      "step": 1985
+    },
+    {
+      "epoch": 2.302108237339709,
+      "grad_norm": 0.25721946358680725,
+      "learning_rate": 3.858620689655172e-05,
+      "loss": 0.1117,
+      "step": 1986
+    },
+    {
+      "epoch": 2.303267405636456,
+      "grad_norm": 0.24373944103717804,
+      "learning_rate": 3.8580459770114945e-05,
+      "loss": 0.1043,
+      "step": 1987
+    },
+    {
+      "epoch": 2.304426573933203,
+      "grad_norm": 0.2496458888053894,
+      "learning_rate": 3.857471264367816e-05,
+      "loss": 0.1147,
+      "step": 1988
+    },
+    {
+      "epoch": 2.30558574222995,
+      "grad_norm": 0.21120178699493408,
+      "learning_rate": 3.856896551724138e-05,
+      "loss": 0.1069,
+      "step": 1989
+    },
+    {
+      "epoch": 2.3067449105266973,
+      "grad_norm": 0.30783671140670776,
+      "learning_rate": 3.85632183908046e-05,
+      "loss": 0.125,
+      "step": 1990
+    },
+    {
+      "epoch": 2.3079040788234444,
+      "grad_norm": 0.22991196811199188,
+      "learning_rate": 3.855747126436782e-05,
+      "loss": 0.1206,
+      "step": 1991
+    },
+    {
+      "epoch": 2.309063247120191,
+      "grad_norm": 0.2827235758304596,
+      "learning_rate": 3.855172413793104e-05,
+      "loss": 0.1333,
+      "step": 1992
+    },
+    {
+      "epoch": 2.310222415416938,
+      "grad_norm": 0.3185621500015259,
+      "learning_rate": 3.8545977011494254e-05,
+      "loss": 0.135,
+      "step": 1993
+    },
+    {
+      "epoch": 2.3113815837136853,
+      "grad_norm": 0.25081896781921387,
+      "learning_rate": 3.854022988505747e-05,
+      "loss": 0.1197,
+      "step": 1994
+    },
+    {
+      "epoch": 2.3125407520104324,
+      "grad_norm": 0.3022123873233795,
+      "learning_rate": 3.853448275862069e-05,
+      "loss": 0.1232,
+      "step": 1995
+    },
+    {
+      "epoch": 2.3136999203071795,
+      "grad_norm": 0.3585253357887268,
+      "learning_rate": 3.852873563218391e-05,
+      "loss": 0.1371,
+      "step": 1996
+    },
+    {
+      "epoch": 2.3148590886039266,
+      "grad_norm": 0.22381120920181274,
+      "learning_rate": 3.852298850574713e-05,
+      "loss": 0.1093,
+      "step": 1997
+    },
+    {
+      "epoch": 2.3160182569006738,
+      "grad_norm": 0.3148757219314575,
+      "learning_rate": 3.851724137931035e-05,
+      "loss": 0.1487,
+      "step": 1998
+    },
+    {
+      "epoch": 2.317177425197421,
+      "grad_norm": 0.298365980386734,
+      "learning_rate": 3.8511494252873564e-05,
+      "loss": 0.1225,
+      "step": 1999
+    },
+    {
+      "epoch": 2.318336593494168,
+      "grad_norm": 0.27845144271850586,
+      "learning_rate": 3.850574712643678e-05,
+      "loss": 0.1202,
+      "step": 2000
+    },
+    {
+      "epoch": 2.319495761790915,
+      "grad_norm": 0.2849871814250946,
+      "learning_rate": 3.85e-05,
+      "loss": 0.1209,
+      "step": 2001
+    },
+    {
+      "epoch": 2.320654930087662,
+      "grad_norm": 0.30878233909606934,
+      "learning_rate": 3.849425287356322e-05,
+      "loss": 0.1256,
+      "step": 2002
+    },
+    {
+      "epoch": 2.3218140983844093,
+      "grad_norm": 0.27193957567214966,
+      "learning_rate": 3.8488505747126444e-05,
+      "loss": 0.1206,
+      "step": 2003
+    },
+    {
+      "epoch": 2.3229732666811564,
+      "grad_norm": 0.23535265028476715,
+      "learning_rate": 3.848275862068966e-05,
+      "loss": 0.1188,
+      "step": 2004
+    },
+    {
+      "epoch": 2.324132434977903,
+      "grad_norm": 0.2573367953300476,
+      "learning_rate": 3.8477011494252874e-05,
+      "loss": 0.1165,
+      "step": 2005
+    },
+    {
+      "epoch": 2.3252916032746507,
+      "grad_norm": 0.2889256179332733,
+      "learning_rate": 3.8471264367816096e-05,
+      "loss": 0.1293,
+      "step": 2006
+    },
+    {
+      "epoch": 2.3264507715713973,
+      "grad_norm": 0.29052430391311646,
+      "learning_rate": 3.846551724137931e-05,
+      "loss": 0.1194,
+      "step": 2007
+    },
+    {
+      "epoch": 2.3276099398681445,
+      "grad_norm": 0.2844145596027374,
+      "learning_rate": 3.8459770114942525e-05,
+      "loss": 0.1342,
+      "step": 2008
+    },
+    {
+      "epoch": 2.3287691081648916,
+      "grad_norm": 0.2795712649822235,
+      "learning_rate": 3.845402298850575e-05,
+      "loss": 0.1216,
+      "step": 2009
+    },
+    {
+      "epoch": 2.3299282764616387,
+      "grad_norm": 0.27312591671943665,
+      "learning_rate": 3.844827586206897e-05,
+      "loss": 0.1224,
+      "step": 2010
+    },
+    {
+      "epoch": 2.331087444758386,
+      "grad_norm": 0.20868080854415894,
+      "learning_rate": 3.844252873563219e-05,
+      "loss": 0.1051,
+      "step": 2011
+    },
+    {
+      "epoch": 2.332246613055133,
+      "grad_norm": 0.27325716614723206,
+      "learning_rate": 3.8436781609195405e-05,
+      "loss": 0.1154,
+      "step": 2012
+    },
+    {
+      "epoch": 2.33340578135188,
+      "grad_norm": 0.2669987082481384,
+      "learning_rate": 3.843103448275862e-05,
+      "loss": 0.1131,
+      "step": 2013
+    },
+    {
+      "epoch": 2.334564949648627,
+      "grad_norm": 0.25275173783302307,
+      "learning_rate": 3.842528735632184e-05,
+      "loss": 0.1209,
+      "step": 2014
+    },
+    {
+      "epoch": 2.3357241179453743,
+      "grad_norm": 0.2534005045890808,
+      "learning_rate": 3.841954022988506e-05,
+      "loss": 0.1266,
+      "step": 2015
+    },
+    {
+      "epoch": 2.3368832862421214,
+      "grad_norm": 0.26923778653144836,
+      "learning_rate": 3.841379310344828e-05,
+      "loss": 0.1289,
+      "step": 2016
+    },
+    {
+      "epoch": 2.3380424545388685,
+      "grad_norm": 0.23470233380794525,
+      "learning_rate": 3.84080459770115e-05,
+      "loss": 0.1097,
+      "step": 2017
+    },
+    {
+      "epoch": 2.3392016228356156,
+      "grad_norm": 0.29632678627967834,
+      "learning_rate": 3.8402298850574715e-05,
+      "loss": 0.1136,
+      "step": 2018
+    },
+    {
+      "epoch": 2.3403607911323627,
+      "grad_norm": 0.22054457664489746,
+      "learning_rate": 3.839655172413793e-05,
+      "loss": 0.1087,
+      "step": 2019
+    },
+    {
+      "epoch": 2.3415199594291094,
+      "grad_norm": 0.29391518235206604,
+      "learning_rate": 3.839080459770115e-05,
+      "loss": 0.1152,
+      "step": 2020
+    },
+    {
+      "epoch": 2.3426791277258565,
+      "grad_norm": 0.2556963264942169,
+      "learning_rate": 3.8385057471264366e-05,
+      "loss": 0.1298,
+      "step": 2021
+    },
+    {
+      "epoch": 2.3438382960226036,
+      "grad_norm": 0.20713378489017487,
+      "learning_rate": 3.837931034482759e-05,
+      "loss": 0.1121,
+      "step": 2022
+    },
+    {
+      "epoch": 2.3449974643193507,
+      "grad_norm": 0.30257120728492737,
+      "learning_rate": 3.837356321839081e-05,
+      "loss": 0.1267,
+      "step": 2023
+    },
+    {
+      "epoch": 2.346156632616098,
+      "grad_norm": 0.3050404489040375,
+      "learning_rate": 3.8367816091954025e-05,
+      "loss": 0.1361,
+      "step": 2024
+    },
+    {
+      "epoch": 2.347315800912845,
+      "grad_norm": 0.1983451396226883,
+      "learning_rate": 3.8362068965517246e-05,
+      "loss": 0.1103,
+      "step": 2025
+    },
+    {
+      "epoch": 2.348474969209592,
+      "grad_norm": 0.23524372279644012,
+      "learning_rate": 3.835632183908046e-05,
+      "loss": 0.1147,
+      "step": 2026
+    },
+    {
+      "epoch": 2.349634137506339,
+      "grad_norm": 0.2787211537361145,
+      "learning_rate": 3.8350574712643676e-05,
+      "loss": 0.1175,
+      "step": 2027
+    },
+    {
+      "epoch": 2.3507933058030863,
+      "grad_norm": 0.2609887421131134,
+      "learning_rate": 3.83448275862069e-05,
+      "loss": 0.118,
+      "step": 2028
+    },
+    {
+      "epoch": 2.3519524740998334,
+      "grad_norm": 0.2685818076133728,
+      "learning_rate": 3.833908045977011e-05,
+      "loss": 0.1245,
+      "step": 2029
+    },
+    {
+      "epoch": 2.3531116423965805,
+      "grad_norm": 0.25744444131851196,
+      "learning_rate": 3.8333333333333334e-05,
+      "loss": 0.1141,
+      "step": 2030
+    },
+    {
+      "epoch": 2.3542708106933277,
+      "grad_norm": 0.22216452658176422,
+      "learning_rate": 3.8327586206896556e-05,
+      "loss": 0.1097,
+      "step": 2031
+    },
+    {
+      "epoch": 2.355429978990075,
+      "grad_norm": 0.23163087666034698,
+      "learning_rate": 3.832183908045977e-05,
+      "loss": 0.1042,
+      "step": 2032
+    },
+    {
+      "epoch": 2.356589147286822,
+      "grad_norm": 0.3125256896018982,
+      "learning_rate": 3.831609195402299e-05,
+      "loss": 0.1074,
+      "step": 2033
+    },
+    {
+      "epoch": 2.357748315583569,
+      "grad_norm": 0.35050761699676514,
+      "learning_rate": 3.831034482758621e-05,
+      "loss": 0.1332,
+      "step": 2034
+    },
+    {
+      "epoch": 2.3589074838803157,
+      "grad_norm": 0.2466825395822525,
+      "learning_rate": 3.830459770114942e-05,
+      "loss": 0.1192,
+      "step": 2035
+    },
+    {
+      "epoch": 2.360066652177063,
+      "grad_norm": 0.26629340648651123,
+      "learning_rate": 3.8298850574712644e-05,
+      "loss": 0.1227,
+      "step": 2036
+    },
+    {
+      "epoch": 2.36122582047381,
+      "grad_norm": 0.3505549728870392,
+      "learning_rate": 3.8293103448275866e-05,
+      "loss": 0.1138,
+      "step": 2037
+    },
+    {
+      "epoch": 2.362384988770557,
+      "grad_norm": 0.22414235770702362,
+      "learning_rate": 3.828735632183908e-05,
+      "loss": 0.1087,
+      "step": 2038
+    },
+    {
+      "epoch": 2.363544157067304,
+      "grad_norm": 0.23031394183635712,
+      "learning_rate": 3.82816091954023e-05,
+      "loss": 0.1178,
+      "step": 2039
+    },
+    {
+      "epoch": 2.3647033253640513,
+      "grad_norm": 0.37229323387145996,
+      "learning_rate": 3.827586206896552e-05,
+      "loss": 0.1398,
+      "step": 2040
+    },
+    {
+      "epoch": 2.3658624936607984,
+      "grad_norm": 0.2622588872909546,
+      "learning_rate": 3.827011494252874e-05,
+      "loss": 0.1189,
+      "step": 2041
+    },
+    {
+      "epoch": 2.3670216619575455,
+      "grad_norm": 0.2855938673019409,
+      "learning_rate": 3.8264367816091954e-05,
+      "loss": 0.1256,
+      "step": 2042
+    },
+    {
+      "epoch": 2.3681808302542926,
+      "grad_norm": 0.3094041049480438,
+      "learning_rate": 3.8258620689655176e-05,
+      "loss": 0.1268,
+      "step": 2043
+    },
+    {
+      "epoch": 2.3693399985510397,
+      "grad_norm": 0.30594077706336975,
+      "learning_rate": 3.82528735632184e-05,
+      "loss": 0.1206,
+      "step": 2044
+    },
+    {
+      "epoch": 2.370499166847787,
+      "grad_norm": 0.3137442469596863,
+      "learning_rate": 3.824712643678161e-05,
+      "loss": 0.1346,
+      "step": 2045
+    },
+    {
+      "epoch": 2.371658335144534,
+      "grad_norm": 0.25371280312538147,
+      "learning_rate": 3.824137931034483e-05,
+      "loss": 0.121,
+      "step": 2046
+    },
+    {
+      "epoch": 2.372817503441281,
+      "grad_norm": 0.24225755035877228,
+      "learning_rate": 3.823563218390805e-05,
+      "loss": 0.1124,
+      "step": 2047
+    },
+    {
+      "epoch": 2.3739766717380277,
+      "grad_norm": 0.2342030107975006,
+      "learning_rate": 3.8229885057471264e-05,
+      "loss": 0.1014,
+      "step": 2048
+    },
+    {
+      "epoch": 2.3751358400347753,
+      "grad_norm": 0.3285464942455292,
+      "learning_rate": 3.8224137931034485e-05,
+      "loss": 0.122,
+      "step": 2049
+    },
+    {
+      "epoch": 2.376295008331522,
+      "grad_norm": 0.2837320864200592,
+      "learning_rate": 3.82183908045977e-05,
+      "loss": 0.1265,
+      "step": 2050
+    },
+    {
+      "epoch": 2.377454176628269,
+      "grad_norm": 0.30253443121910095,
+      "learning_rate": 3.821264367816092e-05,
+      "loss": 0.1191,
+      "step": 2051
+    },
+    {
+      "epoch": 2.378613344925016,
+      "grad_norm": 0.24019919335842133,
+      "learning_rate": 3.8206896551724144e-05,
+      "loss": 0.1035,
+      "step": 2052
+    },
+    {
+      "epoch": 2.3797725132217633,
+      "grad_norm": 0.2411479502916336,
+      "learning_rate": 3.820114942528736e-05,
+      "loss": 0.1238,
+      "step": 2053
+    },
+    {
+      "epoch": 2.3809316815185104,
+      "grad_norm": 0.2749464213848114,
+      "learning_rate": 3.8195402298850573e-05,
+      "loss": 0.1259,
+      "step": 2054
+    },
+    {
+      "epoch": 2.3820908498152575,
+      "grad_norm": 0.2516106963157654,
+      "learning_rate": 3.8189655172413795e-05,
+      "loss": 0.1173,
+      "step": 2055
+    },
+    {
+      "epoch": 2.3832500181120047,
+      "grad_norm": 0.2396615743637085,
+      "learning_rate": 3.818390804597701e-05,
+      "loss": 0.1057,
+      "step": 2056
+    },
+    {
+      "epoch": 2.3844091864087518,
+      "grad_norm": 0.23807305097579956,
+      "learning_rate": 3.817816091954023e-05,
+      "loss": 0.1216,
+      "step": 2057
+    },
+    {
+      "epoch": 2.385568354705499,
+      "grad_norm": 0.24773019552230835,
+      "learning_rate": 3.8172413793103453e-05,
+      "loss": 0.1243,
+      "step": 2058
+    },
+    {
+      "epoch": 2.386727523002246,
+      "grad_norm": 0.21717150509357452,
+      "learning_rate": 3.816666666666667e-05,
+      "loss": 0.1101,
+      "step": 2059
+    },
+    {
+      "epoch": 2.387886691298993,
+      "grad_norm": 0.28333330154418945,
+      "learning_rate": 3.816091954022989e-05,
+      "loss": 0.1183,
+      "step": 2060
+    },
+    {
+      "epoch": 2.3890458595957402,
+      "grad_norm": 0.22554628551006317,
+      "learning_rate": 3.8155172413793105e-05,
+      "loss": 0.1276,
+      "step": 2061
+    },
+    {
+      "epoch": 2.3902050278924873,
+      "grad_norm": 0.2866622507572174,
+      "learning_rate": 3.814942528735632e-05,
+      "loss": 0.1378,
+      "step": 2062
+    },
+    {
+      "epoch": 2.391364196189234,
+      "grad_norm": 0.25763562321662903,
+      "learning_rate": 3.814367816091954e-05,
+      "loss": 0.1174,
+      "step": 2063
+    },
+    {
+      "epoch": 2.392523364485981,
+      "grad_norm": 0.24066491425037384,
+      "learning_rate": 3.813793103448276e-05,
+      "loss": 0.1193,
+      "step": 2064
+    },
+    {
+      "epoch": 2.3936825327827282,
+      "grad_norm": 0.27270743250846863,
+      "learning_rate": 3.813218390804598e-05,
+      "loss": 0.1219,
+      "step": 2065
+    },
+    {
+      "epoch": 2.3948417010794754,
+      "grad_norm": 0.28559228777885437,
+      "learning_rate": 3.81264367816092e-05,
+      "loss": 0.1345,
+      "step": 2066
+    },
+    {
+      "epoch": 2.3960008693762225,
+      "grad_norm": 0.27294817566871643,
+      "learning_rate": 3.8120689655172415e-05,
+      "loss": 0.1181,
+      "step": 2067
+    },
+    {
+      "epoch": 2.3971600376729696,
+      "grad_norm": 0.2862413227558136,
+      "learning_rate": 3.8114942528735636e-05,
+      "loss": 0.1141,
+      "step": 2068
+    },
+    {
+      "epoch": 2.3983192059697167,
+      "grad_norm": 0.26329490542411804,
+      "learning_rate": 3.810919540229885e-05,
+      "loss": 0.1233,
+      "step": 2069
+    },
+    {
+      "epoch": 2.399478374266464,
+      "grad_norm": 0.25338056683540344,
+      "learning_rate": 3.8103448275862066e-05,
+      "loss": 0.1157,
+      "step": 2070
+    },
+    {
+      "epoch": 2.400637542563211,
+      "grad_norm": 0.251355916261673,
+      "learning_rate": 3.809770114942529e-05,
+      "loss": 0.123,
+      "step": 2071
+    },
+    {
+      "epoch": 2.401796710859958,
+      "grad_norm": 0.32914426922798157,
+      "learning_rate": 3.809195402298851e-05,
+      "loss": 0.1366,
+      "step": 2072
+    },
+    {
+      "epoch": 2.402955879156705,
+      "grad_norm": 0.3044224977493286,
+      "learning_rate": 3.8086206896551724e-05,
+      "loss": 0.1273,
+      "step": 2073
+    },
+    {
+      "epoch": 2.4041150474534523,
+      "grad_norm": 0.2599372863769531,
+      "learning_rate": 3.8080459770114946e-05,
+      "loss": 0.1143,
+      "step": 2074
+    },
+    {
+      "epoch": 2.4052742157501994,
+      "grad_norm": 0.26074209809303284,
+      "learning_rate": 3.807471264367816e-05,
+      "loss": 0.121,
+      "step": 2075
+    },
+    {
+      "epoch": 2.406433384046946,
+      "grad_norm": 0.2514488399028778,
+      "learning_rate": 3.8068965517241376e-05,
+      "loss": 0.123,
+      "step": 2076
+    },
+    {
+      "epoch": 2.4075925523436936,
+      "grad_norm": 0.26624441146850586,
+      "learning_rate": 3.80632183908046e-05,
+      "loss": 0.1219,
+      "step": 2077
+    },
+    {
+      "epoch": 2.4087517206404403,
+      "grad_norm": 0.2705528140068054,
+      "learning_rate": 3.805747126436782e-05,
+      "loss": 0.1155,
+      "step": 2078
+    },
+    {
+      "epoch": 2.4099108889371874,
+      "grad_norm": 0.2879337668418884,
+      "learning_rate": 3.805172413793104e-05,
+      "loss": 0.1241,
+      "step": 2079
+    },
+    {
+      "epoch": 2.4110700572339345,
+      "grad_norm": 0.2518971264362335,
+      "learning_rate": 3.8045977011494256e-05,
+      "loss": 0.1157,
+      "step": 2080
+    },
+    {
+      "epoch": 2.4122292255306816,
+      "grad_norm": 0.3543863296508789,
+      "learning_rate": 3.804022988505747e-05,
+      "loss": 0.1295,
+      "step": 2081
+    },
+    {
+      "epoch": 2.4133883938274288,
+      "grad_norm": 0.2539769113063812,
+      "learning_rate": 3.803448275862069e-05,
+      "loss": 0.1139,
+      "step": 2082
+    },
+    {
+      "epoch": 2.414547562124176,
+      "grad_norm": 0.271513432264328,
+      "learning_rate": 3.802873563218391e-05,
+      "loss": 0.1255,
+      "step": 2083
+    },
+    {
+      "epoch": 2.415706730420923,
+      "grad_norm": 0.3522006869316101,
+      "learning_rate": 3.802298850574713e-05,
+      "loss": 0.1294,
+      "step": 2084
+    },
+    {
+      "epoch": 2.41686589871767,
+      "grad_norm": 0.27951931953430176,
+      "learning_rate": 3.801724137931035e-05,
+      "loss": 0.1331,
+      "step": 2085
+    },
+    {
+      "epoch": 2.418025067014417,
+      "grad_norm": 0.2659072279930115,
+      "learning_rate": 3.8011494252873566e-05,
+      "loss": 0.1187,
+      "step": 2086
+    },
+    {
+      "epoch": 2.4191842353111643,
+      "grad_norm": 0.23229147493839264,
+      "learning_rate": 3.800574712643679e-05,
+      "loss": 0.1146,
+      "step": 2087
+    },
+    {
+      "epoch": 2.4203434036079114,
+      "grad_norm": 0.28749972581863403,
+      "learning_rate": 3.8e-05,
+      "loss": 0.123,
+      "step": 2088
+    },
+    {
+      "epoch": 2.4215025719046586,
+      "grad_norm": 0.3175642788410187,
+      "learning_rate": 3.799425287356322e-05,
+      "loss": 0.1264,
+      "step": 2089
+    },
+    {
+      "epoch": 2.4226617402014057,
+      "grad_norm": 0.25972867012023926,
+      "learning_rate": 3.798850574712644e-05,
+      "loss": 0.1193,
+      "step": 2090
+    },
+    {
+      "epoch": 2.4238209084981523,
+      "grad_norm": 0.26062315702438354,
+      "learning_rate": 3.7982758620689654e-05,
+      "loss": 0.121,
+      "step": 2091
+    },
+    {
+      "epoch": 2.4249800767949,
+      "grad_norm": 0.25162971019744873,
+      "learning_rate": 3.7977011494252875e-05,
+      "loss": 0.1115,
+      "step": 2092
+    },
+    {
+      "epoch": 2.4261392450916466,
+      "grad_norm": 0.2526095509529114,
+      "learning_rate": 3.79712643678161e-05,
+      "loss": 0.1135,
+      "step": 2093
+    },
+    {
+      "epoch": 2.4272984133883937,
+      "grad_norm": 0.25769245624542236,
+      "learning_rate": 3.796551724137931e-05,
+      "loss": 0.116,
+      "step": 2094
+    },
+    {
+      "epoch": 2.428457581685141,
+      "grad_norm": 0.2587513029575348,
+      "learning_rate": 3.7959770114942534e-05,
+      "loss": 0.1046,
+      "step": 2095
+    },
+    {
+      "epoch": 2.429616749981888,
+      "grad_norm": 0.2877984941005707,
+      "learning_rate": 3.795402298850575e-05,
+      "loss": 0.1147,
+      "step": 2096
+    },
+    {
+      "epoch": 2.430775918278635,
+      "grad_norm": 0.28424739837646484,
+      "learning_rate": 3.794827586206896e-05,
+      "loss": 0.1188,
+      "step": 2097
+    },
+    {
+      "epoch": 2.431935086575382,
+      "grad_norm": 0.38208964467048645,
+      "learning_rate": 3.7942528735632185e-05,
+      "loss": 0.1128,
+      "step": 2098
+    },
+    {
+      "epoch": 2.4330942548721293,
+      "grad_norm": 0.27642908692359924,
+      "learning_rate": 3.793678160919541e-05,
+      "loss": 0.1336,
+      "step": 2099
+    },
+    {
+      "epoch": 2.4342534231688764,
+      "grad_norm": 0.2712388336658478,
+      "learning_rate": 3.793103448275862e-05,
+      "loss": 0.1335,
+      "step": 2100
+    },
+    {
+      "epoch": 2.4354125914656235,
+      "grad_norm": 0.3354480564594269,
+      "learning_rate": 3.792528735632184e-05,
+      "loss": 0.1292,
+      "step": 2101
+    },
+    {
+      "epoch": 2.4365717597623706,
+      "grad_norm": 0.2718063294887543,
+      "learning_rate": 3.791954022988506e-05,
+      "loss": 0.1318,
+      "step": 2102
+    },
+    {
+      "epoch": 2.4377309280591177,
+      "grad_norm": 0.2873290777206421,
+      "learning_rate": 3.791379310344827e-05,
+      "loss": 0.1184,
+      "step": 2103
+    },
+    {
+      "epoch": 2.438890096355865,
+      "grad_norm": 0.2832820415496826,
+      "learning_rate": 3.7908045977011495e-05,
+      "loss": 0.1181,
+      "step": 2104
+    },
+    {
+      "epoch": 2.440049264652612,
+      "grad_norm": 0.27394813299179077,
+      "learning_rate": 3.7902298850574716e-05,
+      "loss": 0.1117,
+      "step": 2105
+    },
+    {
+      "epoch": 2.4412084329493586,
+      "grad_norm": 0.2313212752342224,
+      "learning_rate": 3.789655172413794e-05,
+      "loss": 0.1154,
+      "step": 2106
+    },
+    {
+      "epoch": 2.4423676012461057,
+      "grad_norm": 0.2213948369026184,
+      "learning_rate": 3.789080459770115e-05,
+      "loss": 0.1214,
+      "step": 2107
+    },
+    {
+      "epoch": 2.443526769542853,
+      "grad_norm": 0.2309620976448059,
+      "learning_rate": 3.788505747126437e-05,
+      "loss": 0.1104,
+      "step": 2108
+    },
+    {
+      "epoch": 2.4446859378396,
+      "grad_norm": 0.24604758620262146,
+      "learning_rate": 3.787931034482759e-05,
+      "loss": 0.1153,
+      "step": 2109
+    },
+    {
+      "epoch": 2.445845106136347,
+      "grad_norm": 0.21898497641086578,
+      "learning_rate": 3.7873563218390804e-05,
+      "loss": 0.1093,
+      "step": 2110
+    },
+    {
+      "epoch": 2.447004274433094,
+      "grad_norm": 0.24178272485733032,
+      "learning_rate": 3.786781609195402e-05,
+      "loss": 0.1178,
+      "step": 2111
+    },
+    {
+      "epoch": 2.4481634427298413,
+      "grad_norm": 0.28789666295051575,
+      "learning_rate": 3.786206896551725e-05,
+      "loss": 0.1074,
+      "step": 2112
+    },
+    {
+      "epoch": 2.4493226110265884,
+      "grad_norm": 0.22450456023216248,
+      "learning_rate": 3.785632183908046e-05,
+      "loss": 0.1111,
+      "step": 2113
+    },
+    {
+      "epoch": 2.4504817793233356,
+      "grad_norm": 0.23580452799797058,
+      "learning_rate": 3.7850574712643684e-05,
+      "loss": 0.1088,
+      "step": 2114
+    },
+    {
+      "epoch": 2.4516409476200827,
+      "grad_norm": 0.26990747451782227,
+      "learning_rate": 3.78448275862069e-05,
+      "loss": 0.1343,
+      "step": 2115
+    },
+    {
+      "epoch": 2.45280011591683,
+      "grad_norm": 0.3045923709869385,
+      "learning_rate": 3.7839080459770114e-05,
+      "loss": 0.1226,
+      "step": 2116
+    },
+    {
+      "epoch": 2.453959284213577,
+      "grad_norm": 0.24597370624542236,
+      "learning_rate": 3.7833333333333336e-05,
+      "loss": 0.1329,
+      "step": 2117
+    },
+    {
+      "epoch": 2.455118452510324,
+      "grad_norm": 0.32915717363357544,
+      "learning_rate": 3.782758620689655e-05,
+      "loss": 0.1257,
+      "step": 2118
+    },
+    {
+      "epoch": 2.4562776208070707,
+      "grad_norm": 0.2542544901371002,
+      "learning_rate": 3.782183908045977e-05,
+      "loss": 0.1222,
+      "step": 2119
+    },
+    {
+      "epoch": 2.4574367891038182,
+      "grad_norm": 0.28636133670806885,
+      "learning_rate": 3.7816091954022994e-05,
+      "loss": 0.1311,
+      "step": 2120
+    },
+    {
+      "epoch": 2.458595957400565,
+      "grad_norm": 0.2496955841779709,
+      "learning_rate": 3.781034482758621e-05,
+      "loss": 0.1221,
+      "step": 2121
+    },
+    {
+      "epoch": 2.459755125697312,
+      "grad_norm": 0.31861692667007446,
+      "learning_rate": 3.7804597701149424e-05,
+      "loss": 0.115,
+      "step": 2122
+    },
+    {
+      "epoch": 2.460914293994059,
+      "grad_norm": 0.27043020725250244,
+      "learning_rate": 3.7798850574712646e-05,
+      "loss": 0.1132,
+      "step": 2123
+    },
+    {
+      "epoch": 2.4620734622908063,
+      "grad_norm": 0.22890667617321014,
+      "learning_rate": 3.779310344827586e-05,
+      "loss": 0.1082,
+      "step": 2124
+    },
+    {
+      "epoch": 2.4632326305875534,
+      "grad_norm": 0.27371424436569214,
+      "learning_rate": 3.778735632183908e-05,
+      "loss": 0.1231,
+      "step": 2125
+    },
+    {
+      "epoch": 2.4643917988843005,
+      "grad_norm": 0.2712050974369049,
+      "learning_rate": 3.7781609195402304e-05,
+      "loss": 0.141,
+      "step": 2126
+    },
+    {
+      "epoch": 2.4655509671810476,
+      "grad_norm": 0.26597630977630615,
+      "learning_rate": 3.777586206896552e-05,
+      "loss": 0.1193,
+      "step": 2127
+    },
+    {
+      "epoch": 2.4667101354777947,
+      "grad_norm": 0.28506308794021606,
+      "learning_rate": 3.777011494252874e-05,
+      "loss": 0.1172,
+      "step": 2128
+    },
+    {
+      "epoch": 2.467869303774542,
+      "grad_norm": 0.2775918245315552,
+      "learning_rate": 3.7764367816091955e-05,
+      "loss": 0.1131,
+      "step": 2129
+    },
+    {
+      "epoch": 2.469028472071289,
+      "grad_norm": 0.2832791209220886,
+      "learning_rate": 3.775862068965517e-05,
+      "loss": 0.13,
+      "step": 2130
+    },
+    {
+      "epoch": 2.470187640368036,
+      "grad_norm": 0.2610412836074829,
+      "learning_rate": 3.775287356321839e-05,
+      "loss": 0.1195,
+      "step": 2131
+    },
+    {
+      "epoch": 2.471346808664783,
+      "grad_norm": 0.2954767346382141,
+      "learning_rate": 3.774712643678161e-05,
+      "loss": 0.1197,
+      "step": 2132
+    },
+    {
+      "epoch": 2.4725059769615303,
+      "grad_norm": 0.30459877848625183,
+      "learning_rate": 3.7741379310344835e-05,
+      "loss": 0.1233,
+      "step": 2133
+    },
+    {
+      "epoch": 2.473665145258277,
+      "grad_norm": 0.2544924318790436,
+      "learning_rate": 3.773563218390805e-05,
+      "loss": 0.1124,
+      "step": 2134
+    },
+    {
+      "epoch": 2.474824313555024,
+      "grad_norm": 0.3522142469882965,
+      "learning_rate": 3.7729885057471265e-05,
+      "loss": 0.1125,
+      "step": 2135
+    },
+    {
+      "epoch": 2.475983481851771,
+      "grad_norm": 0.2694511115550995,
+      "learning_rate": 3.772413793103449e-05,
+      "loss": 0.117,
+      "step": 2136
+    },
+    {
+      "epoch": 2.4771426501485183,
+      "grad_norm": 0.261785626411438,
+      "learning_rate": 3.77183908045977e-05,
+      "loss": 0.112,
+      "step": 2137
+    },
+    {
+      "epoch": 2.4783018184452654,
+      "grad_norm": 0.24989941716194153,
+      "learning_rate": 3.7712643678160917e-05,
+      "loss": 0.1075,
+      "step": 2138
+    },
+    {
+      "epoch": 2.4794609867420125,
+      "grad_norm": 0.3909321427345276,
+      "learning_rate": 3.770689655172414e-05,
+      "loss": 0.1222,
+      "step": 2139
+    },
+    {
+      "epoch": 2.4806201550387597,
+      "grad_norm": 0.2468433380126953,
+      "learning_rate": 3.770114942528736e-05,
+      "loss": 0.1193,
+      "step": 2140
+    },
+    {
+      "epoch": 2.4817793233355068,
+      "grad_norm": 0.27097734808921814,
+      "learning_rate": 3.7695402298850575e-05,
+      "loss": 0.1379,
+      "step": 2141
+    },
+    {
+      "epoch": 2.482938491632254,
+      "grad_norm": 0.26425981521606445,
+      "learning_rate": 3.7689655172413797e-05,
+      "loss": 0.1156,
+      "step": 2142
+    },
+    {
+      "epoch": 2.484097659929001,
+      "grad_norm": 0.23942025005817413,
+      "learning_rate": 3.768390804597701e-05,
+      "loss": 0.1217,
+      "step": 2143
+    },
+    {
+      "epoch": 2.485256828225748,
+      "grad_norm": 0.27993398904800415,
+      "learning_rate": 3.767816091954023e-05,
+      "loss": 0.1239,
+      "step": 2144
+    },
+    {
+      "epoch": 2.4864159965224952,
+      "grad_norm": 0.2943548262119293,
+      "learning_rate": 3.767241379310345e-05,
+      "loss": 0.1265,
+      "step": 2145
+    },
+    {
+      "epoch": 2.4875751648192423,
+      "grad_norm": 0.33951184153556824,
+      "learning_rate": 3.766666666666667e-05,
+      "loss": 0.1267,
+      "step": 2146
+    },
+    {
+      "epoch": 2.4887343331159895,
+      "grad_norm": 0.2834022641181946,
+      "learning_rate": 3.766091954022989e-05,
+      "loss": 0.1261,
+      "step": 2147
+    },
+    {
+      "epoch": 2.4898935014127366,
+      "grad_norm": 0.2666892111301422,
+      "learning_rate": 3.7655172413793106e-05,
+      "loss": 0.1148,
+      "step": 2148
+    },
+    {
+      "epoch": 2.4910526697094832,
+      "grad_norm": 0.22419312596321106,
+      "learning_rate": 3.764942528735632e-05,
+      "loss": 0.1095,
+      "step": 2149
+    },
+    {
+      "epoch": 2.4922118380062304,
+      "grad_norm": 0.2352280616760254,
+      "learning_rate": 3.764367816091954e-05,
+      "loss": 0.1183,
+      "step": 2150
+    },
+    {
+      "epoch": 2.4933710063029775,
+      "grad_norm": 0.2600967288017273,
+      "learning_rate": 3.763793103448276e-05,
+      "loss": 0.1197,
+      "step": 2151
+    },
+    {
+      "epoch": 2.4945301745997246,
+      "grad_norm": 0.25782260298728943,
+      "learning_rate": 3.763218390804598e-05,
+      "loss": 0.1201,
+      "step": 2152
+    },
+    {
+      "epoch": 2.4956893428964717,
+      "grad_norm": 0.2279644012451172,
+      "learning_rate": 3.76264367816092e-05,
+      "loss": 0.1141,
+      "step": 2153
+    },
+    {
+      "epoch": 2.496848511193219,
+      "grad_norm": 0.3026202917098999,
+      "learning_rate": 3.7620689655172416e-05,
+      "loss": 0.1371,
+      "step": 2154
+    },
+    {
+      "epoch": 2.498007679489966,
+      "grad_norm": 0.18554648756980896,
+      "learning_rate": 3.761494252873564e-05,
+      "loss": 0.0954,
+      "step": 2155
+    },
+    {
+      "epoch": 2.499166847786713,
+      "grad_norm": 0.22495919466018677,
+      "learning_rate": 3.760919540229885e-05,
+      "loss": 0.1215,
+      "step": 2156
+    },
+    {
+      "epoch": 2.50032601608346,
+      "grad_norm": 0.26901552081108093,
+      "learning_rate": 3.760344827586207e-05,
+      "loss": 0.1227,
+      "step": 2157
+    },
+    {
+      "epoch": 2.5014851843802073,
+      "grad_norm": 0.2286049723625183,
+      "learning_rate": 3.759770114942529e-05,
+      "loss": 0.1266,
+      "step": 2158
+    },
+    {
+      "epoch": 2.5026443526769544,
+      "grad_norm": 0.2322368025779724,
+      "learning_rate": 3.7591954022988504e-05,
+      "loss": 0.1073,
+      "step": 2159
+    },
+    {
+      "epoch": 2.5038035209737015,
+      "grad_norm": 0.24827679991722107,
+      "learning_rate": 3.7586206896551726e-05,
+      "loss": 0.1278,
+      "step": 2160
+    },
+    {
+      "epoch": 2.5049626892704486,
+      "grad_norm": 0.24690257012844086,
+      "learning_rate": 3.758045977011495e-05,
+      "loss": 0.1126,
+      "step": 2161
+    },
+    {
+      "epoch": 2.5061218575671953,
+      "grad_norm": 0.2480195313692093,
+      "learning_rate": 3.757471264367816e-05,
+      "loss": 0.119,
+      "step": 2162
+    },
+    {
+      "epoch": 2.507281025863943,
+      "grad_norm": 0.21804195642471313,
+      "learning_rate": 3.7568965517241384e-05,
+      "loss": 0.1119,
+      "step": 2163
+    },
+    {
+      "epoch": 2.5084401941606895,
+      "grad_norm": 0.25382915139198303,
+      "learning_rate": 3.75632183908046e-05,
+      "loss": 0.1198,
+      "step": 2164
+    },
+    {
+      "epoch": 2.5095993624574366,
+      "grad_norm": 0.24101321399211884,
+      "learning_rate": 3.7557471264367814e-05,
+      "loss": 0.1187,
+      "step": 2165
+    },
+    {
+      "epoch": 2.5107585307541838,
+      "grad_norm": 0.28679129481315613,
+      "learning_rate": 3.7551724137931035e-05,
+      "loss": 0.127,
+      "step": 2166
+    },
+    {
+      "epoch": 2.511917699050931,
+      "grad_norm": 0.2896977365016937,
+      "learning_rate": 3.754597701149426e-05,
+      "loss": 0.1246,
+      "step": 2167
+    },
+    {
+      "epoch": 2.513076867347678,
+      "grad_norm": 0.31614774465560913,
+      "learning_rate": 3.754022988505747e-05,
+      "loss": 0.1283,
+      "step": 2168
+    },
+    {
+      "epoch": 2.514236035644425,
+      "grad_norm": 0.29304152727127075,
+      "learning_rate": 3.7534482758620694e-05,
+      "loss": 0.1263,
+      "step": 2169
+    },
+    {
+      "epoch": 2.515395203941172,
+      "grad_norm": 0.29353946447372437,
+      "learning_rate": 3.752873563218391e-05,
+      "loss": 0.1148,
+      "step": 2170
+    },
+    {
+      "epoch": 2.5165543722379193,
+      "grad_norm": 0.28586602210998535,
+      "learning_rate": 3.752298850574713e-05,
+      "loss": 0.1202,
+      "step": 2171
+    },
+    {
+      "epoch": 2.5177135405346664,
+      "grad_norm": 0.29421526193618774,
+      "learning_rate": 3.7517241379310345e-05,
+      "loss": 0.1329,
+      "step": 2172
+    },
+    {
+      "epoch": 2.5188727088314136,
+      "grad_norm": 0.23773986101150513,
+      "learning_rate": 3.751149425287357e-05,
+      "loss": 0.1212,
+      "step": 2173
+    },
+    {
+      "epoch": 2.5200318771281607,
+      "grad_norm": 0.2569189965724945,
+      "learning_rate": 3.750574712643679e-05,
+      "loss": 0.1142,
+      "step": 2174
+    },
+    {
+      "epoch": 2.521191045424908,
+      "grad_norm": 0.2963407337665558,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.1409,
+      "step": 2175
+    },
+    {
+      "epoch": 2.522350213721655,
+      "grad_norm": 0.24243849515914917,
+      "learning_rate": 3.749425287356322e-05,
+      "loss": 0.1304,
+      "step": 2176
+    },
+    {
+      "epoch": 2.5235093820184016,
+      "grad_norm": 0.27366575598716736,
+      "learning_rate": 3.748850574712644e-05,
+      "loss": 0.1381,
+      "step": 2177
+    },
+    {
+      "epoch": 2.524668550315149,
+      "grad_norm": 0.2454218864440918,
+      "learning_rate": 3.7482758620689655e-05,
+      "loss": 0.1145,
+      "step": 2178
+    },
+    {
+      "epoch": 2.525827718611896,
+      "grad_norm": 0.23390009999275208,
+      "learning_rate": 3.747701149425287e-05,
+      "loss": 0.1131,
+      "step": 2179
+    },
+    {
+      "epoch": 2.526986886908643,
+      "grad_norm": 0.24678659439086914,
+      "learning_rate": 3.747126436781609e-05,
+      "loss": 0.1215,
+      "step": 2180
+    },
+    {
+      "epoch": 2.52814605520539,
+      "grad_norm": 0.2842147648334503,
+      "learning_rate": 3.746551724137931e-05,
+      "loss": 0.1298,
+      "step": 2181
+    },
+    {
+      "epoch": 2.529305223502137,
+      "grad_norm": 0.3027459979057312,
+      "learning_rate": 3.7459770114942535e-05,
+      "loss": 0.1212,
+      "step": 2182
+    },
+    {
+      "epoch": 2.5304643917988843,
+      "grad_norm": 0.31403622031211853,
+      "learning_rate": 3.745402298850575e-05,
+      "loss": 0.1231,
+      "step": 2183
+    },
+    {
+      "epoch": 2.5316235600956314,
+      "grad_norm": 0.2930856943130493,
+      "learning_rate": 3.7448275862068965e-05,
+      "loss": 0.13,
+      "step": 2184
+    },
+    {
+      "epoch": 2.5327827283923785,
+      "grad_norm": 0.2218947857618332,
+      "learning_rate": 3.7442528735632186e-05,
+      "loss": 0.1079,
+      "step": 2185
+    },
+    {
+      "epoch": 2.5339418966891256,
+      "grad_norm": 0.2198045253753662,
+      "learning_rate": 3.74367816091954e-05,
+      "loss": 0.1077,
+      "step": 2186
+    },
+    {
+      "epoch": 2.5351010649858727,
+      "grad_norm": 0.198323592543602,
+      "learning_rate": 3.743103448275862e-05,
+      "loss": 0.11,
+      "step": 2187
+    },
+    {
+      "epoch": 2.53626023328262,
+      "grad_norm": 0.258024662733078,
+      "learning_rate": 3.7425287356321845e-05,
+      "loss": 0.1162,
+      "step": 2188
+    },
+    {
+      "epoch": 2.537419401579367,
+      "grad_norm": 0.3027409315109253,
+      "learning_rate": 3.741954022988506e-05,
+      "loss": 0.1383,
+      "step": 2189
+    },
+    {
+      "epoch": 2.5385785698761136,
+      "grad_norm": 0.27792859077453613,
+      "learning_rate": 3.741379310344828e-05,
+      "loss": 0.1138,
+      "step": 2190
+    },
+    {
+      "epoch": 2.539737738172861,
+      "grad_norm": 0.2331332266330719,
+      "learning_rate": 3.7408045977011496e-05,
+      "loss": 0.1065,
+      "step": 2191
+    },
+    {
+      "epoch": 2.540896906469608,
+      "grad_norm": 0.21128974854946136,
+      "learning_rate": 3.740229885057471e-05,
+      "loss": 0.1197,
+      "step": 2192
+    },
+    {
+      "epoch": 2.542056074766355,
+      "grad_norm": 0.2796282172203064,
+      "learning_rate": 3.739655172413793e-05,
+      "loss": 0.1131,
+      "step": 2193
+    },
+    {
+      "epoch": 2.543215243063102,
+      "grad_norm": 0.24139229953289032,
+      "learning_rate": 3.7390804597701154e-05,
+      "loss": 0.1165,
+      "step": 2194
+    },
+    {
+      "epoch": 2.544374411359849,
+      "grad_norm": 0.22860391438007355,
+      "learning_rate": 3.738505747126437e-05,
+      "loss": 0.1176,
+      "step": 2195
+    },
+    {
+      "epoch": 2.5455335796565963,
+      "grad_norm": 0.2661796510219574,
+      "learning_rate": 3.737931034482759e-05,
+      "loss": 0.1295,
+      "step": 2196
+    },
+    {
+      "epoch": 2.5466927479533434,
+      "grad_norm": 0.28988024592399597,
+      "learning_rate": 3.7373563218390806e-05,
+      "loss": 0.1363,
+      "step": 2197
+    },
+    {
+      "epoch": 2.5478519162500906,
+      "grad_norm": 0.26467469334602356,
+      "learning_rate": 3.736781609195402e-05,
+      "loss": 0.1078,
+      "step": 2198
+    },
+    {
+      "epoch": 2.5490110845468377,
+      "grad_norm": 0.3318430483341217,
+      "learning_rate": 3.736206896551724e-05,
+      "loss": 0.1299,
+      "step": 2199
+    },
+    {
+      "epoch": 2.550170252843585,
+      "grad_norm": 0.24909447133541107,
+      "learning_rate": 3.735632183908046e-05,
+      "loss": 0.1019,
+      "step": 2200
+    },
+    {
+      "epoch": 2.551329421140332,
+      "grad_norm": 0.3391667604446411,
+      "learning_rate": 3.735057471264368e-05,
+      "loss": 0.1336,
+      "step": 2201
+    },
+    {
+      "epoch": 2.552488589437079,
+      "grad_norm": 0.2599574327468872,
+      "learning_rate": 3.73448275862069e-05,
+      "loss": 0.1154,
+      "step": 2202
+    },
+    {
+      "epoch": 2.553647757733826,
+      "grad_norm": 0.22761644423007965,
+      "learning_rate": 3.7339080459770116e-05,
+      "loss": 0.0995,
+      "step": 2203
+    },
+    {
+      "epoch": 2.5548069260305732,
+      "grad_norm": 0.2874188721179962,
+      "learning_rate": 3.733333333333334e-05,
+      "loss": 0.1172,
+      "step": 2204
+    },
+    {
+      "epoch": 2.55596609432732,
+      "grad_norm": 0.2598547041416168,
+      "learning_rate": 3.732758620689655e-05,
+      "loss": 0.1315,
+      "step": 2205
+    },
+    {
+      "epoch": 2.5571252626240675,
+      "grad_norm": 0.2864150106906891,
+      "learning_rate": 3.732183908045977e-05,
+      "loss": 0.1163,
+      "step": 2206
+    },
+    {
+      "epoch": 2.558284430920814,
+      "grad_norm": 0.3307492434978485,
+      "learning_rate": 3.731609195402299e-05,
+      "loss": 0.1267,
+      "step": 2207
+    },
+    {
+      "epoch": 2.5594435992175613,
+      "grad_norm": 0.2648185193538666,
+      "learning_rate": 3.731034482758621e-05,
+      "loss": 0.1218,
+      "step": 2208
+    },
+    {
+      "epoch": 2.5606027675143084,
+      "grad_norm": 0.21011538803577423,
+      "learning_rate": 3.730459770114943e-05,
+      "loss": 0.1069,
+      "step": 2209
+    },
+    {
+      "epoch": 2.5617619358110555,
+      "grad_norm": 0.31918808817863464,
+      "learning_rate": 3.729885057471265e-05,
+      "loss": 0.1256,
+      "step": 2210
+    },
+    {
+      "epoch": 2.5629211041078026,
+      "grad_norm": 0.28522172570228577,
+      "learning_rate": 3.729310344827586e-05,
+      "loss": 0.1294,
+      "step": 2211
+    },
+    {
+      "epoch": 2.5640802724045497,
+      "grad_norm": 0.21739451587200165,
+      "learning_rate": 3.7287356321839084e-05,
+      "loss": 0.1001,
+      "step": 2212
+    },
+    {
+      "epoch": 2.565239440701297,
+      "grad_norm": 0.23575380444526672,
+      "learning_rate": 3.72816091954023e-05,
+      "loss": 0.1192,
+      "step": 2213
+    },
+    {
+      "epoch": 2.566398608998044,
+      "grad_norm": 0.20924952626228333,
+      "learning_rate": 3.727586206896552e-05,
+      "loss": 0.1079,
+      "step": 2214
+    },
+    {
+      "epoch": 2.567557777294791,
+      "grad_norm": 0.31391969323158264,
+      "learning_rate": 3.727011494252874e-05,
+      "loss": 0.1236,
+      "step": 2215
+    },
+    {
+      "epoch": 2.568716945591538,
+      "grad_norm": 0.2599903643131256,
+      "learning_rate": 3.726436781609196e-05,
+      "loss": 0.1175,
+      "step": 2216
+    },
+    {
+      "epoch": 2.5698761138882853,
+      "grad_norm": 0.2094232141971588,
+      "learning_rate": 3.725862068965517e-05,
+      "loss": 0.1147,
+      "step": 2217
+    },
+    {
+      "epoch": 2.571035282185032,
+      "grad_norm": 0.2692113518714905,
+      "learning_rate": 3.725287356321839e-05,
+      "loss": 0.125,
+      "step": 2218
+    },
+    {
+      "epoch": 2.5721944504817795,
+      "grad_norm": 0.28311142325401306,
+      "learning_rate": 3.724712643678161e-05,
+      "loss": 0.1233,
+      "step": 2219
+    },
+    {
+      "epoch": 2.573353618778526,
+      "grad_norm": 0.31285127997398376,
+      "learning_rate": 3.724137931034483e-05,
+      "loss": 0.1282,
+      "step": 2220
+    },
+    {
+      "epoch": 2.5745127870752738,
+      "grad_norm": 0.2766589820384979,
+      "learning_rate": 3.7235632183908045e-05,
+      "loss": 0.1163,
+      "step": 2221
+    },
+    {
+      "epoch": 2.5756719553720204,
+      "grad_norm": 0.26953697204589844,
+      "learning_rate": 3.7229885057471267e-05,
+      "loss": 0.1296,
+      "step": 2222
+    },
+    {
+      "epoch": 2.5768311236687675,
+      "grad_norm": 0.23072504997253418,
+      "learning_rate": 3.722413793103449e-05,
+      "loss": 0.1202,
+      "step": 2223
+    },
+    {
+      "epoch": 2.5779902919655147,
+      "grad_norm": 0.22109727561473846,
+      "learning_rate": 3.72183908045977e-05,
+      "loss": 0.1187,
+      "step": 2224
+    },
+    {
+      "epoch": 2.5791494602622618,
+      "grad_norm": 0.28811052441596985,
+      "learning_rate": 3.721264367816092e-05,
+      "loss": 0.1201,
+      "step": 2225
+    },
+    {
+      "epoch": 2.580308628559009,
+      "grad_norm": 0.31841617822647095,
+      "learning_rate": 3.720689655172414e-05,
+      "loss": 0.1092,
+      "step": 2226
+    },
+    {
+      "epoch": 2.581467796855756,
+      "grad_norm": 0.27956047654151917,
+      "learning_rate": 3.7201149425287355e-05,
+      "loss": 0.1238,
+      "step": 2227
+    },
+    {
+      "epoch": 2.582626965152503,
+      "grad_norm": 0.28410202264785767,
+      "learning_rate": 3.7195402298850576e-05,
+      "loss": 0.122,
+      "step": 2228
+    },
+    {
+      "epoch": 2.5837861334492502,
+      "grad_norm": 0.23132778704166412,
+      "learning_rate": 3.71896551724138e-05,
+      "loss": 0.1221,
+      "step": 2229
+    },
+    {
+      "epoch": 2.5849453017459973,
+      "grad_norm": 0.27684658765792847,
+      "learning_rate": 3.718390804597701e-05,
+      "loss": 0.1099,
+      "step": 2230
+    },
+    {
+      "epoch": 2.5861044700427445,
+      "grad_norm": 0.241298645734787,
+      "learning_rate": 3.7178160919540235e-05,
+      "loss": 0.11,
+      "step": 2231
+    },
+    {
+      "epoch": 2.5872636383394916,
+      "grad_norm": 0.23437266051769257,
+      "learning_rate": 3.717241379310345e-05,
+      "loss": 0.1234,
+      "step": 2232
+    },
+    {
+      "epoch": 2.5884228066362382,
+      "grad_norm": 0.27701497077941895,
+      "learning_rate": 3.7166666666666664e-05,
+      "loss": 0.1105,
+      "step": 2233
+    },
+    {
+      "epoch": 2.589581974932986,
+      "grad_norm": 0.24334724247455597,
+      "learning_rate": 3.7160919540229886e-05,
+      "loss": 0.1087,
+      "step": 2234
+    },
+    {
+      "epoch": 2.5907411432297325,
+      "grad_norm": 0.25163525342941284,
+      "learning_rate": 3.715517241379311e-05,
+      "loss": 0.1084,
+      "step": 2235
+    },
+    {
+      "epoch": 2.5919003115264796,
+      "grad_norm": 0.313559353351593,
+      "learning_rate": 3.714942528735632e-05,
+      "loss": 0.1397,
+      "step": 2236
+    },
+    {
+      "epoch": 2.5930594798232267,
+      "grad_norm": 0.2449546456336975,
+      "learning_rate": 3.7143678160919544e-05,
+      "loss": 0.1127,
+      "step": 2237
+    },
+    {
+      "epoch": 2.594218648119974,
+      "grad_norm": 0.2583812475204468,
+      "learning_rate": 3.713793103448276e-05,
+      "loss": 0.1089,
+      "step": 2238
+    },
+    {
+      "epoch": 2.595377816416721,
+      "grad_norm": 0.2844179570674896,
+      "learning_rate": 3.713218390804598e-05,
+      "loss": 0.1166,
+      "step": 2239
+    },
+    {
+      "epoch": 2.596536984713468,
+      "grad_norm": 0.2738819718360901,
+      "learning_rate": 3.7126436781609196e-05,
+      "loss": 0.1199,
+      "step": 2240
+    },
+    {
+      "epoch": 2.597696153010215,
+      "grad_norm": 0.253930002450943,
+      "learning_rate": 3.712068965517241e-05,
+      "loss": 0.1134,
+      "step": 2241
+    },
+    {
+      "epoch": 2.5988553213069623,
+      "grad_norm": 0.2768709659576416,
+      "learning_rate": 3.711494252873563e-05,
+      "loss": 0.1121,
+      "step": 2242
+    },
+    {
+      "epoch": 2.6000144896037094,
+      "grad_norm": 0.2729703187942505,
+      "learning_rate": 3.7109195402298854e-05,
+      "loss": 0.1197,
+      "step": 2243
+    },
+    {
+      "epoch": 2.6011736579004565,
+      "grad_norm": 0.3085278570652008,
+      "learning_rate": 3.710344827586207e-05,
+      "loss": 0.1363,
+      "step": 2244
+    },
+    {
+      "epoch": 2.6023328261972036,
+      "grad_norm": 0.2976699471473694,
+      "learning_rate": 3.709770114942529e-05,
+      "loss": 0.1215,
+      "step": 2245
+    },
+    {
+      "epoch": 2.6034919944939507,
+      "grad_norm": 0.24314960837364197,
+      "learning_rate": 3.7091954022988505e-05,
+      "loss": 0.1124,
+      "step": 2246
+    },
+    {
+      "epoch": 2.604651162790698,
+      "grad_norm": 0.20558379590511322,
+      "learning_rate": 3.708620689655173e-05,
+      "loss": 0.1064,
+      "step": 2247
+    },
+    {
+      "epoch": 2.6058103310874445,
+      "grad_norm": 0.2385900467634201,
+      "learning_rate": 3.708045977011494e-05,
+      "loss": 0.1125,
+      "step": 2248
+    },
+    {
+      "epoch": 2.606969499384192,
+      "grad_norm": 0.3381904363632202,
+      "learning_rate": 3.7074712643678164e-05,
+      "loss": 0.1222,
+      "step": 2249
+    },
+    {
+      "epoch": 2.6081286676809388,
+      "grad_norm": 0.24625416100025177,
+      "learning_rate": 3.7068965517241385e-05,
+      "loss": 0.1172,
+      "step": 2250
+    },
+    {
+      "epoch": 2.609287835977686,
+      "grad_norm": 0.2692759931087494,
+      "learning_rate": 3.70632183908046e-05,
+      "loss": 0.1275,
+      "step": 2251
+    },
+    {
+      "epoch": 2.610447004274433,
+      "grad_norm": 0.2261049598455429,
+      "learning_rate": 3.7057471264367815e-05,
+      "loss": 0.1132,
+      "step": 2252
+    },
+    {
+      "epoch": 2.61160617257118,
+      "grad_norm": 0.2826228141784668,
+      "learning_rate": 3.705172413793104e-05,
+      "loss": 0.1276,
+      "step": 2253
+    },
+    {
+      "epoch": 2.612765340867927,
+      "grad_norm": 0.24635249376296997,
+      "learning_rate": 3.704597701149425e-05,
+      "loss": 0.1231,
+      "step": 2254
+    },
+    {
+      "epoch": 2.6139245091646743,
+      "grad_norm": 0.2234896570444107,
+      "learning_rate": 3.7040229885057473e-05,
+      "loss": 0.1048,
+      "step": 2255
+    },
+    {
+      "epoch": 2.6150836774614215,
+      "grad_norm": 0.2619444727897644,
+      "learning_rate": 3.7034482758620695e-05,
+      "loss": 0.1306,
+      "step": 2256
+    },
+    {
+      "epoch": 2.6162428457581686,
+      "grad_norm": 0.3317265808582306,
+      "learning_rate": 3.702873563218391e-05,
+      "loss": 0.1275,
+      "step": 2257
+    },
+    {
+      "epoch": 2.6174020140549157,
+      "grad_norm": 0.2708195447921753,
+      "learning_rate": 3.702298850574713e-05,
+      "loss": 0.1133,
+      "step": 2258
+    },
+    {
+      "epoch": 2.618561182351663,
+      "grad_norm": 0.2827620208263397,
+      "learning_rate": 3.701724137931035e-05,
+      "loss": 0.1307,
+      "step": 2259
+    },
+    {
+      "epoch": 2.61972035064841,
+      "grad_norm": 0.24215327203273773,
+      "learning_rate": 3.701149425287356e-05,
+      "loss": 0.1187,
+      "step": 2260
+    },
+    {
+      "epoch": 2.6208795189451566,
+      "grad_norm": 0.2541255056858063,
+      "learning_rate": 3.700574712643678e-05,
+      "loss": 0.1215,
+      "step": 2261
+    },
+    {
+      "epoch": 2.622038687241904,
+      "grad_norm": 0.2253810614347458,
+      "learning_rate": 3.7e-05,
+      "loss": 0.1169,
+      "step": 2262
+    },
+    {
+      "epoch": 2.623197855538651,
+      "grad_norm": 0.27330759167671204,
+      "learning_rate": 3.699425287356322e-05,
+      "loss": 0.1285,
+      "step": 2263
+    },
+    {
+      "epoch": 2.6243570238353984,
+      "grad_norm": 0.2241210639476776,
+      "learning_rate": 3.698850574712644e-05,
+      "loss": 0.1143,
+      "step": 2264
+    },
+    {
+      "epoch": 2.625516192132145,
+      "grad_norm": 0.28012943267822266,
+      "learning_rate": 3.6982758620689656e-05,
+      "loss": 0.1303,
+      "step": 2265
+    },
+    {
+      "epoch": 2.626675360428892,
+      "grad_norm": 0.2770480215549469,
+      "learning_rate": 3.697701149425288e-05,
+      "loss": 0.1294,
+      "step": 2266
+    },
+    {
+      "epoch": 2.6278345287256393,
+      "grad_norm": 0.22335301339626312,
+      "learning_rate": 3.697126436781609e-05,
+      "loss": 0.1223,
+      "step": 2267
+    },
+    {
+      "epoch": 2.6289936970223864,
+      "grad_norm": 0.2822023332118988,
+      "learning_rate": 3.696551724137931e-05,
+      "loss": 0.123,
+      "step": 2268
+    },
+    {
+      "epoch": 2.6301528653191335,
+      "grad_norm": 0.2653454840183258,
+      "learning_rate": 3.695977011494253e-05,
+      "loss": 0.1153,
+      "step": 2269
+    },
+    {
+      "epoch": 2.6313120336158806,
+      "grad_norm": 0.2543680965900421,
+      "learning_rate": 3.695402298850575e-05,
+      "loss": 0.1211,
+      "step": 2270
+    },
+    {
+      "epoch": 2.6324712019126277,
+      "grad_norm": 0.3247624635696411,
+      "learning_rate": 3.6948275862068966e-05,
+      "loss": 0.1232,
+      "step": 2271
+    },
+    {
+      "epoch": 2.633630370209375,
+      "grad_norm": 0.30818063020706177,
+      "learning_rate": 3.694252873563219e-05,
+      "loss": 0.1255,
+      "step": 2272
+    },
+    {
+      "epoch": 2.634789538506122,
+      "grad_norm": 0.2480584979057312,
+      "learning_rate": 3.69367816091954e-05,
+      "loss": 0.1266,
+      "step": 2273
+    },
+    {
+      "epoch": 2.635948706802869,
+      "grad_norm": 0.2562277615070343,
+      "learning_rate": 3.6931034482758624e-05,
+      "loss": 0.1245,
+      "step": 2274
+    },
+    {
+      "epoch": 2.637107875099616,
+      "grad_norm": 0.21452297270298004,
+      "learning_rate": 3.692528735632184e-05,
+      "loss": 0.112,
+      "step": 2275
+    },
+    {
+      "epoch": 2.638267043396363,
+      "grad_norm": 0.20973417162895203,
+      "learning_rate": 3.691954022988506e-05,
+      "loss": 0.1148,
+      "step": 2276
+    },
+    {
+      "epoch": 2.6394262116931104,
+      "grad_norm": 0.2719014286994934,
+      "learning_rate": 3.691379310344828e-05,
+      "loss": 0.1165,
+      "step": 2277
+    },
+    {
+      "epoch": 2.640585379989857,
+      "grad_norm": 0.2603360414505005,
+      "learning_rate": 3.69080459770115e-05,
+      "loss": 0.1217,
+      "step": 2278
+    },
+    {
+      "epoch": 2.641744548286604,
+      "grad_norm": 0.2589358985424042,
+      "learning_rate": 3.690229885057471e-05,
+      "loss": 0.1261,
+      "step": 2279
+    },
+    {
+      "epoch": 2.6429037165833513,
+      "grad_norm": 0.2366466522216797,
+      "learning_rate": 3.6896551724137934e-05,
+      "loss": 0.112,
+      "step": 2280
+    },
+    {
+      "epoch": 2.6440628848800984,
+      "grad_norm": 0.3076455891132355,
+      "learning_rate": 3.689080459770115e-05,
+      "loss": 0.1337,
+      "step": 2281
+    },
+    {
+      "epoch": 2.6452220531768456,
+      "grad_norm": 0.24393005669116974,
+      "learning_rate": 3.6885057471264364e-05,
+      "loss": 0.1198,
+      "step": 2282
+    },
+    {
+      "epoch": 2.6463812214735927,
+      "grad_norm": 0.23066796362400055,
+      "learning_rate": 3.6879310344827586e-05,
+      "loss": 0.1197,
+      "step": 2283
+    },
+    {
+      "epoch": 2.64754038977034,
+      "grad_norm": 0.22284121811389923,
+      "learning_rate": 3.687356321839081e-05,
+      "loss": 0.1076,
+      "step": 2284
+    },
+    {
+      "epoch": 2.648699558067087,
+      "grad_norm": 0.2574465870857239,
+      "learning_rate": 3.686781609195403e-05,
+      "loss": 0.1185,
+      "step": 2285
+    },
+    {
+      "epoch": 2.649858726363834,
+      "grad_norm": 0.2645076811313629,
+      "learning_rate": 3.6862068965517244e-05,
+      "loss": 0.1175,
+      "step": 2286
+    },
+    {
+      "epoch": 2.651017894660581,
+      "grad_norm": 0.25155895948410034,
+      "learning_rate": 3.685632183908046e-05,
+      "loss": 0.1228,
+      "step": 2287
+    },
+    {
+      "epoch": 2.6521770629573282,
+      "grad_norm": 0.26526379585266113,
+      "learning_rate": 3.685057471264368e-05,
+      "loss": 0.1196,
+      "step": 2288
+    },
+    {
+      "epoch": 2.6533362312540754,
+      "grad_norm": 0.25388583540916443,
+      "learning_rate": 3.6844827586206895e-05,
+      "loss": 0.1181,
+      "step": 2289
+    },
+    {
+      "epoch": 2.6544953995508225,
+      "grad_norm": 0.31535226106643677,
+      "learning_rate": 3.683908045977012e-05,
+      "loss": 0.1282,
+      "step": 2290
+    },
+    {
+      "epoch": 2.655654567847569,
+      "grad_norm": 0.2896163761615753,
+      "learning_rate": 3.683333333333334e-05,
+      "loss": 0.1271,
+      "step": 2291
+    },
+    {
+      "epoch": 2.6568137361443167,
+      "grad_norm": 0.26265934109687805,
+      "learning_rate": 3.6827586206896554e-05,
+      "loss": 0.1289,
+      "step": 2292
+    },
+    {
+      "epoch": 2.6579729044410634,
+      "grad_norm": 0.25363272428512573,
+      "learning_rate": 3.6821839080459775e-05,
+      "loss": 0.1212,
+      "step": 2293
+    },
+    {
+      "epoch": 2.6591320727378105,
+      "grad_norm": 0.304082453250885,
+      "learning_rate": 3.681609195402299e-05,
+      "loss": 0.1121,
+      "step": 2294
+    },
+    {
+      "epoch": 2.6602912410345576,
+      "grad_norm": 0.21810661256313324,
+      "learning_rate": 3.6810344827586205e-05,
+      "loss": 0.1177,
+      "step": 2295
+    },
+    {
+      "epoch": 2.6614504093313047,
+      "grad_norm": 0.24036329984664917,
+      "learning_rate": 3.680459770114943e-05,
+      "loss": 0.109,
+      "step": 2296
+    },
+    {
+      "epoch": 2.662609577628052,
+      "grad_norm": 0.2967372238636017,
+      "learning_rate": 3.679885057471265e-05,
+      "loss": 0.1218,
+      "step": 2297
+    },
+    {
+      "epoch": 2.663768745924799,
+      "grad_norm": 0.3007439374923706,
+      "learning_rate": 3.679310344827586e-05,
+      "loss": 0.1286,
+      "step": 2298
+    },
+    {
+      "epoch": 2.664927914221546,
+      "grad_norm": 0.26587292551994324,
+      "learning_rate": 3.6787356321839085e-05,
+      "loss": 0.1242,
+      "step": 2299
+    },
+    {
+      "epoch": 2.666087082518293,
+      "grad_norm": 0.2876141369342804,
+      "learning_rate": 3.67816091954023e-05,
+      "loss": 0.1322,
+      "step": 2300
+    },
+    {
+      "epoch": 2.6672462508150403,
+      "grad_norm": 0.27488911151885986,
+      "learning_rate": 3.6775862068965515e-05,
+      "loss": 0.1218,
+      "step": 2301
+    },
+    {
+      "epoch": 2.6684054191117874,
+      "grad_norm": 0.2519826292991638,
+      "learning_rate": 3.6770114942528736e-05,
+      "loss": 0.1138,
+      "step": 2302
+    },
+    {
+      "epoch": 2.6695645874085345,
+      "grad_norm": 0.26412853598594666,
+      "learning_rate": 3.676436781609195e-05,
+      "loss": 0.1184,
+      "step": 2303
+    },
+    {
+      "epoch": 2.670723755705281,
+      "grad_norm": 0.24107958376407623,
+      "learning_rate": 3.675862068965518e-05,
+      "loss": 0.1159,
+      "step": 2304
+    },
+    {
+      "epoch": 2.6718829240020288,
+      "grad_norm": 0.25105899572372437,
+      "learning_rate": 3.6752873563218395e-05,
+      "loss": 0.1075,
+      "step": 2305
+    },
+    {
+      "epoch": 2.6730420922987754,
+      "grad_norm": 0.23559515178203583,
+      "learning_rate": 3.674712643678161e-05,
+      "loss": 0.1225,
+      "step": 2306
+    },
+    {
+      "epoch": 2.674201260595523,
+      "grad_norm": 0.23702873289585114,
+      "learning_rate": 3.674137931034483e-05,
+      "loss": 0.1221,
+      "step": 2307
+    },
+    {
+      "epoch": 2.6753604288922697,
+      "grad_norm": 0.21996693313121796,
+      "learning_rate": 3.6735632183908046e-05,
+      "loss": 0.1152,
+      "step": 2308
+    },
+    {
+      "epoch": 2.6765195971890168,
+      "grad_norm": 0.2361491173505783,
+      "learning_rate": 3.672988505747126e-05,
+      "loss": 0.115,
+      "step": 2309
+    },
+    {
+      "epoch": 2.677678765485764,
+      "grad_norm": 0.24125604331493378,
+      "learning_rate": 3.672413793103448e-05,
+      "loss": 0.1197,
+      "step": 2310
+    },
+    {
+      "epoch": 2.678837933782511,
+      "grad_norm": 0.23669910430908203,
+      "learning_rate": 3.6718390804597704e-05,
+      "loss": 0.1091,
+      "step": 2311
+    },
+    {
+      "epoch": 2.679997102079258,
+      "grad_norm": 0.26078295707702637,
+      "learning_rate": 3.6712643678160926e-05,
+      "loss": 0.1059,
+      "step": 2312
+    },
+    {
+      "epoch": 2.6811562703760052,
+      "grad_norm": 0.30441415309906006,
+      "learning_rate": 3.670689655172414e-05,
+      "loss": 0.1154,
+      "step": 2313
+    },
+    {
+      "epoch": 2.6823154386727524,
+      "grad_norm": 0.22218701243400574,
+      "learning_rate": 3.6701149425287356e-05,
+      "loss": 0.1235,
+      "step": 2314
+    },
+    {
+      "epoch": 2.6834746069694995,
+      "grad_norm": 0.25640806555747986,
+      "learning_rate": 3.669540229885058e-05,
+      "loss": 0.1218,
+      "step": 2315
+    },
+    {
+      "epoch": 2.6846337752662466,
+      "grad_norm": 0.2934137284755707,
+      "learning_rate": 3.668965517241379e-05,
+      "loss": 0.1164,
+      "step": 2316
+    },
+    {
+      "epoch": 2.6857929435629937,
+      "grad_norm": 0.27650073170661926,
+      "learning_rate": 3.6683908045977014e-05,
+      "loss": 0.1183,
+      "step": 2317
+    },
+    {
+      "epoch": 2.686952111859741,
+      "grad_norm": 0.24524487555027008,
+      "learning_rate": 3.6678160919540236e-05,
+      "loss": 0.1197,
+      "step": 2318
+    },
+    {
+      "epoch": 2.6881112801564875,
+      "grad_norm": 0.3019874095916748,
+      "learning_rate": 3.667241379310345e-05,
+      "loss": 0.1294,
+      "step": 2319
+    },
+    {
+      "epoch": 2.689270448453235,
+      "grad_norm": 0.321593314409256,
+      "learning_rate": 3.6666666666666666e-05,
+      "loss": 0.1279,
+      "step": 2320
+    },
+    {
+      "epoch": 2.6904296167499817,
+      "grad_norm": 0.26887890696525574,
+      "learning_rate": 3.666091954022989e-05,
+      "loss": 0.1225,
+      "step": 2321
+    },
+    {
+      "epoch": 2.691588785046729,
+      "grad_norm": 0.24992501735687256,
+      "learning_rate": 3.66551724137931e-05,
+      "loss": 0.1136,
+      "step": 2322
+    },
+    {
+      "epoch": 2.692747953343476,
+      "grad_norm": 0.21639297902584076,
+      "learning_rate": 3.6649425287356324e-05,
+      "loss": 0.1074,
+      "step": 2323
+    },
+    {
+      "epoch": 2.693907121640223,
+      "grad_norm": 0.3038938045501709,
+      "learning_rate": 3.6643678160919546e-05,
+      "loss": 0.1307,
+      "step": 2324
+    },
+    {
+      "epoch": 2.69506628993697,
+      "grad_norm": 0.31261420249938965,
+      "learning_rate": 3.663793103448276e-05,
+      "loss": 0.1392,
+      "step": 2325
+    },
+    {
+      "epoch": 2.6962254582337173,
+      "grad_norm": 0.22381141781806946,
+      "learning_rate": 3.663218390804598e-05,
+      "loss": 0.1043,
+      "step": 2326
+    },
+    {
+      "epoch": 2.6973846265304644,
+      "grad_norm": 0.28085264563560486,
+      "learning_rate": 3.66264367816092e-05,
+      "loss": 0.1145,
+      "step": 2327
+    },
+    {
+      "epoch": 2.6985437948272115,
+      "grad_norm": 0.2581823468208313,
+      "learning_rate": 3.662068965517241e-05,
+      "loss": 0.1333,
+      "step": 2328
+    },
+    {
+      "epoch": 2.6997029631239586,
+      "grad_norm": 0.25244686007499695,
+      "learning_rate": 3.6614942528735634e-05,
+      "loss": 0.1249,
+      "step": 2329
+    },
+    {
+      "epoch": 2.7008621314207057,
+      "grad_norm": 0.24646331369876862,
+      "learning_rate": 3.660919540229885e-05,
+      "loss": 0.1142,
+      "step": 2330
+    },
+    {
+      "epoch": 2.702021299717453,
+      "grad_norm": 0.2165023535490036,
+      "learning_rate": 3.660344827586207e-05,
+      "loss": 0.1225,
+      "step": 2331
+    },
+    {
+      "epoch": 2.7031804680142,
+      "grad_norm": 0.25640419125556946,
+      "learning_rate": 3.659770114942529e-05,
+      "loss": 0.1242,
+      "step": 2332
+    },
+    {
+      "epoch": 2.704339636310947,
+      "grad_norm": 0.33699721097946167,
+      "learning_rate": 3.659195402298851e-05,
+      "loss": 0.1284,
+      "step": 2333
+    },
+    {
+      "epoch": 2.7054988046076938,
+      "grad_norm": 0.2463027685880661,
+      "learning_rate": 3.658620689655173e-05,
+      "loss": 0.1252,
+      "step": 2334
+    },
+    {
+      "epoch": 2.7066579729044413,
+      "grad_norm": 0.2819848954677582,
+      "learning_rate": 3.6580459770114943e-05,
+      "loss": 0.1097,
+      "step": 2335
+    },
+    {
+      "epoch": 2.707817141201188,
+      "grad_norm": 0.2654036283493042,
+      "learning_rate": 3.657471264367816e-05,
+      "loss": 0.1187,
+      "step": 2336
+    },
+    {
+      "epoch": 2.708976309497935,
+      "grad_norm": 0.2798532545566559,
+      "learning_rate": 3.656896551724138e-05,
+      "loss": 0.1169,
+      "step": 2337
+    },
+    {
+      "epoch": 2.7101354777946822,
+      "grad_norm": 0.24664916098117828,
+      "learning_rate": 3.65632183908046e-05,
+      "loss": 0.1337,
+      "step": 2338
+    },
+    {
+      "epoch": 2.7112946460914293,
+      "grad_norm": 0.2289317101240158,
+      "learning_rate": 3.655747126436782e-05,
+      "loss": 0.1255,
+      "step": 2339
+    },
+    {
+      "epoch": 2.7124538143881765,
+      "grad_norm": 0.28109264373779297,
+      "learning_rate": 3.655172413793104e-05,
+      "loss": 0.1219,
+      "step": 2340
+    },
+    {
+      "epoch": 2.7136129826849236,
+      "grad_norm": 0.2645872235298157,
+      "learning_rate": 3.654597701149425e-05,
+      "loss": 0.1365,
+      "step": 2341
+    },
+    {
+      "epoch": 2.7147721509816707,
+      "grad_norm": 0.2895812690258026,
+      "learning_rate": 3.6540229885057475e-05,
+      "loss": 0.1297,
+      "step": 2342
+    },
+    {
+      "epoch": 2.715931319278418,
+      "grad_norm": 0.2900197207927704,
+      "learning_rate": 3.653448275862069e-05,
+      "loss": 0.1319,
+      "step": 2343
+    },
+    {
+      "epoch": 2.717090487575165,
+      "grad_norm": 0.22161021828651428,
+      "learning_rate": 3.6528735632183905e-05,
+      "loss": 0.1153,
+      "step": 2344
+    },
+    {
+      "epoch": 2.718249655871912,
+      "grad_norm": 0.28364619612693787,
+      "learning_rate": 3.652298850574713e-05,
+      "loss": 0.1261,
+      "step": 2345
+    },
+    {
+      "epoch": 2.719408824168659,
+      "grad_norm": 0.2544122636318207,
+      "learning_rate": 3.651724137931035e-05,
+      "loss": 0.1175,
+      "step": 2346
+    },
+    {
+      "epoch": 2.720567992465406,
+      "grad_norm": 0.21084167063236237,
+      "learning_rate": 3.651149425287356e-05,
+      "loss": 0.1011,
+      "step": 2347
+    },
+    {
+      "epoch": 2.7217271607621534,
+      "grad_norm": 0.24158698320388794,
+      "learning_rate": 3.6505747126436785e-05,
+      "loss": 0.1125,
+      "step": 2348
+    },
+    {
+      "epoch": 2.7228863290589,
+      "grad_norm": 0.2733836770057678,
+      "learning_rate": 3.65e-05,
+      "loss": 0.1187,
+      "step": 2349
+    },
+    {
+      "epoch": 2.724045497355647,
+      "grad_norm": 0.32197657227516174,
+      "learning_rate": 3.649425287356322e-05,
+      "loss": 0.1252,
+      "step": 2350
+    },
+    {
+      "epoch": 2.7252046656523943,
+      "grad_norm": 0.2622753381729126,
+      "learning_rate": 3.6488505747126436e-05,
+      "loss": 0.1251,
+      "step": 2351
+    },
+    {
+      "epoch": 2.7263638339491414,
+      "grad_norm": 0.2564488351345062,
+      "learning_rate": 3.648275862068966e-05,
+      "loss": 0.1175,
+      "step": 2352
+    },
+    {
+      "epoch": 2.7275230022458885,
+      "grad_norm": 0.29768943786621094,
+      "learning_rate": 3.647701149425288e-05,
+      "loss": 0.1133,
+      "step": 2353
+    },
+    {
+      "epoch": 2.7286821705426356,
+      "grad_norm": 0.297397643327713,
+      "learning_rate": 3.6471264367816094e-05,
+      "loss": 0.121,
+      "step": 2354
+    },
+    {
+      "epoch": 2.7298413388393827,
+      "grad_norm": 0.29082873463630676,
+      "learning_rate": 3.646551724137931e-05,
+      "loss": 0.1182,
+      "step": 2355
+    },
+    {
+      "epoch": 2.73100050713613,
+      "grad_norm": 0.22852347791194916,
+      "learning_rate": 3.645977011494253e-05,
+      "loss": 0.1106,
+      "step": 2356
+    },
+    {
+      "epoch": 2.732159675432877,
+      "grad_norm": 0.24832133948802948,
+      "learning_rate": 3.6454022988505746e-05,
+      "loss": 0.1103,
+      "step": 2357
+    },
+    {
+      "epoch": 2.733318843729624,
+      "grad_norm": 0.2785024046897888,
+      "learning_rate": 3.644827586206897e-05,
+      "loss": 0.1094,
+      "step": 2358
+    },
+    {
+      "epoch": 2.734478012026371,
+      "grad_norm": 0.25696471333503723,
+      "learning_rate": 3.644252873563219e-05,
+      "loss": 0.1202,
+      "step": 2359
+    },
+    {
+      "epoch": 2.7356371803231183,
+      "grad_norm": 0.264342337846756,
+      "learning_rate": 3.6436781609195404e-05,
+      "loss": 0.1151,
+      "step": 2360
+    },
+    {
+      "epoch": 2.7367963486198654,
+      "grad_norm": 0.21554605662822723,
+      "learning_rate": 3.6431034482758626e-05,
+      "loss": 0.1109,
+      "step": 2361
+    },
+    {
+      "epoch": 2.737955516916612,
+      "grad_norm": 0.26814398169517517,
+      "learning_rate": 3.642528735632184e-05,
+      "loss": 0.1184,
+      "step": 2362
+    },
+    {
+      "epoch": 2.7391146852133597,
+      "grad_norm": 0.24080701172351837,
+      "learning_rate": 3.6419540229885056e-05,
+      "loss": 0.1151,
+      "step": 2363
+    },
+    {
+      "epoch": 2.7402738535101063,
+      "grad_norm": 0.2593724727630615,
+      "learning_rate": 3.641379310344828e-05,
+      "loss": 0.1201,
+      "step": 2364
+    },
+    {
+      "epoch": 2.7414330218068534,
+      "grad_norm": 0.21561947464942932,
+      "learning_rate": 3.64080459770115e-05,
+      "loss": 0.1213,
+      "step": 2365
+    },
+    {
+      "epoch": 2.7425921901036006,
+      "grad_norm": 0.3276143968105316,
+      "learning_rate": 3.6402298850574714e-05,
+      "loss": 0.1139,
+      "step": 2366
+    },
+    {
+      "epoch": 2.7437513584003477,
+      "grad_norm": 0.26314690709114075,
+      "learning_rate": 3.6396551724137936e-05,
+      "loss": 0.1104,
+      "step": 2367
+    },
+    {
+      "epoch": 2.744910526697095,
+      "grad_norm": 0.2540189027786255,
+      "learning_rate": 3.639080459770115e-05,
+      "loss": 0.1218,
+      "step": 2368
+    },
+    {
+      "epoch": 2.746069694993842,
+      "grad_norm": 0.23746971786022186,
+      "learning_rate": 3.638505747126437e-05,
+      "loss": 0.1109,
+      "step": 2369
+    },
+    {
+      "epoch": 2.747228863290589,
+      "grad_norm": 0.25298699736595154,
+      "learning_rate": 3.637931034482759e-05,
+      "loss": 0.1159,
+      "step": 2370
+    },
+    {
+      "epoch": 2.748388031587336,
+      "grad_norm": 0.25313448905944824,
+      "learning_rate": 3.63735632183908e-05,
+      "loss": 0.1115,
+      "step": 2371
+    },
+    {
+      "epoch": 2.7495471998840832,
+      "grad_norm": 0.2032177448272705,
+      "learning_rate": 3.6367816091954024e-05,
+      "loss": 0.1203,
+      "step": 2372
+    },
+    {
+      "epoch": 2.7507063681808304,
+      "grad_norm": 0.3549204468727112,
+      "learning_rate": 3.6362068965517245e-05,
+      "loss": 0.1234,
+      "step": 2373
+    },
+    {
+      "epoch": 2.7518655364775775,
+      "grad_norm": 0.2937655746936798,
+      "learning_rate": 3.635632183908046e-05,
+      "loss": 0.1222,
+      "step": 2374
+    },
+    {
+      "epoch": 2.753024704774324,
+      "grad_norm": 0.23844723403453827,
+      "learning_rate": 3.635057471264368e-05,
+      "loss": 0.1168,
+      "step": 2375
+    },
+    {
+      "epoch": 2.7541838730710717,
+      "grad_norm": 0.3607582747936249,
+      "learning_rate": 3.63448275862069e-05,
+      "loss": 0.1224,
+      "step": 2376
+    },
+    {
+      "epoch": 2.7553430413678184,
+      "grad_norm": 0.29361268877983093,
+      "learning_rate": 3.633908045977011e-05,
+      "loss": 0.1209,
+      "step": 2377
+    },
+    {
+      "epoch": 2.756502209664566,
+      "grad_norm": 0.2490907460451126,
+      "learning_rate": 3.633333333333333e-05,
+      "loss": 0.1157,
+      "step": 2378
+    },
+    {
+      "epoch": 2.7576613779613126,
+      "grad_norm": 0.24479351937770844,
+      "learning_rate": 3.6327586206896555e-05,
+      "loss": 0.1194,
+      "step": 2379
+    },
+    {
+      "epoch": 2.7588205462580597,
+      "grad_norm": 0.33146244287490845,
+      "learning_rate": 3.632183908045978e-05,
+      "loss": 0.1374,
+      "step": 2380
+    },
+    {
+      "epoch": 2.759979714554807,
+      "grad_norm": 0.2993434965610504,
+      "learning_rate": 3.631609195402299e-05,
+      "loss": 0.1125,
+      "step": 2381
+    },
+    {
+      "epoch": 2.761138882851554,
+      "grad_norm": 0.25163406133651733,
+      "learning_rate": 3.6310344827586206e-05,
+      "loss": 0.1168,
+      "step": 2382
+    },
+    {
+      "epoch": 2.762298051148301,
+      "grad_norm": 0.2083575427532196,
+      "learning_rate": 3.630459770114943e-05,
+      "loss": 0.1033,
+      "step": 2383
+    },
+    {
+      "epoch": 2.763457219445048,
+      "grad_norm": 0.3714704215526581,
+      "learning_rate": 3.629885057471264e-05,
+      "loss": 0.1262,
+      "step": 2384
+    },
+    {
+      "epoch": 2.7646163877417953,
+      "grad_norm": 0.3778088688850403,
+      "learning_rate": 3.6293103448275865e-05,
+      "loss": 0.1243,
+      "step": 2385
+    },
+    {
+      "epoch": 2.7657755560385424,
+      "grad_norm": 0.28580304980278015,
+      "learning_rate": 3.6287356321839086e-05,
+      "loss": 0.1188,
+      "step": 2386
+    },
+    {
+      "epoch": 2.7669347243352895,
+      "grad_norm": 0.3345694839954376,
+      "learning_rate": 3.62816091954023e-05,
+      "loss": 0.1078,
+      "step": 2387
+    },
+    {
+      "epoch": 2.7680938926320366,
+      "grad_norm": 0.2846716642379761,
+      "learning_rate": 3.627586206896552e-05,
+      "loss": 0.1142,
+      "step": 2388
+    },
+    {
+      "epoch": 2.7692530609287838,
+      "grad_norm": 0.30462977290153503,
+      "learning_rate": 3.627011494252874e-05,
+      "loss": 0.1327,
+      "step": 2389
+    },
+    {
+      "epoch": 2.7704122292255304,
+      "grad_norm": 0.2605188190937042,
+      "learning_rate": 3.626436781609195e-05,
+      "loss": 0.1099,
+      "step": 2390
+    },
+    {
+      "epoch": 2.771571397522278,
+      "grad_norm": 0.2525080740451813,
+      "learning_rate": 3.6258620689655174e-05,
+      "loss": 0.1149,
+      "step": 2391
+    },
+    {
+      "epoch": 2.7727305658190247,
+      "grad_norm": 0.2309272438287735,
+      "learning_rate": 3.625287356321839e-05,
+      "loss": 0.1232,
+      "step": 2392
+    },
+    {
+      "epoch": 2.7738897341157718,
+      "grad_norm": 0.25411146879196167,
+      "learning_rate": 3.624712643678161e-05,
+      "loss": 0.138,
+      "step": 2393
+    },
+    {
+      "epoch": 2.775048902412519,
+      "grad_norm": 0.2844621241092682,
+      "learning_rate": 3.624137931034483e-05,
+      "loss": 0.1282,
+      "step": 2394
+    },
+    {
+      "epoch": 2.776208070709266,
+      "grad_norm": 0.29295527935028076,
+      "learning_rate": 3.623563218390805e-05,
+      "loss": 0.1213,
+      "step": 2395
+    },
+    {
+      "epoch": 2.777367239006013,
+      "grad_norm": 0.3641514778137207,
+      "learning_rate": 3.622988505747126e-05,
+      "loss": 0.1313,
+      "step": 2396
+    },
+    {
+      "epoch": 2.7785264073027602,
+      "grad_norm": 0.24175724387168884,
+      "learning_rate": 3.6224137931034484e-05,
+      "loss": 0.1199,
+      "step": 2397
+    },
+    {
+      "epoch": 2.7796855755995074,
+      "grad_norm": 0.3517894744873047,
+      "learning_rate": 3.62183908045977e-05,
+      "loss": 0.1308,
+      "step": 2398
+    },
+    {
+      "epoch": 2.7808447438962545,
+      "grad_norm": 0.24343693256378174,
+      "learning_rate": 3.621264367816092e-05,
+      "loss": 0.1162,
+      "step": 2399
+    },
+    {
+      "epoch": 2.7820039121930016,
+      "grad_norm": 0.2250448763370514,
+      "learning_rate": 3.620689655172414e-05,
+      "loss": 0.112,
+      "step": 2400
+    },
+    {
+      "epoch": 2.7831630804897487,
+      "grad_norm": 0.23824208974838257,
+      "learning_rate": 3.620114942528736e-05,
+      "loss": 0.1222,
+      "step": 2401
+    },
+    {
+      "epoch": 2.784322248786496,
+      "grad_norm": 0.22008417546749115,
+      "learning_rate": 3.619540229885058e-05,
+      "loss": 0.1213,
+      "step": 2402
+    },
+    {
+      "epoch": 2.785481417083243,
+      "grad_norm": 0.3020162582397461,
+      "learning_rate": 3.6189655172413794e-05,
+      "loss": 0.1329,
+      "step": 2403
+    },
+    {
+      "epoch": 2.78664058537999,
+      "grad_norm": 0.21677182614803314,
+      "learning_rate": 3.618390804597701e-05,
+      "loss": 0.1186,
+      "step": 2404
+    },
+    {
+      "epoch": 2.7877997536767367,
+      "grad_norm": 0.25215545296669006,
+      "learning_rate": 3.617816091954023e-05,
+      "loss": 0.117,
+      "step": 2405
+    },
+    {
+      "epoch": 2.7889589219734843,
+      "grad_norm": 0.24186648428440094,
+      "learning_rate": 3.617241379310345e-05,
+      "loss": 0.1181,
+      "step": 2406
+    },
+    {
+      "epoch": 2.790118090270231,
+      "grad_norm": 0.2735014259815216,
+      "learning_rate": 3.6166666666666674e-05,
+      "loss": 0.1276,
+      "step": 2407
+    },
+    {
+      "epoch": 2.791277258566978,
+      "grad_norm": 0.20679566264152527,
+      "learning_rate": 3.616091954022989e-05,
+      "loss": 0.1146,
+      "step": 2408
+    },
+    {
+      "epoch": 2.792436426863725,
+      "grad_norm": 0.25650107860565186,
+      "learning_rate": 3.6155172413793104e-05,
+      "loss": 0.1223,
+      "step": 2409
+    },
+    {
+      "epoch": 2.7935955951604723,
+      "grad_norm": 0.22591929137706757,
+      "learning_rate": 3.6149425287356325e-05,
+      "loss": 0.1167,
+      "step": 2410
+    },
+    {
+      "epoch": 2.7947547634572194,
+      "grad_norm": 0.2470989227294922,
+      "learning_rate": 3.614367816091954e-05,
+      "loss": 0.1116,
+      "step": 2411
+    },
+    {
+      "epoch": 2.7959139317539665,
+      "grad_norm": 0.24131318926811218,
+      "learning_rate": 3.6137931034482755e-05,
+      "loss": 0.117,
+      "step": 2412
+    },
+    {
+      "epoch": 2.7970731000507136,
+      "grad_norm": 0.2714328467845917,
+      "learning_rate": 3.613218390804598e-05,
+      "loss": 0.1246,
+      "step": 2413
+    },
+    {
+      "epoch": 2.7982322683474607,
+      "grad_norm": 0.27446985244750977,
+      "learning_rate": 3.61264367816092e-05,
+      "loss": 0.1134,
+      "step": 2414
+    },
+    {
+      "epoch": 2.799391436644208,
+      "grad_norm": 0.23050987720489502,
+      "learning_rate": 3.6120689655172413e-05,
+      "loss": 0.1099,
+      "step": 2415
+    },
+    {
+      "epoch": 2.800550604940955,
+      "grad_norm": 0.27243009209632874,
+      "learning_rate": 3.6114942528735635e-05,
+      "loss": 0.1351,
+      "step": 2416
+    },
+    {
+      "epoch": 2.801709773237702,
+      "grad_norm": 0.2836304306983948,
+      "learning_rate": 3.610919540229885e-05,
+      "loss": 0.1248,
+      "step": 2417
+    },
+    {
+      "epoch": 2.8028689415344488,
+      "grad_norm": 0.27356666326522827,
+      "learning_rate": 3.610344827586207e-05,
+      "loss": 0.1193,
+      "step": 2418
+    },
+    {
+      "epoch": 2.8040281098311963,
+      "grad_norm": 0.2828332185745239,
+      "learning_rate": 3.6097701149425287e-05,
+      "loss": 0.1253,
+      "step": 2419
+    },
+    {
+      "epoch": 2.805187278127943,
+      "grad_norm": 0.2918018102645874,
+      "learning_rate": 3.609195402298851e-05,
+      "loss": 0.1309,
+      "step": 2420
+    },
+    {
+      "epoch": 2.8063464464246906,
+      "grad_norm": 0.22664931416511536,
+      "learning_rate": 3.608620689655173e-05,
+      "loss": 0.1153,
+      "step": 2421
+    },
+    {
+      "epoch": 2.8075056147214372,
+      "grad_norm": 0.2359132617712021,
+      "learning_rate": 3.6080459770114945e-05,
+      "loss": 0.1218,
+      "step": 2422
+    },
+    {
+      "epoch": 2.8086647830181843,
+      "grad_norm": 0.26460394263267517,
+      "learning_rate": 3.607471264367816e-05,
+      "loss": 0.1309,
+      "step": 2423
+    },
+    {
+      "epoch": 2.8098239513149315,
+      "grad_norm": 0.25006482005119324,
+      "learning_rate": 3.606896551724138e-05,
+      "loss": 0.1078,
+      "step": 2424
+    },
+    {
+      "epoch": 2.8109831196116786,
+      "grad_norm": 0.25657373666763306,
+      "learning_rate": 3.6063218390804596e-05,
+      "loss": 0.1084,
+      "step": 2425
+    },
+    {
+      "epoch": 2.8121422879084257,
+      "grad_norm": 0.3210379183292389,
+      "learning_rate": 3.605747126436782e-05,
+      "loss": 0.1272,
+      "step": 2426
+    },
+    {
+      "epoch": 2.813301456205173,
+      "grad_norm": 0.23152370750904083,
+      "learning_rate": 3.605172413793104e-05,
+      "loss": 0.1124,
+      "step": 2427
+    },
+    {
+      "epoch": 2.81446062450192,
+      "grad_norm": 0.24926093220710754,
+      "learning_rate": 3.6045977011494255e-05,
+      "loss": 0.1181,
+      "step": 2428
+    },
+    {
+      "epoch": 2.815619792798667,
+      "grad_norm": 0.25474682450294495,
+      "learning_rate": 3.6040229885057476e-05,
+      "loss": 0.1287,
+      "step": 2429
+    },
+    {
+      "epoch": 2.816778961095414,
+      "grad_norm": 0.27318352460861206,
+      "learning_rate": 3.603448275862069e-05,
+      "loss": 0.1203,
+      "step": 2430
+    },
+    {
+      "epoch": 2.8179381293921613,
+      "grad_norm": 0.2800019383430481,
+      "learning_rate": 3.6028735632183906e-05,
+      "loss": 0.1196,
+      "step": 2431
+    },
+    {
+      "epoch": 2.8190972976889084,
+      "grad_norm": 0.20935480296611786,
+      "learning_rate": 3.602298850574713e-05,
+      "loss": 0.1105,
+      "step": 2432
+    },
+    {
+      "epoch": 2.820256465985655,
+      "grad_norm": 0.24666768312454224,
+      "learning_rate": 3.601724137931034e-05,
+      "loss": 0.1182,
+      "step": 2433
+    },
+    {
+      "epoch": 2.8214156342824026,
+      "grad_norm": 0.2466108649969101,
+      "learning_rate": 3.6011494252873564e-05,
+      "loss": 0.1183,
+      "step": 2434
+    },
+    {
+      "epoch": 2.8225748025791493,
+      "grad_norm": 0.2528558671474457,
+      "learning_rate": 3.6005747126436786e-05,
+      "loss": 0.1188,
+      "step": 2435
+    },
+    {
+      "epoch": 2.8237339708758964,
+      "grad_norm": 0.29567837715148926,
+      "learning_rate": 3.6e-05,
+      "loss": 0.1443,
+      "step": 2436
+    },
+    {
+      "epoch": 2.8248931391726435,
+      "grad_norm": 0.31635522842407227,
+      "learning_rate": 3.599425287356322e-05,
+      "loss": 0.107,
+      "step": 2437
+    },
+    {
+      "epoch": 2.8260523074693906,
+      "grad_norm": 0.23318928480148315,
+      "learning_rate": 3.598850574712644e-05,
+      "loss": 0.1231,
+      "step": 2438
+    },
+    {
+      "epoch": 2.8272114757661377,
+      "grad_norm": 0.29607391357421875,
+      "learning_rate": 3.598275862068965e-05,
+      "loss": 0.1257,
+      "step": 2439
+    },
+    {
+      "epoch": 2.828370644062885,
+      "grad_norm": 0.2570621371269226,
+      "learning_rate": 3.5977011494252874e-05,
+      "loss": 0.1128,
+      "step": 2440
+    },
+    {
+      "epoch": 2.829529812359632,
+      "grad_norm": 0.24534910917282104,
+      "learning_rate": 3.5971264367816096e-05,
+      "loss": 0.111,
+      "step": 2441
+    },
+    {
+      "epoch": 2.830688980656379,
+      "grad_norm": 0.26997894048690796,
+      "learning_rate": 3.596551724137931e-05,
+      "loss": 0.1114,
+      "step": 2442
+    },
+    {
+      "epoch": 2.831848148953126,
+      "grad_norm": 0.29635146260261536,
+      "learning_rate": 3.595977011494253e-05,
+      "loss": 0.1228,
+      "step": 2443
+    },
+    {
+      "epoch": 2.8330073172498733,
+      "grad_norm": 0.2879303991794586,
+      "learning_rate": 3.595402298850575e-05,
+      "loss": 0.1248,
+      "step": 2444
+    },
+    {
+      "epoch": 2.8341664855466204,
+      "grad_norm": 0.2170044481754303,
+      "learning_rate": 3.594827586206897e-05,
+      "loss": 0.11,
+      "step": 2445
+    },
+    {
+      "epoch": 2.8353256538433675,
+      "grad_norm": 0.2590021789073944,
+      "learning_rate": 3.5942528735632184e-05,
+      "loss": 0.1257,
+      "step": 2446
+    },
+    {
+      "epoch": 2.8364848221401147,
+      "grad_norm": 0.25677573680877686,
+      "learning_rate": 3.5936781609195405e-05,
+      "loss": 0.1266,
+      "step": 2447
+    },
+    {
+      "epoch": 2.8376439904368613,
+      "grad_norm": 0.2618117928504944,
+      "learning_rate": 3.593103448275863e-05,
+      "loss": 0.1143,
+      "step": 2448
+    },
+    {
+      "epoch": 2.838803158733609,
+      "grad_norm": 0.23307304084300995,
+      "learning_rate": 3.592528735632184e-05,
+      "loss": 0.1126,
+      "step": 2449
+    },
+    {
+      "epoch": 2.8399623270303556,
+      "grad_norm": 0.2070130705833435,
+      "learning_rate": 3.591954022988506e-05,
+      "loss": 0.1141,
+      "step": 2450
+    },
+    {
+      "epoch": 2.8411214953271027,
+      "grad_norm": 0.25860536098480225,
+      "learning_rate": 3.591379310344828e-05,
+      "loss": 0.1102,
+      "step": 2451
+    },
+    {
+      "epoch": 2.84228066362385,
+      "grad_norm": 0.22545821964740753,
+      "learning_rate": 3.5908045977011494e-05,
+      "loss": 0.1037,
+      "step": 2452
+    },
+    {
+      "epoch": 2.843439831920597,
+      "grad_norm": 0.27270445227622986,
+      "learning_rate": 3.5902298850574715e-05,
+      "loss": 0.1171,
+      "step": 2453
+    },
+    {
+      "epoch": 2.844599000217344,
+      "grad_norm": 0.23583996295928955,
+      "learning_rate": 3.589655172413793e-05,
+      "loss": 0.099,
+      "step": 2454
+    },
+    {
+      "epoch": 2.845758168514091,
+      "grad_norm": 0.27526620030403137,
+      "learning_rate": 3.589080459770115e-05,
+      "loss": 0.1124,
+      "step": 2455
+    },
+    {
+      "epoch": 2.8469173368108383,
+      "grad_norm": 0.23859523236751556,
+      "learning_rate": 3.5885057471264373e-05,
+      "loss": 0.1142,
+      "step": 2456
+    },
+    {
+      "epoch": 2.8480765051075854,
+      "grad_norm": 0.3173846900463104,
+      "learning_rate": 3.587931034482759e-05,
+      "loss": 0.1249,
+      "step": 2457
+    },
+    {
+      "epoch": 2.8492356734043325,
+      "grad_norm": 0.2413249909877777,
+      "learning_rate": 3.58735632183908e-05,
+      "loss": 0.1165,
+      "step": 2458
+    },
+    {
+      "epoch": 2.8503948417010796,
+      "grad_norm": 0.2488858848810196,
+      "learning_rate": 3.5867816091954025e-05,
+      "loss": 0.1176,
+      "step": 2459
+    },
+    {
+      "epoch": 2.8515540099978267,
+      "grad_norm": 0.272587388753891,
+      "learning_rate": 3.586206896551724e-05,
+      "loss": 0.1212,
+      "step": 2460
+    },
+    {
+      "epoch": 2.8527131782945734,
+      "grad_norm": 0.29471564292907715,
+      "learning_rate": 3.585632183908046e-05,
+      "loss": 0.1298,
+      "step": 2461
+    },
+    {
+      "epoch": 2.853872346591321,
+      "grad_norm": 0.25901687145233154,
+      "learning_rate": 3.585057471264368e-05,
+      "loss": 0.1231,
+      "step": 2462
+    },
+    {
+      "epoch": 2.8550315148880676,
+      "grad_norm": 0.2517445981502533,
+      "learning_rate": 3.58448275862069e-05,
+      "loss": 0.1169,
+      "step": 2463
+    },
+    {
+      "epoch": 2.856190683184815,
+      "grad_norm": 0.23075416684150696,
+      "learning_rate": 3.583908045977012e-05,
+      "loss": 0.1025,
+      "step": 2464
+    },
+    {
+      "epoch": 2.857349851481562,
+      "grad_norm": 0.2388884723186493,
+      "learning_rate": 3.5833333333333335e-05,
+      "loss": 0.1107,
+      "step": 2465
+    },
+    {
+      "epoch": 2.858509019778309,
+      "grad_norm": 0.3148244619369507,
+      "learning_rate": 3.582758620689655e-05,
+      "loss": 0.1179,
+      "step": 2466
+    },
+    {
+      "epoch": 2.859668188075056,
+      "grad_norm": 0.28545230627059937,
+      "learning_rate": 3.582183908045977e-05,
+      "loss": 0.1237,
+      "step": 2467
+    },
+    {
+      "epoch": 2.860827356371803,
+      "grad_norm": 0.2898615002632141,
+      "learning_rate": 3.581609195402299e-05,
+      "loss": 0.1331,
+      "step": 2468
+    },
+    {
+      "epoch": 2.8619865246685503,
+      "grad_norm": 0.2582353353500366,
+      "learning_rate": 3.581034482758621e-05,
+      "loss": 0.1204,
+      "step": 2469
+    },
+    {
+      "epoch": 2.8631456929652974,
+      "grad_norm": 0.2813541293144226,
+      "learning_rate": 3.580459770114943e-05,
+      "loss": 0.1232,
+      "step": 2470
+    },
+    {
+      "epoch": 2.8643048612620445,
+      "grad_norm": 0.2520640194416046,
+      "learning_rate": 3.5798850574712644e-05,
+      "loss": 0.1137,
+      "step": 2471
+    },
+    {
+      "epoch": 2.8654640295587916,
+      "grad_norm": 0.29542282223701477,
+      "learning_rate": 3.5793103448275866e-05,
+      "loss": 0.121,
+      "step": 2472
+    },
+    {
+      "epoch": 2.8666231978555388,
+      "grad_norm": 0.21729327738285065,
+      "learning_rate": 3.578735632183908e-05,
+      "loss": 0.1108,
+      "step": 2473
+    },
+    {
+      "epoch": 2.867782366152286,
+      "grad_norm": 0.285705029964447,
+      "learning_rate": 3.5781609195402296e-05,
+      "loss": 0.1234,
+      "step": 2474
+    },
+    {
+      "epoch": 2.868941534449033,
+      "grad_norm": 0.2638833224773407,
+      "learning_rate": 3.5775862068965524e-05,
+      "loss": 0.1361,
+      "step": 2475
+    },
+    {
+      "epoch": 2.8701007027457797,
+      "grad_norm": 0.25807875394821167,
+      "learning_rate": 3.577011494252874e-05,
+      "loss": 0.1067,
+      "step": 2476
+    },
+    {
+      "epoch": 2.8712598710425272,
+      "grad_norm": 0.20497660338878632,
+      "learning_rate": 3.5764367816091954e-05,
+      "loss": 0.1072,
+      "step": 2477
+    },
+    {
+      "epoch": 2.872419039339274,
+      "grad_norm": 0.23752939701080322,
+      "learning_rate": 3.5758620689655176e-05,
+      "loss": 0.1231,
+      "step": 2478
+    },
+    {
+      "epoch": 2.873578207636021,
+      "grad_norm": 0.2198176383972168,
+      "learning_rate": 3.575287356321839e-05,
+      "loss": 0.1212,
+      "step": 2479
+    },
+    {
+      "epoch": 2.874737375932768,
+      "grad_norm": 0.28689807653427124,
+      "learning_rate": 3.5747126436781606e-05,
+      "loss": 0.1292,
+      "step": 2480
+    },
+    {
+      "epoch": 2.8758965442295152,
+      "grad_norm": 0.21287685632705688,
+      "learning_rate": 3.574137931034483e-05,
+      "loss": 0.106,
+      "step": 2481
+    },
+    {
+      "epoch": 2.8770557125262624,
+      "grad_norm": 0.3098215162754059,
+      "learning_rate": 3.573563218390805e-05,
+      "loss": 0.1212,
+      "step": 2482
+    },
+    {
+      "epoch": 2.8782148808230095,
+      "grad_norm": 0.1953476518392563,
+      "learning_rate": 3.572988505747127e-05,
+      "loss": 0.1001,
+      "step": 2483
+    },
+    {
+      "epoch": 2.8793740491197566,
+      "grad_norm": 0.22873975336551666,
+      "learning_rate": 3.5724137931034486e-05,
+      "loss": 0.1115,
+      "step": 2484
+    },
+    {
+      "epoch": 2.8805332174165037,
+      "grad_norm": 0.21951954066753387,
+      "learning_rate": 3.57183908045977e-05,
+      "loss": 0.1197,
+      "step": 2485
+    },
+    {
+      "epoch": 2.881692385713251,
+      "grad_norm": 0.25903722643852234,
+      "learning_rate": 3.571264367816092e-05,
+      "loss": 0.1082,
+      "step": 2486
+    },
+    {
+      "epoch": 2.882851554009998,
+      "grad_norm": 0.2247413992881775,
+      "learning_rate": 3.570689655172414e-05,
+      "loss": 0.1138,
+      "step": 2487
+    },
+    {
+      "epoch": 2.884010722306745,
+      "grad_norm": 0.2930486798286438,
+      "learning_rate": 3.570114942528736e-05,
+      "loss": 0.1327,
+      "step": 2488
+    },
+    {
+      "epoch": 2.885169890603492,
+      "grad_norm": 0.32218900322914124,
+      "learning_rate": 3.569540229885058e-05,
+      "loss": 0.1261,
+      "step": 2489
+    },
+    {
+      "epoch": 2.8863290589002393,
+      "grad_norm": 0.2783204913139343,
+      "learning_rate": 3.5689655172413795e-05,
+      "loss": 0.1132,
+      "step": 2490
+    },
+    {
+      "epoch": 2.887488227196986,
+      "grad_norm": 0.214439257979393,
+      "learning_rate": 3.568390804597702e-05,
+      "loss": 0.1179,
+      "step": 2491
+    },
+    {
+      "epoch": 2.8886473954937335,
+      "grad_norm": 0.25692418217658997,
+      "learning_rate": 3.567816091954023e-05,
+      "loss": 0.1147,
+      "step": 2492
+    },
+    {
+      "epoch": 2.88980656379048,
+      "grad_norm": 0.27473074197769165,
+      "learning_rate": 3.567241379310345e-05,
+      "loss": 0.1351,
+      "step": 2493
+    },
+    {
+      "epoch": 2.8909657320872273,
+      "grad_norm": 0.27782875299453735,
+      "learning_rate": 3.566666666666667e-05,
+      "loss": 0.1246,
+      "step": 2494
+    },
+    {
+      "epoch": 2.8921249003839744,
+      "grad_norm": 0.22847121953964233,
+      "learning_rate": 3.5660919540229883e-05,
+      "loss": 0.1118,
+      "step": 2495
+    },
+    {
+      "epoch": 2.8932840686807215,
+      "grad_norm": 0.22717076539993286,
+      "learning_rate": 3.5655172413793105e-05,
+      "loss": 0.1233,
+      "step": 2496
+    },
+    {
+      "epoch": 2.8944432369774686,
+      "grad_norm": 0.2825745642185211,
+      "learning_rate": 3.564942528735633e-05,
+      "loss": 0.1282,
+      "step": 2497
+    },
+    {
+      "epoch": 2.8956024052742158,
+      "grad_norm": 0.25072160363197327,
+      "learning_rate": 3.564367816091954e-05,
+      "loss": 0.1239,
+      "step": 2498
+    },
+    {
+      "epoch": 2.896761573570963,
+      "grad_norm": 0.2573794424533844,
+      "learning_rate": 3.5637931034482757e-05,
+      "loss": 0.1083,
+      "step": 2499
+    },
+    {
+      "epoch": 2.89792074186771,
+      "grad_norm": 0.29869526624679565,
+      "learning_rate": 3.563218390804598e-05,
+      "loss": 0.132,
+      "step": 2500
+    },
+    {
+      "epoch": 2.899079910164457,
+      "grad_norm": 0.24335411190986633,
+      "learning_rate": 3.562643678160919e-05,
+      "loss": 0.1187,
+      "step": 2501
+    },
+    {
+      "epoch": 2.900239078461204,
+      "grad_norm": 0.25249624252319336,
+      "learning_rate": 3.5620689655172415e-05,
+      "loss": 0.1157,
+      "step": 2502
+    },
+    {
+      "epoch": 2.9013982467579513,
+      "grad_norm": 0.2802266776561737,
+      "learning_rate": 3.5614942528735637e-05,
+      "loss": 0.1267,
+      "step": 2503
+    },
+    {
+      "epoch": 2.902557415054698,
+      "grad_norm": 0.3055930435657501,
+      "learning_rate": 3.560919540229885e-05,
+      "loss": 0.1287,
+      "step": 2504
+    },
+    {
+      "epoch": 2.9037165833514456,
+      "grad_norm": 0.2905328571796417,
+      "learning_rate": 3.560344827586207e-05,
+      "loss": 0.1317,
+      "step": 2505
+    },
+    {
+      "epoch": 2.9048757516481922,
+      "grad_norm": 0.21209411323070526,
+      "learning_rate": 3.559770114942529e-05,
+      "loss": 0.1131,
+      "step": 2506
+    },
+    {
+      "epoch": 2.9060349199449393,
+      "grad_norm": 0.2763073444366455,
+      "learning_rate": 3.55919540229885e-05,
+      "loss": 0.1205,
+      "step": 2507
+    },
+    {
+      "epoch": 2.9071940882416865,
+      "grad_norm": 0.23870214819908142,
+      "learning_rate": 3.5586206896551725e-05,
+      "loss": 0.1237,
+      "step": 2508
+    },
+    {
+      "epoch": 2.9083532565384336,
+      "grad_norm": 0.2398124486207962,
+      "learning_rate": 3.5580459770114946e-05,
+      "loss": 0.1161,
+      "step": 2509
+    },
+    {
+      "epoch": 2.9095124248351807,
+      "grad_norm": 0.23274962604045868,
+      "learning_rate": 3.557471264367817e-05,
+      "loss": 0.1094,
+      "step": 2510
+    },
+    {
+      "epoch": 2.910671593131928,
+      "grad_norm": 0.2622717022895813,
+      "learning_rate": 3.556896551724138e-05,
+      "loss": 0.1189,
+      "step": 2511
+    },
+    {
+      "epoch": 2.911830761428675,
+      "grad_norm": 0.335947185754776,
+      "learning_rate": 3.55632183908046e-05,
+      "loss": 0.1237,
+      "step": 2512
+    },
+    {
+      "epoch": 2.912989929725422,
+      "grad_norm": 0.2339511215686798,
+      "learning_rate": 3.555747126436782e-05,
+      "loss": 0.1195,
+      "step": 2513
+    },
+    {
+      "epoch": 2.914149098022169,
+      "grad_norm": 0.2983540892601013,
+      "learning_rate": 3.5551724137931034e-05,
+      "loss": 0.1184,
+      "step": 2514
+    },
+    {
+      "epoch": 2.9153082663189163,
+      "grad_norm": 0.29137369990348816,
+      "learning_rate": 3.554597701149425e-05,
+      "loss": 0.1074,
+      "step": 2515
+    },
+    {
+      "epoch": 2.9164674346156634,
+      "grad_norm": 0.3541145622730255,
+      "learning_rate": 3.554022988505748e-05,
+      "loss": 0.1187,
+      "step": 2516
+    },
+    {
+      "epoch": 2.9176266029124105,
+      "grad_norm": 0.2662143111228943,
+      "learning_rate": 3.553448275862069e-05,
+      "loss": 0.1227,
+      "step": 2517
+    },
+    {
+      "epoch": 2.9187857712091576,
+      "grad_norm": 0.2711283266544342,
+      "learning_rate": 3.552873563218391e-05,
+      "loss": 0.1304,
+      "step": 2518
+    },
+    {
+      "epoch": 2.9199449395059043,
+      "grad_norm": 0.2492939978837967,
+      "learning_rate": 3.552298850574713e-05,
+      "loss": 0.1191,
+      "step": 2519
+    },
+    {
+      "epoch": 2.921104107802652,
+      "grad_norm": 0.25983232259750366,
+      "learning_rate": 3.5517241379310344e-05,
+      "loss": 0.1129,
+      "step": 2520
+    },
+    {
+      "epoch": 2.9222632760993985,
+      "grad_norm": 0.24803534150123596,
+      "learning_rate": 3.5511494252873566e-05,
+      "loss": 0.116,
+      "step": 2521
+    },
+    {
+      "epoch": 2.9234224443961456,
+      "grad_norm": 0.2651211619377136,
+      "learning_rate": 3.550574712643678e-05,
+      "loss": 0.1361,
+      "step": 2522
+    },
+    {
+      "epoch": 2.9245816126928927,
+      "grad_norm": 0.2923348844051361,
+      "learning_rate": 3.55e-05,
+      "loss": 0.135,
+      "step": 2523
+    },
+    {
+      "epoch": 2.92574078098964,
+      "grad_norm": 0.26953768730163574,
+      "learning_rate": 3.5494252873563224e-05,
+      "loss": 0.1183,
+      "step": 2524
+    },
+    {
+      "epoch": 2.926899949286387,
+      "grad_norm": 0.235351100564003,
+      "learning_rate": 3.548850574712644e-05,
+      "loss": 0.1216,
+      "step": 2525
+    },
+    {
+      "epoch": 2.928059117583134,
+      "grad_norm": 0.2557763457298279,
+      "learning_rate": 3.5482758620689654e-05,
+      "loss": 0.1138,
+      "step": 2526
+    },
+    {
+      "epoch": 2.929218285879881,
+      "grad_norm": 0.2516639232635498,
+      "learning_rate": 3.5477011494252875e-05,
+      "loss": 0.144,
+      "step": 2527
+    },
+    {
+      "epoch": 2.9303774541766283,
+      "grad_norm": 0.19669529795646667,
+      "learning_rate": 3.547126436781609e-05,
+      "loss": 0.1112,
+      "step": 2528
+    },
+    {
+      "epoch": 2.9315366224733754,
+      "grad_norm": 0.21730567514896393,
+      "learning_rate": 3.546551724137931e-05,
+      "loss": 0.1113,
+      "step": 2529
+    },
+    {
+      "epoch": 2.9326957907701225,
+      "grad_norm": 0.20566688477993011,
+      "learning_rate": 3.5459770114942534e-05,
+      "loss": 0.1146,
+      "step": 2530
+    },
+    {
+      "epoch": 2.9338549590668697,
+      "grad_norm": 0.26135411858558655,
+      "learning_rate": 3.545402298850575e-05,
+      "loss": 0.1083,
+      "step": 2531
+    },
+    {
+      "epoch": 2.9350141273636163,
+      "grad_norm": 0.23038803040981293,
+      "learning_rate": 3.544827586206897e-05,
+      "loss": 0.1212,
+      "step": 2532
+    },
+    {
+      "epoch": 2.936173295660364,
+      "grad_norm": 0.2405983805656433,
+      "learning_rate": 3.5442528735632185e-05,
+      "loss": 0.1085,
+      "step": 2533
+    },
+    {
+      "epoch": 2.9373324639571106,
+      "grad_norm": 0.2417573630809784,
+      "learning_rate": 3.54367816091954e-05,
+      "loss": 0.1173,
+      "step": 2534
+    },
+    {
+      "epoch": 2.938491632253858,
+      "grad_norm": 0.21391847729682922,
+      "learning_rate": 3.543103448275862e-05,
+      "loss": 0.1161,
+      "step": 2535
+    },
+    {
+      "epoch": 2.939650800550605,
+      "grad_norm": 0.2825719714164734,
+      "learning_rate": 3.5425287356321843e-05,
+      "loss": 0.1332,
+      "step": 2536
+    },
+    {
+      "epoch": 2.940809968847352,
+      "grad_norm": 0.23542939126491547,
+      "learning_rate": 3.541954022988506e-05,
+      "loss": 0.1184,
+      "step": 2537
+    },
+    {
+      "epoch": 2.941969137144099,
+      "grad_norm": 0.2510528862476349,
+      "learning_rate": 3.541379310344828e-05,
+      "loss": 0.1238,
+      "step": 2538
+    },
+    {
+      "epoch": 2.943128305440846,
+      "grad_norm": 0.2628154456615448,
+      "learning_rate": 3.5408045977011495e-05,
+      "loss": 0.1224,
+      "step": 2539
+    },
+    {
+      "epoch": 2.9442874737375933,
+      "grad_norm": 0.18918563425540924,
+      "learning_rate": 3.540229885057472e-05,
+      "loss": 0.1026,
+      "step": 2540
+    },
+    {
+      "epoch": 2.9454466420343404,
+      "grad_norm": 0.299092561006546,
+      "learning_rate": 3.539655172413793e-05,
+      "loss": 0.1332,
+      "step": 2541
+    },
+    {
+      "epoch": 2.9466058103310875,
+      "grad_norm": 0.22174027562141418,
+      "learning_rate": 3.5390804597701146e-05,
+      "loss": 0.1071,
+      "step": 2542
+    },
+    {
+      "epoch": 2.9477649786278346,
+      "grad_norm": 0.30895039439201355,
+      "learning_rate": 3.538505747126437e-05,
+      "loss": 0.1243,
+      "step": 2543
+    },
+    {
+      "epoch": 2.9489241469245817,
+      "grad_norm": 0.30053114891052246,
+      "learning_rate": 3.537931034482759e-05,
+      "loss": 0.1197,
+      "step": 2544
+    },
+    {
+      "epoch": 2.950083315221329,
+      "grad_norm": 0.276027113199234,
+      "learning_rate": 3.5373563218390805e-05,
+      "loss": 0.1166,
+      "step": 2545
+    },
+    {
+      "epoch": 2.951242483518076,
+      "grad_norm": 0.2735234797000885,
+      "learning_rate": 3.5367816091954026e-05,
+      "loss": 0.1244,
+      "step": 2546
+    },
+    {
+      "epoch": 2.9524016518148226,
+      "grad_norm": 0.3124696612358093,
+      "learning_rate": 3.536206896551724e-05,
+      "loss": 0.1115,
+      "step": 2547
+    },
+    {
+      "epoch": 2.95356082011157,
+      "grad_norm": 0.21694187819957733,
+      "learning_rate": 3.535632183908046e-05,
+      "loss": 0.1146,
+      "step": 2548
+    },
+    {
+      "epoch": 2.954719988408317,
+      "grad_norm": 0.28036484122276306,
+      "learning_rate": 3.535057471264368e-05,
+      "loss": 0.1027,
+      "step": 2549
+    },
+    {
+      "epoch": 2.955879156705064,
+      "grad_norm": 0.31232985854148865,
+      "learning_rate": 3.53448275862069e-05,
+      "loss": 0.1165,
+      "step": 2550
+    },
+    {
+      "epoch": 2.957038325001811,
+      "grad_norm": 0.28552737832069397,
+      "learning_rate": 3.533908045977012e-05,
+      "loss": 0.1225,
+      "step": 2551
+    },
+    {
+      "epoch": 2.958197493298558,
+      "grad_norm": 0.27900782227516174,
+      "learning_rate": 3.5333333333333336e-05,
+      "loss": 0.1191,
+      "step": 2552
+    },
+    {
+      "epoch": 2.9593566615953053,
+      "grad_norm": 0.25439637899398804,
+      "learning_rate": 3.532758620689655e-05,
+      "loss": 0.1324,
+      "step": 2553
+    },
+    {
+      "epoch": 2.9605158298920524,
+      "grad_norm": 0.24893079698085785,
+      "learning_rate": 3.532183908045977e-05,
+      "loss": 0.1161,
+      "step": 2554
+    },
+    {
+      "epoch": 2.9616749981887995,
+      "grad_norm": 0.2921384274959564,
+      "learning_rate": 3.531609195402299e-05,
+      "loss": 0.1209,
+      "step": 2555
+    },
+    {
+      "epoch": 2.9628341664855466,
+      "grad_norm": 0.21801045536994934,
+      "learning_rate": 3.53103448275862e-05,
+      "loss": 0.118,
+      "step": 2556
+    },
+    {
+      "epoch": 2.9639933347822938,
+      "grad_norm": 0.26922228932380676,
+      "learning_rate": 3.530459770114943e-05,
+      "loss": 0.1315,
+      "step": 2557
+    },
+    {
+      "epoch": 2.965152503079041,
+      "grad_norm": 0.2805965840816498,
+      "learning_rate": 3.5298850574712646e-05,
+      "loss": 0.1291,
+      "step": 2558
+    },
+    {
+      "epoch": 2.966311671375788,
+      "grad_norm": 0.3655463457107544,
+      "learning_rate": 3.529310344827587e-05,
+      "loss": 0.1303,
+      "step": 2559
+    },
+    {
+      "epoch": 2.967470839672535,
+      "grad_norm": 0.2937909960746765,
+      "learning_rate": 3.528735632183908e-05,
+      "loss": 0.1259,
+      "step": 2560
+    },
+    {
+      "epoch": 2.9686300079692822,
+      "grad_norm": 0.3068729043006897,
+      "learning_rate": 3.52816091954023e-05,
+      "loss": 0.1348,
+      "step": 2561
+    },
+    {
+      "epoch": 2.969789176266029,
+      "grad_norm": 0.2570815980434418,
+      "learning_rate": 3.527586206896552e-05,
+      "loss": 0.1261,
+      "step": 2562
+    },
+    {
+      "epoch": 2.9709483445627765,
+      "grad_norm": 0.22781704366207123,
+      "learning_rate": 3.5270114942528734e-05,
+      "loss": 0.1177,
+      "step": 2563
+    },
+    {
+      "epoch": 2.972107512859523,
+      "grad_norm": 0.2892128527164459,
+      "learning_rate": 3.5264367816091956e-05,
+      "loss": 0.1181,
+      "step": 2564
+    },
+    {
+      "epoch": 2.9732666811562702,
+      "grad_norm": 0.24574293196201324,
+      "learning_rate": 3.525862068965518e-05,
+      "loss": 0.1309,
+      "step": 2565
+    },
+    {
+      "epoch": 2.9744258494530174,
+      "grad_norm": 0.2564846873283386,
+      "learning_rate": 3.525287356321839e-05,
+      "loss": 0.1201,
+      "step": 2566
+    },
+    {
+      "epoch": 2.9755850177497645,
+      "grad_norm": 0.3438556492328644,
+      "learning_rate": 3.5247126436781614e-05,
+      "loss": 0.1292,
+      "step": 2567
+    },
+    {
+      "epoch": 2.9767441860465116,
+      "grad_norm": 0.25281572341918945,
+      "learning_rate": 3.524137931034483e-05,
+      "loss": 0.1202,
+      "step": 2568
+    },
+    {
+      "epoch": 2.9779033543432587,
+      "grad_norm": 0.20820648968219757,
+      "learning_rate": 3.5235632183908044e-05,
+      "loss": 0.1221,
+      "step": 2569
+    },
+    {
+      "epoch": 2.979062522640006,
+      "grad_norm": 0.23150451481342316,
+      "learning_rate": 3.5229885057471265e-05,
+      "loss": 0.1269,
+      "step": 2570
+    },
+    {
+      "epoch": 2.980221690936753,
+      "grad_norm": 0.25065869092941284,
+      "learning_rate": 3.522413793103449e-05,
+      "loss": 0.1268,
+      "step": 2571
+    },
+    {
+      "epoch": 2.9813808592335,
+      "grad_norm": 0.25164109468460083,
+      "learning_rate": 3.52183908045977e-05,
+      "loss": 0.1156,
+      "step": 2572
+    },
+    {
+      "epoch": 2.982540027530247,
+      "grad_norm": 0.2479361742734909,
+      "learning_rate": 3.5212643678160924e-05,
+      "loss": 0.1204,
+      "step": 2573
+    },
+    {
+      "epoch": 2.9836991958269943,
+      "grad_norm": 0.21308045089244843,
+      "learning_rate": 3.520689655172414e-05,
+      "loss": 0.1161,
+      "step": 2574
+    },
+    {
+      "epoch": 2.984858364123741,
+      "grad_norm": 0.2212274670600891,
+      "learning_rate": 3.5201149425287353e-05,
+      "loss": 0.1138,
+      "step": 2575
+    },
+    {
+      "epoch": 2.9860175324204885,
+      "grad_norm": 0.26298844814300537,
+      "learning_rate": 3.5195402298850575e-05,
+      "loss": 0.1218,
+      "step": 2576
+    },
+    {
+      "epoch": 2.987176700717235,
+      "grad_norm": 0.20742398500442505,
+      "learning_rate": 3.51896551724138e-05,
+      "loss": 0.1156,
+      "step": 2577
+    },
+    {
+      "epoch": 2.9883358690139827,
+      "grad_norm": 0.23608221113681793,
+      "learning_rate": 3.518390804597702e-05,
+      "loss": 0.1058,
+      "step": 2578
+    },
+    {
+      "epoch": 2.9894950373107294,
+      "grad_norm": 0.25990620255470276,
+      "learning_rate": 3.517816091954023e-05,
+      "loss": 0.1249,
+      "step": 2579
+    },
+    {
+      "epoch": 2.9906542056074765,
+      "grad_norm": 0.2265261709690094,
+      "learning_rate": 3.517241379310345e-05,
+      "loss": 0.1177,
+      "step": 2580
+    },
+    {
+      "epoch": 2.9918133739042236,
+      "grad_norm": 0.21493545174598694,
+      "learning_rate": 3.516666666666667e-05,
+      "loss": 0.1042,
+      "step": 2581
+    },
+    {
+      "epoch": 2.9929725422009708,
+      "grad_norm": 0.2955082058906555,
+      "learning_rate": 3.5160919540229885e-05,
+      "loss": 0.1213,
+      "step": 2582
+    },
+    {
+      "epoch": 2.994131710497718,
+      "grad_norm": 0.25260546803474426,
+      "learning_rate": 3.51551724137931e-05,
+      "loss": 0.1231,
+      "step": 2583
+    },
+    {
+      "epoch": 2.995290878794465,
+      "grad_norm": 0.22875334322452545,
+      "learning_rate": 3.514942528735632e-05,
+      "loss": 0.1313,
+      "step": 2584
+    },
+    {
+      "epoch": 2.996450047091212,
+      "grad_norm": 0.24003781378269196,
+      "learning_rate": 3.514367816091954e-05,
+      "loss": 0.1311,
+      "step": 2585
+    },
+    {
+      "epoch": 2.997609215387959,
+      "grad_norm": 0.23002997040748596,
+      "learning_rate": 3.5137931034482765e-05,
+      "loss": 0.1153,
+      "step": 2586
+    },
+    {
+      "epoch": 2.9987683836847063,
+      "grad_norm": 0.28721874952316284,
+      "learning_rate": 3.513218390804598e-05,
+      "loss": 0.1178,
+      "step": 2587
+    },
+    {
+      "epoch": 2.9999275519814534,
+      "grad_norm": 0.2430833876132965,
+      "learning_rate": 3.5126436781609195e-05,
+      "loss": 0.1155,
+      "step": 2588
+    },
+    {
+      "epoch": 2.9999275519814534,
+      "eval_loss": 0.13075338304042816,
+      "eval_runtime": 279.8691,
+      "eval_samples_per_second": 5.481,
+      "eval_steps_per_second": 5.481,
+      "step": 2588
+    },
+    {
+      "epoch": 3.0010867202782006,
+      "grad_norm": 0.23414799571037292,
+      "learning_rate": 3.5120689655172416e-05,
+      "loss": 0.1143,
+      "step": 2589
+    },
+    {
+      "epoch": 3.0022458885749477,
+      "grad_norm": 0.25894778966903687,
+      "learning_rate": 3.511494252873563e-05,
+      "loss": 0.1089,
+      "step": 2590
+    },
+    {
+      "epoch": 3.0034050568716943,
+      "grad_norm": 0.23415379226207733,
+      "learning_rate": 3.510919540229885e-05,
+      "loss": 0.1097,
+      "step": 2591
+    },
+    {
+      "epoch": 3.0045642251684415,
+      "grad_norm": 0.21930260956287384,
+      "learning_rate": 3.5103448275862074e-05,
+      "loss": 0.1111,
+      "step": 2592
+    },
+    {
+      "epoch": 3.0057233934651886,
+      "grad_norm": 0.21448297798633575,
+      "learning_rate": 3.509770114942529e-05,
+      "loss": 0.1116,
+      "step": 2593
+    },
+    {
+      "epoch": 3.0068825617619357,
+      "grad_norm": 0.23125120997428894,
+      "learning_rate": 3.509195402298851e-05,
+      "loss": 0.116,
+      "step": 2594
+    },
+    {
+      "epoch": 3.008041730058683,
+      "grad_norm": 0.25538870692253113,
+      "learning_rate": 3.5086206896551726e-05,
+      "loss": 0.1146,
+      "step": 2595
+    },
+    {
+      "epoch": 3.00920089835543,
+      "grad_norm": 0.2790210247039795,
+      "learning_rate": 3.508045977011494e-05,
+      "loss": 0.1138,
+      "step": 2596
+    },
+    {
+      "epoch": 3.010360066652177,
+      "grad_norm": 0.22470445930957794,
+      "learning_rate": 3.507471264367816e-05,
+      "loss": 0.119,
+      "step": 2597
+    },
+    {
+      "epoch": 3.011519234948924,
+      "grad_norm": 0.2440544217824936,
+      "learning_rate": 3.5068965517241384e-05,
+      "loss": 0.1027,
+      "step": 2598
+    },
+    {
+      "epoch": 3.0126784032456713,
+      "grad_norm": 0.3405747413635254,
+      "learning_rate": 3.50632183908046e-05,
+      "loss": 0.113,
+      "step": 2599
+    },
+    {
+      "epoch": 3.0138375715424184,
+      "grad_norm": 0.22075963020324707,
+      "learning_rate": 3.505747126436782e-05,
+      "loss": 0.1043,
+      "step": 2600
+    },
+    {
+      "epoch": 3.0149967398391655,
+      "grad_norm": 0.28365811705589294,
+      "learning_rate": 3.5051724137931036e-05,
+      "loss": 0.1105,
+      "step": 2601
+    },
+    {
+      "epoch": 3.0161559081359126,
+      "grad_norm": 0.38459697365760803,
+      "learning_rate": 3.504597701149425e-05,
+      "loss": 0.1043,
+      "step": 2602
+    },
+    {
+      "epoch": 3.0173150764326597,
+      "grad_norm": 0.3330821394920349,
+      "learning_rate": 3.504022988505747e-05,
+      "loss": 0.1052,
+      "step": 2603
+    },
+    {
+      "epoch": 3.018474244729407,
+      "grad_norm": 0.25718963146209717,
+      "learning_rate": 3.503448275862069e-05,
+      "loss": 0.1091,
+      "step": 2604
+    },
+    {
+      "epoch": 3.0196334130261535,
+      "grad_norm": 0.2592739462852478,
+      "learning_rate": 3.502873563218391e-05,
+      "loss": 0.0919,
+      "step": 2605
+    },
+    {
+      "epoch": 3.0207925813229006,
+      "grad_norm": 0.24682362377643585,
+      "learning_rate": 3.502298850574713e-05,
+      "loss": 0.1094,
+      "step": 2606
+    },
+    {
+      "epoch": 3.0219517496196477,
+      "grad_norm": 0.4260392487049103,
+      "learning_rate": 3.5017241379310345e-05,
+      "loss": 0.1122,
+      "step": 2607
+    },
+    {
+      "epoch": 3.023110917916395,
+      "grad_norm": 0.35808852314949036,
+      "learning_rate": 3.501149425287357e-05,
+      "loss": 0.1215,
+      "step": 2608
+    },
+    {
+      "epoch": 3.024270086213142,
+      "grad_norm": 0.31842851638793945,
+      "learning_rate": 3.500574712643678e-05,
+      "loss": 0.1083,
+      "step": 2609
+    },
+    {
+      "epoch": 3.025429254509889,
+      "grad_norm": 0.29208144545555115,
+      "learning_rate": 3.5e-05,
+      "loss": 0.1146,
+      "step": 2610
+    },
+    {
+      "epoch": 3.026588422806636,
+      "grad_norm": 0.2514886260032654,
+      "learning_rate": 3.499425287356322e-05,
+      "loss": 0.0972,
+      "step": 2611
+    },
+    {
+      "epoch": 3.0277475911033833,
+      "grad_norm": 0.3120005428791046,
+      "learning_rate": 3.498850574712644e-05,
+      "loss": 0.1072,
+      "step": 2612
+    },
+    {
+      "epoch": 3.0289067594001304,
+      "grad_norm": 0.30786648392677307,
+      "learning_rate": 3.498275862068966e-05,
+      "loss": 0.1142,
+      "step": 2613
+    },
+    {
+      "epoch": 3.0300659276968775,
+      "grad_norm": 0.3202109634876251,
+      "learning_rate": 3.497701149425288e-05,
+      "loss": 0.1153,
+      "step": 2614
+    },
+    {
+      "epoch": 3.0312250959936247,
+      "grad_norm": 0.3748902976512909,
+      "learning_rate": 3.497126436781609e-05,
+      "loss": 0.1064,
+      "step": 2615
+    },
+    {
+      "epoch": 3.032384264290372,
+      "grad_norm": 0.3179676830768585,
+      "learning_rate": 3.4965517241379313e-05,
+      "loss": 0.1193,
+      "step": 2616
+    },
+    {
+      "epoch": 3.033543432587119,
+      "grad_norm": 0.31394878029823303,
+      "learning_rate": 3.495977011494253e-05,
+      "loss": 0.1158,
+      "step": 2617
+    },
+    {
+      "epoch": 3.034702600883866,
+      "grad_norm": 0.2936527132987976,
+      "learning_rate": 3.495402298850575e-05,
+      "loss": 0.1094,
+      "step": 2618
+    },
+    {
+      "epoch": 3.035861769180613,
+      "grad_norm": 0.252445787191391,
+      "learning_rate": 3.494827586206897e-05,
+      "loss": 0.1038,
+      "step": 2619
+    },
+    {
+      "epoch": 3.03702093747736,
+      "grad_norm": 0.29907506704330444,
+      "learning_rate": 3.4942528735632187e-05,
+      "loss": 0.112,
+      "step": 2620
+    },
+    {
+      "epoch": 3.038180105774107,
+      "grad_norm": 0.24381442368030548,
+      "learning_rate": 3.49367816091954e-05,
+      "loss": 0.1026,
+      "step": 2621
+    },
+    {
+      "epoch": 3.039339274070854,
+      "grad_norm": 0.3066432476043701,
+      "learning_rate": 3.493103448275862e-05,
+      "loss": 0.1018,
+      "step": 2622
+    },
+    {
+      "epoch": 3.040498442367601,
+      "grad_norm": 0.2872762680053711,
+      "learning_rate": 3.492528735632184e-05,
+      "loss": 0.1105,
+      "step": 2623
+    },
+    {
+      "epoch": 3.0416576106643483,
+      "grad_norm": 0.2535305917263031,
+      "learning_rate": 3.491954022988506e-05,
+      "loss": 0.1013,
+      "step": 2624
+    },
+    {
+      "epoch": 3.0428167789610954,
+      "grad_norm": 0.22548238933086395,
+      "learning_rate": 3.4913793103448275e-05,
+      "loss": 0.1034,
+      "step": 2625
+    },
+    {
+      "epoch": 3.0439759472578425,
+      "grad_norm": 0.3591626286506653,
+      "learning_rate": 3.4908045977011496e-05,
+      "loss": 0.1124,
+      "step": 2626
+    },
+    {
+      "epoch": 3.0451351155545896,
+      "grad_norm": 0.3023275136947632,
+      "learning_rate": 3.490229885057472e-05,
+      "loss": 0.1029,
+      "step": 2627
+    },
+    {
+      "epoch": 3.0462942838513367,
+      "grad_norm": 0.25562551617622375,
+      "learning_rate": 3.489655172413793e-05,
+      "loss": 0.1056,
+      "step": 2628
+    },
+    {
+      "epoch": 3.047453452148084,
+      "grad_norm": 0.32490625977516174,
+      "learning_rate": 3.489080459770115e-05,
+      "loss": 0.1061,
+      "step": 2629
+    },
+    {
+      "epoch": 3.048612620444831,
+      "grad_norm": 0.30418357253074646,
+      "learning_rate": 3.488505747126437e-05,
+      "loss": 0.1139,
+      "step": 2630
+    },
+    {
+      "epoch": 3.049771788741578,
+      "grad_norm": 0.3882814645767212,
+      "learning_rate": 3.4879310344827584e-05,
+      "loss": 0.1037,
+      "step": 2631
+    },
+    {
+      "epoch": 3.050930957038325,
+      "grad_norm": 0.4713079631328583,
+      "learning_rate": 3.4873563218390806e-05,
+      "loss": 0.1157,
+      "step": 2632
+    },
+    {
+      "epoch": 3.0520901253350723,
+      "grad_norm": 0.32217907905578613,
+      "learning_rate": 3.486781609195403e-05,
+      "loss": 0.1135,
+      "step": 2633
+    },
+    {
+      "epoch": 3.053249293631819,
+      "grad_norm": 0.24278047680854797,
+      "learning_rate": 3.486206896551724e-05,
+      "loss": 0.1121,
+      "step": 2634
+    },
+    {
+      "epoch": 3.054408461928566,
+      "grad_norm": 0.32192277908325195,
+      "learning_rate": 3.4856321839080464e-05,
+      "loss": 0.1019,
+      "step": 2635
+    },
+    {
+      "epoch": 3.055567630225313,
+      "grad_norm": 0.317410945892334,
+      "learning_rate": 3.485057471264368e-05,
+      "loss": 0.1081,
+      "step": 2636
+    },
+    {
+      "epoch": 3.0567267985220603,
+      "grad_norm": 0.30316248536109924,
+      "learning_rate": 3.4844827586206894e-05,
+      "loss": 0.0999,
+      "step": 2637
+    },
+    {
+      "epoch": 3.0578859668188074,
+      "grad_norm": 0.4302930235862732,
+      "learning_rate": 3.4839080459770116e-05,
+      "loss": 0.113,
+      "step": 2638
+    },
+    {
+      "epoch": 3.0590451351155545,
+      "grad_norm": 0.29626715183258057,
+      "learning_rate": 3.483333333333334e-05,
+      "loss": 0.1057,
+      "step": 2639
+    },
+    {
+      "epoch": 3.0602043034123017,
+      "grad_norm": 0.38166263699531555,
+      "learning_rate": 3.482758620689655e-05,
+      "loss": 0.1216,
+      "step": 2640
+    },
+    {
+      "epoch": 3.0613634717090488,
+      "grad_norm": 0.3923417031764984,
+      "learning_rate": 3.4821839080459774e-05,
+      "loss": 0.1105,
+      "step": 2641
+    },
+    {
+      "epoch": 3.062522640005796,
+      "grad_norm": 0.27287542819976807,
+      "learning_rate": 3.481609195402299e-05,
+      "loss": 0.1096,
+      "step": 2642
+    },
+    {
+      "epoch": 3.063681808302543,
+      "grad_norm": 0.2712195813655853,
+      "learning_rate": 3.481034482758621e-05,
+      "loss": 0.1141,
+      "step": 2643
+    },
+    {
+      "epoch": 3.06484097659929,
+      "grad_norm": 0.2444847673177719,
+      "learning_rate": 3.4804597701149426e-05,
+      "loss": 0.1045,
+      "step": 2644
+    },
+    {
+      "epoch": 3.0660001448960372,
+      "grad_norm": 0.26770317554473877,
+      "learning_rate": 3.479885057471264e-05,
+      "loss": 0.1044,
+      "step": 2645
+    },
+    {
+      "epoch": 3.0671593131927843,
+      "grad_norm": 0.27558213472366333,
+      "learning_rate": 3.479310344827587e-05,
+      "loss": 0.1089,
+      "step": 2646
+    },
+    {
+      "epoch": 3.0683184814895315,
+      "grad_norm": 0.26137158274650574,
+      "learning_rate": 3.4787356321839084e-05,
+      "loss": 0.1059,
+      "step": 2647
+    },
+    {
+      "epoch": 3.069477649786278,
+      "grad_norm": 0.3349991738796234,
+      "learning_rate": 3.47816091954023e-05,
+      "loss": 0.1185,
+      "step": 2648
+    },
+    {
+      "epoch": 3.0706368180830252,
+      "grad_norm": 0.32622119784355164,
+      "learning_rate": 3.477586206896552e-05,
+      "loss": 0.1101,
+      "step": 2649
+    },
+    {
+      "epoch": 3.0717959863797724,
+      "grad_norm": 0.3372982442378998,
+      "learning_rate": 3.4770114942528735e-05,
+      "loss": 0.113,
+      "step": 2650
+    },
+    {
+      "epoch": 3.0729551546765195,
+      "grad_norm": 0.31593719124794006,
+      "learning_rate": 3.476436781609196e-05,
+      "loss": 0.1188,
+      "step": 2651
+    },
+    {
+      "epoch": 3.0741143229732666,
+      "grad_norm": 0.26254966855049133,
+      "learning_rate": 3.475862068965517e-05,
+      "loss": 0.1185,
+      "step": 2652
+    },
+    {
+      "epoch": 3.0752734912700137,
+      "grad_norm": 0.22806406021118164,
+      "learning_rate": 3.4752873563218394e-05,
+      "loss": 0.0998,
+      "step": 2653
+    },
+    {
+      "epoch": 3.076432659566761,
+      "grad_norm": 0.26122182607650757,
+      "learning_rate": 3.4747126436781615e-05,
+      "loss": 0.1111,
+      "step": 2654
+    },
+    {
+      "epoch": 3.077591827863508,
+      "grad_norm": 0.2905748188495636,
+      "learning_rate": 3.474137931034483e-05,
+      "loss": 0.1127,
+      "step": 2655
+    },
+    {
+      "epoch": 3.078750996160255,
+      "grad_norm": 0.32908645272254944,
+      "learning_rate": 3.4735632183908045e-05,
+      "loss": 0.1165,
+      "step": 2656
+    },
+    {
+      "epoch": 3.079910164457002,
+      "grad_norm": 0.32499000430107117,
+      "learning_rate": 3.472988505747127e-05,
+      "loss": 0.1151,
+      "step": 2657
+    },
+    {
+      "epoch": 3.0810693327537493,
+      "grad_norm": 0.30107995867729187,
+      "learning_rate": 3.472413793103448e-05,
+      "loss": 0.1142,
+      "step": 2658
+    },
+    {
+      "epoch": 3.0822285010504964,
+      "grad_norm": 0.29092562198638916,
+      "learning_rate": 3.47183908045977e-05,
+      "loss": 0.1015,
+      "step": 2659
+    },
+    {
+      "epoch": 3.0833876693472435,
+      "grad_norm": 0.23830758035182953,
+      "learning_rate": 3.4712643678160925e-05,
+      "loss": 0.1084,
+      "step": 2660
+    },
+    {
+      "epoch": 3.0845468376439906,
+      "grad_norm": 0.254711776971817,
+      "learning_rate": 3.470689655172414e-05,
+      "loss": 0.106,
+      "step": 2661
+    },
+    {
+      "epoch": 3.0857060059407377,
+      "grad_norm": 0.24054522812366486,
+      "learning_rate": 3.470114942528736e-05,
+      "loss": 0.0995,
+      "step": 2662
+    },
+    {
+      "epoch": 3.0868651742374844,
+      "grad_norm": 0.2799350917339325,
+      "learning_rate": 3.4695402298850576e-05,
+      "loss": 0.1114,
+      "step": 2663
+    },
+    {
+      "epoch": 3.0880243425342315,
+      "grad_norm": 0.3153596818447113,
+      "learning_rate": 3.468965517241379e-05,
+      "loss": 0.1107,
+      "step": 2664
+    },
+    {
+      "epoch": 3.0891835108309786,
+      "grad_norm": 0.2980448603630066,
+      "learning_rate": 3.468390804597701e-05,
+      "loss": 0.1169,
+      "step": 2665
+    },
+    {
+      "epoch": 3.0903426791277258,
+      "grad_norm": 0.26500749588012695,
+      "learning_rate": 3.467816091954023e-05,
+      "loss": 0.1071,
+      "step": 2666
+    },
+    {
+      "epoch": 3.091501847424473,
+      "grad_norm": 0.308677077293396,
+      "learning_rate": 3.467241379310345e-05,
+      "loss": 0.112,
+      "step": 2667
+    },
+    {
+      "epoch": 3.09266101572122,
+      "grad_norm": 0.3179532289505005,
+      "learning_rate": 3.466666666666667e-05,
+      "loss": 0.1219,
+      "step": 2668
+    },
+    {
+      "epoch": 3.093820184017967,
+      "grad_norm": 0.28046876192092896,
+      "learning_rate": 3.4660919540229886e-05,
+      "loss": 0.1197,
+      "step": 2669
+    },
+    {
+      "epoch": 3.094979352314714,
+      "grad_norm": 0.3597003221511841,
+      "learning_rate": 3.465517241379311e-05,
+      "loss": 0.0995,
+      "step": 2670
+    },
+    {
+      "epoch": 3.0961385206114613,
+      "grad_norm": 0.2430323362350464,
+      "learning_rate": 3.464942528735632e-05,
+      "loss": 0.1078,
+      "step": 2671
+    },
+    {
+      "epoch": 3.0972976889082084,
+      "grad_norm": 0.3916395902633667,
+      "learning_rate": 3.464367816091954e-05,
+      "loss": 0.1289,
+      "step": 2672
+    },
+    {
+      "epoch": 3.0984568572049556,
+      "grad_norm": 0.28007376194000244,
+      "learning_rate": 3.463793103448276e-05,
+      "loss": 0.1082,
+      "step": 2673
+    },
+    {
+      "epoch": 3.0996160255017027,
+      "grad_norm": 0.370697945356369,
+      "learning_rate": 3.463218390804598e-05,
+      "loss": 0.1157,
+      "step": 2674
+    },
+    {
+      "epoch": 3.10077519379845,
+      "grad_norm": 0.27986010909080505,
+      "learning_rate": 3.4626436781609196e-05,
+      "loss": 0.1067,
+      "step": 2675
+    },
+    {
+      "epoch": 3.101934362095197,
+      "grad_norm": 0.34155765175819397,
+      "learning_rate": 3.462068965517242e-05,
+      "loss": 0.1124,
+      "step": 2676
+    },
+    {
+      "epoch": 3.1030935303919436,
+      "grad_norm": 0.27783599495887756,
+      "learning_rate": 3.461494252873563e-05,
+      "loss": 0.1084,
+      "step": 2677
+    },
+    {
+      "epoch": 3.1042526986886907,
+      "grad_norm": 0.32850217819213867,
+      "learning_rate": 3.460919540229885e-05,
+      "loss": 0.1029,
+      "step": 2678
+    },
+    {
+      "epoch": 3.105411866985438,
+      "grad_norm": 0.21651075780391693,
+      "learning_rate": 3.460344827586207e-05,
+      "loss": 0.0985,
+      "step": 2679
+    },
+    {
+      "epoch": 3.106571035282185,
+      "grad_norm": 0.29915618896484375,
+      "learning_rate": 3.459770114942529e-05,
+      "loss": 0.1114,
+      "step": 2680
+    },
+    {
+      "epoch": 3.107730203578932,
+      "grad_norm": 0.2699195146560669,
+      "learning_rate": 3.459195402298851e-05,
+      "loss": 0.108,
+      "step": 2681
+    },
+    {
+      "epoch": 3.108889371875679,
+      "grad_norm": 0.33292192220687866,
+      "learning_rate": 3.458620689655173e-05,
+      "loss": 0.121,
+      "step": 2682
+    },
+    {
+      "epoch": 3.1100485401724263,
+      "grad_norm": 0.48338431119918823,
+      "learning_rate": 3.458045977011494e-05,
+      "loss": 0.1148,
+      "step": 2683
+    },
+    {
+      "epoch": 3.1112077084691734,
+      "grad_norm": 0.5552323460578918,
+      "learning_rate": 3.4574712643678164e-05,
+      "loss": 0.1119,
+      "step": 2684
+    },
+    {
+      "epoch": 3.1123668767659205,
+      "grad_norm": 0.26004117727279663,
+      "learning_rate": 3.456896551724138e-05,
+      "loss": 0.1115,
+      "step": 2685
+    },
+    {
+      "epoch": 3.1135260450626676,
+      "grad_norm": 0.35093238949775696,
+      "learning_rate": 3.4563218390804594e-05,
+      "loss": 0.1196,
+      "step": 2686
+    },
+    {
+      "epoch": 3.1146852133594147,
+      "grad_norm": 0.26992592215538025,
+      "learning_rate": 3.455747126436782e-05,
+      "loss": 0.1114,
+      "step": 2687
+    },
+    {
+      "epoch": 3.115844381656162,
+      "grad_norm": 0.23782843351364136,
+      "learning_rate": 3.455172413793104e-05,
+      "loss": 0.112,
+      "step": 2688
+    },
+    {
+      "epoch": 3.117003549952909,
+      "grad_norm": 0.37826716899871826,
+      "learning_rate": 3.454597701149426e-05,
+      "loss": 0.1154,
+      "step": 2689
+    },
+    {
+      "epoch": 3.118162718249656,
+      "grad_norm": 0.4076231122016907,
+      "learning_rate": 3.4540229885057474e-05,
+      "loss": 0.1181,
+      "step": 2690
+    },
+    {
+      "epoch": 3.1193218865464027,
+      "grad_norm": 0.27670371532440186,
+      "learning_rate": 3.453448275862069e-05,
+      "loss": 0.1192,
+      "step": 2691
+    },
+    {
+      "epoch": 3.12048105484315,
+      "grad_norm": 0.3018391728401184,
+      "learning_rate": 3.452873563218391e-05,
+      "loss": 0.112,
+      "step": 2692
+    },
+    {
+      "epoch": 3.121640223139897,
+      "grad_norm": 0.3222883641719818,
+      "learning_rate": 3.4522988505747125e-05,
+      "loss": 0.1108,
+      "step": 2693
+    },
+    {
+      "epoch": 3.122799391436644,
+      "grad_norm": 0.32334110140800476,
+      "learning_rate": 3.451724137931035e-05,
+      "loss": 0.1059,
+      "step": 2694
+    },
+    {
+      "epoch": 3.123958559733391,
+      "grad_norm": 0.3832739293575287,
+      "learning_rate": 3.451149425287357e-05,
+      "loss": 0.1176,
+      "step": 2695
+    },
+    {
+      "epoch": 3.1251177280301383,
+      "grad_norm": 0.30942925810813904,
+      "learning_rate": 3.4505747126436783e-05,
+      "loss": 0.1249,
+      "step": 2696
+    },
+    {
+      "epoch": 3.1262768963268854,
+      "grad_norm": 0.25625792145729065,
+      "learning_rate": 3.45e-05,
+      "loss": 0.1054,
+      "step": 2697
+    },
+    {
+      "epoch": 3.1274360646236326,
+      "grad_norm": 0.28217336535453796,
+      "learning_rate": 3.449425287356322e-05,
+      "loss": 0.1095,
+      "step": 2698
+    },
+    {
+      "epoch": 3.1285952329203797,
+      "grad_norm": 0.2541414499282837,
+      "learning_rate": 3.4488505747126435e-05,
+      "loss": 0.108,
+      "step": 2699
+    },
+    {
+      "epoch": 3.129754401217127,
+      "grad_norm": 0.25402045249938965,
+      "learning_rate": 3.4482758620689657e-05,
+      "loss": 0.1133,
+      "step": 2700
+    },
+    {
+      "epoch": 3.130913569513874,
+      "grad_norm": 0.23192809522151947,
+      "learning_rate": 3.447701149425288e-05,
+      "loss": 0.1102,
+      "step": 2701
+    },
+    {
+      "epoch": 3.132072737810621,
+      "grad_norm": 0.3315061926841736,
+      "learning_rate": 3.447126436781609e-05,
+      "loss": 0.1214,
+      "step": 2702
+    },
+    {
+      "epoch": 3.133231906107368,
+      "grad_norm": 0.2749609053134918,
+      "learning_rate": 3.4465517241379315e-05,
+      "loss": 0.1005,
+      "step": 2703
+    },
+    {
+      "epoch": 3.1343910744041152,
+      "grad_norm": 0.36684563755989075,
+      "learning_rate": 3.445977011494253e-05,
+      "loss": 0.1055,
+      "step": 2704
+    },
+    {
+      "epoch": 3.1355502427008624,
+      "grad_norm": 0.30152618885040283,
+      "learning_rate": 3.4454022988505745e-05,
+      "loss": 0.1053,
+      "step": 2705
+    },
+    {
+      "epoch": 3.136709410997609,
+      "grad_norm": 0.24934984743595123,
+      "learning_rate": 3.4448275862068966e-05,
+      "loss": 0.1231,
+      "step": 2706
+    },
+    {
+      "epoch": 3.137868579294356,
+      "grad_norm": 0.3130822777748108,
+      "learning_rate": 3.444252873563219e-05,
+      "loss": 0.1075,
+      "step": 2707
+    },
+    {
+      "epoch": 3.1390277475911033,
+      "grad_norm": 0.28072500228881836,
+      "learning_rate": 3.443678160919541e-05,
+      "loss": 0.1016,
+      "step": 2708
+    },
+    {
+      "epoch": 3.1401869158878504,
+      "grad_norm": 0.36836108565330505,
+      "learning_rate": 3.4431034482758625e-05,
+      "loss": 0.1132,
+      "step": 2709
+    },
+    {
+      "epoch": 3.1413460841845975,
+      "grad_norm": 0.40246373414993286,
+      "learning_rate": 3.442528735632184e-05,
+      "loss": 0.1265,
+      "step": 2710
+    },
+    {
+      "epoch": 3.1425052524813446,
+      "grad_norm": 0.3051183223724365,
+      "learning_rate": 3.441954022988506e-05,
+      "loss": 0.0992,
+      "step": 2711
+    },
+    {
+      "epoch": 3.1436644207780917,
+      "grad_norm": 0.2942332625389099,
+      "learning_rate": 3.4413793103448276e-05,
+      "loss": 0.1132,
+      "step": 2712
+    },
+    {
+      "epoch": 3.144823589074839,
+      "grad_norm": 0.34289440512657166,
+      "learning_rate": 3.440804597701149e-05,
+      "loss": 0.112,
+      "step": 2713
+    },
+    {
+      "epoch": 3.145982757371586,
+      "grad_norm": 0.273400217294693,
+      "learning_rate": 3.440229885057471e-05,
+      "loss": 0.0977,
+      "step": 2714
+    },
+    {
+      "epoch": 3.147141925668333,
+      "grad_norm": 0.3162487745285034,
+      "learning_rate": 3.4396551724137934e-05,
+      "loss": 0.1127,
+      "step": 2715
+    },
+    {
+      "epoch": 3.14830109396508,
+      "grad_norm": 0.33499714732170105,
+      "learning_rate": 3.439080459770115e-05,
+      "loss": 0.112,
+      "step": 2716
+    },
+    {
+      "epoch": 3.1494602622618273,
+      "grad_norm": 0.28966301679611206,
+      "learning_rate": 3.438505747126437e-05,
+      "loss": 0.1062,
+      "step": 2717
+    },
+    {
+      "epoch": 3.1506194305585744,
+      "grad_norm": 0.2868296205997467,
+      "learning_rate": 3.4379310344827586e-05,
+      "loss": 0.1149,
+      "step": 2718
+    },
+    {
+      "epoch": 3.151778598855321,
+      "grad_norm": 0.4284828007221222,
+      "learning_rate": 3.437356321839081e-05,
+      "loss": 0.1244,
+      "step": 2719
+    },
+    {
+      "epoch": 3.152937767152068,
+      "grad_norm": 0.3205833435058594,
+      "learning_rate": 3.436781609195402e-05,
+      "loss": 0.0985,
+      "step": 2720
+    },
+    {
+      "epoch": 3.1540969354488153,
+      "grad_norm": 0.29444050788879395,
+      "learning_rate": 3.4362068965517244e-05,
+      "loss": 0.1103,
+      "step": 2721
+    },
+    {
+      "epoch": 3.1552561037455624,
+      "grad_norm": 0.2642725706100464,
+      "learning_rate": 3.4356321839080466e-05,
+      "loss": 0.1047,
+      "step": 2722
+    },
+    {
+      "epoch": 3.1564152720423095,
+      "grad_norm": 0.34533828496932983,
+      "learning_rate": 3.435057471264368e-05,
+      "loss": 0.1131,
+      "step": 2723
+    },
+    {
+      "epoch": 3.1575744403390567,
+      "grad_norm": 0.31518617272377014,
+      "learning_rate": 3.4344827586206896e-05,
+      "loss": 0.1065,
+      "step": 2724
+    },
+    {
+      "epoch": 3.1587336086358038,
+      "grad_norm": 0.3125956058502197,
+      "learning_rate": 3.433908045977012e-05,
+      "loss": 0.1039,
+      "step": 2725
+    },
+    {
+      "epoch": 3.159892776932551,
+      "grad_norm": 0.24645163118839264,
+      "learning_rate": 3.433333333333333e-05,
+      "loss": 0.1016,
+      "step": 2726
+    },
+    {
+      "epoch": 3.161051945229298,
+      "grad_norm": 0.42728784680366516,
+      "learning_rate": 3.4327586206896554e-05,
+      "loss": 0.101,
+      "step": 2727
+    },
+    {
+      "epoch": 3.162211113526045,
+      "grad_norm": 0.5982040166854858,
+      "learning_rate": 3.4321839080459775e-05,
+      "loss": 0.1286,
+      "step": 2728
+    },
+    {
+      "epoch": 3.1633702818227922,
+      "grad_norm": 0.3353968560695648,
+      "learning_rate": 3.431609195402299e-05,
+      "loss": 0.1232,
+      "step": 2729
+    },
+    {
+      "epoch": 3.1645294501195393,
+      "grad_norm": 0.3435153663158417,
+      "learning_rate": 3.431034482758621e-05,
+      "loss": 0.1188,
+      "step": 2730
+    },
+    {
+      "epoch": 3.1656886184162865,
+      "grad_norm": 0.39836829900741577,
+      "learning_rate": 3.430459770114943e-05,
+      "loss": 0.121,
+      "step": 2731
+    },
+    {
+      "epoch": 3.1668477867130336,
+      "grad_norm": 0.30429473519325256,
+      "learning_rate": 3.429885057471264e-05,
+      "loss": 0.1089,
+      "step": 2732
+    },
+    {
+      "epoch": 3.1680069550097807,
+      "grad_norm": 0.3069549798965454,
+      "learning_rate": 3.4293103448275864e-05,
+      "loss": 0.1196,
+      "step": 2733
+    },
+    {
+      "epoch": 3.1691661233065274,
+      "grad_norm": 0.2685747444629669,
+      "learning_rate": 3.428735632183908e-05,
+      "loss": 0.1202,
+      "step": 2734
+    },
+    {
+      "epoch": 3.1703252916032745,
+      "grad_norm": 0.2461479902267456,
+      "learning_rate": 3.42816091954023e-05,
+      "loss": 0.1145,
+      "step": 2735
+    },
+    {
+      "epoch": 3.1714844599000216,
+      "grad_norm": 0.29422634840011597,
+      "learning_rate": 3.427586206896552e-05,
+      "loss": 0.118,
+      "step": 2736
+    },
+    {
+      "epoch": 3.1726436281967687,
+      "grad_norm": 0.3010058104991913,
+      "learning_rate": 3.427011494252874e-05,
+      "loss": 0.1168,
+      "step": 2737
+    },
+    {
+      "epoch": 3.173802796493516,
+      "grad_norm": 0.2945718765258789,
+      "learning_rate": 3.426436781609196e-05,
+      "loss": 0.109,
+      "step": 2738
+    },
+    {
+      "epoch": 3.174961964790263,
+      "grad_norm": 0.2734999358654022,
+      "learning_rate": 3.425862068965517e-05,
+      "loss": 0.1117,
+      "step": 2739
+    },
+    {
+      "epoch": 3.17612113308701,
+      "grad_norm": 0.25881072878837585,
+      "learning_rate": 3.425287356321839e-05,
+      "loss": 0.1062,
+      "step": 2740
+    },
+    {
+      "epoch": 3.177280301383757,
+      "grad_norm": 0.2748000919818878,
+      "learning_rate": 3.424712643678161e-05,
+      "loss": 0.1093,
+      "step": 2741
+    },
+    {
+      "epoch": 3.1784394696805043,
+      "grad_norm": 0.37148886919021606,
+      "learning_rate": 3.424137931034483e-05,
+      "loss": 0.1227,
+      "step": 2742
+    },
+    {
+      "epoch": 3.1795986379772514,
+      "grad_norm": 0.32973724603652954,
+      "learning_rate": 3.4235632183908046e-05,
+      "loss": 0.1057,
+      "step": 2743
+    },
+    {
+      "epoch": 3.1807578062739985,
+      "grad_norm": 0.27131950855255127,
+      "learning_rate": 3.422988505747127e-05,
+      "loss": 0.106,
+      "step": 2744
+    },
+    {
+      "epoch": 3.1819169745707456,
+      "grad_norm": 0.25479647517204285,
+      "learning_rate": 3.422413793103448e-05,
+      "loss": 0.1154,
+      "step": 2745
+    },
+    {
+      "epoch": 3.1830761428674927,
+      "grad_norm": 0.331696480512619,
+      "learning_rate": 3.4218390804597705e-05,
+      "loss": 0.1166,
+      "step": 2746
+    },
+    {
+      "epoch": 3.18423531116424,
+      "grad_norm": 0.30211329460144043,
+      "learning_rate": 3.421264367816092e-05,
+      "loss": 0.1158,
+      "step": 2747
+    },
+    {
+      "epoch": 3.185394479460987,
+      "grad_norm": 0.265440434217453,
+      "learning_rate": 3.420689655172414e-05,
+      "loss": 0.0981,
+      "step": 2748
+    },
+    {
+      "epoch": 3.1865536477577336,
+      "grad_norm": 0.34413856267929077,
+      "learning_rate": 3.420114942528736e-05,
+      "loss": 0.1058,
+      "step": 2749
+    },
+    {
+      "epoch": 3.1877128160544808,
+      "grad_norm": 0.3049376308917999,
+      "learning_rate": 3.419540229885058e-05,
+      "loss": 0.107,
+      "step": 2750
+    },
+    {
+      "epoch": 3.188871984351228,
+      "grad_norm": 0.24063755571842194,
+      "learning_rate": 3.418965517241379e-05,
+      "loss": 0.1063,
+      "step": 2751
+    },
+    {
+      "epoch": 3.190031152647975,
+      "grad_norm": 0.40382859110832214,
+      "learning_rate": 3.4183908045977014e-05,
+      "loss": 0.1112,
+      "step": 2752
+    },
+    {
+      "epoch": 3.191190320944722,
+      "grad_norm": 0.3165836036205292,
+      "learning_rate": 3.417816091954023e-05,
+      "loss": 0.1113,
+      "step": 2753
+    },
+    {
+      "epoch": 3.192349489241469,
+      "grad_norm": 0.3423636555671692,
+      "learning_rate": 3.4172413793103444e-05,
+      "loss": 0.1036,
+      "step": 2754
+    },
+    {
+      "epoch": 3.1935086575382163,
+      "grad_norm": 0.3611285388469696,
+      "learning_rate": 3.4166666666666666e-05,
+      "loss": 0.1057,
+      "step": 2755
+    },
+    {
+      "epoch": 3.1946678258349634,
+      "grad_norm": 0.2877699136734009,
+      "learning_rate": 3.416091954022989e-05,
+      "loss": 0.1124,
+      "step": 2756
+    },
+    {
+      "epoch": 3.1958269941317106,
+      "grad_norm": 0.32133352756500244,
+      "learning_rate": 3.415517241379311e-05,
+      "loss": 0.1136,
+      "step": 2757
+    },
+    {
+      "epoch": 3.1969861624284577,
+      "grad_norm": 0.28176063299179077,
+      "learning_rate": 3.4149425287356324e-05,
+      "loss": 0.1036,
+      "step": 2758
+    },
+    {
+      "epoch": 3.198145330725205,
+      "grad_norm": 0.35929253697395325,
+      "learning_rate": 3.414367816091954e-05,
+      "loss": 0.1226,
+      "step": 2759
+    },
+    {
+      "epoch": 3.199304499021952,
+      "grad_norm": 0.37938886880874634,
+      "learning_rate": 3.413793103448276e-05,
+      "loss": 0.1133,
+      "step": 2760
+    },
+    {
+      "epoch": 3.200463667318699,
+      "grad_norm": 0.27585509419441223,
+      "learning_rate": 3.4132183908045976e-05,
+      "loss": 0.0993,
+      "step": 2761
+    },
+    {
+      "epoch": 3.2016228356154457,
+      "grad_norm": 0.33102548122406006,
+      "learning_rate": 3.41264367816092e-05,
+      "loss": 0.1055,
+      "step": 2762
+    },
+    {
+      "epoch": 3.202782003912193,
+      "grad_norm": 0.4190862774848938,
+      "learning_rate": 3.412068965517242e-05,
+      "loss": 0.1232,
+      "step": 2763
+    },
+    {
+      "epoch": 3.20394117220894,
+      "grad_norm": 0.3287636935710907,
+      "learning_rate": 3.4114942528735634e-05,
+      "loss": 0.1189,
+      "step": 2764
+    },
+    {
+      "epoch": 3.205100340505687,
+      "grad_norm": 0.27761226892471313,
+      "learning_rate": 3.4109195402298856e-05,
+      "loss": 0.1102,
+      "step": 2765
+    },
+    {
+      "epoch": 3.206259508802434,
+      "grad_norm": 0.271768718957901,
+      "learning_rate": 3.410344827586207e-05,
+      "loss": 0.1042,
+      "step": 2766
+    },
+    {
+      "epoch": 3.2074186770991813,
+      "grad_norm": 0.44091716408729553,
+      "learning_rate": 3.4097701149425285e-05,
+      "loss": 0.1098,
+      "step": 2767
+    },
+    {
+      "epoch": 3.2085778453959284,
+      "grad_norm": 0.32181066274642944,
+      "learning_rate": 3.409195402298851e-05,
+      "loss": 0.1077,
+      "step": 2768
+    },
+    {
+      "epoch": 3.2097370136926755,
+      "grad_norm": 0.33620011806488037,
+      "learning_rate": 3.408620689655173e-05,
+      "loss": 0.1113,
+      "step": 2769
+    },
+    {
+      "epoch": 3.2108961819894226,
+      "grad_norm": 0.3269021511077881,
+      "learning_rate": 3.4080459770114944e-05,
+      "loss": 0.1266,
+      "step": 2770
+    },
+    {
+      "epoch": 3.2120553502861697,
+      "grad_norm": 0.3237850069999695,
+      "learning_rate": 3.4074712643678165e-05,
+      "loss": 0.1132,
+      "step": 2771
+    },
+    {
+      "epoch": 3.213214518582917,
+      "grad_norm": 0.37990424036979675,
+      "learning_rate": 3.406896551724138e-05,
+      "loss": 0.1028,
+      "step": 2772
+    },
+    {
+      "epoch": 3.214373686879664,
+      "grad_norm": 0.28017759323120117,
+      "learning_rate": 3.40632183908046e-05,
+      "loss": 0.108,
+      "step": 2773
+    },
+    {
+      "epoch": 3.215532855176411,
+      "grad_norm": 0.3556600511074066,
+      "learning_rate": 3.405747126436782e-05,
+      "loss": 0.121,
+      "step": 2774
+    },
+    {
+      "epoch": 3.216692023473158,
+      "grad_norm": 0.26597216725349426,
+      "learning_rate": 3.405172413793103e-05,
+      "loss": 0.1024,
+      "step": 2775
+    },
+    {
+      "epoch": 3.2178511917699053,
+      "grad_norm": 0.35249412059783936,
+      "learning_rate": 3.4045977011494253e-05,
+      "loss": 0.1202,
+      "step": 2776
+    },
+    {
+      "epoch": 3.219010360066652,
+      "grad_norm": 0.35307571291923523,
+      "learning_rate": 3.4040229885057475e-05,
+      "loss": 0.1139,
+      "step": 2777
+    },
+    {
+      "epoch": 3.220169528363399,
+      "grad_norm": 0.25899243354797363,
+      "learning_rate": 3.403448275862069e-05,
+      "loss": 0.1079,
+      "step": 2778
+    },
+    {
+      "epoch": 3.221328696660146,
+      "grad_norm": 0.3125319480895996,
+      "learning_rate": 3.402873563218391e-05,
+      "loss": 0.1137,
+      "step": 2779
+    },
+    {
+      "epoch": 3.2224878649568933,
+      "grad_norm": 0.3008607029914856,
+      "learning_rate": 3.4022988505747127e-05,
+      "loss": 0.1013,
+      "step": 2780
+    },
+    {
+      "epoch": 3.2236470332536404,
+      "grad_norm": 0.2770404517650604,
+      "learning_rate": 3.401724137931034e-05,
+      "loss": 0.1137,
+      "step": 2781
+    },
+    {
+      "epoch": 3.2248062015503876,
+      "grad_norm": 0.35955601930618286,
+      "learning_rate": 3.401149425287356e-05,
+      "loss": 0.1108,
+      "step": 2782
+    },
+    {
+      "epoch": 3.2259653698471347,
+      "grad_norm": 0.2946624159812927,
+      "learning_rate": 3.4005747126436785e-05,
+      "loss": 0.1023,
+      "step": 2783
+    },
+    {
+      "epoch": 3.227124538143882,
+      "grad_norm": 0.34001263976097107,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 0.1049,
+      "step": 2784
+    },
+    {
+      "epoch": 3.228283706440629,
+      "grad_norm": 0.2924330234527588,
+      "learning_rate": 3.399425287356322e-05,
+      "loss": 0.1097,
+      "step": 2785
+    },
+    {
+      "epoch": 3.229442874737376,
+      "grad_norm": 0.3034241199493408,
+      "learning_rate": 3.3988505747126436e-05,
+      "loss": 0.1148,
+      "step": 2786
+    },
+    {
+      "epoch": 3.230602043034123,
+      "grad_norm": 0.3855332136154175,
+      "learning_rate": 3.398275862068966e-05,
+      "loss": 0.0945,
+      "step": 2787
+    },
+    {
+      "epoch": 3.2317612113308702,
+      "grad_norm": 0.24068623781204224,
+      "learning_rate": 3.397701149425287e-05,
+      "loss": 0.1092,
+      "step": 2788
+    },
+    {
+      "epoch": 3.2329203796276174,
+      "grad_norm": 0.2632269859313965,
+      "learning_rate": 3.3971264367816095e-05,
+      "loss": 0.1123,
+      "step": 2789
+    },
+    {
+      "epoch": 3.2340795479243645,
+      "grad_norm": 0.2876719832420349,
+      "learning_rate": 3.3965517241379316e-05,
+      "loss": 0.1158,
+      "step": 2790
+    },
+    {
+      "epoch": 3.2352387162211116,
+      "grad_norm": 0.3035225570201874,
+      "learning_rate": 3.395977011494253e-05,
+      "loss": 0.1187,
+      "step": 2791
+    },
+    {
+      "epoch": 3.2363978845178583,
+      "grad_norm": 0.27335041761398315,
+      "learning_rate": 3.395402298850575e-05,
+      "loss": 0.1005,
+      "step": 2792
+    },
+    {
+      "epoch": 3.2375570528146054,
+      "grad_norm": 0.2715912461280823,
+      "learning_rate": 3.394827586206897e-05,
+      "loss": 0.1102,
+      "step": 2793
+    },
+    {
+      "epoch": 3.2387162211113525,
+      "grad_norm": 0.2653818726539612,
+      "learning_rate": 3.394252873563218e-05,
+      "loss": 0.1119,
+      "step": 2794
+    },
+    {
+      "epoch": 3.2398753894080996,
+      "grad_norm": 0.3413951098918915,
+      "learning_rate": 3.3936781609195404e-05,
+      "loss": 0.1209,
+      "step": 2795
+    },
+    {
+      "epoch": 3.2410345577048467,
+      "grad_norm": 0.3429463505744934,
+      "learning_rate": 3.393103448275862e-05,
+      "loss": 0.1285,
+      "step": 2796
+    },
+    {
+      "epoch": 3.242193726001594,
+      "grad_norm": 0.3493320345878601,
+      "learning_rate": 3.392528735632184e-05,
+      "loss": 0.1169,
+      "step": 2797
+    },
+    {
+      "epoch": 3.243352894298341,
+      "grad_norm": 0.31924867630004883,
+      "learning_rate": 3.391954022988506e-05,
+      "loss": 0.1119,
+      "step": 2798
+    },
+    {
+      "epoch": 3.244512062595088,
+      "grad_norm": 0.42040061950683594,
+      "learning_rate": 3.391379310344828e-05,
+      "loss": 0.1106,
+      "step": 2799
+    },
+    {
+      "epoch": 3.245671230891835,
+      "grad_norm": 0.3470522463321686,
+      "learning_rate": 3.390804597701149e-05,
+      "loss": 0.1122,
+      "step": 2800
+    },
+    {
+      "epoch": 3.2468303991885823,
+      "grad_norm": 0.356770783662796,
+      "learning_rate": 3.3902298850574714e-05,
+      "loss": 0.1191,
+      "step": 2801
+    },
+    {
+      "epoch": 3.2479895674853294,
+      "grad_norm": 0.40925726294517517,
+      "learning_rate": 3.389655172413793e-05,
+      "loss": 0.1073,
+      "step": 2802
+    },
+    {
+      "epoch": 3.2491487357820765,
+      "grad_norm": 0.34631669521331787,
+      "learning_rate": 3.389080459770115e-05,
+      "loss": 0.1188,
+      "step": 2803
+    },
+    {
+      "epoch": 3.2503079040788236,
+      "grad_norm": 0.3311026394367218,
+      "learning_rate": 3.388505747126437e-05,
+      "loss": 0.1099,
+      "step": 2804
+    },
+    {
+      "epoch": 3.2514670723755703,
+      "grad_norm": 0.34615299105644226,
+      "learning_rate": 3.387931034482759e-05,
+      "loss": 0.1197,
+      "step": 2805
+    },
+    {
+      "epoch": 3.2526262406723174,
+      "grad_norm": 0.33618980646133423,
+      "learning_rate": 3.387356321839081e-05,
+      "loss": 0.1158,
+      "step": 2806
+    },
+    {
+      "epoch": 3.2537854089690645,
+      "grad_norm": 0.30485883355140686,
+      "learning_rate": 3.3867816091954024e-05,
+      "loss": 0.1141,
+      "step": 2807
+    },
+    {
+      "epoch": 3.2549445772658117,
+      "grad_norm": 0.35726383328437805,
+      "learning_rate": 3.386206896551724e-05,
+      "loss": 0.1059,
+      "step": 2808
+    },
+    {
+      "epoch": 3.2561037455625588,
+      "grad_norm": 0.45298418402671814,
+      "learning_rate": 3.385632183908046e-05,
+      "loss": 0.1142,
+      "step": 2809
+    },
+    {
+      "epoch": 3.257262913859306,
+      "grad_norm": 0.2797136902809143,
+      "learning_rate": 3.385057471264368e-05,
+      "loss": 0.104,
+      "step": 2810
+    },
+    {
+      "epoch": 3.258422082156053,
+      "grad_norm": 0.2695922553539276,
+      "learning_rate": 3.3844827586206904e-05,
+      "loss": 0.1047,
+      "step": 2811
+    },
+    {
+      "epoch": 3.2595812504528,
+      "grad_norm": 0.2859964668750763,
+      "learning_rate": 3.383908045977012e-05,
+      "loss": 0.1057,
+      "step": 2812
+    },
+    {
+      "epoch": 3.2607404187495472,
+      "grad_norm": 0.26754042506217957,
+      "learning_rate": 3.3833333333333334e-05,
+      "loss": 0.1138,
+      "step": 2813
+    },
+    {
+      "epoch": 3.2618995870462943,
+      "grad_norm": 0.3192819058895111,
+      "learning_rate": 3.3827586206896555e-05,
+      "loss": 0.1159,
+      "step": 2814
+    },
+    {
+      "epoch": 3.2630587553430415,
+      "grad_norm": 0.33752667903900146,
+      "learning_rate": 3.382183908045977e-05,
+      "loss": 0.1124,
+      "step": 2815
+    },
+    {
+      "epoch": 3.2642179236397886,
+      "grad_norm": 0.2708832025527954,
+      "learning_rate": 3.3816091954022985e-05,
+      "loss": 0.1181,
+      "step": 2816
+    },
+    {
+      "epoch": 3.2653770919365357,
+      "grad_norm": 0.3345796465873718,
+      "learning_rate": 3.381034482758621e-05,
+      "loss": 0.1123,
+      "step": 2817
+    },
+    {
+      "epoch": 3.266536260233283,
+      "grad_norm": 0.33695465326309204,
+      "learning_rate": 3.380459770114943e-05,
+      "loss": 0.0986,
+      "step": 2818
+    },
+    {
+      "epoch": 3.26769542853003,
+      "grad_norm": 0.2809793949127197,
+      "learning_rate": 3.379885057471264e-05,
+      "loss": 0.1151,
+      "step": 2819
+    },
+    {
+      "epoch": 3.2688545968267766,
+      "grad_norm": 0.35505884885787964,
+      "learning_rate": 3.3793103448275865e-05,
+      "loss": 0.1203,
+      "step": 2820
+    },
+    {
+      "epoch": 3.2700137651235237,
+      "grad_norm": 0.2393438071012497,
+      "learning_rate": 3.378735632183908e-05,
+      "loss": 0.1109,
+      "step": 2821
+    },
+    {
+      "epoch": 3.271172933420271,
+      "grad_norm": 0.35280153155326843,
+      "learning_rate": 3.37816091954023e-05,
+      "loss": 0.1217,
+      "step": 2822
+    },
+    {
+      "epoch": 3.272332101717018,
+      "grad_norm": 0.3676847815513611,
+      "learning_rate": 3.3775862068965516e-05,
+      "loss": 0.1182,
+      "step": 2823
+    },
+    {
+      "epoch": 3.273491270013765,
+      "grad_norm": 0.26780447363853455,
+      "learning_rate": 3.377011494252874e-05,
+      "loss": 0.1093,
+      "step": 2824
+    },
+    {
+      "epoch": 3.274650438310512,
+      "grad_norm": 0.3136596381664276,
+      "learning_rate": 3.376436781609196e-05,
+      "loss": 0.1126,
+      "step": 2825
+    },
+    {
+      "epoch": 3.2758096066072593,
+      "grad_norm": 0.26697683334350586,
+      "learning_rate": 3.3758620689655175e-05,
+      "loss": 0.1163,
+      "step": 2826
+    },
+    {
+      "epoch": 3.2769687749040064,
+      "grad_norm": 0.32463008165359497,
+      "learning_rate": 3.375287356321839e-05,
+      "loss": 0.1113,
+      "step": 2827
+    },
+    {
+      "epoch": 3.2781279432007535,
+      "grad_norm": 0.3022761344909668,
+      "learning_rate": 3.374712643678161e-05,
+      "loss": 0.113,
+      "step": 2828
+    },
+    {
+      "epoch": 3.2792871114975006,
+      "grad_norm": 0.2690412700176239,
+      "learning_rate": 3.3741379310344826e-05,
+      "loss": 0.1018,
+      "step": 2829
+    },
+    {
+      "epoch": 3.2804462797942477,
+      "grad_norm": 0.2957005798816681,
+      "learning_rate": 3.373563218390805e-05,
+      "loss": 0.1163,
+      "step": 2830
+    },
+    {
+      "epoch": 3.281605448090995,
+      "grad_norm": 0.3116534352302551,
+      "learning_rate": 3.372988505747127e-05,
+      "loss": 0.1263,
+      "step": 2831
+    },
+    {
+      "epoch": 3.282764616387742,
+      "grad_norm": 0.2315097600221634,
+      "learning_rate": 3.3724137931034484e-05,
+      "loss": 0.1036,
+      "step": 2832
+    },
+    {
+      "epoch": 3.2839237846844886,
+      "grad_norm": 0.5004782676696777,
+      "learning_rate": 3.3718390804597706e-05,
+      "loss": 0.1086,
+      "step": 2833
+    },
+    {
+      "epoch": 3.285082952981236,
+      "grad_norm": 0.3481523096561432,
+      "learning_rate": 3.371264367816092e-05,
+      "loss": 0.1112,
+      "step": 2834
+    },
+    {
+      "epoch": 3.286242121277983,
+      "grad_norm": 0.27010461688041687,
+      "learning_rate": 3.3706896551724136e-05,
+      "loss": 0.1145,
+      "step": 2835
+    },
+    {
+      "epoch": 3.28740128957473,
+      "grad_norm": 0.30692821741104126,
+      "learning_rate": 3.370114942528736e-05,
+      "loss": 0.1043,
+      "step": 2836
+    },
+    {
+      "epoch": 3.288560457871477,
+      "grad_norm": 0.33265191316604614,
+      "learning_rate": 3.369540229885057e-05,
+      "loss": 0.1151,
+      "step": 2837
+    },
+    {
+      "epoch": 3.289719626168224,
+      "grad_norm": 0.328311562538147,
+      "learning_rate": 3.3689655172413794e-05,
+      "loss": 0.1136,
+      "step": 2838
+    },
+    {
+      "epoch": 3.2908787944649713,
+      "grad_norm": 0.2806066870689392,
+      "learning_rate": 3.3683908045977016e-05,
+      "loss": 0.114,
+      "step": 2839
+    },
+    {
+      "epoch": 3.2920379627617185,
+      "grad_norm": 0.30125707387924194,
+      "learning_rate": 3.367816091954023e-05,
+      "loss": 0.128,
+      "step": 2840
+    },
+    {
+      "epoch": 3.2931971310584656,
+      "grad_norm": 0.31410592794418335,
+      "learning_rate": 3.367241379310345e-05,
+      "loss": 0.1133,
+      "step": 2841
+    },
+    {
+      "epoch": 3.2943562993552127,
+      "grad_norm": 0.24413692951202393,
+      "learning_rate": 3.366666666666667e-05,
+      "loss": 0.1056,
+      "step": 2842
+    },
+    {
+      "epoch": 3.29551546765196,
+      "grad_norm": 0.3842843472957611,
+      "learning_rate": 3.366091954022988e-05,
+      "loss": 0.1107,
+      "step": 2843
+    },
+    {
+      "epoch": 3.296674635948707,
+      "grad_norm": 0.2507694661617279,
+      "learning_rate": 3.3655172413793104e-05,
+      "loss": 0.108,
+      "step": 2844
+    },
+    {
+      "epoch": 3.297833804245454,
+      "grad_norm": 0.28499695658683777,
+      "learning_rate": 3.3649425287356326e-05,
+      "loss": 0.1169,
+      "step": 2845
+    },
+    {
+      "epoch": 3.298992972542201,
+      "grad_norm": 0.2902618646621704,
+      "learning_rate": 3.364367816091954e-05,
+      "loss": 0.118,
+      "step": 2846
+    },
+    {
+      "epoch": 3.3001521408389483,
+      "grad_norm": 0.26456037163734436,
+      "learning_rate": 3.363793103448276e-05,
+      "loss": 0.1105,
+      "step": 2847
+    },
+    {
+      "epoch": 3.301311309135695,
+      "grad_norm": 0.3139413595199585,
+      "learning_rate": 3.363218390804598e-05,
+      "loss": 0.1272,
+      "step": 2848
+    },
+    {
+      "epoch": 3.302470477432442,
+      "grad_norm": 0.282294899225235,
+      "learning_rate": 3.36264367816092e-05,
+      "loss": 0.1098,
+      "step": 2849
+    },
+    {
+      "epoch": 3.303629645729189,
+      "grad_norm": 0.2812632918357849,
+      "learning_rate": 3.3620689655172414e-05,
+      "loss": 0.1124,
+      "step": 2850
+    },
+    {
+      "epoch": 3.3047888140259363,
+      "grad_norm": 0.27748167514801025,
+      "learning_rate": 3.3614942528735635e-05,
+      "loss": 0.1028,
+      "step": 2851
+    },
+    {
+      "epoch": 3.3059479823226834,
+      "grad_norm": 0.39685526490211487,
+      "learning_rate": 3.360919540229886e-05,
+      "loss": 0.1191,
+      "step": 2852
+    },
+    {
+      "epoch": 3.3071071506194305,
+      "grad_norm": 0.33615151047706604,
+      "learning_rate": 3.360344827586207e-05,
+      "loss": 0.1085,
+      "step": 2853
+    },
+    {
+      "epoch": 3.3082663189161776,
+      "grad_norm": 0.2919673025608063,
+      "learning_rate": 3.359770114942529e-05,
+      "loss": 0.1045,
+      "step": 2854
+    },
+    {
+      "epoch": 3.3094254872129247,
+      "grad_norm": 0.2599186301231384,
+      "learning_rate": 3.359195402298851e-05,
+      "loss": 0.1106,
+      "step": 2855
+    },
+    {
+      "epoch": 3.310584655509672,
+      "grad_norm": 0.3396240174770355,
+      "learning_rate": 3.3586206896551723e-05,
+      "loss": 0.1063,
+      "step": 2856
+    },
+    {
+      "epoch": 3.311743823806419,
+      "grad_norm": 0.2739236354827881,
+      "learning_rate": 3.358045977011494e-05,
+      "loss": 0.1033,
+      "step": 2857
+    },
+    {
+      "epoch": 3.312902992103166,
+      "grad_norm": 0.5088075399398804,
+      "learning_rate": 3.357471264367817e-05,
+      "loss": 0.1295,
+      "step": 2858
+    },
+    {
+      "epoch": 3.314062160399913,
+      "grad_norm": 0.3902330994606018,
+      "learning_rate": 3.356896551724138e-05,
+      "loss": 0.1233,
+      "step": 2859
+    },
+    {
+      "epoch": 3.3152213286966603,
+      "grad_norm": 0.33790311217308044,
+      "learning_rate": 3.35632183908046e-05,
+      "loss": 0.1204,
+      "step": 2860
+    },
+    {
+      "epoch": 3.3163804969934074,
+      "grad_norm": 0.2987493872642517,
+      "learning_rate": 3.355747126436782e-05,
+      "loss": 0.1068,
+      "step": 2861
+    },
+    {
+      "epoch": 3.3175396652901545,
+      "grad_norm": 0.3023374080657959,
+      "learning_rate": 3.355172413793103e-05,
+      "loss": 0.1109,
+      "step": 2862
+    },
+    {
+      "epoch": 3.318698833586901,
+      "grad_norm": 0.26257169246673584,
+      "learning_rate": 3.3545977011494255e-05,
+      "loss": 0.1092,
+      "step": 2863
+    },
+    {
+      "epoch": 3.3198580018836483,
+      "grad_norm": 0.22975148260593414,
+      "learning_rate": 3.354022988505747e-05,
+      "loss": 0.1036,
+      "step": 2864
+    },
+    {
+      "epoch": 3.3210171701803954,
+      "grad_norm": 0.29056644439697266,
+      "learning_rate": 3.353448275862069e-05,
+      "loss": 0.1094,
+      "step": 2865
+    },
+    {
+      "epoch": 3.3221763384771426,
+      "grad_norm": 0.3132513761520386,
+      "learning_rate": 3.352873563218391e-05,
+      "loss": 0.1257,
+      "step": 2866
+    },
+    {
+      "epoch": 3.3233355067738897,
+      "grad_norm": 0.23093485832214355,
+      "learning_rate": 3.352298850574713e-05,
+      "loss": 0.1044,
+      "step": 2867
+    },
+    {
+      "epoch": 3.324494675070637,
+      "grad_norm": 0.3373148441314697,
+      "learning_rate": 3.351724137931035e-05,
+      "loss": 0.1174,
+      "step": 2868
+    },
+    {
+      "epoch": 3.325653843367384,
+      "grad_norm": 0.2774425148963928,
+      "learning_rate": 3.3511494252873565e-05,
+      "loss": 0.1137,
+      "step": 2869
+    },
+    {
+      "epoch": 3.326813011664131,
+      "grad_norm": 0.2680792808532715,
+      "learning_rate": 3.350574712643678e-05,
+      "loss": 0.1045,
+      "step": 2870
+    },
+    {
+      "epoch": 3.327972179960878,
+      "grad_norm": 0.39170289039611816,
+      "learning_rate": 3.35e-05,
+      "loss": 0.1121,
+      "step": 2871
+    },
+    {
+      "epoch": 3.3291313482576252,
+      "grad_norm": 0.2982361614704132,
+      "learning_rate": 3.349425287356322e-05,
+      "loss": 0.1169,
+      "step": 2872
+    },
+    {
+      "epoch": 3.3302905165543724,
+      "grad_norm": 0.2685178816318512,
+      "learning_rate": 3.348850574712644e-05,
+      "loss": 0.1001,
+      "step": 2873
+    },
+    {
+      "epoch": 3.3314496848511195,
+      "grad_norm": 0.35294216871261597,
+      "learning_rate": 3.348275862068966e-05,
+      "loss": 0.1087,
+      "step": 2874
+    },
+    {
+      "epoch": 3.3326088531478666,
+      "grad_norm": 0.3147522509098053,
+      "learning_rate": 3.3477011494252874e-05,
+      "loss": 0.1244,
+      "step": 2875
+    },
+    {
+      "epoch": 3.3337680214446133,
+      "grad_norm": 0.2905179560184479,
+      "learning_rate": 3.347126436781609e-05,
+      "loss": 0.1017,
+      "step": 2876
+    },
+    {
+      "epoch": 3.334927189741361,
+      "grad_norm": 0.28454288840293884,
+      "learning_rate": 3.346551724137931e-05,
+      "loss": 0.117,
+      "step": 2877
+    },
+    {
+      "epoch": 3.3360863580381075,
+      "grad_norm": 0.24592730402946472,
+      "learning_rate": 3.3459770114942526e-05,
+      "loss": 0.105,
+      "step": 2878
+    },
+    {
+      "epoch": 3.3372455263348546,
+      "grad_norm": 0.2803703248500824,
+      "learning_rate": 3.3454022988505754e-05,
+      "loss": 0.1251,
+      "step": 2879
+    },
+    {
+      "epoch": 3.3384046946316017,
+      "grad_norm": 0.2251860499382019,
+      "learning_rate": 3.344827586206897e-05,
+      "loss": 0.0974,
+      "step": 2880
+    },
+    {
+      "epoch": 3.339563862928349,
+      "grad_norm": 0.30766502022743225,
+      "learning_rate": 3.3442528735632184e-05,
+      "loss": 0.1055,
+      "step": 2881
+    },
+    {
+      "epoch": 3.340723031225096,
+      "grad_norm": 0.31499797105789185,
+      "learning_rate": 3.3436781609195406e-05,
+      "loss": 0.1112,
+      "step": 2882
+    },
+    {
+      "epoch": 3.341882199521843,
+      "grad_norm": 0.3749004900455475,
+      "learning_rate": 3.343103448275862e-05,
+      "loss": 0.1056,
+      "step": 2883
+    },
+    {
+      "epoch": 3.34304136781859,
+      "grad_norm": 0.3857807517051697,
+      "learning_rate": 3.3425287356321836e-05,
+      "loss": 0.1111,
+      "step": 2884
+    },
+    {
+      "epoch": 3.3442005361153373,
+      "grad_norm": 0.31204038858413696,
+      "learning_rate": 3.341954022988506e-05,
+      "loss": 0.1077,
+      "step": 2885
+    },
+    {
+      "epoch": 3.3453597044120844,
+      "grad_norm": 0.3444509506225586,
+      "learning_rate": 3.341379310344828e-05,
+      "loss": 0.1068,
+      "step": 2886
+    },
+    {
+      "epoch": 3.3465188727088315,
+      "grad_norm": 0.34033796191215515,
+      "learning_rate": 3.34080459770115e-05,
+      "loss": 0.1187,
+      "step": 2887
+    },
+    {
+      "epoch": 3.3476780410055786,
+      "grad_norm": 0.3054271936416626,
+      "learning_rate": 3.3402298850574715e-05,
+      "loss": 0.118,
+      "step": 2888
+    },
+    {
+      "epoch": 3.3488372093023258,
+      "grad_norm": 0.2640574276447296,
+      "learning_rate": 3.339655172413793e-05,
+      "loss": 0.1113,
+      "step": 2889
+    },
+    {
+      "epoch": 3.349996377599073,
+      "grad_norm": 0.9027268886566162,
+      "learning_rate": 3.339080459770115e-05,
+      "loss": 0.1232,
+      "step": 2890
+    },
+    {
+      "epoch": 3.3511555458958195,
+      "grad_norm": 0.2914552390575409,
+      "learning_rate": 3.338505747126437e-05,
+      "loss": 0.1053,
+      "step": 2891
+    },
+    {
+      "epoch": 3.3523147141925667,
+      "grad_norm": 0.2577556371688843,
+      "learning_rate": 3.337931034482759e-05,
+      "loss": 0.1057,
+      "step": 2892
+    },
+    {
+      "epoch": 3.3534738824893138,
+      "grad_norm": 0.2784647047519684,
+      "learning_rate": 3.337356321839081e-05,
+      "loss": 0.1012,
+      "step": 2893
+    },
+    {
+      "epoch": 3.354633050786061,
+      "grad_norm": 0.2601286470890045,
+      "learning_rate": 3.3367816091954025e-05,
+      "loss": 0.1066,
+      "step": 2894
+    },
+    {
+      "epoch": 3.355792219082808,
+      "grad_norm": 0.25884032249450684,
+      "learning_rate": 3.336206896551724e-05,
+      "loss": 0.1121,
+      "step": 2895
+    },
+    {
+      "epoch": 3.356951387379555,
+      "grad_norm": 0.3100210428237915,
+      "learning_rate": 3.335632183908046e-05,
+      "loss": 0.1129,
+      "step": 2896
+    },
+    {
+      "epoch": 3.3581105556763022,
+      "grad_norm": 0.3463558256626129,
+      "learning_rate": 3.335057471264368e-05,
+      "loss": 0.1178,
+      "step": 2897
+    },
+    {
+      "epoch": 3.3592697239730493,
+      "grad_norm": 0.3152965009212494,
+      "learning_rate": 3.33448275862069e-05,
+      "loss": 0.1214,
+      "step": 2898
+    },
+    {
+      "epoch": 3.3604288922697965,
+      "grad_norm": 0.32749250531196594,
+      "learning_rate": 3.333908045977012e-05,
+      "loss": 0.1165,
+      "step": 2899
+    },
+    {
+      "epoch": 3.3615880605665436,
+      "grad_norm": 0.3287135660648346,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.124,
+      "step": 2900
+    },
+    {
+      "epoch": 3.3627472288632907,
+      "grad_norm": 0.2673903703689575,
+      "learning_rate": 3.3327586206896557e-05,
+      "loss": 0.1162,
+      "step": 2901
+    },
+    {
+      "epoch": 3.363906397160038,
+      "grad_norm": 0.29948368668556213,
+      "learning_rate": 3.332183908045977e-05,
+      "loss": 0.1116,
+      "step": 2902
+    },
+    {
+      "epoch": 3.365065565456785,
+      "grad_norm": 0.4442507028579712,
+      "learning_rate": 3.3316091954022986e-05,
+      "loss": 0.1093,
+      "step": 2903
+    },
+    {
+      "epoch": 3.3662247337535316,
+      "grad_norm": 0.2337656021118164,
+      "learning_rate": 3.331034482758621e-05,
+      "loss": 0.0952,
+      "step": 2904
+    },
+    {
+      "epoch": 3.367383902050279,
+      "grad_norm": 0.2944662272930145,
+      "learning_rate": 3.330459770114942e-05,
+      "loss": 0.1011,
+      "step": 2905
+    },
+    {
+      "epoch": 3.368543070347026,
+      "grad_norm": 0.24339404702186584,
+      "learning_rate": 3.3298850574712645e-05,
+      "loss": 0.098,
+      "step": 2906
+    },
+    {
+      "epoch": 3.369702238643773,
+      "grad_norm": 0.33986908197402954,
+      "learning_rate": 3.3293103448275866e-05,
+      "loss": 0.1111,
+      "step": 2907
+    },
+    {
+      "epoch": 3.37086140694052,
+      "grad_norm": 0.333913117647171,
+      "learning_rate": 3.328735632183908e-05,
+      "loss": 0.1163,
+      "step": 2908
+    },
+    {
+      "epoch": 3.372020575237267,
+      "grad_norm": 0.2759074568748474,
+      "learning_rate": 3.32816091954023e-05,
+      "loss": 0.1083,
+      "step": 2909
+    },
+    {
+      "epoch": 3.3731797435340143,
+      "grad_norm": 0.264443039894104,
+      "learning_rate": 3.327586206896552e-05,
+      "loss": 0.1069,
+      "step": 2910
+    },
+    {
+      "epoch": 3.3743389118307614,
+      "grad_norm": 0.33492133021354675,
+      "learning_rate": 3.327011494252873e-05,
+      "loss": 0.1081,
+      "step": 2911
+    },
+    {
+      "epoch": 3.3754980801275085,
+      "grad_norm": 0.37320229411125183,
+      "learning_rate": 3.3264367816091954e-05,
+      "loss": 0.1177,
+      "step": 2912
+    },
+    {
+      "epoch": 3.3766572484242556,
+      "grad_norm": 0.3110753297805786,
+      "learning_rate": 3.3258620689655176e-05,
+      "loss": 0.1062,
+      "step": 2913
+    },
+    {
+      "epoch": 3.3778164167210027,
+      "grad_norm": 0.35673680901527405,
+      "learning_rate": 3.325287356321839e-05,
+      "loss": 0.1095,
+      "step": 2914
+    },
+    {
+      "epoch": 3.37897558501775,
+      "grad_norm": 0.3656083643436432,
+      "learning_rate": 3.324712643678161e-05,
+      "loss": 0.1075,
+      "step": 2915
+    },
+    {
+      "epoch": 3.380134753314497,
+      "grad_norm": 0.29794302582740784,
+      "learning_rate": 3.324137931034483e-05,
+      "loss": 0.1062,
+      "step": 2916
+    },
+    {
+      "epoch": 3.381293921611244,
+      "grad_norm": 0.3165527284145355,
+      "learning_rate": 3.323563218390805e-05,
+      "loss": 0.1035,
+      "step": 2917
+    },
+    {
+      "epoch": 3.382453089907991,
+      "grad_norm": 0.44483765959739685,
+      "learning_rate": 3.3229885057471264e-05,
+      "loss": 0.1039,
+      "step": 2918
+    },
+    {
+      "epoch": 3.383612258204738,
+      "grad_norm": 0.3703659474849701,
+      "learning_rate": 3.3224137931034486e-05,
+      "loss": 0.1176,
+      "step": 2919
+    },
+    {
+      "epoch": 3.384771426501485,
+      "grad_norm": 0.2532442510128021,
+      "learning_rate": 3.321839080459771e-05,
+      "loss": 0.1028,
+      "step": 2920
+    },
+    {
+      "epoch": 3.385930594798232,
+      "grad_norm": 0.2925904393196106,
+      "learning_rate": 3.321264367816092e-05,
+      "loss": 0.1037,
+      "step": 2921
+    },
+    {
+      "epoch": 3.3870897630949792,
+      "grad_norm": 0.45509782433509827,
+      "learning_rate": 3.320689655172414e-05,
+      "loss": 0.1159,
+      "step": 2922
+    },
+    {
+      "epoch": 3.3882489313917263,
+      "grad_norm": 0.36248907446861267,
+      "learning_rate": 3.320114942528736e-05,
+      "loss": 0.114,
+      "step": 2923
+    },
+    {
+      "epoch": 3.3894080996884735,
+      "grad_norm": 0.3373396098613739,
+      "learning_rate": 3.3195402298850574e-05,
+      "loss": 0.1131,
+      "step": 2924
+    },
+    {
+      "epoch": 3.3905672679852206,
+      "grad_norm": 0.44164901971817017,
+      "learning_rate": 3.3189655172413796e-05,
+      "loss": 0.125,
+      "step": 2925
+    },
+    {
+      "epoch": 3.3917264362819677,
+      "grad_norm": 0.32218971848487854,
+      "learning_rate": 3.318390804597701e-05,
+      "loss": 0.1103,
+      "step": 2926
+    },
+    {
+      "epoch": 3.392885604578715,
+      "grad_norm": 0.25145384669303894,
+      "learning_rate": 3.317816091954023e-05,
+      "loss": 0.1172,
+      "step": 2927
+    },
+    {
+      "epoch": 3.394044772875462,
+      "grad_norm": 0.3067503571510315,
+      "learning_rate": 3.3172413793103454e-05,
+      "loss": 0.1042,
+      "step": 2928
+    },
+    {
+      "epoch": 3.395203941172209,
+      "grad_norm": 0.31007614731788635,
+      "learning_rate": 3.316666666666667e-05,
+      "loss": 0.1022,
+      "step": 2929
+    },
+    {
+      "epoch": 3.396363109468956,
+      "grad_norm": 0.398503839969635,
+      "learning_rate": 3.3160919540229884e-05,
+      "loss": 0.1158,
+      "step": 2930
+    },
+    {
+      "epoch": 3.3975222777657033,
+      "grad_norm": 0.32212406396865845,
+      "learning_rate": 3.3155172413793105e-05,
+      "loss": 0.1178,
+      "step": 2931
+    },
+    {
+      "epoch": 3.3986814460624504,
+      "grad_norm": 0.40230342745780945,
+      "learning_rate": 3.314942528735632e-05,
+      "loss": 0.1165,
+      "step": 2932
+    },
+    {
+      "epoch": 3.3998406143591975,
+      "grad_norm": 0.46653199195861816,
+      "learning_rate": 3.314367816091954e-05,
+      "loss": 0.1123,
+      "step": 2933
+    },
+    {
+      "epoch": 3.400999782655944,
+      "grad_norm": 0.217983677983284,
+      "learning_rate": 3.3137931034482764e-05,
+      "loss": 0.0943,
+      "step": 2934
+    },
+    {
+      "epoch": 3.4021589509526913,
+      "grad_norm": 0.3591748774051666,
+      "learning_rate": 3.313218390804598e-05,
+      "loss": 0.1121,
+      "step": 2935
+    },
+    {
+      "epoch": 3.4033181192494384,
+      "grad_norm": 0.28786206245422363,
+      "learning_rate": 3.31264367816092e-05,
+      "loss": 0.104,
+      "step": 2936
+    },
+    {
+      "epoch": 3.4044772875461855,
+      "grad_norm": 0.23771996796131134,
+      "learning_rate": 3.3120689655172415e-05,
+      "loss": 0.1036,
+      "step": 2937
+    },
+    {
+      "epoch": 3.4056364558429326,
+      "grad_norm": 0.2682191729545593,
+      "learning_rate": 3.311494252873563e-05,
+      "loss": 0.1108,
+      "step": 2938
+    },
+    {
+      "epoch": 3.4067956241396797,
+      "grad_norm": 0.2601390480995178,
+      "learning_rate": 3.310919540229885e-05,
+      "loss": 0.1215,
+      "step": 2939
+    },
+    {
+      "epoch": 3.407954792436427,
+      "grad_norm": 0.3103736639022827,
+      "learning_rate": 3.310344827586207e-05,
+      "loss": 0.1083,
+      "step": 2940
+    },
+    {
+      "epoch": 3.409113960733174,
+      "grad_norm": 0.2568131983280182,
+      "learning_rate": 3.309770114942529e-05,
+      "loss": 0.107,
+      "step": 2941
+    },
+    {
+      "epoch": 3.410273129029921,
+      "grad_norm": 0.29044461250305176,
+      "learning_rate": 3.309195402298851e-05,
+      "loss": 0.1074,
+      "step": 2942
+    },
+    {
+      "epoch": 3.411432297326668,
+      "grad_norm": 0.30530881881713867,
+      "learning_rate": 3.3086206896551725e-05,
+      "loss": 0.1095,
+      "step": 2943
+    },
+    {
+      "epoch": 3.4125914656234153,
+      "grad_norm": 0.30023837089538574,
+      "learning_rate": 3.3080459770114946e-05,
+      "loss": 0.1057,
+      "step": 2944
+    },
+    {
+      "epoch": 3.4137506339201624,
+      "grad_norm": 0.3304160237312317,
+      "learning_rate": 3.307471264367816e-05,
+      "loss": 0.1015,
+      "step": 2945
+    },
+    {
+      "epoch": 3.4149098022169095,
+      "grad_norm": 0.3712501525878906,
+      "learning_rate": 3.3068965517241376e-05,
+      "loss": 0.1215,
+      "step": 2946
+    },
+    {
+      "epoch": 3.416068970513656,
+      "grad_norm": 0.36417120695114136,
+      "learning_rate": 3.30632183908046e-05,
+      "loss": 0.1185,
+      "step": 2947
+    },
+    {
+      "epoch": 3.4172281388104038,
+      "grad_norm": 0.401915580034256,
+      "learning_rate": 3.305747126436782e-05,
+      "loss": 0.1118,
+      "step": 2948
+    },
+    {
+      "epoch": 3.4183873071071504,
+      "grad_norm": 0.27606403827667236,
+      "learning_rate": 3.3051724137931035e-05,
+      "loss": 0.1035,
+      "step": 2949
+    },
+    {
+      "epoch": 3.4195464754038976,
+      "grad_norm": 0.28905513882637024,
+      "learning_rate": 3.3045977011494256e-05,
+      "loss": 0.1059,
+      "step": 2950
+    },
+    {
+      "epoch": 3.4207056437006447,
+      "grad_norm": 0.30320367217063904,
+      "learning_rate": 3.304022988505747e-05,
+      "loss": 0.1162,
+      "step": 2951
+    },
+    {
+      "epoch": 3.421864811997392,
+      "grad_norm": 0.36751824617385864,
+      "learning_rate": 3.303448275862069e-05,
+      "loss": 0.1222,
+      "step": 2952
+    },
+    {
+      "epoch": 3.423023980294139,
+      "grad_norm": 0.5594837665557861,
+      "learning_rate": 3.302873563218391e-05,
+      "loss": 0.0988,
+      "step": 2953
+    },
+    {
+      "epoch": 3.424183148590886,
+      "grad_norm": 0.2814250588417053,
+      "learning_rate": 3.302298850574713e-05,
+      "loss": 0.1089,
+      "step": 2954
+    },
+    {
+      "epoch": 3.425342316887633,
+      "grad_norm": 0.30620330572128296,
+      "learning_rate": 3.301724137931035e-05,
+      "loss": 0.1238,
+      "step": 2955
+    },
+    {
+      "epoch": 3.4265014851843802,
+      "grad_norm": 0.3259787857532501,
+      "learning_rate": 3.3011494252873566e-05,
+      "loss": 0.1061,
+      "step": 2956
+    },
+    {
+      "epoch": 3.4276606534811274,
+      "grad_norm": 0.3204762041568756,
+      "learning_rate": 3.300574712643678e-05,
+      "loss": 0.1081,
+      "step": 2957
+    },
+    {
+      "epoch": 3.4288198217778745,
+      "grad_norm": 0.32642972469329834,
+      "learning_rate": 3.3e-05,
+      "loss": 0.1097,
+      "step": 2958
+    },
+    {
+      "epoch": 3.4299789900746216,
+      "grad_norm": 0.3611741065979004,
+      "learning_rate": 3.299425287356322e-05,
+      "loss": 0.1153,
+      "step": 2959
+    },
+    {
+      "epoch": 3.4311381583713687,
+      "grad_norm": 0.24560561776161194,
+      "learning_rate": 3.298850574712644e-05,
+      "loss": 0.1122,
+      "step": 2960
+    },
+    {
+      "epoch": 3.432297326668116,
+      "grad_norm": 0.2226334512233734,
+      "learning_rate": 3.298275862068966e-05,
+      "loss": 0.1046,
+      "step": 2961
+    },
+    {
+      "epoch": 3.4334564949648625,
+      "grad_norm": 0.3255598545074463,
+      "learning_rate": 3.2977011494252876e-05,
+      "loss": 0.1003,
+      "step": 2962
+    },
+    {
+      "epoch": 3.4346156632616096,
+      "grad_norm": 0.2790594696998596,
+      "learning_rate": 3.29712643678161e-05,
+      "loss": 0.1162,
+      "step": 2963
+    },
+    {
+      "epoch": 3.4357748315583567,
+      "grad_norm": 0.3118756413459778,
+      "learning_rate": 3.296551724137931e-05,
+      "loss": 0.1025,
+      "step": 2964
+    },
+    {
+      "epoch": 3.436933999855104,
+      "grad_norm": 0.3635786473751068,
+      "learning_rate": 3.295977011494253e-05,
+      "loss": 0.1311,
+      "step": 2965
+    },
+    {
+      "epoch": 3.438093168151851,
+      "grad_norm": 0.2933517098426819,
+      "learning_rate": 3.295402298850575e-05,
+      "loss": 0.1107,
+      "step": 2966
+    },
+    {
+      "epoch": 3.439252336448598,
+      "grad_norm": 0.37896665930747986,
+      "learning_rate": 3.2948275862068964e-05,
+      "loss": 0.1058,
+      "step": 2967
+    },
+    {
+      "epoch": 3.440411504745345,
+      "grad_norm": 0.23080338537693024,
+      "learning_rate": 3.2942528735632185e-05,
+      "loss": 0.0987,
+      "step": 2968
+    },
+    {
+      "epoch": 3.4415706730420923,
+      "grad_norm": 0.262114554643631,
+      "learning_rate": 3.293678160919541e-05,
+      "loss": 0.1152,
+      "step": 2969
+    },
+    {
+      "epoch": 3.4427298413388394,
+      "grad_norm": 0.30394595861434937,
+      "learning_rate": 3.293103448275862e-05,
+      "loss": 0.1076,
+      "step": 2970
+    },
+    {
+      "epoch": 3.4438890096355865,
+      "grad_norm": 0.30665647983551025,
+      "learning_rate": 3.2925287356321844e-05,
+      "loss": 0.1169,
+      "step": 2971
+    },
+    {
+      "epoch": 3.4450481779323336,
+      "grad_norm": 0.35701873898506165,
+      "learning_rate": 3.291954022988506e-05,
+      "loss": 0.1186,
+      "step": 2972
+    },
+    {
+      "epoch": 3.4462073462290808,
+      "grad_norm": 0.3494682013988495,
+      "learning_rate": 3.2913793103448273e-05,
+      "loss": 0.1188,
+      "step": 2973
+    },
+    {
+      "epoch": 3.447366514525828,
+      "grad_norm": 0.28734609484672546,
+      "learning_rate": 3.2908045977011495e-05,
+      "loss": 0.1164,
+      "step": 2974
+    },
+    {
+      "epoch": 3.448525682822575,
+      "grad_norm": 0.25234439969062805,
+      "learning_rate": 3.290229885057472e-05,
+      "loss": 0.107,
+      "step": 2975
+    },
+    {
+      "epoch": 3.449684851119322,
+      "grad_norm": 0.3193553388118744,
+      "learning_rate": 3.289655172413793e-05,
+      "loss": 0.1101,
+      "step": 2976
+    },
+    {
+      "epoch": 3.4508440194160688,
+      "grad_norm": 0.32543477416038513,
+      "learning_rate": 3.2890804597701153e-05,
+      "loss": 0.1204,
+      "step": 2977
+    },
+    {
+      "epoch": 3.452003187712816,
+      "grad_norm": 0.34769558906555176,
+      "learning_rate": 3.288505747126437e-05,
+      "loss": 0.1161,
+      "step": 2978
+    },
+    {
+      "epoch": 3.453162356009563,
+      "grad_norm": 0.315849244594574,
+      "learning_rate": 3.287931034482758e-05,
+      "loss": 0.1074,
+      "step": 2979
+    },
+    {
+      "epoch": 3.45432152430631,
+      "grad_norm": 0.3821331262588501,
+      "learning_rate": 3.2873563218390805e-05,
+      "loss": 0.1193,
+      "step": 2980
+    },
+    {
+      "epoch": 3.4554806926030572,
+      "grad_norm": 0.2808358669281006,
+      "learning_rate": 3.2867816091954027e-05,
+      "loss": 0.1031,
+      "step": 2981
+    },
+    {
+      "epoch": 3.4566398608998044,
+      "grad_norm": 0.5840047001838684,
+      "learning_rate": 3.286206896551725e-05,
+      "loss": 0.1185,
+      "step": 2982
+    },
+    {
+      "epoch": 3.4577990291965515,
+      "grad_norm": 0.3200035095214844,
+      "learning_rate": 3.285632183908046e-05,
+      "loss": 0.1226,
+      "step": 2983
+    },
+    {
+      "epoch": 3.4589581974932986,
+      "grad_norm": 0.3244805335998535,
+      "learning_rate": 3.285057471264368e-05,
+      "loss": 0.1222,
+      "step": 2984
+    },
+    {
+      "epoch": 3.4601173657900457,
+      "grad_norm": 0.3481799364089966,
+      "learning_rate": 3.28448275862069e-05,
+      "loss": 0.1145,
+      "step": 2985
+    },
+    {
+      "epoch": 3.461276534086793,
+      "grad_norm": 0.40199366211891174,
+      "learning_rate": 3.2839080459770115e-05,
+      "loss": 0.1198,
+      "step": 2986
+    },
+    {
+      "epoch": 3.46243570238354,
+      "grad_norm": 0.297984778881073,
+      "learning_rate": 3.283333333333333e-05,
+      "loss": 0.1072,
+      "step": 2987
+    },
+    {
+      "epoch": 3.463594870680287,
+      "grad_norm": 0.272242933511734,
+      "learning_rate": 3.282758620689655e-05,
+      "loss": 0.1262,
+      "step": 2988
+    },
+    {
+      "epoch": 3.464754038977034,
+      "grad_norm": 0.5346304178237915,
+      "learning_rate": 3.282183908045977e-05,
+      "loss": 0.1111,
+      "step": 2989
+    },
+    {
+      "epoch": 3.465913207273781,
+      "grad_norm": 0.25313276052474976,
+      "learning_rate": 3.2816091954022995e-05,
+      "loss": 0.0966,
+      "step": 2990
+    },
+    {
+      "epoch": 3.4670723755705284,
+      "grad_norm": 0.2953258156776428,
+      "learning_rate": 3.281034482758621e-05,
+      "loss": 0.11,
+      "step": 2991
+    },
+    {
+      "epoch": 3.468231543867275,
+      "grad_norm": 0.2898904085159302,
+      "learning_rate": 3.2804597701149424e-05,
+      "loss": 0.1163,
+      "step": 2992
+    },
+    {
+      "epoch": 3.469390712164022,
+      "grad_norm": 0.2814153730869293,
+      "learning_rate": 3.2798850574712646e-05,
+      "loss": 0.1005,
+      "step": 2993
+    },
+    {
+      "epoch": 3.4705498804607693,
+      "grad_norm": 0.25759944319725037,
+      "learning_rate": 3.279310344827586e-05,
+      "loss": 0.1051,
+      "step": 2994
+    },
+    {
+      "epoch": 3.4717090487575164,
+      "grad_norm": 0.2664312720298767,
+      "learning_rate": 3.278735632183908e-05,
+      "loss": 0.1084,
+      "step": 2995
+    },
+    {
+      "epoch": 3.4728682170542635,
+      "grad_norm": 0.26857104897499084,
+      "learning_rate": 3.2781609195402304e-05,
+      "loss": 0.1059,
+      "step": 2996
+    },
+    {
+      "epoch": 3.4740273853510106,
+      "grad_norm": 0.6132874488830566,
+      "learning_rate": 3.277586206896552e-05,
+      "loss": 0.1036,
+      "step": 2997
+    },
+    {
+      "epoch": 3.4751865536477577,
+      "grad_norm": 0.2552327513694763,
+      "learning_rate": 3.2770114942528734e-05,
+      "loss": 0.0961,
+      "step": 2998
+    },
+    {
+      "epoch": 3.476345721944505,
+      "grad_norm": 0.3144066035747528,
+      "learning_rate": 3.2764367816091956e-05,
+      "loss": 0.1203,
+      "step": 2999
+    },
+    {
+      "epoch": 3.477504890241252,
+      "grad_norm": 0.2916877567768097,
+      "learning_rate": 3.275862068965517e-05,
+      "loss": 0.1169,
+      "step": 3000
+    },
+    {
+      "epoch": 3.478664058537999,
+      "grad_norm": 0.2755250036716461,
+      "learning_rate": 3.275287356321839e-05,
+      "loss": 0.1092,
+      "step": 3001
+    },
+    {
+      "epoch": 3.479823226834746,
+      "grad_norm": 0.30885928869247437,
+      "learning_rate": 3.2747126436781614e-05,
+      "loss": 0.1072,
+      "step": 3002
+    },
+    {
+      "epoch": 3.4809823951314933,
+      "grad_norm": 0.26637914776802063,
+      "learning_rate": 3.274137931034483e-05,
+      "loss": 0.1115,
+      "step": 3003
+    },
+    {
+      "epoch": 3.4821415634282404,
+      "grad_norm": 0.25272154808044434,
+      "learning_rate": 3.273563218390805e-05,
+      "loss": 0.1062,
+      "step": 3004
+    },
+    {
+      "epoch": 3.483300731724987,
+      "grad_norm": 0.24959425628185272,
+      "learning_rate": 3.2729885057471266e-05,
+      "loss": 0.1034,
+      "step": 3005
+    },
+    {
+      "epoch": 3.4844599000217342,
+      "grad_norm": 0.29696011543273926,
+      "learning_rate": 3.272413793103448e-05,
+      "loss": 0.129,
+      "step": 3006
+    },
+    {
+      "epoch": 3.4856190683184813,
+      "grad_norm": 0.24814733862876892,
+      "learning_rate": 3.27183908045977e-05,
+      "loss": 0.1057,
+      "step": 3007
+    },
+    {
+      "epoch": 3.4867782366152285,
+      "grad_norm": 0.2551887035369873,
+      "learning_rate": 3.271264367816092e-05,
+      "loss": 0.1072,
+      "step": 3008
+    },
+    {
+      "epoch": 3.4879374049119756,
+      "grad_norm": 0.2791741192340851,
+      "learning_rate": 3.2706896551724145e-05,
+      "loss": 0.1059,
+      "step": 3009
+    },
+    {
+      "epoch": 3.4890965732087227,
+      "grad_norm": 0.3654879033565521,
+      "learning_rate": 3.270114942528736e-05,
+      "loss": 0.1077,
+      "step": 3010
+    },
+    {
+      "epoch": 3.49025574150547,
+      "grad_norm": 0.21658289432525635,
+      "learning_rate": 3.2695402298850575e-05,
+      "loss": 0.0997,
+      "step": 3011
+    },
+    {
+      "epoch": 3.491414909802217,
+      "grad_norm": 0.2808610796928406,
+      "learning_rate": 3.26896551724138e-05,
+      "loss": 0.1127,
+      "step": 3012
+    },
+    {
+      "epoch": 3.492574078098964,
+      "grad_norm": 0.327570378780365,
+      "learning_rate": 3.268390804597701e-05,
+      "loss": 0.1062,
+      "step": 3013
+    },
+    {
+      "epoch": 3.493733246395711,
+      "grad_norm": 0.24436767399311066,
+      "learning_rate": 3.267816091954023e-05,
+      "loss": 0.1024,
+      "step": 3014
+    },
+    {
+      "epoch": 3.4948924146924583,
+      "grad_norm": 0.33249321579933167,
+      "learning_rate": 3.267241379310345e-05,
+      "loss": 0.1109,
+      "step": 3015
+    },
+    {
+      "epoch": 3.4960515829892054,
+      "grad_norm": 0.31921130418777466,
+      "learning_rate": 3.266666666666667e-05,
+      "loss": 0.1264,
+      "step": 3016
+    },
+    {
+      "epoch": 3.4972107512859525,
+      "grad_norm": 0.2958783805370331,
+      "learning_rate": 3.2660919540229885e-05,
+      "loss": 0.1125,
+      "step": 3017
+    },
+    {
+      "epoch": 3.4983699195826996,
+      "grad_norm": 0.2842378318309784,
+      "learning_rate": 3.265517241379311e-05,
+      "loss": 0.1187,
+      "step": 3018
+    },
+    {
+      "epoch": 3.4995290878794467,
+      "grad_norm": 0.37194153666496277,
+      "learning_rate": 3.264942528735632e-05,
+      "loss": 0.1131,
+      "step": 3019
+    },
+    {
+      "epoch": 3.5006882561761934,
+      "grad_norm": 0.342540979385376,
+      "learning_rate": 3.264367816091954e-05,
+      "loss": 0.1059,
+      "step": 3020
+    },
+    {
+      "epoch": 3.5018474244729405,
+      "grad_norm": 0.517421305179596,
+      "learning_rate": 3.263793103448276e-05,
+      "loss": 0.1226,
+      "step": 3021
+    },
+    {
+      "epoch": 3.5030065927696876,
+      "grad_norm": 0.2997574508190155,
+      "learning_rate": 3.263218390804598e-05,
+      "loss": 0.0962,
+      "step": 3022
+    },
+    {
+      "epoch": 3.5041657610664347,
+      "grad_norm": 0.3273774981498718,
+      "learning_rate": 3.26264367816092e-05,
+      "loss": 0.1205,
+      "step": 3023
+    },
+    {
+      "epoch": 3.505324929363182,
+      "grad_norm": 0.25956934690475464,
+      "learning_rate": 3.2620689655172416e-05,
+      "loss": 0.1186,
+      "step": 3024
+    },
+    {
+      "epoch": 3.506484097659929,
+      "grad_norm": 0.306949645280838,
+      "learning_rate": 3.261494252873563e-05,
+      "loss": 0.1115,
+      "step": 3025
+    },
+    {
+      "epoch": 3.507643265956676,
+      "grad_norm": 0.3232230842113495,
+      "learning_rate": 3.260919540229885e-05,
+      "loss": 0.1155,
+      "step": 3026
+    },
+    {
+      "epoch": 3.508802434253423,
+      "grad_norm": 0.2607809007167816,
+      "learning_rate": 3.260344827586207e-05,
+      "loss": 0.1142,
+      "step": 3027
+    },
+    {
+      "epoch": 3.5099616025501703,
+      "grad_norm": 0.31110328435897827,
+      "learning_rate": 3.259770114942529e-05,
+      "loss": 0.1102,
+      "step": 3028
+    },
+    {
+      "epoch": 3.5111207708469174,
+      "grad_norm": 0.24277092516422272,
+      "learning_rate": 3.2591954022988505e-05,
+      "loss": 0.1149,
+      "step": 3029
+    },
+    {
+      "epoch": 3.5122799391436645,
+      "grad_norm": 0.29219481348991394,
+      "learning_rate": 3.2586206896551726e-05,
+      "loss": 0.1141,
+      "step": 3030
+    },
+    {
+      "epoch": 3.5134391074404117,
+      "grad_norm": 0.24297522008419037,
+      "learning_rate": 3.258045977011495e-05,
+      "loss": 0.1015,
+      "step": 3031
+    },
+    {
+      "epoch": 3.5145982757371588,
+      "grad_norm": 0.2821162939071655,
+      "learning_rate": 3.257471264367816e-05,
+      "loss": 0.1153,
+      "step": 3032
+    },
+    {
+      "epoch": 3.5157574440339054,
+      "grad_norm": 0.5800468325614929,
+      "learning_rate": 3.256896551724138e-05,
+      "loss": 0.1193,
+      "step": 3033
+    },
+    {
+      "epoch": 3.516916612330653,
+      "grad_norm": 0.3560374677181244,
+      "learning_rate": 3.25632183908046e-05,
+      "loss": 0.1076,
+      "step": 3034
+    },
+    {
+      "epoch": 3.5180757806273997,
+      "grad_norm": 0.4053660035133362,
+      "learning_rate": 3.2557471264367814e-05,
+      "loss": 0.1098,
+      "step": 3035
+    },
+    {
+      "epoch": 3.519234948924147,
+      "grad_norm": 0.36452099680900574,
+      "learning_rate": 3.2551724137931036e-05,
+      "loss": 0.1074,
+      "step": 3036
+    },
+    {
+      "epoch": 3.520394117220894,
+      "grad_norm": 0.2924714982509613,
+      "learning_rate": 3.254597701149426e-05,
+      "loss": 0.1131,
+      "step": 3037
+    },
+    {
+      "epoch": 3.521553285517641,
+      "grad_norm": 0.30950403213500977,
+      "learning_rate": 3.254022988505747e-05,
+      "loss": 0.1121,
+      "step": 3038
+    },
+    {
+      "epoch": 3.522712453814388,
+      "grad_norm": 0.3289961814880371,
+      "learning_rate": 3.2534482758620694e-05,
+      "loss": 0.1153,
+      "step": 3039
+    },
+    {
+      "epoch": 3.5238716221111352,
+      "grad_norm": 0.2730542719364166,
+      "learning_rate": 3.252873563218391e-05,
+      "loss": 0.1181,
+      "step": 3040
+    },
+    {
+      "epoch": 3.5250307904078824,
+      "grad_norm": 0.4356801211833954,
+      "learning_rate": 3.2522988505747124e-05,
+      "loss": 0.1049,
+      "step": 3041
+    },
+    {
+      "epoch": 3.5261899587046295,
+      "grad_norm": 0.2651293873786926,
+      "learning_rate": 3.2517241379310346e-05,
+      "loss": 0.1194,
+      "step": 3042
+    },
+    {
+      "epoch": 3.5273491270013766,
+      "grad_norm": 0.3155745267868042,
+      "learning_rate": 3.251149425287357e-05,
+      "loss": 0.1106,
+      "step": 3043
+    },
+    {
+      "epoch": 3.5285082952981237,
+      "grad_norm": 0.2664744257926941,
+      "learning_rate": 3.250574712643678e-05,
+      "loss": 0.1166,
+      "step": 3044
+    },
+    {
+      "epoch": 3.529667463594871,
+      "grad_norm": 0.3852986693382263,
+      "learning_rate": 3.2500000000000004e-05,
+      "loss": 0.1129,
+      "step": 3045
+    },
+    {
+      "epoch": 3.5308266318916175,
+      "grad_norm": 0.31656765937805176,
+      "learning_rate": 3.249425287356322e-05,
+      "loss": 0.1202,
+      "step": 3046
+    },
+    {
+      "epoch": 3.531985800188365,
+      "grad_norm": 0.2873196303844452,
+      "learning_rate": 3.248850574712644e-05,
+      "loss": 0.1228,
+      "step": 3047
+    },
+    {
+      "epoch": 3.5331449684851117,
+      "grad_norm": 0.29903656244277954,
+      "learning_rate": 3.2482758620689655e-05,
+      "loss": 0.1052,
+      "step": 3048
+    },
+    {
+      "epoch": 3.5343041367818593,
+      "grad_norm": 0.2554726302623749,
+      "learning_rate": 3.247701149425287e-05,
+      "loss": 0.1036,
+      "step": 3049
+    },
+    {
+      "epoch": 3.535463305078606,
+      "grad_norm": 0.33585837483406067,
+      "learning_rate": 3.24712643678161e-05,
+      "loss": 0.117,
+      "step": 3050
+    },
+    {
+      "epoch": 3.536622473375353,
+      "grad_norm": 0.28238433599472046,
+      "learning_rate": 3.2465517241379314e-05,
+      "loss": 0.1106,
+      "step": 3051
+    },
+    {
+      "epoch": 3.5377816416721,
+      "grad_norm": 0.33771008253097534,
+      "learning_rate": 3.245977011494253e-05,
+      "loss": 0.1117,
+      "step": 3052
+    },
+    {
+      "epoch": 3.5389408099688473,
+      "grad_norm": 0.34234920144081116,
+      "learning_rate": 3.245402298850575e-05,
+      "loss": 0.1119,
+      "step": 3053
+    },
+    {
+      "epoch": 3.5400999782655944,
+      "grad_norm": 0.29577329754829407,
+      "learning_rate": 3.2448275862068965e-05,
+      "loss": 0.1119,
+      "step": 3054
+    },
+    {
+      "epoch": 3.5412591465623415,
+      "grad_norm": 0.5385765433311462,
+      "learning_rate": 3.244252873563218e-05,
+      "loss": 0.1013,
+      "step": 3055
+    },
+    {
+      "epoch": 3.5424183148590886,
+      "grad_norm": 0.31815600395202637,
+      "learning_rate": 3.24367816091954e-05,
+      "loss": 0.1136,
+      "step": 3056
+    },
+    {
+      "epoch": 3.5435774831558358,
+      "grad_norm": 0.3204317092895508,
+      "learning_rate": 3.2431034482758623e-05,
+      "loss": 0.1209,
+      "step": 3057
+    },
+    {
+      "epoch": 3.544736651452583,
+      "grad_norm": 0.34729644656181335,
+      "learning_rate": 3.2425287356321845e-05,
+      "loss": 0.1186,
+      "step": 3058
+    },
+    {
+      "epoch": 3.54589581974933,
+      "grad_norm": 0.35338544845581055,
+      "learning_rate": 3.241954022988506e-05,
+      "loss": 0.1165,
+      "step": 3059
+    },
+    {
+      "epoch": 3.547054988046077,
+      "grad_norm": 0.36400482058525085,
+      "learning_rate": 3.2413793103448275e-05,
+      "loss": 0.1285,
+      "step": 3060
+    },
+    {
+      "epoch": 3.5482141563428238,
+      "grad_norm": 0.2854449450969696,
+      "learning_rate": 3.2408045977011497e-05,
+      "loss": 0.1064,
+      "step": 3061
+    },
+    {
+      "epoch": 3.5493733246395713,
+      "grad_norm": 0.2662312090396881,
+      "learning_rate": 3.240229885057471e-05,
+      "loss": 0.105,
+      "step": 3062
+    },
+    {
+      "epoch": 3.550532492936318,
+      "grad_norm": 0.30098652839660645,
+      "learning_rate": 3.239655172413793e-05,
+      "loss": 0.1025,
+      "step": 3063
+    },
+    {
+      "epoch": 3.551691661233065,
+      "grad_norm": 0.35262545943260193,
+      "learning_rate": 3.2390804597701155e-05,
+      "loss": 0.1148,
+      "step": 3064
+    },
+    {
+      "epoch": 3.5528508295298122,
+      "grad_norm": 0.2994409501552582,
+      "learning_rate": 3.238505747126437e-05,
+      "loss": 0.1072,
+      "step": 3065
+    },
+    {
+      "epoch": 3.5540099978265594,
+      "grad_norm": 0.3354470133781433,
+      "learning_rate": 3.237931034482759e-05,
+      "loss": 0.1112,
+      "step": 3066
+    },
+    {
+      "epoch": 3.5551691661233065,
+      "grad_norm": 0.30493679642677307,
+      "learning_rate": 3.2373563218390806e-05,
+      "loss": 0.1071,
+      "step": 3067
+    },
+    {
+      "epoch": 3.5563283344200536,
+      "grad_norm": 0.24388879537582397,
+      "learning_rate": 3.236781609195402e-05,
+      "loss": 0.1135,
+      "step": 3068
+    },
+    {
+      "epoch": 3.5574875027168007,
+      "grad_norm": 0.37016189098358154,
+      "learning_rate": 3.236206896551724e-05,
+      "loss": 0.1184,
+      "step": 3069
+    },
+    {
+      "epoch": 3.558646671013548,
+      "grad_norm": 0.3760676681995392,
+      "learning_rate": 3.2356321839080465e-05,
+      "loss": 0.1099,
+      "step": 3070
+    },
+    {
+      "epoch": 3.559805839310295,
+      "grad_norm": 0.25357261300086975,
+      "learning_rate": 3.235057471264368e-05,
+      "loss": 0.1016,
+      "step": 3071
+    },
+    {
+      "epoch": 3.560965007607042,
+      "grad_norm": 0.3364717960357666,
+      "learning_rate": 3.23448275862069e-05,
+      "loss": 0.1095,
+      "step": 3072
+    },
+    {
+      "epoch": 3.562124175903789,
+      "grad_norm": 0.316201776266098,
+      "learning_rate": 3.2339080459770116e-05,
+      "loss": 0.1217,
+      "step": 3073
+    },
+    {
+      "epoch": 3.5632833442005363,
+      "grad_norm": 0.2470405250787735,
+      "learning_rate": 3.233333333333333e-05,
+      "loss": 0.1078,
+      "step": 3074
+    },
+    {
+      "epoch": 3.5644425124972834,
+      "grad_norm": 0.25586774945259094,
+      "learning_rate": 3.232758620689655e-05,
+      "loss": 0.1138,
+      "step": 3075
+    },
+    {
+      "epoch": 3.56560168079403,
+      "grad_norm": 0.37907102704048157,
+      "learning_rate": 3.232183908045977e-05,
+      "loss": 0.1062,
+      "step": 3076
+    },
+    {
+      "epoch": 3.5667608490907776,
+      "grad_norm": 0.36059078574180603,
+      "learning_rate": 3.231609195402299e-05,
+      "loss": 0.1178,
+      "step": 3077
+    },
+    {
+      "epoch": 3.5679200173875243,
+      "grad_norm": 0.24348652362823486,
+      "learning_rate": 3.231034482758621e-05,
+      "loss": 0.1101,
+      "step": 3078
+    },
+    {
+      "epoch": 3.5690791856842714,
+      "grad_norm": 0.31023815274238586,
+      "learning_rate": 3.2304597701149426e-05,
+      "loss": 0.1072,
+      "step": 3079
+    },
+    {
+      "epoch": 3.5702383539810185,
+      "grad_norm": 0.2670823931694031,
+      "learning_rate": 3.229885057471265e-05,
+      "loss": 0.1137,
+      "step": 3080
+    },
+    {
+      "epoch": 3.5713975222777656,
+      "grad_norm": 0.4615989029407501,
+      "learning_rate": 3.229310344827586e-05,
+      "loss": 0.1186,
+      "step": 3081
+    },
+    {
+      "epoch": 3.5725566905745127,
+      "grad_norm": 0.2701643407344818,
+      "learning_rate": 3.228735632183908e-05,
+      "loss": 0.1039,
+      "step": 3082
+    },
+    {
+      "epoch": 3.57371585887126,
+      "grad_norm": 0.28233602643013,
+      "learning_rate": 3.22816091954023e-05,
+      "loss": 0.1227,
+      "step": 3083
+    },
+    {
+      "epoch": 3.574875027168007,
+      "grad_norm": 0.3075142204761505,
+      "learning_rate": 3.227586206896552e-05,
+      "loss": 0.115,
+      "step": 3084
+    },
+    {
+      "epoch": 3.576034195464754,
+      "grad_norm": 0.31267890334129333,
+      "learning_rate": 3.227011494252874e-05,
+      "loss": 0.1172,
+      "step": 3085
+    },
+    {
+      "epoch": 3.577193363761501,
+      "grad_norm": 0.27376192808151245,
+      "learning_rate": 3.226436781609196e-05,
+      "loss": 0.1111,
+      "step": 3086
+    },
+    {
+      "epoch": 3.5783525320582483,
+      "grad_norm": 0.2703624963760376,
+      "learning_rate": 3.225862068965517e-05,
+      "loss": 0.1122,
+      "step": 3087
+    },
+    {
+      "epoch": 3.5795117003549954,
+      "grad_norm": 0.32042473554611206,
+      "learning_rate": 3.2252873563218394e-05,
+      "loss": 0.1001,
+      "step": 3088
+    },
+    {
+      "epoch": 3.580670868651742,
+      "grad_norm": 0.28498318791389465,
+      "learning_rate": 3.224712643678161e-05,
+      "loss": 0.1022,
+      "step": 3089
+    },
+    {
+      "epoch": 3.5818300369484897,
+      "grad_norm": 0.2931499183177948,
+      "learning_rate": 3.2241379310344824e-05,
+      "loss": 0.1109,
+      "step": 3090
+    },
+    {
+      "epoch": 3.5829892052452363,
+      "grad_norm": 0.34564685821533203,
+      "learning_rate": 3.223563218390805e-05,
+      "loss": 0.1222,
+      "step": 3091
+    },
+    {
+      "epoch": 3.584148373541984,
+      "grad_norm": 0.40452006459236145,
+      "learning_rate": 3.222988505747127e-05,
+      "loss": 0.1101,
+      "step": 3092
+    },
+    {
+      "epoch": 3.5853075418387306,
+      "grad_norm": 0.40536507964134216,
+      "learning_rate": 3.222413793103448e-05,
+      "loss": 0.1166,
+      "step": 3093
+    },
+    {
+      "epoch": 3.5864667101354777,
+      "grad_norm": 0.30855557322502136,
+      "learning_rate": 3.2218390804597704e-05,
+      "loss": 0.1273,
+      "step": 3094
+    },
+    {
+      "epoch": 3.587625878432225,
+      "grad_norm": 0.2754945755004883,
+      "learning_rate": 3.221264367816092e-05,
+      "loss": 0.1053,
+      "step": 3095
+    },
+    {
+      "epoch": 3.588785046728972,
+      "grad_norm": 0.3107283413410187,
+      "learning_rate": 3.220689655172414e-05,
+      "loss": 0.1072,
+      "step": 3096
+    },
+    {
+      "epoch": 3.589944215025719,
+      "grad_norm": 0.30340340733528137,
+      "learning_rate": 3.2201149425287355e-05,
+      "loss": 0.1079,
+      "step": 3097
+    },
+    {
+      "epoch": 3.591103383322466,
+      "grad_norm": 0.32089850306510925,
+      "learning_rate": 3.219540229885058e-05,
+      "loss": 0.1076,
+      "step": 3098
+    },
+    {
+      "epoch": 3.5922625516192133,
+      "grad_norm": 0.3080294728279114,
+      "learning_rate": 3.21896551724138e-05,
+      "loss": 0.1093,
+      "step": 3099
+    },
+    {
+      "epoch": 3.5934217199159604,
+      "grad_norm": 0.28277286887168884,
+      "learning_rate": 3.218390804597701e-05,
+      "loss": 0.1045,
+      "step": 3100
+    },
+    {
+      "epoch": 3.5945808882127075,
+      "grad_norm": 0.30949509143829346,
+      "learning_rate": 3.217816091954023e-05,
+      "loss": 0.1089,
+      "step": 3101
+    },
+    {
+      "epoch": 3.5957400565094546,
+      "grad_norm": 0.24771907925605774,
+      "learning_rate": 3.217241379310345e-05,
+      "loss": 0.1061,
+      "step": 3102
+    },
+    {
+      "epoch": 3.5968992248062017,
+      "grad_norm": 0.37252581119537354,
+      "learning_rate": 3.2166666666666665e-05,
+      "loss": 0.12,
+      "step": 3103
+    },
+    {
+      "epoch": 3.5980583931029484,
+      "grad_norm": 0.3569035232067108,
+      "learning_rate": 3.2160919540229886e-05,
+      "loss": 0.1188,
+      "step": 3104
+    },
+    {
+      "epoch": 3.599217561399696,
+      "grad_norm": 0.46223220229148865,
+      "learning_rate": 3.215517241379311e-05,
+      "loss": 0.1251,
+      "step": 3105
+    },
+    {
+      "epoch": 3.6003767296964426,
+      "grad_norm": 0.42109113931655884,
+      "learning_rate": 3.214942528735632e-05,
+      "loss": 0.1117,
+      "step": 3106
+    },
+    {
+      "epoch": 3.6015358979931897,
+      "grad_norm": 0.3445476293563843,
+      "learning_rate": 3.2143678160919545e-05,
+      "loss": 0.1015,
+      "step": 3107
+    },
+    {
+      "epoch": 3.602695066289937,
+      "grad_norm": 0.3141631484031677,
+      "learning_rate": 3.213793103448276e-05,
+      "loss": 0.1086,
+      "step": 3108
+    },
+    {
+      "epoch": 3.603854234586684,
+      "grad_norm": 0.4695309102535248,
+      "learning_rate": 3.2132183908045974e-05,
+      "loss": 0.1293,
+      "step": 3109
+    },
+    {
+      "epoch": 3.605013402883431,
+      "grad_norm": 0.2591305077075958,
+      "learning_rate": 3.2126436781609196e-05,
+      "loss": 0.1013,
+      "step": 3110
+    },
+    {
+      "epoch": 3.606172571180178,
+      "grad_norm": 0.26712268590927124,
+      "learning_rate": 3.212068965517242e-05,
+      "loss": 0.1071,
+      "step": 3111
+    },
+    {
+      "epoch": 3.6073317394769253,
+      "grad_norm": 0.30240774154663086,
+      "learning_rate": 3.211494252873564e-05,
+      "loss": 0.1096,
+      "step": 3112
+    },
+    {
+      "epoch": 3.6084909077736724,
+      "grad_norm": 0.3110927641391754,
+      "learning_rate": 3.2109195402298854e-05,
+      "loss": 0.1147,
+      "step": 3113
+    },
+    {
+      "epoch": 3.6096500760704195,
+      "grad_norm": 0.3136695325374603,
+      "learning_rate": 3.210344827586207e-05,
+      "loss": 0.1206,
+      "step": 3114
+    },
+    {
+      "epoch": 3.6108092443671667,
+      "grad_norm": 0.23492947220802307,
+      "learning_rate": 3.209770114942529e-05,
+      "loss": 0.1091,
+      "step": 3115
+    },
+    {
+      "epoch": 3.6119684126639138,
+      "grad_norm": 0.29082900285720825,
+      "learning_rate": 3.2091954022988506e-05,
+      "loss": 0.1052,
+      "step": 3116
+    },
+    {
+      "epoch": 3.613127580960661,
+      "grad_norm": 0.2898150086402893,
+      "learning_rate": 3.208620689655172e-05,
+      "loss": 0.1239,
+      "step": 3117
+    },
+    {
+      "epoch": 3.614286749257408,
+      "grad_norm": 0.25189781188964844,
+      "learning_rate": 3.208045977011494e-05,
+      "loss": 0.1027,
+      "step": 3118
+    },
+    {
+      "epoch": 3.6154459175541547,
+      "grad_norm": 0.2974126636981964,
+      "learning_rate": 3.2074712643678164e-05,
+      "loss": 0.0982,
+      "step": 3119
+    },
+    {
+      "epoch": 3.6166050858509022,
+      "grad_norm": 0.2558945119380951,
+      "learning_rate": 3.206896551724138e-05,
+      "loss": 0.1072,
+      "step": 3120
+    },
+    {
+      "epoch": 3.617764254147649,
+      "grad_norm": 0.2936224937438965,
+      "learning_rate": 3.20632183908046e-05,
+      "loss": 0.1153,
+      "step": 3121
+    },
+    {
+      "epoch": 3.618923422444396,
+      "grad_norm": 0.24101215600967407,
+      "learning_rate": 3.2057471264367816e-05,
+      "loss": 0.1005,
+      "step": 3122
+    },
+    {
+      "epoch": 3.620082590741143,
+      "grad_norm": 0.3117283582687378,
+      "learning_rate": 3.205172413793104e-05,
+      "loss": 0.1159,
+      "step": 3123
+    },
+    {
+      "epoch": 3.6212417590378903,
+      "grad_norm": 0.3123790919780731,
+      "learning_rate": 3.204597701149425e-05,
+      "loss": 0.1102,
+      "step": 3124
+    },
+    {
+      "epoch": 3.6224009273346374,
+      "grad_norm": 0.32129356265068054,
+      "learning_rate": 3.2040229885057474e-05,
+      "loss": 0.1247,
+      "step": 3125
+    },
+    {
+      "epoch": 3.6235600956313845,
+      "grad_norm": 0.287790983915329,
+      "learning_rate": 3.2034482758620696e-05,
+      "loss": 0.1225,
+      "step": 3126
+    },
+    {
+      "epoch": 3.6247192639281316,
+      "grad_norm": 0.33491361141204834,
+      "learning_rate": 3.202873563218391e-05,
+      "loss": 0.1189,
+      "step": 3127
+    },
+    {
+      "epoch": 3.6258784322248787,
+      "grad_norm": 0.3275797963142395,
+      "learning_rate": 3.2022988505747125e-05,
+      "loss": 0.1246,
+      "step": 3128
+    },
+    {
+      "epoch": 3.627037600521626,
+      "grad_norm": 0.30787503719329834,
+      "learning_rate": 3.201724137931035e-05,
+      "loss": 0.1204,
+      "step": 3129
+    },
+    {
+      "epoch": 3.628196768818373,
+      "grad_norm": 0.254406601190567,
+      "learning_rate": 3.201149425287356e-05,
+      "loss": 0.1033,
+      "step": 3130
+    },
+    {
+      "epoch": 3.62935593711512,
+      "grad_norm": 0.2660934329032898,
+      "learning_rate": 3.2005747126436784e-05,
+      "loss": 0.1096,
+      "step": 3131
+    },
+    {
+      "epoch": 3.6305151054118667,
+      "grad_norm": 0.30187851190567017,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.1064,
+      "step": 3132
+    },
+    {
+      "epoch": 3.6316742737086143,
+      "grad_norm": 0.2814703583717346,
+      "learning_rate": 3.199425287356322e-05,
+      "loss": 0.1209,
+      "step": 3133
+    },
+    {
+      "epoch": 3.632833442005361,
+      "grad_norm": 0.31418412923812866,
+      "learning_rate": 3.198850574712644e-05,
+      "loss": 0.1147,
+      "step": 3134
+    },
+    {
+      "epoch": 3.633992610302108,
+      "grad_norm": 0.25230860710144043,
+      "learning_rate": 3.198275862068966e-05,
+      "loss": 0.1094,
+      "step": 3135
+    },
+    {
+      "epoch": 3.635151778598855,
+      "grad_norm": 0.2896537482738495,
+      "learning_rate": 3.197701149425287e-05,
+      "loss": 0.1166,
+      "step": 3136
+    },
+    {
+      "epoch": 3.6363109468956023,
+      "grad_norm": 0.2596437931060791,
+      "learning_rate": 3.1971264367816093e-05,
+      "loss": 0.1141,
+      "step": 3137
+    },
+    {
+      "epoch": 3.6374701151923494,
+      "grad_norm": 0.29989323019981384,
+      "learning_rate": 3.196551724137931e-05,
+      "loss": 0.1154,
+      "step": 3138
+    },
+    {
+      "epoch": 3.6386292834890965,
+      "grad_norm": 0.303406685590744,
+      "learning_rate": 3.195977011494253e-05,
+      "loss": 0.1147,
+      "step": 3139
+    },
+    {
+      "epoch": 3.6397884517858436,
+      "grad_norm": 0.2839892506599426,
+      "learning_rate": 3.195402298850575e-05,
+      "loss": 0.1164,
+      "step": 3140
+    },
+    {
+      "epoch": 3.6409476200825908,
+      "grad_norm": 0.29430249333381653,
+      "learning_rate": 3.1948275862068967e-05,
+      "loss": 0.1083,
+      "step": 3141
+    },
+    {
+      "epoch": 3.642106788379338,
+      "grad_norm": 0.24778763949871063,
+      "learning_rate": 3.194252873563219e-05,
+      "loss": 0.1037,
+      "step": 3142
+    },
+    {
+      "epoch": 3.643265956676085,
+      "grad_norm": 0.3377165198326111,
+      "learning_rate": 3.19367816091954e-05,
+      "loss": 0.1178,
+      "step": 3143
+    },
+    {
+      "epoch": 3.644425124972832,
+      "grad_norm": 0.38875332474708557,
+      "learning_rate": 3.193103448275862e-05,
+      "loss": 0.1159,
+      "step": 3144
+    },
+    {
+      "epoch": 3.6455842932695792,
+      "grad_norm": 0.3728177547454834,
+      "learning_rate": 3.192528735632184e-05,
+      "loss": 0.1088,
+      "step": 3145
+    },
+    {
+      "epoch": 3.6467434615663263,
+      "grad_norm": 0.31668561697006226,
+      "learning_rate": 3.191954022988506e-05,
+      "loss": 0.1055,
+      "step": 3146
+    },
+    {
+      "epoch": 3.647902629863073,
+      "grad_norm": 0.3226422965526581,
+      "learning_rate": 3.1913793103448276e-05,
+      "loss": 0.1252,
+      "step": 3147
+    },
+    {
+      "epoch": 3.6490617981598206,
+      "grad_norm": 0.31225472688674927,
+      "learning_rate": 3.19080459770115e-05,
+      "loss": 0.1131,
+      "step": 3148
+    },
+    {
+      "epoch": 3.6502209664565672,
+      "grad_norm": 0.3561670482158661,
+      "learning_rate": 3.190229885057471e-05,
+      "loss": 0.1341,
+      "step": 3149
+    },
+    {
+      "epoch": 3.6513801347533144,
+      "grad_norm": 0.29987701773643494,
+      "learning_rate": 3.1896551724137935e-05,
+      "loss": 0.1263,
+      "step": 3150
+    },
+    {
+      "epoch": 3.6525393030500615,
+      "grad_norm": 0.2811523675918579,
+      "learning_rate": 3.189080459770115e-05,
+      "loss": 0.1225,
+      "step": 3151
+    },
+    {
+      "epoch": 3.6536984713468086,
+      "grad_norm": 0.31684601306915283,
+      "learning_rate": 3.188505747126437e-05,
+      "loss": 0.1086,
+      "step": 3152
+    },
+    {
+      "epoch": 3.6548576396435557,
+      "grad_norm": 0.28780969977378845,
+      "learning_rate": 3.187931034482759e-05,
+      "loss": 0.1085,
+      "step": 3153
+    },
+    {
+      "epoch": 3.656016807940303,
+      "grad_norm": 0.31450673937797546,
+      "learning_rate": 3.187356321839081e-05,
+      "loss": 0.109,
+      "step": 3154
+    },
+    {
+      "epoch": 3.65717597623705,
+      "grad_norm": 0.2759770452976227,
+      "learning_rate": 3.186781609195402e-05,
+      "loss": 0.1049,
+      "step": 3155
+    },
+    {
+      "epoch": 3.658335144533797,
+      "grad_norm": 0.3448818624019623,
+      "learning_rate": 3.1862068965517244e-05,
+      "loss": 0.1283,
+      "step": 3156
+    },
+    {
+      "epoch": 3.659494312830544,
+      "grad_norm": 0.2501785457134247,
+      "learning_rate": 3.185632183908046e-05,
+      "loss": 0.1056,
+      "step": 3157
+    },
+    {
+      "epoch": 3.6606534811272913,
+      "grad_norm": 0.288875550031662,
+      "learning_rate": 3.1850574712643674e-05,
+      "loss": 0.1018,
+      "step": 3158
+    },
+    {
+      "epoch": 3.6618126494240384,
+      "grad_norm": 0.29204368591308594,
+      "learning_rate": 3.1844827586206896e-05,
+      "loss": 0.1082,
+      "step": 3159
+    },
+    {
+      "epoch": 3.662971817720785,
+      "grad_norm": 0.3856884837150574,
+      "learning_rate": 3.183908045977012e-05,
+      "loss": 0.0979,
+      "step": 3160
+    },
+    {
+      "epoch": 3.6641309860175326,
+      "grad_norm": 0.30575424432754517,
+      "learning_rate": 3.183333333333334e-05,
+      "loss": 0.1151,
+      "step": 3161
+    },
+    {
+      "epoch": 3.6652901543142793,
+      "grad_norm": 0.3039374053478241,
+      "learning_rate": 3.1827586206896554e-05,
+      "loss": 0.1203,
+      "step": 3162
+    },
+    {
+      "epoch": 3.666449322611027,
+      "grad_norm": 0.2760923206806183,
+      "learning_rate": 3.182183908045977e-05,
+      "loss": 0.1114,
+      "step": 3163
+    },
+    {
+      "epoch": 3.6676084909077735,
+      "grad_norm": 0.3372345566749573,
+      "learning_rate": 3.181609195402299e-05,
+      "loss": 0.1204,
+      "step": 3164
+    },
+    {
+      "epoch": 3.6687676592045206,
+      "grad_norm": 0.34602412581443787,
+      "learning_rate": 3.1810344827586206e-05,
+      "loss": 0.1149,
+      "step": 3165
+    },
+    {
+      "epoch": 3.6699268275012678,
+      "grad_norm": 0.3075660467147827,
+      "learning_rate": 3.180459770114943e-05,
+      "loss": 0.109,
+      "step": 3166
+    },
+    {
+      "epoch": 3.671085995798015,
+      "grad_norm": 0.4555627405643463,
+      "learning_rate": 3.179885057471265e-05,
+      "loss": 0.1119,
+      "step": 3167
+    },
+    {
+      "epoch": 3.672245164094762,
+      "grad_norm": 0.3053813576698303,
+      "learning_rate": 3.1793103448275864e-05,
+      "loss": 0.117,
+      "step": 3168
+    },
+    {
+      "epoch": 3.673404332391509,
+      "grad_norm": 0.34098416566848755,
+      "learning_rate": 3.1787356321839085e-05,
+      "loss": 0.1131,
+      "step": 3169
+    },
+    {
+      "epoch": 3.674563500688256,
+      "grad_norm": 0.2684711813926697,
+      "learning_rate": 3.17816091954023e-05,
+      "loss": 0.1205,
+      "step": 3170
+    },
+    {
+      "epoch": 3.6757226689850033,
+      "grad_norm": 0.2564740777015686,
+      "learning_rate": 3.1775862068965515e-05,
+      "loss": 0.1063,
+      "step": 3171
+    },
+    {
+      "epoch": 3.6768818372817504,
+      "grad_norm": 0.27508318424224854,
+      "learning_rate": 3.177011494252874e-05,
+      "loss": 0.1106,
+      "step": 3172
+    },
+    {
+      "epoch": 3.6780410055784976,
+      "grad_norm": 0.23721154034137726,
+      "learning_rate": 3.176436781609196e-05,
+      "loss": 0.107,
+      "step": 3173
+    },
+    {
+      "epoch": 3.6792001738752447,
+      "grad_norm": 0.2993345260620117,
+      "learning_rate": 3.1758620689655174e-05,
+      "loss": 0.1075,
+      "step": 3174
+    },
+    {
+      "epoch": 3.6803593421719913,
+      "grad_norm": 0.3475472927093506,
+      "learning_rate": 3.1752873563218395e-05,
+      "loss": 0.1034,
+      "step": 3175
+    },
+    {
+      "epoch": 3.681518510468739,
+      "grad_norm": 0.25363993644714355,
+      "learning_rate": 3.174712643678161e-05,
+      "loss": 0.113,
+      "step": 3176
+    },
+    {
+      "epoch": 3.6826776787654856,
+      "grad_norm": 0.27115947008132935,
+      "learning_rate": 3.1741379310344825e-05,
+      "loss": 0.1231,
+      "step": 3177
+    },
+    {
+      "epoch": 3.6838368470622327,
+      "grad_norm": 0.2857735753059387,
+      "learning_rate": 3.173563218390805e-05,
+      "loss": 0.1171,
+      "step": 3178
+    },
+    {
+      "epoch": 3.68499601535898,
+      "grad_norm": 0.2827852666378021,
+      "learning_rate": 3.172988505747126e-05,
+      "loss": 0.1089,
+      "step": 3179
+    },
+    {
+      "epoch": 3.686155183655727,
+      "grad_norm": 0.31064677238464355,
+      "learning_rate": 3.172413793103448e-05,
+      "loss": 0.1281,
+      "step": 3180
+    },
+    {
+      "epoch": 3.687314351952474,
+      "grad_norm": 0.38487324118614197,
+      "learning_rate": 3.1718390804597705e-05,
+      "loss": 0.1147,
+      "step": 3181
+    },
+    {
+      "epoch": 3.688473520249221,
+      "grad_norm": 0.2798865735530853,
+      "learning_rate": 3.171264367816092e-05,
+      "loss": 0.1065,
+      "step": 3182
+    },
+    {
+      "epoch": 3.6896326885459683,
+      "grad_norm": 0.3226175308227539,
+      "learning_rate": 3.170689655172414e-05,
+      "loss": 0.1131,
+      "step": 3183
+    },
+    {
+      "epoch": 3.6907918568427154,
+      "grad_norm": 0.2735946774482727,
+      "learning_rate": 3.1701149425287356e-05,
+      "loss": 0.1149,
+      "step": 3184
+    },
+    {
+      "epoch": 3.6919510251394625,
+      "grad_norm": 0.3574925661087036,
+      "learning_rate": 3.169540229885057e-05,
+      "loss": 0.1081,
+      "step": 3185
+    },
+    {
+      "epoch": 3.6931101934362096,
+      "grad_norm": 0.24798470735549927,
+      "learning_rate": 3.168965517241379e-05,
+      "loss": 0.1078,
+      "step": 3186
+    },
+    {
+      "epoch": 3.6942693617329567,
+      "grad_norm": 0.27632611989974976,
+      "learning_rate": 3.1683908045977015e-05,
+      "loss": 0.1247,
+      "step": 3187
+    },
+    {
+      "epoch": 3.695428530029704,
+      "grad_norm": 0.21540036797523499,
+      "learning_rate": 3.1678160919540236e-05,
+      "loss": 0.0948,
+      "step": 3188
+    },
+    {
+      "epoch": 3.696587698326451,
+      "grad_norm": 0.34586644172668457,
+      "learning_rate": 3.167241379310345e-05,
+      "loss": 0.1299,
+      "step": 3189
+    },
+    {
+      "epoch": 3.6977468666231976,
+      "grad_norm": 0.3047373592853546,
+      "learning_rate": 3.1666666666666666e-05,
+      "loss": 0.1185,
+      "step": 3190
+    },
+    {
+      "epoch": 3.698906034919945,
+      "grad_norm": 0.30293378233909607,
+      "learning_rate": 3.166091954022989e-05,
+      "loss": 0.114,
+      "step": 3191
+    },
+    {
+      "epoch": 3.700065203216692,
+      "grad_norm": 0.3187353014945984,
+      "learning_rate": 3.16551724137931e-05,
+      "loss": 0.1255,
+      "step": 3192
+    },
+    {
+      "epoch": 3.701224371513439,
+      "grad_norm": 0.25981852412223816,
+      "learning_rate": 3.1649425287356324e-05,
+      "loss": 0.114,
+      "step": 3193
+    },
+    {
+      "epoch": 3.702383539810186,
+      "grad_norm": 0.2848886549472809,
+      "learning_rate": 3.1643678160919546e-05,
+      "loss": 0.104,
+      "step": 3194
+    },
+    {
+      "epoch": 3.703542708106933,
+      "grad_norm": 0.2802824079990387,
+      "learning_rate": 3.163793103448276e-05,
+      "loss": 0.0887,
+      "step": 3195
+    },
+    {
+      "epoch": 3.7047018764036803,
+      "grad_norm": 0.2874505817890167,
+      "learning_rate": 3.1632183908045976e-05,
+      "loss": 0.1067,
+      "step": 3196
+    },
+    {
+      "epoch": 3.7058610447004274,
+      "grad_norm": 0.4384983479976654,
+      "learning_rate": 3.16264367816092e-05,
+      "loss": 0.1229,
+      "step": 3197
+    },
+    {
+      "epoch": 3.7070202129971745,
+      "grad_norm": 0.3324441909790039,
+      "learning_rate": 3.162068965517241e-05,
+      "loss": 0.1089,
+      "step": 3198
+    },
+    {
+      "epoch": 3.7081793812939217,
+      "grad_norm": 0.311897337436676,
+      "learning_rate": 3.1614942528735634e-05,
+      "loss": 0.1182,
+      "step": 3199
+    },
+    {
+      "epoch": 3.7093385495906688,
+      "grad_norm": 0.28386661410331726,
+      "learning_rate": 3.160919540229885e-05,
+      "loss": 0.1092,
+      "step": 3200
+    },
+    {
+      "epoch": 3.710497717887416,
+      "grad_norm": 0.27510008215904236,
+      "learning_rate": 3.160344827586207e-05,
+      "loss": 0.1098,
+      "step": 3201
+    },
+    {
+      "epoch": 3.711656886184163,
+      "grad_norm": 0.25964272022247314,
+      "learning_rate": 3.159770114942529e-05,
+      "loss": 0.0969,
+      "step": 3202
+    },
+    {
+      "epoch": 3.7128160544809097,
+      "grad_norm": 0.31421464681625366,
+      "learning_rate": 3.159195402298851e-05,
+      "loss": 0.1123,
+      "step": 3203
+    },
+    {
+      "epoch": 3.7139752227776572,
+      "grad_norm": 0.2845355272293091,
+      "learning_rate": 3.158620689655172e-05,
+      "loss": 0.1089,
+      "step": 3204
+    },
+    {
+      "epoch": 3.715134391074404,
+      "grad_norm": 0.5368481874465942,
+      "learning_rate": 3.1580459770114944e-05,
+      "loss": 0.1038,
+      "step": 3205
+    },
+    {
+      "epoch": 3.7162935593711515,
+      "grad_norm": 0.5967825651168823,
+      "learning_rate": 3.157471264367816e-05,
+      "loss": 0.105,
+      "step": 3206
+    },
+    {
+      "epoch": 3.717452727667898,
+      "grad_norm": 0.29082855582237244,
+      "learning_rate": 3.156896551724138e-05,
+      "loss": 0.1157,
+      "step": 3207
+    },
+    {
+      "epoch": 3.7186118959646453,
+      "grad_norm": 0.337176114320755,
+      "learning_rate": 3.15632183908046e-05,
+      "loss": 0.1105,
+      "step": 3208
+    },
+    {
+      "epoch": 3.7197710642613924,
+      "grad_norm": 0.4318813979625702,
+      "learning_rate": 3.155747126436782e-05,
+      "loss": 0.1172,
+      "step": 3209
+    },
+    {
+      "epoch": 3.7209302325581395,
+      "grad_norm": 0.2882601320743561,
+      "learning_rate": 3.155172413793104e-05,
+      "loss": 0.1016,
+      "step": 3210
+    },
+    {
+      "epoch": 3.7220894008548866,
+      "grad_norm": 0.2961920499801636,
+      "learning_rate": 3.1545977011494254e-05,
+      "loss": 0.1317,
+      "step": 3211
+    },
+    {
+      "epoch": 3.7232485691516337,
+      "grad_norm": 0.2166760265827179,
+      "learning_rate": 3.154022988505747e-05,
+      "loss": 0.1111,
+      "step": 3212
+    },
+    {
+      "epoch": 3.724407737448381,
+      "grad_norm": 0.28654518723487854,
+      "learning_rate": 3.153448275862069e-05,
+      "loss": 0.1187,
+      "step": 3213
+    },
+    {
+      "epoch": 3.725566905745128,
+      "grad_norm": 0.34299778938293457,
+      "learning_rate": 3.152873563218391e-05,
+      "loss": 0.1278,
+      "step": 3214
+    },
+    {
+      "epoch": 3.726726074041875,
+      "grad_norm": 0.25071898102760315,
+      "learning_rate": 3.152298850574713e-05,
+      "loss": 0.1067,
+      "step": 3215
+    },
+    {
+      "epoch": 3.727885242338622,
+      "grad_norm": 0.2629477083683014,
+      "learning_rate": 3.151724137931035e-05,
+      "loss": 0.1121,
+      "step": 3216
+    },
+    {
+      "epoch": 3.7290444106353693,
+      "grad_norm": 0.37242335081100464,
+      "learning_rate": 3.151149425287356e-05,
+      "loss": 0.1083,
+      "step": 3217
+    },
+    {
+      "epoch": 3.730203578932116,
+      "grad_norm": 0.27490779757499695,
+      "learning_rate": 3.1505747126436785e-05,
+      "loss": 0.1174,
+      "step": 3218
+    },
+    {
+      "epoch": 3.7313627472288635,
+      "grad_norm": 0.2524685263633728,
+      "learning_rate": 3.15e-05,
+      "loss": 0.118,
+      "step": 3219
+    },
+    {
+      "epoch": 3.73252191552561,
+      "grad_norm": 0.27024635672569275,
+      "learning_rate": 3.1494252873563215e-05,
+      "loss": 0.1198,
+      "step": 3220
+    },
+    {
+      "epoch": 3.7336810838223573,
+      "grad_norm": 0.2741260528564453,
+      "learning_rate": 3.148850574712644e-05,
+      "loss": 0.1144,
+      "step": 3221
+    },
+    {
+      "epoch": 3.7348402521191044,
+      "grad_norm": 0.321684867143631,
+      "learning_rate": 3.148275862068966e-05,
+      "loss": 0.0993,
+      "step": 3222
+    },
+    {
+      "epoch": 3.7359994204158515,
+      "grad_norm": 0.19403128325939178,
+      "learning_rate": 3.147701149425287e-05,
+      "loss": 0.0992,
+      "step": 3223
+    },
+    {
+      "epoch": 3.7371585887125987,
+      "grad_norm": 0.26611411571502686,
+      "learning_rate": 3.1471264367816095e-05,
+      "loss": 0.1176,
+      "step": 3224
+    },
+    {
+      "epoch": 3.7383177570093458,
+      "grad_norm": 0.36536505818367004,
+      "learning_rate": 3.146551724137931e-05,
+      "loss": 0.1099,
+      "step": 3225
+    },
+    {
+      "epoch": 3.739476925306093,
+      "grad_norm": 0.2902340590953827,
+      "learning_rate": 3.145977011494253e-05,
+      "loss": 0.1109,
+      "step": 3226
+    },
+    {
+      "epoch": 3.74063609360284,
+      "grad_norm": 0.24061763286590576,
+      "learning_rate": 3.1454022988505746e-05,
+      "loss": 0.111,
+      "step": 3227
+    },
+    {
+      "epoch": 3.741795261899587,
+      "grad_norm": 0.32074034214019775,
+      "learning_rate": 3.144827586206897e-05,
+      "loss": 0.1056,
+      "step": 3228
+    },
+    {
+      "epoch": 3.7429544301963342,
+      "grad_norm": 0.2975504994392395,
+      "learning_rate": 3.144252873563219e-05,
+      "loss": 0.1241,
+      "step": 3229
+    },
+    {
+      "epoch": 3.7441135984930813,
+      "grad_norm": 0.33813774585723877,
+      "learning_rate": 3.1436781609195405e-05,
+      "loss": 0.11,
+      "step": 3230
+    },
+    {
+      "epoch": 3.7452727667898285,
+      "grad_norm": 0.29403892159461975,
+      "learning_rate": 3.143103448275862e-05,
+      "loss": 0.1118,
+      "step": 3231
+    },
+    {
+      "epoch": 3.7464319350865756,
+      "grad_norm": 0.2610683739185333,
+      "learning_rate": 3.142528735632184e-05,
+      "loss": 0.1055,
+      "step": 3232
+    },
+    {
+      "epoch": 3.7475911033833222,
+      "grad_norm": 0.2696630358695984,
+      "learning_rate": 3.1419540229885056e-05,
+      "loss": 0.1062,
+      "step": 3233
+    },
+    {
+      "epoch": 3.74875027168007,
+      "grad_norm": 0.3874567747116089,
+      "learning_rate": 3.141379310344828e-05,
+      "loss": 0.1243,
+      "step": 3234
+    },
+    {
+      "epoch": 3.7499094399768165,
+      "grad_norm": 0.39800602197647095,
+      "learning_rate": 3.14080459770115e-05,
+      "loss": 0.124,
+      "step": 3235
+    },
+    {
+      "epoch": 3.7510686082735636,
+      "grad_norm": 0.3324180245399475,
+      "learning_rate": 3.1402298850574714e-05,
+      "loss": 0.1292,
+      "step": 3236
+    },
+    {
+      "epoch": 3.7522277765703107,
+      "grad_norm": 0.3287622630596161,
+      "learning_rate": 3.1396551724137936e-05,
+      "loss": 0.1152,
+      "step": 3237
+    },
+    {
+      "epoch": 3.753386944867058,
+      "grad_norm": 0.22335346043109894,
+      "learning_rate": 3.139080459770115e-05,
+      "loss": 0.1006,
+      "step": 3238
+    },
+    {
+      "epoch": 3.754546113163805,
+      "grad_norm": 0.32788801193237305,
+      "learning_rate": 3.1385057471264366e-05,
+      "loss": 0.1072,
+      "step": 3239
+    },
+    {
+      "epoch": 3.755705281460552,
+      "grad_norm": 0.3683432340621948,
+      "learning_rate": 3.137931034482759e-05,
+      "loss": 0.1085,
+      "step": 3240
+    },
+    {
+      "epoch": 3.756864449757299,
+      "grad_norm": 0.33938443660736084,
+      "learning_rate": 3.13735632183908e-05,
+      "loss": 0.1115,
+      "step": 3241
+    },
+    {
+      "epoch": 3.7580236180540463,
+      "grad_norm": 0.27710798382759094,
+      "learning_rate": 3.1367816091954024e-05,
+      "loss": 0.1152,
+      "step": 3242
+    },
+    {
+      "epoch": 3.7591827863507934,
+      "grad_norm": 0.4090668261051178,
+      "learning_rate": 3.1362068965517246e-05,
+      "loss": 0.1064,
+      "step": 3243
+    },
+    {
+      "epoch": 3.7603419546475405,
+      "grad_norm": 0.33840107917785645,
+      "learning_rate": 3.135632183908046e-05,
+      "loss": 0.1124,
+      "step": 3244
+    },
+    {
+      "epoch": 3.7615011229442876,
+      "grad_norm": 0.2494470775127411,
+      "learning_rate": 3.135057471264368e-05,
+      "loss": 0.1,
+      "step": 3245
+    },
+    {
+      "epoch": 3.7626602912410343,
+      "grad_norm": 0.2987157106399536,
+      "learning_rate": 3.13448275862069e-05,
+      "loss": 0.1159,
+      "step": 3246
+    },
+    {
+      "epoch": 3.763819459537782,
+      "grad_norm": 0.3439898192882538,
+      "learning_rate": 3.133908045977011e-05,
+      "loss": 0.1171,
+      "step": 3247
+    },
+    {
+      "epoch": 3.7649786278345285,
+      "grad_norm": 0.4079365134239197,
+      "learning_rate": 3.1333333333333334e-05,
+      "loss": 0.1184,
+      "step": 3248
+    },
+    {
+      "epoch": 3.766137796131276,
+      "grad_norm": 0.25229400396347046,
+      "learning_rate": 3.1327586206896555e-05,
+      "loss": 0.107,
+      "step": 3249
+    },
+    {
+      "epoch": 3.7672969644280228,
+      "grad_norm": 0.28538668155670166,
+      "learning_rate": 3.132183908045977e-05,
+      "loss": 0.1143,
+      "step": 3250
+    },
+    {
+      "epoch": 3.76845613272477,
+      "grad_norm": 0.3976428806781769,
+      "learning_rate": 3.131609195402299e-05,
+      "loss": 0.1152,
+      "step": 3251
+    },
+    {
+      "epoch": 3.769615301021517,
+      "grad_norm": 0.3206281363964081,
+      "learning_rate": 3.131034482758621e-05,
+      "loss": 0.1219,
+      "step": 3252
+    },
+    {
+      "epoch": 3.770774469318264,
+      "grad_norm": 0.34073176980018616,
+      "learning_rate": 3.130459770114942e-05,
+      "loss": 0.1111,
+      "step": 3253
+    },
+    {
+      "epoch": 3.771933637615011,
+      "grad_norm": 0.2973473370075226,
+      "learning_rate": 3.1298850574712643e-05,
+      "loss": 0.1155,
+      "step": 3254
+    },
+    {
+      "epoch": 3.7730928059117583,
+      "grad_norm": 0.268359512090683,
+      "learning_rate": 3.1293103448275865e-05,
+      "loss": 0.1207,
+      "step": 3255
+    },
+    {
+      "epoch": 3.7742519742085054,
+      "grad_norm": 0.24991478025913239,
+      "learning_rate": 3.128735632183909e-05,
+      "loss": 0.1108,
+      "step": 3256
+    },
+    {
+      "epoch": 3.7754111425052526,
+      "grad_norm": 0.28493621945381165,
+      "learning_rate": 3.12816091954023e-05,
+      "loss": 0.1153,
+      "step": 3257
+    },
+    {
+      "epoch": 3.7765703108019997,
+      "grad_norm": 0.25810399651527405,
+      "learning_rate": 3.127586206896552e-05,
+      "loss": 0.1134,
+      "step": 3258
+    },
+    {
+      "epoch": 3.777729479098747,
+      "grad_norm": 0.33807915449142456,
+      "learning_rate": 3.127011494252874e-05,
+      "loss": 0.1156,
+      "step": 3259
+    },
+    {
+      "epoch": 3.778888647395494,
+      "grad_norm": 0.37843331694602966,
+      "learning_rate": 3.126436781609195e-05,
+      "loss": 0.1112,
+      "step": 3260
+    },
+    {
+      "epoch": 3.7800478156922406,
+      "grad_norm": 0.3349437713623047,
+      "learning_rate": 3.125862068965517e-05,
+      "loss": 0.1004,
+      "step": 3261
+    },
+    {
+      "epoch": 3.781206983988988,
+      "grad_norm": 0.24173638224601746,
+      "learning_rate": 3.1252873563218397e-05,
+      "loss": 0.1016,
+      "step": 3262
+    },
+    {
+      "epoch": 3.782366152285735,
+      "grad_norm": 0.27250927686691284,
+      "learning_rate": 3.124712643678161e-05,
+      "loss": 0.1076,
+      "step": 3263
+    },
+    {
+      "epoch": 3.783525320582482,
+      "grad_norm": 0.3555189073085785,
+      "learning_rate": 3.124137931034483e-05,
+      "loss": 0.1234,
+      "step": 3264
+    },
+    {
+      "epoch": 3.784684488879229,
+      "grad_norm": 0.3194526433944702,
+      "learning_rate": 3.123563218390805e-05,
+      "loss": 0.1168,
+      "step": 3265
+    },
+    {
+      "epoch": 3.785843657175976,
+      "grad_norm": 0.3455429673194885,
+      "learning_rate": 3.122988505747126e-05,
+      "loss": 0.1243,
+      "step": 3266
+    },
+    {
+      "epoch": 3.7870028254727233,
+      "grad_norm": 0.31268778443336487,
+      "learning_rate": 3.1224137931034485e-05,
+      "loss": 0.1013,
+      "step": 3267
+    },
+    {
+      "epoch": 3.7881619937694704,
+      "grad_norm": 0.3976519703865051,
+      "learning_rate": 3.12183908045977e-05,
+      "loss": 0.1106,
+      "step": 3268
+    },
+    {
+      "epoch": 3.7893211620662175,
+      "grad_norm": 0.37077805399894714,
+      "learning_rate": 3.121264367816092e-05,
+      "loss": 0.115,
+      "step": 3269
+    },
+    {
+      "epoch": 3.7904803303629646,
+      "grad_norm": 0.26564523577690125,
+      "learning_rate": 3.120689655172414e-05,
+      "loss": 0.1025,
+      "step": 3270
+    },
+    {
+      "epoch": 3.7916394986597117,
+      "grad_norm": 0.30875927209854126,
+      "learning_rate": 3.120114942528736e-05,
+      "loss": 0.1061,
+      "step": 3271
+    },
+    {
+      "epoch": 3.792798666956459,
+      "grad_norm": 0.2600082755088806,
+      "learning_rate": 3.119540229885058e-05,
+      "loss": 0.0994,
+      "step": 3272
+    },
+    {
+      "epoch": 3.793957835253206,
+      "grad_norm": 0.2653372585773468,
+      "learning_rate": 3.1189655172413794e-05,
+      "loss": 0.1006,
+      "step": 3273
+    },
+    {
+      "epoch": 3.795117003549953,
+      "grad_norm": 0.23784595727920532,
+      "learning_rate": 3.118390804597701e-05,
+      "loss": 0.1064,
+      "step": 3274
+    },
+    {
+      "epoch": 3.7962761718467,
+      "grad_norm": 0.3346762955188751,
+      "learning_rate": 3.117816091954023e-05,
+      "loss": 0.1174,
+      "step": 3275
+    },
+    {
+      "epoch": 3.797435340143447,
+      "grad_norm": 0.28256985545158386,
+      "learning_rate": 3.117241379310345e-05,
+      "loss": 0.1233,
+      "step": 3276
+    },
+    {
+      "epoch": 3.7985945084401944,
+      "grad_norm": 0.3030017614364624,
+      "learning_rate": 3.116666666666667e-05,
+      "loss": 0.1022,
+      "step": 3277
+    },
+    {
+      "epoch": 3.799753676736941,
+      "grad_norm": 0.28801146149635315,
+      "learning_rate": 3.116091954022989e-05,
+      "loss": 0.1084,
+      "step": 3278
+    },
+    {
+      "epoch": 3.800912845033688,
+      "grad_norm": 0.3173084259033203,
+      "learning_rate": 3.1155172413793104e-05,
+      "loss": 0.1118,
+      "step": 3279
+    },
+    {
+      "epoch": 3.8020720133304353,
+      "grad_norm": 0.22037595510482788,
+      "learning_rate": 3.114942528735632e-05,
+      "loss": 0.1013,
+      "step": 3280
+    },
+    {
+      "epoch": 3.8032311816271824,
+      "grad_norm": 0.4173763394355774,
+      "learning_rate": 3.114367816091954e-05,
+      "loss": 0.1147,
+      "step": 3281
+    },
+    {
+      "epoch": 3.8043903499239295,
+      "grad_norm": 0.38189518451690674,
+      "learning_rate": 3.113793103448276e-05,
+      "loss": 0.1191,
+      "step": 3282
+    },
+    {
+      "epoch": 3.8055495182206767,
+      "grad_norm": 0.28138333559036255,
+      "learning_rate": 3.1132183908045984e-05,
+      "loss": 0.118,
+      "step": 3283
+    },
+    {
+      "epoch": 3.806708686517424,
+      "grad_norm": 0.28683844208717346,
+      "learning_rate": 3.11264367816092e-05,
+      "loss": 0.1156,
+      "step": 3284
+    },
+    {
+      "epoch": 3.807867854814171,
+      "grad_norm": 0.28898361325263977,
+      "learning_rate": 3.1120689655172414e-05,
+      "loss": 0.1054,
+      "step": 3285
+    },
+    {
+      "epoch": 3.809027023110918,
+      "grad_norm": 0.2948157787322998,
+      "learning_rate": 3.1114942528735636e-05,
+      "loss": 0.0977,
+      "step": 3286
+    },
+    {
+      "epoch": 3.810186191407665,
+      "grad_norm": 0.27834632992744446,
+      "learning_rate": 3.110919540229885e-05,
+      "loss": 0.1206,
+      "step": 3287
+    },
+    {
+      "epoch": 3.8113453597044122,
+      "grad_norm": 0.2628282606601715,
+      "learning_rate": 3.1103448275862065e-05,
+      "loss": 0.1112,
+      "step": 3288
+    },
+    {
+      "epoch": 3.812504528001159,
+      "grad_norm": 0.2561165988445282,
+      "learning_rate": 3.109770114942529e-05,
+      "loss": 0.1138,
+      "step": 3289
+    },
+    {
+      "epoch": 3.8136636962979065,
+      "grad_norm": 0.23604638874530792,
+      "learning_rate": 3.109195402298851e-05,
+      "loss": 0.1081,
+      "step": 3290
+    },
+    {
+      "epoch": 3.814822864594653,
+      "grad_norm": 0.2508888840675354,
+      "learning_rate": 3.108620689655173e-05,
+      "loss": 0.1002,
+      "step": 3291
+    },
+    {
+      "epoch": 3.8159820328914003,
+      "grad_norm": 0.3218640089035034,
+      "learning_rate": 3.1080459770114945e-05,
+      "loss": 0.1064,
+      "step": 3292
+    },
+    {
+      "epoch": 3.8171412011881474,
+      "grad_norm": 0.22713243961334229,
+      "learning_rate": 3.107471264367816e-05,
+      "loss": 0.1074,
+      "step": 3293
+    },
+    {
+      "epoch": 3.8183003694848945,
+      "grad_norm": 0.2802816927433014,
+      "learning_rate": 3.106896551724138e-05,
+      "loss": 0.1089,
+      "step": 3294
+    },
+    {
+      "epoch": 3.8194595377816416,
+      "grad_norm": 0.25849002599716187,
+      "learning_rate": 3.10632183908046e-05,
+      "loss": 0.1132,
+      "step": 3295
+    },
+    {
+      "epoch": 3.8206187060783887,
+      "grad_norm": 0.3340066373348236,
+      "learning_rate": 3.105747126436782e-05,
+      "loss": 0.1287,
+      "step": 3296
+    },
+    {
+      "epoch": 3.821777874375136,
+      "grad_norm": 0.3333747684955597,
+      "learning_rate": 3.105172413793104e-05,
+      "loss": 0.1157,
+      "step": 3297
+    },
+    {
+      "epoch": 3.822937042671883,
+      "grad_norm": 0.26639991998672485,
+      "learning_rate": 3.1045977011494255e-05,
+      "loss": 0.1089,
+      "step": 3298
+    },
+    {
+      "epoch": 3.82409621096863,
+      "grad_norm": 0.2500338852405548,
+      "learning_rate": 3.104022988505747e-05,
+      "loss": 0.1146,
+      "step": 3299
+    },
+    {
+      "epoch": 3.825255379265377,
+      "grad_norm": 0.3188420534133911,
+      "learning_rate": 3.103448275862069e-05,
+      "loss": 0.1142,
+      "step": 3300
+    },
+    {
+      "epoch": 3.8264145475621243,
+      "grad_norm": 0.22074191272258759,
+      "learning_rate": 3.1028735632183907e-05,
+      "loss": 0.1022,
+      "step": 3301
+    },
+    {
+      "epoch": 3.8275737158588714,
+      "grad_norm": 0.32169198989868164,
+      "learning_rate": 3.102298850574713e-05,
+      "loss": 0.1234,
+      "step": 3302
+    },
+    {
+      "epoch": 3.8287328841556185,
+      "grad_norm": 0.23068860173225403,
+      "learning_rate": 3.101724137931035e-05,
+      "loss": 0.1052,
+      "step": 3303
+    },
+    {
+      "epoch": 3.829892052452365,
+      "grad_norm": 0.4056016206741333,
+      "learning_rate": 3.1011494252873565e-05,
+      "loss": 0.1122,
+      "step": 3304
+    },
+    {
+      "epoch": 3.8310512207491128,
+      "grad_norm": 0.7055277228355408,
+      "learning_rate": 3.1005747126436786e-05,
+      "loss": 0.1154,
+      "step": 3305
+    },
+    {
+      "epoch": 3.8322103890458594,
+      "grad_norm": 0.3167892098426819,
+      "learning_rate": 3.1e-05,
+      "loss": 0.1151,
+      "step": 3306
+    },
+    {
+      "epoch": 3.8333695573426065,
+      "grad_norm": 0.3068336546421051,
+      "learning_rate": 3.0994252873563216e-05,
+      "loss": 0.1093,
+      "step": 3307
+    },
+    {
+      "epoch": 3.8345287256393537,
+      "grad_norm": 0.28874102234840393,
+      "learning_rate": 3.098850574712644e-05,
+      "loss": 0.101,
+      "step": 3308
+    },
+    {
+      "epoch": 3.8356878939361008,
+      "grad_norm": 0.30974501371383667,
+      "learning_rate": 3.098275862068965e-05,
+      "loss": 0.1292,
+      "step": 3309
+    },
+    {
+      "epoch": 3.836847062232848,
+      "grad_norm": 0.2519693672657013,
+      "learning_rate": 3.0977011494252875e-05,
+      "loss": 0.0984,
+      "step": 3310
+    },
+    {
+      "epoch": 3.838006230529595,
+      "grad_norm": 0.3312533497810364,
+      "learning_rate": 3.0971264367816096e-05,
+      "loss": 0.1092,
+      "step": 3311
+    },
+    {
+      "epoch": 3.839165398826342,
+      "grad_norm": 0.33562129735946655,
+      "learning_rate": 3.096551724137931e-05,
+      "loss": 0.1045,
+      "step": 3312
+    },
+    {
+      "epoch": 3.8403245671230892,
+      "grad_norm": 0.22888796031475067,
+      "learning_rate": 3.095977011494253e-05,
+      "loss": 0.1107,
+      "step": 3313
+    },
+    {
+      "epoch": 3.8414837354198363,
+      "grad_norm": 0.338222861289978,
+      "learning_rate": 3.095402298850575e-05,
+      "loss": 0.1227,
+      "step": 3314
+    },
+    {
+      "epoch": 3.8426429037165835,
+      "grad_norm": 0.3241053521633148,
+      "learning_rate": 3.094827586206896e-05,
+      "loss": 0.1243,
+      "step": 3315
+    },
+    {
+      "epoch": 3.8438020720133306,
+      "grad_norm": 0.2689477503299713,
+      "learning_rate": 3.0942528735632184e-05,
+      "loss": 0.1091,
+      "step": 3316
+    },
+    {
+      "epoch": 3.8449612403100772,
+      "grad_norm": 0.2917823791503906,
+      "learning_rate": 3.0936781609195406e-05,
+      "loss": 0.099,
+      "step": 3317
+    },
+    {
+      "epoch": 3.846120408606825,
+      "grad_norm": 0.2831590473651886,
+      "learning_rate": 3.093103448275862e-05,
+      "loss": 0.0987,
+      "step": 3318
+    },
+    {
+      "epoch": 3.8472795769035715,
+      "grad_norm": 0.2956966161727905,
+      "learning_rate": 3.092528735632184e-05,
+      "loss": 0.1038,
+      "step": 3319
+    },
+    {
+      "epoch": 3.848438745200319,
+      "grad_norm": 0.41253408789634705,
+      "learning_rate": 3.091954022988506e-05,
+      "loss": 0.1173,
+      "step": 3320
+    },
+    {
+      "epoch": 3.8495979134970657,
+      "grad_norm": 0.37665021419525146,
+      "learning_rate": 3.091379310344828e-05,
+      "loss": 0.1211,
+      "step": 3321
+    },
+    {
+      "epoch": 3.850757081793813,
+      "grad_norm": 0.29172924160957336,
+      "learning_rate": 3.0908045977011494e-05,
+      "loss": 0.1087,
+      "step": 3322
+    },
+    {
+      "epoch": 3.85191625009056,
+      "grad_norm": 0.4337025284767151,
+      "learning_rate": 3.0902298850574716e-05,
+      "loss": 0.1089,
+      "step": 3323
+    },
+    {
+      "epoch": 3.853075418387307,
+      "grad_norm": 0.36892715096473694,
+      "learning_rate": 3.089655172413794e-05,
+      "loss": 0.108,
+      "step": 3324
+    },
+    {
+      "epoch": 3.854234586684054,
+      "grad_norm": 0.34563952684402466,
+      "learning_rate": 3.089080459770115e-05,
+      "loss": 0.116,
+      "step": 3325
+    },
+    {
+      "epoch": 3.8553937549808013,
+      "grad_norm": 0.3011876344680786,
+      "learning_rate": 3.088505747126437e-05,
+      "loss": 0.1105,
+      "step": 3326
+    },
+    {
+      "epoch": 3.8565529232775484,
+      "grad_norm": 0.36122533679008484,
+      "learning_rate": 3.087931034482759e-05,
+      "loss": 0.1171,
+      "step": 3327
+    },
+    {
+      "epoch": 3.8577120915742955,
+      "grad_norm": 0.31418511271476746,
+      "learning_rate": 3.0873563218390804e-05,
+      "loss": 0.1189,
+      "step": 3328
+    },
+    {
+      "epoch": 3.8588712598710426,
+      "grad_norm": 0.2561214566230774,
+      "learning_rate": 3.0867816091954025e-05,
+      "loss": 0.1061,
+      "step": 3329
+    },
+    {
+      "epoch": 3.8600304281677897,
+      "grad_norm": 0.28758811950683594,
+      "learning_rate": 3.086206896551724e-05,
+      "loss": 0.1142,
+      "step": 3330
+    },
+    {
+      "epoch": 3.861189596464537,
+      "grad_norm": 0.2774584889411926,
+      "learning_rate": 3.085632183908046e-05,
+      "loss": 0.1097,
+      "step": 3331
+    },
+    {
+      "epoch": 3.8623487647612835,
+      "grad_norm": 0.3186490535736084,
+      "learning_rate": 3.0850574712643684e-05,
+      "loss": 0.117,
+      "step": 3332
+    },
+    {
+      "epoch": 3.863507933058031,
+      "grad_norm": 0.3841571807861328,
+      "learning_rate": 3.08448275862069e-05,
+      "loss": 0.1224,
+      "step": 3333
+    },
+    {
+      "epoch": 3.8646671013547778,
+      "grad_norm": 0.2925528585910797,
+      "learning_rate": 3.0839080459770113e-05,
+      "loss": 0.1049,
+      "step": 3334
+    },
+    {
+      "epoch": 3.865826269651525,
+      "grad_norm": 0.35688725113868713,
+      "learning_rate": 3.0833333333333335e-05,
+      "loss": 0.1102,
+      "step": 3335
+    },
+    {
+      "epoch": 3.866985437948272,
+      "grad_norm": 0.27751150727272034,
+      "learning_rate": 3.082758620689655e-05,
+      "loss": 0.1114,
+      "step": 3336
+    },
+    {
+      "epoch": 3.868144606245019,
+      "grad_norm": 0.25139451026916504,
+      "learning_rate": 3.082183908045977e-05,
+      "loss": 0.1159,
+      "step": 3337
+    },
+    {
+      "epoch": 3.869303774541766,
+      "grad_norm": 0.4053381085395813,
+      "learning_rate": 3.0816091954022993e-05,
+      "loss": 0.1268,
+      "step": 3338
+    },
+    {
+      "epoch": 3.8704629428385133,
+      "grad_norm": 0.22363197803497314,
+      "learning_rate": 3.081034482758621e-05,
+      "loss": 0.106,
+      "step": 3339
+    },
+    {
+      "epoch": 3.8716221111352604,
+      "grad_norm": 0.310592919588089,
+      "learning_rate": 3.080459770114943e-05,
+      "loss": 0.1168,
+      "step": 3340
+    },
+    {
+      "epoch": 3.8727812794320076,
+      "grad_norm": 0.29735296964645386,
+      "learning_rate": 3.0798850574712645e-05,
+      "loss": 0.1257,
+      "step": 3341
+    },
+    {
+      "epoch": 3.8739404477287547,
+      "grad_norm": 0.2444659024477005,
+      "learning_rate": 3.079310344827586e-05,
+      "loss": 0.109,
+      "step": 3342
+    },
+    {
+      "epoch": 3.875099616025502,
+      "grad_norm": 0.32433512806892395,
+      "learning_rate": 3.078735632183908e-05,
+      "loss": 0.1177,
+      "step": 3343
+    },
+    {
+      "epoch": 3.876258784322249,
+      "grad_norm": 0.2731485366821289,
+      "learning_rate": 3.07816091954023e-05,
+      "loss": 0.1176,
+      "step": 3344
+    },
+    {
+      "epoch": 3.877417952618996,
+      "grad_norm": 0.25933563709259033,
+      "learning_rate": 3.077586206896552e-05,
+      "loss": 0.0964,
+      "step": 3345
+    },
+    {
+      "epoch": 3.878577120915743,
+      "grad_norm": 0.34049245715141296,
+      "learning_rate": 3.077011494252874e-05,
+      "loss": 0.1237,
+      "step": 3346
+    },
+    {
+      "epoch": 3.87973628921249,
+      "grad_norm": 0.2941095232963562,
+      "learning_rate": 3.0764367816091955e-05,
+      "loss": 0.1159,
+      "step": 3347
+    },
+    {
+      "epoch": 3.8808954575092374,
+      "grad_norm": 0.2893029451370239,
+      "learning_rate": 3.0758620689655176e-05,
+      "loss": 0.1072,
+      "step": 3348
+    },
+    {
+      "epoch": 3.882054625805984,
+      "grad_norm": 0.2857878506183624,
+      "learning_rate": 3.075287356321839e-05,
+      "loss": 0.1117,
+      "step": 3349
+    },
+    {
+      "epoch": 3.883213794102731,
+      "grad_norm": 0.3128105103969574,
+      "learning_rate": 3.0747126436781606e-05,
+      "loss": 0.1154,
+      "step": 3350
+    },
+    {
+      "epoch": 3.8843729623994783,
+      "grad_norm": 0.34108099341392517,
+      "learning_rate": 3.074137931034483e-05,
+      "loss": 0.1083,
+      "step": 3351
+    },
+    {
+      "epoch": 3.8855321306962254,
+      "grad_norm": 0.29815927147865295,
+      "learning_rate": 3.073563218390805e-05,
+      "loss": 0.1092,
+      "step": 3352
+    },
+    {
+      "epoch": 3.8866912989929725,
+      "grad_norm": 0.32374799251556396,
+      "learning_rate": 3.0729885057471264e-05,
+      "loss": 0.1138,
+      "step": 3353
+    },
+    {
+      "epoch": 3.8878504672897196,
+      "grad_norm": 0.30145663022994995,
+      "learning_rate": 3.0724137931034486e-05,
+      "loss": 0.1105,
+      "step": 3354
+    },
+    {
+      "epoch": 3.8890096355864667,
+      "grad_norm": 0.27380499243736267,
+      "learning_rate": 3.07183908045977e-05,
+      "loss": 0.1038,
+      "step": 3355
+    },
+    {
+      "epoch": 3.890168803883214,
+      "grad_norm": 0.33210647106170654,
+      "learning_rate": 3.0712643678160916e-05,
+      "loss": 0.1312,
+      "step": 3356
+    },
+    {
+      "epoch": 3.891327972179961,
+      "grad_norm": 0.27347707748413086,
+      "learning_rate": 3.070689655172414e-05,
+      "loss": 0.104,
+      "step": 3357
+    },
+    {
+      "epoch": 3.892487140476708,
+      "grad_norm": 0.31707629561424255,
+      "learning_rate": 3.070114942528736e-05,
+      "loss": 0.1097,
+      "step": 3358
+    },
+    {
+      "epoch": 3.893646308773455,
+      "grad_norm": 0.3601592481136322,
+      "learning_rate": 3.069540229885058e-05,
+      "loss": 0.1209,
+      "step": 3359
+    },
+    {
+      "epoch": 3.894805477070202,
+      "grad_norm": 0.29122745990753174,
+      "learning_rate": 3.0689655172413796e-05,
+      "loss": 0.1059,
+      "step": 3360
+    },
+    {
+      "epoch": 3.8959646453669494,
+      "grad_norm": 0.3259906470775604,
+      "learning_rate": 3.068390804597701e-05,
+      "loss": 0.1168,
+      "step": 3361
+    },
+    {
+      "epoch": 3.897123813663696,
+      "grad_norm": 0.31396666169166565,
+      "learning_rate": 3.067816091954023e-05,
+      "loss": 0.1048,
+      "step": 3362
+    },
+    {
+      "epoch": 3.8982829819604436,
+      "grad_norm": 0.28336969017982483,
+      "learning_rate": 3.067241379310345e-05,
+      "loss": 0.1083,
+      "step": 3363
+    },
+    {
+      "epoch": 3.8994421502571903,
+      "grad_norm": 0.29647666215896606,
+      "learning_rate": 3.066666666666667e-05,
+      "loss": 0.1163,
+      "step": 3364
+    },
+    {
+      "epoch": 3.9006013185539374,
+      "grad_norm": 0.4049275815486908,
+      "learning_rate": 3.066091954022989e-05,
+      "loss": 0.1132,
+      "step": 3365
+    },
+    {
+      "epoch": 3.9017604868506846,
+      "grad_norm": 0.2655840516090393,
+      "learning_rate": 3.0655172413793106e-05,
+      "loss": 0.0983,
+      "step": 3366
+    },
+    {
+      "epoch": 3.9029196551474317,
+      "grad_norm": 0.5942867994308472,
+      "learning_rate": 3.064942528735633e-05,
+      "loss": 0.1326,
+      "step": 3367
+    },
+    {
+      "epoch": 3.904078823444179,
+      "grad_norm": 0.24042394757270813,
+      "learning_rate": 3.064367816091954e-05,
+      "loss": 0.1083,
+      "step": 3368
+    },
+    {
+      "epoch": 3.905237991740926,
+      "grad_norm": 0.24177686870098114,
+      "learning_rate": 3.063793103448276e-05,
+      "loss": 0.1151,
+      "step": 3369
+    },
+    {
+      "epoch": 3.906397160037673,
+      "grad_norm": 0.33689844608306885,
+      "learning_rate": 3.063218390804598e-05,
+      "loss": 0.1128,
+      "step": 3370
+    },
+    {
+      "epoch": 3.90755632833442,
+      "grad_norm": 0.3097141683101654,
+      "learning_rate": 3.0626436781609194e-05,
+      "loss": 0.1209,
+      "step": 3371
+    },
+    {
+      "epoch": 3.9087154966311672,
+      "grad_norm": 0.2580316364765167,
+      "learning_rate": 3.0620689655172415e-05,
+      "loss": 0.1079,
+      "step": 3372
+    },
+    {
+      "epoch": 3.9098746649279144,
+      "grad_norm": 0.2943916320800781,
+      "learning_rate": 3.061494252873564e-05,
+      "loss": 0.124,
+      "step": 3373
+    },
+    {
+      "epoch": 3.9110338332246615,
+      "grad_norm": 0.27771899104118347,
+      "learning_rate": 3.060919540229885e-05,
+      "loss": 0.1084,
+      "step": 3374
+    },
+    {
+      "epoch": 3.912193001521408,
+      "grad_norm": 0.30981943011283875,
+      "learning_rate": 3.060344827586207e-05,
+      "loss": 0.1076,
+      "step": 3375
+    },
+    {
+      "epoch": 3.9133521698181557,
+      "grad_norm": 0.3062245547771454,
+      "learning_rate": 3.059770114942529e-05,
+      "loss": 0.1273,
+      "step": 3376
+    },
+    {
+      "epoch": 3.9145113381149024,
+      "grad_norm": 0.25687843561172485,
+      "learning_rate": 3.05919540229885e-05,
+      "loss": 0.1053,
+      "step": 3377
+    },
+    {
+      "epoch": 3.9156705064116495,
+      "grad_norm": 0.25983986258506775,
+      "learning_rate": 3.0586206896551725e-05,
+      "loss": 0.1062,
+      "step": 3378
+    },
+    {
+      "epoch": 3.9168296747083966,
+      "grad_norm": 0.7442601323127747,
+      "learning_rate": 3.058045977011495e-05,
+      "loss": 0.1203,
+      "step": 3379
+    },
+    {
+      "epoch": 3.9179888430051437,
+      "grad_norm": 0.289230078458786,
+      "learning_rate": 3.057471264367816e-05,
+      "loss": 0.1178,
+      "step": 3380
+    },
+    {
+      "epoch": 3.919148011301891,
+      "grad_norm": 0.25480005145072937,
+      "learning_rate": 3.056896551724138e-05,
+      "loss": 0.1119,
+      "step": 3381
+    },
+    {
+      "epoch": 3.920307179598638,
+      "grad_norm": 0.2439074069261551,
+      "learning_rate": 3.05632183908046e-05,
+      "loss": 0.103,
+      "step": 3382
+    },
+    {
+      "epoch": 3.921466347895385,
+      "grad_norm": 0.32492339611053467,
+      "learning_rate": 3.055747126436781e-05,
+      "loss": 0.1203,
+      "step": 3383
+    },
+    {
+      "epoch": 3.922625516192132,
+      "grad_norm": 0.2885267436504364,
+      "learning_rate": 3.0551724137931035e-05,
+      "loss": 0.1092,
+      "step": 3384
+    },
+    {
+      "epoch": 3.9237846844888793,
+      "grad_norm": 0.5572437644004822,
+      "learning_rate": 3.0545977011494256e-05,
+      "loss": 0.1181,
+      "step": 3385
+    },
+    {
+      "epoch": 3.9249438527856264,
+      "grad_norm": 0.3531005382537842,
+      "learning_rate": 3.054022988505748e-05,
+      "loss": 0.1129,
+      "step": 3386
+    },
+    {
+      "epoch": 3.9261030210823735,
+      "grad_norm": 0.35553064942359924,
+      "learning_rate": 3.053448275862069e-05,
+      "loss": 0.1238,
+      "step": 3387
+    },
+    {
+      "epoch": 3.9272621893791206,
+      "grad_norm": 0.31617647409439087,
+      "learning_rate": 3.052873563218391e-05,
+      "loss": 0.0986,
+      "step": 3388
+    },
+    {
+      "epoch": 3.9284213576758678,
+      "grad_norm": 0.3103896379470825,
+      "learning_rate": 3.052298850574713e-05,
+      "loss": 0.1315,
+      "step": 3389
+    },
+    {
+      "epoch": 3.9295805259726144,
+      "grad_norm": 0.31774574518203735,
+      "learning_rate": 3.0517241379310348e-05,
+      "loss": 0.1116,
+      "step": 3390
+    },
+    {
+      "epoch": 3.930739694269362,
+      "grad_norm": 0.2169465720653534,
+      "learning_rate": 3.0511494252873563e-05,
+      "loss": 0.0987,
+      "step": 3391
+    },
+    {
+      "epoch": 3.9318988625661087,
+      "grad_norm": 0.27531975507736206,
+      "learning_rate": 3.0505747126436784e-05,
+      "loss": 0.1173,
+      "step": 3392
+    },
+    {
+      "epoch": 3.9330580308628558,
+      "grad_norm": 0.34132128953933716,
+      "learning_rate": 3.05e-05,
+      "loss": 0.1246,
+      "step": 3393
+    },
+    {
+      "epoch": 3.934217199159603,
+      "grad_norm": 0.3186652362346649,
+      "learning_rate": 3.0494252873563218e-05,
+      "loss": 0.1182,
+      "step": 3394
+    },
+    {
+      "epoch": 3.93537636745635,
+      "grad_norm": 0.2878797948360443,
+      "learning_rate": 3.048850574712644e-05,
+      "loss": 0.1155,
+      "step": 3395
+    },
+    {
+      "epoch": 3.936535535753097,
+      "grad_norm": 0.27938562631607056,
+      "learning_rate": 3.0482758620689654e-05,
+      "loss": 0.1024,
+      "step": 3396
+    },
+    {
+      "epoch": 3.9376947040498442,
+      "grad_norm": 0.3184402883052826,
+      "learning_rate": 3.0477011494252876e-05,
+      "loss": 0.1094,
+      "step": 3397
+    },
+    {
+      "epoch": 3.9388538723465913,
+      "grad_norm": 0.2606091797351837,
+      "learning_rate": 3.0471264367816094e-05,
+      "loss": 0.1115,
+      "step": 3398
+    },
+    {
+      "epoch": 3.9400130406433385,
+      "grad_norm": 0.3425968885421753,
+      "learning_rate": 3.046551724137931e-05,
+      "loss": 0.1085,
+      "step": 3399
+    },
+    {
+      "epoch": 3.9411722089400856,
+      "grad_norm": 0.30267462134361267,
+      "learning_rate": 3.045977011494253e-05,
+      "loss": 0.1239,
+      "step": 3400
+    },
+    {
+      "epoch": 3.9423313772368327,
+      "grad_norm": 0.31355491280555725,
+      "learning_rate": 3.045402298850575e-05,
+      "loss": 0.1179,
+      "step": 3401
+    },
+    {
+      "epoch": 3.94349054553358,
+      "grad_norm": 0.2975825071334839,
+      "learning_rate": 3.0448275862068964e-05,
+      "loss": 0.1146,
+      "step": 3402
+    },
+    {
+      "epoch": 3.9446497138303265,
+      "grad_norm": 0.3157767653465271,
+      "learning_rate": 3.0442528735632186e-05,
+      "loss": 0.1192,
+      "step": 3403
+    },
+    {
+      "epoch": 3.945808882127074,
+      "grad_norm": 0.2541486620903015,
+      "learning_rate": 3.0436781609195404e-05,
+      "loss": 0.0994,
+      "step": 3404
+    },
+    {
+      "epoch": 3.9469680504238207,
+      "grad_norm": 0.42458853125572205,
+      "learning_rate": 3.0431034482758626e-05,
+      "loss": 0.1295,
+      "step": 3405
+    },
+    {
+      "epoch": 3.9481272187205683,
+      "grad_norm": 0.28080934286117554,
+      "learning_rate": 3.042528735632184e-05,
+      "loss": 0.1086,
+      "step": 3406
+    },
+    {
+      "epoch": 3.949286387017315,
+      "grad_norm": 0.26001524925231934,
+      "learning_rate": 3.041954022988506e-05,
+      "loss": 0.1099,
+      "step": 3407
+    },
+    {
+      "epoch": 3.950445555314062,
+      "grad_norm": 0.2862829267978668,
+      "learning_rate": 3.041379310344828e-05,
+      "loss": 0.1153,
+      "step": 3408
+    },
+    {
+      "epoch": 3.951604723610809,
+      "grad_norm": 0.2606809735298157,
+      "learning_rate": 3.0408045977011495e-05,
+      "loss": 0.1151,
+      "step": 3409
+    },
+    {
+      "epoch": 3.9527638919075563,
+      "grad_norm": 0.3059811294078827,
+      "learning_rate": 3.040229885057471e-05,
+      "loss": 0.1151,
+      "step": 3410
+    },
+    {
+      "epoch": 3.9539230602043034,
+      "grad_norm": 0.30108916759490967,
+      "learning_rate": 3.0396551724137935e-05,
+      "loss": 0.1125,
+      "step": 3411
+    },
+    {
+      "epoch": 3.9550822285010505,
+      "grad_norm": 0.21490095555782318,
+      "learning_rate": 3.039080459770115e-05,
+      "loss": 0.0883,
+      "step": 3412
+    },
+    {
+      "epoch": 3.9562413967977976,
+      "grad_norm": 0.39761850237846375,
+      "learning_rate": 3.0385057471264365e-05,
+      "loss": 0.1313,
+      "step": 3413
+    },
+    {
+      "epoch": 3.9574005650945447,
+      "grad_norm": 0.33936530351638794,
+      "learning_rate": 3.0379310344827587e-05,
+      "loss": 0.1119,
+      "step": 3414
+    },
+    {
+      "epoch": 3.958559733391292,
+      "grad_norm": 0.33767202496528625,
+      "learning_rate": 3.0373563218390805e-05,
+      "loss": 0.1272,
+      "step": 3415
+    },
+    {
+      "epoch": 3.959718901688039,
+      "grad_norm": 0.31474098563194275,
+      "learning_rate": 3.0367816091954027e-05,
+      "loss": 0.1093,
+      "step": 3416
+    },
+    {
+      "epoch": 3.960878069984786,
+      "grad_norm": 0.394424170255661,
+      "learning_rate": 3.0362068965517242e-05,
+      "loss": 0.1262,
+      "step": 3417
+    },
+    {
+      "epoch": 3.9620372382815328,
+      "grad_norm": 0.3407604992389679,
+      "learning_rate": 3.035632183908046e-05,
+      "loss": 0.1184,
+      "step": 3418
+    },
+    {
+      "epoch": 3.9631964065782803,
+      "grad_norm": 0.25149932503700256,
+      "learning_rate": 3.035057471264368e-05,
+      "loss": 0.1003,
+      "step": 3419
+    },
+    {
+      "epoch": 3.964355574875027,
+      "grad_norm": 0.22654342651367188,
+      "learning_rate": 3.0344827586206897e-05,
+      "loss": 0.1024,
+      "step": 3420
+    },
+    {
+      "epoch": 3.965514743171774,
+      "grad_norm": 0.3020685911178589,
+      "learning_rate": 3.0339080459770115e-05,
+      "loss": 0.1096,
+      "step": 3421
+    },
+    {
+      "epoch": 3.966673911468521,
+      "grad_norm": 0.25613948702812195,
+      "learning_rate": 3.0333333333333337e-05,
+      "loss": 0.1232,
+      "step": 3422
+    },
+    {
+      "epoch": 3.9678330797652683,
+      "grad_norm": 0.3310185670852661,
+      "learning_rate": 3.032758620689655e-05,
+      "loss": 0.1222,
+      "step": 3423
+    },
+    {
+      "epoch": 3.9689922480620154,
+      "grad_norm": 0.26435863971710205,
+      "learning_rate": 3.0321839080459773e-05,
+      "loss": 0.1013,
+      "step": 3424
+    },
+    {
+      "epoch": 3.9701514163587626,
+      "grad_norm": 0.3062216341495514,
+      "learning_rate": 3.031609195402299e-05,
+      "loss": 0.1156,
+      "step": 3425
+    },
+    {
+      "epoch": 3.9713105846555097,
+      "grad_norm": 0.3526940643787384,
+      "learning_rate": 3.0310344827586206e-05,
+      "loss": 0.1171,
+      "step": 3426
+    },
+    {
+      "epoch": 3.972469752952257,
+      "grad_norm": 0.2813412547111511,
+      "learning_rate": 3.0304597701149428e-05,
+      "loss": 0.1031,
+      "step": 3427
+    },
+    {
+      "epoch": 3.973628921249004,
+      "grad_norm": 0.3743179738521576,
+      "learning_rate": 3.0298850574712646e-05,
+      "loss": 0.1138,
+      "step": 3428
+    },
+    {
+      "epoch": 3.974788089545751,
+      "grad_norm": 0.30746176838874817,
+      "learning_rate": 3.029310344827586e-05,
+      "loss": 0.1214,
+      "step": 3429
+    },
+    {
+      "epoch": 3.975947257842498,
+      "grad_norm": 0.3143618106842041,
+      "learning_rate": 3.0287356321839083e-05,
+      "loss": 0.1114,
+      "step": 3430
+    },
+    {
+      "epoch": 3.9771064261392453,
+      "grad_norm": 0.33030736446380615,
+      "learning_rate": 3.02816091954023e-05,
+      "loss": 0.1038,
+      "step": 3431
+    },
+    {
+      "epoch": 3.9782655944359924,
+      "grad_norm": 0.40963098406791687,
+      "learning_rate": 3.0275862068965523e-05,
+      "loss": 0.1182,
+      "step": 3432
+    },
+    {
+      "epoch": 3.979424762732739,
+      "grad_norm": 0.31964415311813354,
+      "learning_rate": 3.0270114942528738e-05,
+      "loss": 0.1074,
+      "step": 3433
+    },
+    {
+      "epoch": 3.9805839310294866,
+      "grad_norm": 0.24795931577682495,
+      "learning_rate": 3.0264367816091953e-05,
+      "loss": 0.1048,
+      "step": 3434
+    },
+    {
+      "epoch": 3.9817430993262333,
+      "grad_norm": 0.25050121545791626,
+      "learning_rate": 3.0258620689655178e-05,
+      "loss": 0.1026,
+      "step": 3435
+    },
+    {
+      "epoch": 3.9829022676229804,
+      "grad_norm": 0.27471792697906494,
+      "learning_rate": 3.0252873563218393e-05,
+      "loss": 0.1081,
+      "step": 3436
+    },
+    {
+      "epoch": 3.9840614359197275,
+      "grad_norm": 0.24730391800403595,
+      "learning_rate": 3.0247126436781608e-05,
+      "loss": 0.1041,
+      "step": 3437
+    },
+    {
+      "epoch": 3.9852206042164746,
+      "grad_norm": 0.24231289327144623,
+      "learning_rate": 3.024137931034483e-05,
+      "loss": 0.1031,
+      "step": 3438
+    },
+    {
+      "epoch": 3.9863797725132217,
+      "grad_norm": 0.3932971656322479,
+      "learning_rate": 3.0235632183908047e-05,
+      "loss": 0.1118,
+      "step": 3439
+    },
+    {
+      "epoch": 3.987538940809969,
+      "grad_norm": 0.28917720913887024,
+      "learning_rate": 3.0229885057471262e-05,
+      "loss": 0.1139,
+      "step": 3440
+    },
+    {
+      "epoch": 3.988698109106716,
+      "grad_norm": 0.2533641457557678,
+      "learning_rate": 3.0224137931034484e-05,
+      "loss": 0.0963,
+      "step": 3441
+    },
+    {
+      "epoch": 3.989857277403463,
+      "grad_norm": 0.303200364112854,
+      "learning_rate": 3.0218390804597702e-05,
+      "loss": 0.1199,
+      "step": 3442
+    },
+    {
+      "epoch": 3.99101644570021,
+      "grad_norm": 0.2749335765838623,
+      "learning_rate": 3.0212643678160924e-05,
+      "loss": 0.1025,
+      "step": 3443
+    },
+    {
+      "epoch": 3.9921756139969573,
+      "grad_norm": 0.2617204785346985,
+      "learning_rate": 3.020689655172414e-05,
+      "loss": 0.1123,
+      "step": 3444
+    },
+    {
+      "epoch": 3.9933347822937044,
+      "grad_norm": 0.2848406434059143,
+      "learning_rate": 3.0201149425287357e-05,
+      "loss": 0.1098,
+      "step": 3445
+    },
+    {
+      "epoch": 3.994493950590451,
+      "grad_norm": 0.3450290262699127,
+      "learning_rate": 3.019540229885058e-05,
+      "loss": 0.1087,
+      "step": 3446
+    },
+    {
+      "epoch": 3.9956531188871987,
+      "grad_norm": 0.3187498450279236,
+      "learning_rate": 3.0189655172413794e-05,
+      "loss": 0.1144,
+      "step": 3447
+    },
+    {
+      "epoch": 3.9968122871839453,
+      "grad_norm": 0.3266300857067108,
+      "learning_rate": 3.0183908045977012e-05,
+      "loss": 0.1161,
+      "step": 3448
+    },
+    {
+      "epoch": 3.9979714554806924,
+      "grad_norm": 0.2566640079021454,
+      "learning_rate": 3.0178160919540234e-05,
+      "loss": 0.1034,
+      "step": 3449
+    },
+    {
+      "epoch": 3.9991306237774396,
+      "grad_norm": 0.31494787335395813,
+      "learning_rate": 3.017241379310345e-05,
+      "loss": 0.1086,
+      "step": 3450
+    },
+    {
+      "epoch": 3.9991306237774396,
+      "eval_loss": 0.1329651027917862,
+      "eval_runtime": 265.5812,
+      "eval_samples_per_second": 5.776,
+      "eval_steps_per_second": 5.776,
+      "step": 3450
+    },
+    {
+      "epoch": 4.000289792074187,
+      "grad_norm": 0.3459082245826721,
+      "learning_rate": 3.016666666666667e-05,
+      "loss": 0.1203,
+      "step": 3451
+    },
+    {
+      "epoch": 4.001448960370934,
+      "grad_norm": 0.25177237391471863,
+      "learning_rate": 3.016091954022989e-05,
+      "loss": 0.1,
+      "step": 3452
+    },
+    {
+      "epoch": 4.002608128667681,
+      "grad_norm": 0.2541625201702118,
+      "learning_rate": 3.0155172413793104e-05,
+      "loss": 0.1011,
+      "step": 3453
+    },
+    {
+      "epoch": 4.003767296964428,
+      "grad_norm": 0.23528751730918884,
+      "learning_rate": 3.0149425287356325e-05,
+      "loss": 0.1055,
+      "step": 3454
+    },
+    {
+      "epoch": 4.004926465261175,
+      "grad_norm": 0.25770679116249084,
+      "learning_rate": 3.014367816091954e-05,
+      "loss": 0.0924,
+      "step": 3455
+    },
+    {
+      "epoch": 4.006085633557922,
+      "grad_norm": 0.24531324207782745,
+      "learning_rate": 3.013793103448276e-05,
+      "loss": 0.1031,
+      "step": 3456
+    },
+    {
+      "epoch": 4.007244801854669,
+      "grad_norm": 0.2368677854537964,
+      "learning_rate": 3.013218390804598e-05,
+      "loss": 0.1022,
+      "step": 3457
+    },
+    {
+      "epoch": 4.0084039701514165,
+      "grad_norm": 0.2580307722091675,
+      "learning_rate": 3.0126436781609195e-05,
+      "loss": 0.0883,
+      "step": 3458
+    },
+    {
+      "epoch": 4.009563138448163,
+      "grad_norm": 0.2701059579849243,
+      "learning_rate": 3.0120689655172413e-05,
+      "loss": 0.0988,
+      "step": 3459
+    },
+    {
+      "epoch": 4.010722306744911,
+      "grad_norm": 0.3617304563522339,
+      "learning_rate": 3.0114942528735635e-05,
+      "loss": 0.1142,
+      "step": 3460
+    },
+    {
+      "epoch": 4.011881475041657,
+      "grad_norm": 0.300418883562088,
+      "learning_rate": 3.010919540229885e-05,
+      "loss": 0.0961,
+      "step": 3461
+    },
+    {
+      "epoch": 4.013040643338405,
+      "grad_norm": 0.34037333726882935,
+      "learning_rate": 3.010344827586207e-05,
+      "loss": 0.1101,
+      "step": 3462
+    },
+    {
+      "epoch": 4.014199811635152,
+      "grad_norm": 0.30946558713912964,
+      "learning_rate": 3.009770114942529e-05,
+      "loss": 0.1017,
+      "step": 3463
+    },
+    {
+      "epoch": 4.015358979931899,
+      "grad_norm": 0.2514745891094208,
+      "learning_rate": 3.0091954022988505e-05,
+      "loss": 0.0895,
+      "step": 3464
+    },
+    {
+      "epoch": 4.016518148228646,
+      "grad_norm": 0.3249182403087616,
+      "learning_rate": 3.0086206896551726e-05,
+      "loss": 0.0983,
+      "step": 3465
+    },
+    {
+      "epoch": 4.017677316525393,
+      "grad_norm": 0.30697742104530334,
+      "learning_rate": 3.0080459770114945e-05,
+      "loss": 0.1012,
+      "step": 3466
+    },
+    {
+      "epoch": 4.01883648482214,
+      "grad_norm": 0.5410303473472595,
+      "learning_rate": 3.007471264367816e-05,
+      "loss": 0.101,
+      "step": 3467
+    },
+    {
+      "epoch": 4.019995653118888,
+      "grad_norm": 0.367801308631897,
+      "learning_rate": 3.006896551724138e-05,
+      "loss": 0.0972,
+      "step": 3468
+    },
+    {
+      "epoch": 4.021154821415634,
+      "grad_norm": 0.6262068152427673,
+      "learning_rate": 3.00632183908046e-05,
+      "loss": 0.1085,
+      "step": 3469
+    },
+    {
+      "epoch": 4.022313989712381,
+      "grad_norm": 0.39368757605552673,
+      "learning_rate": 3.005747126436782e-05,
+      "loss": 0.1074,
+      "step": 3470
+    },
+    {
+      "epoch": 4.0234731580091285,
+      "grad_norm": 0.3936510682106018,
+      "learning_rate": 3.0051724137931036e-05,
+      "loss": 0.0996,
+      "step": 3471
+    },
+    {
+      "epoch": 4.024632326305875,
+      "grad_norm": 0.5199762582778931,
+      "learning_rate": 3.0045977011494254e-05,
+      "loss": 0.1011,
+      "step": 3472
+    },
+    {
+      "epoch": 4.025791494602623,
+      "grad_norm": 0.3035762906074524,
+      "learning_rate": 3.0040229885057476e-05,
+      "loss": 0.103,
+      "step": 3473
+    },
+    {
+      "epoch": 4.026950662899369,
+      "grad_norm": 0.34868428111076355,
+      "learning_rate": 3.003448275862069e-05,
+      "loss": 0.1066,
+      "step": 3474
+    },
+    {
+      "epoch": 4.028109831196117,
+      "grad_norm": 0.4163806438446045,
+      "learning_rate": 3.0028735632183906e-05,
+      "loss": 0.103,
+      "step": 3475
+    },
+    {
+      "epoch": 4.029268999492864,
+      "grad_norm": 0.4000799059867859,
+      "learning_rate": 3.002298850574713e-05,
+      "loss": 0.0984,
+      "step": 3476
+    },
+    {
+      "epoch": 4.030428167789611,
+      "grad_norm": 0.3437565267086029,
+      "learning_rate": 3.0017241379310346e-05,
+      "loss": 0.0968,
+      "step": 3477
+    },
+    {
+      "epoch": 4.031587336086358,
+      "grad_norm": 0.2822440564632416,
+      "learning_rate": 3.001149425287356e-05,
+      "loss": 0.1045,
+      "step": 3478
+    },
+    {
+      "epoch": 4.0327465043831054,
+      "grad_norm": 0.44937509298324585,
+      "learning_rate": 3.0005747126436782e-05,
+      "loss": 0.1031,
+      "step": 3479
+    },
+    {
+      "epoch": 4.033905672679852,
+      "grad_norm": 0.667371392250061,
+      "learning_rate": 3e-05,
+      "loss": 0.112,
+      "step": 3480
+    },
+    {
+      "epoch": 4.0350648409766,
+      "grad_norm": 0.44788849353790283,
+      "learning_rate": 2.9994252873563222e-05,
+      "loss": 0.1159,
+      "step": 3481
+    },
+    {
+      "epoch": 4.036224009273346,
+      "grad_norm": 0.3105098009109497,
+      "learning_rate": 2.9988505747126437e-05,
+      "loss": 0.1164,
+      "step": 3482
+    },
+    {
+      "epoch": 4.037383177570093,
+      "grad_norm": 0.3047453463077545,
+      "learning_rate": 2.9982758620689656e-05,
+      "loss": 0.098,
+      "step": 3483
+    },
+    {
+      "epoch": 4.038542345866841,
+      "grad_norm": 0.39169877767562866,
+      "learning_rate": 2.9977011494252877e-05,
+      "loss": 0.0997,
+      "step": 3484
+    },
+    {
+      "epoch": 4.039701514163587,
+      "grad_norm": 0.31317397952079773,
+      "learning_rate": 2.9971264367816092e-05,
+      "loss": 0.1013,
+      "step": 3485
+    },
+    {
+      "epoch": 4.040860682460335,
+      "grad_norm": 0.396670937538147,
+      "learning_rate": 2.996551724137931e-05,
+      "loss": 0.1137,
+      "step": 3486
+    },
+    {
+      "epoch": 4.0420198507570815,
+      "grad_norm": 0.30599015951156616,
+      "learning_rate": 2.9959770114942532e-05,
+      "loss": 0.1096,
+      "step": 3487
+    },
+    {
+      "epoch": 4.043179019053829,
+      "grad_norm": 0.35810816287994385,
+      "learning_rate": 2.9954022988505747e-05,
+      "loss": 0.0991,
+      "step": 3488
+    },
+    {
+      "epoch": 4.044338187350576,
+      "grad_norm": 0.36599722504615784,
+      "learning_rate": 2.994827586206897e-05,
+      "loss": 0.1015,
+      "step": 3489
+    },
+    {
+      "epoch": 4.045497355647323,
+      "grad_norm": 0.30756571888923645,
+      "learning_rate": 2.9942528735632187e-05,
+      "loss": 0.0986,
+      "step": 3490
+    },
+    {
+      "epoch": 4.04665652394407,
+      "grad_norm": 0.42771488428115845,
+      "learning_rate": 2.9936781609195402e-05,
+      "loss": 0.1141,
+      "step": 3491
+    },
+    {
+      "epoch": 4.0478156922408175,
+      "grad_norm": 0.24768565595149994,
+      "learning_rate": 2.9931034482758624e-05,
+      "loss": 0.1063,
+      "step": 3492
+    },
+    {
+      "epoch": 4.048974860537564,
+      "grad_norm": 0.3938906490802765,
+      "learning_rate": 2.9925287356321842e-05,
+      "loss": 0.096,
+      "step": 3493
+    },
+    {
+      "epoch": 4.050134028834312,
+      "grad_norm": 0.3523421883583069,
+      "learning_rate": 2.9919540229885057e-05,
+      "loss": 0.1066,
+      "step": 3494
+    },
+    {
+      "epoch": 4.051293197131058,
+      "grad_norm": 0.32402533292770386,
+      "learning_rate": 2.991379310344828e-05,
+      "loss": 0.106,
+      "step": 3495
+    },
+    {
+      "epoch": 4.052452365427806,
+      "grad_norm": 0.3420311510562897,
+      "learning_rate": 2.9908045977011497e-05,
+      "loss": 0.0934,
+      "step": 3496
+    },
+    {
+      "epoch": 4.053611533724553,
+      "grad_norm": 0.5918967723846436,
+      "learning_rate": 2.990229885057471e-05,
+      "loss": 0.105,
+      "step": 3497
+    },
+    {
+      "epoch": 4.054770702021299,
+      "grad_norm": 0.2808108627796173,
+      "learning_rate": 2.9896551724137933e-05,
+      "loss": 0.1022,
+      "step": 3498
+    },
+    {
+      "epoch": 4.055929870318047,
+      "grad_norm": 0.43918702006340027,
+      "learning_rate": 2.9890804597701148e-05,
+      "loss": 0.1052,
+      "step": 3499
+    },
+    {
+      "epoch": 4.0570890386147935,
+      "grad_norm": 0.32599446177482605,
+      "learning_rate": 2.988505747126437e-05,
+      "loss": 0.0984,
+      "step": 3500
+    },
+    {
+      "epoch": 4.058248206911541,
+      "grad_norm": 0.3194628655910492,
+      "learning_rate": 2.9879310344827588e-05,
+      "loss": 0.0974,
+      "step": 3501
+    },
+    {
+      "epoch": 4.059407375208288,
+      "grad_norm": 0.358102947473526,
+      "learning_rate": 2.9873563218390803e-05,
+      "loss": 0.1023,
+      "step": 3502
+    },
+    {
+      "epoch": 4.060566543505035,
+      "grad_norm": 0.30983057618141174,
+      "learning_rate": 2.9867816091954025e-05,
+      "loss": 0.0888,
+      "step": 3503
+    },
+    {
+      "epoch": 4.061725711801782,
+      "grad_norm": 0.38468021154403687,
+      "learning_rate": 2.9862068965517243e-05,
+      "loss": 0.0959,
+      "step": 3504
+    },
+    {
+      "epoch": 4.0628848800985295,
+      "grad_norm": 0.4226651191711426,
+      "learning_rate": 2.9856321839080458e-05,
+      "loss": 0.1022,
+      "step": 3505
+    },
+    {
+      "epoch": 4.064044048395276,
+      "grad_norm": 0.5442459583282471,
+      "learning_rate": 2.985057471264368e-05,
+      "loss": 0.1106,
+      "step": 3506
+    },
+    {
+      "epoch": 4.065203216692024,
+      "grad_norm": 0.3217725455760956,
+      "learning_rate": 2.9844827586206898e-05,
+      "loss": 0.1021,
+      "step": 3507
+    },
+    {
+      "epoch": 4.0663623849887705,
+      "grad_norm": 0.30229008197784424,
+      "learning_rate": 2.983908045977012e-05,
+      "loss": 0.1031,
+      "step": 3508
+    },
+    {
+      "epoch": 4.067521553285518,
+      "grad_norm": 0.3460632562637329,
+      "learning_rate": 2.9833333333333335e-05,
+      "loss": 0.111,
+      "step": 3509
+    },
+    {
+      "epoch": 4.068680721582265,
+      "grad_norm": 0.38224828243255615,
+      "learning_rate": 2.9827586206896553e-05,
+      "loss": 0.1013,
+      "step": 3510
+    },
+    {
+      "epoch": 4.069839889879012,
+      "grad_norm": 0.35836493968963623,
+      "learning_rate": 2.9821839080459775e-05,
+      "loss": 0.105,
+      "step": 3511
+    },
+    {
+      "epoch": 4.070999058175759,
+      "grad_norm": 0.293371319770813,
+      "learning_rate": 2.981609195402299e-05,
+      "loss": 0.093,
+      "step": 3512
+    },
+    {
+      "epoch": 4.072158226472506,
+      "grad_norm": 0.37686097621917725,
+      "learning_rate": 2.9810344827586208e-05,
+      "loss": 0.1018,
+      "step": 3513
+    },
+    {
+      "epoch": 4.073317394769253,
+      "grad_norm": 0.30317679047584534,
+      "learning_rate": 2.980459770114943e-05,
+      "loss": 0.1095,
+      "step": 3514
+    },
+    {
+      "epoch": 4.074476563066,
+      "grad_norm": 0.3032994866371155,
+      "learning_rate": 2.9798850574712644e-05,
+      "loss": 0.1063,
+      "step": 3515
+    },
+    {
+      "epoch": 4.075635731362747,
+      "grad_norm": 0.29661819338798523,
+      "learning_rate": 2.979310344827586e-05,
+      "loss": 0.1068,
+      "step": 3516
+    },
+    {
+      "epoch": 4.076794899659494,
+      "grad_norm": 0.34293368458747864,
+      "learning_rate": 2.9787356321839084e-05,
+      "loss": 0.1063,
+      "step": 3517
+    },
+    {
+      "epoch": 4.077954067956242,
+      "grad_norm": 0.3552305996417999,
+      "learning_rate": 2.97816091954023e-05,
+      "loss": 0.1002,
+      "step": 3518
+    },
+    {
+      "epoch": 4.079113236252988,
+      "grad_norm": 0.3097629249095917,
+      "learning_rate": 2.977586206896552e-05,
+      "loss": 0.0955,
+      "step": 3519
+    },
+    {
+      "epoch": 4.080272404549736,
+      "grad_norm": 0.35458773374557495,
+      "learning_rate": 2.9770114942528736e-05,
+      "loss": 0.098,
+      "step": 3520
+    },
+    {
+      "epoch": 4.0814315728464825,
+      "grad_norm": 0.30128052830696106,
+      "learning_rate": 2.9764367816091954e-05,
+      "loss": 0.1087,
+      "step": 3521
+    },
+    {
+      "epoch": 4.08259074114323,
+      "grad_norm": 0.407092422246933,
+      "learning_rate": 2.9758620689655176e-05,
+      "loss": 0.0927,
+      "step": 3522
+    },
+    {
+      "epoch": 4.083749909439977,
+      "grad_norm": 0.43123817443847656,
+      "learning_rate": 2.975287356321839e-05,
+      "loss": 0.1045,
+      "step": 3523
+    },
+    {
+      "epoch": 4.084909077736724,
+      "grad_norm": 0.47122377157211304,
+      "learning_rate": 2.974712643678161e-05,
+      "loss": 0.1043,
+      "step": 3524
+    },
+    {
+      "epoch": 4.086068246033471,
+      "grad_norm": 0.4324159324169159,
+      "learning_rate": 2.974137931034483e-05,
+      "loss": 0.1095,
+      "step": 3525
+    },
+    {
+      "epoch": 4.087227414330218,
+      "grad_norm": 0.3203006982803345,
+      "learning_rate": 2.9735632183908045e-05,
+      "loss": 0.0967,
+      "step": 3526
+    },
+    {
+      "epoch": 4.088386582626965,
+      "grad_norm": 0.45111727714538574,
+      "learning_rate": 2.9729885057471267e-05,
+      "loss": 0.0979,
+      "step": 3527
+    },
+    {
+      "epoch": 4.089545750923712,
+      "grad_norm": 0.3182992935180664,
+      "learning_rate": 2.9724137931034485e-05,
+      "loss": 0.0957,
+      "step": 3528
+    },
+    {
+      "epoch": 4.090704919220459,
+      "grad_norm": 0.392347127199173,
+      "learning_rate": 2.97183908045977e-05,
+      "loss": 0.0982,
+      "step": 3529
+    },
+    {
+      "epoch": 4.091864087517206,
+      "grad_norm": 0.41814011335372925,
+      "learning_rate": 2.9712643678160922e-05,
+      "loss": 0.1017,
+      "step": 3530
+    },
+    {
+      "epoch": 4.093023255813954,
+      "grad_norm": 0.3611200749874115,
+      "learning_rate": 2.970689655172414e-05,
+      "loss": 0.0953,
+      "step": 3531
+    },
+    {
+      "epoch": 4.0941824241107,
+      "grad_norm": 0.3499966859817505,
+      "learning_rate": 2.9701149425287355e-05,
+      "loss": 0.1012,
+      "step": 3532
+    },
+    {
+      "epoch": 4.095341592407448,
+      "grad_norm": 0.33209457993507385,
+      "learning_rate": 2.9695402298850577e-05,
+      "loss": 0.0989,
+      "step": 3533
+    },
+    {
+      "epoch": 4.0965007607041946,
+      "grad_norm": 0.421478807926178,
+      "learning_rate": 2.9689655172413795e-05,
+      "loss": 0.1052,
+      "step": 3534
+    },
+    {
+      "epoch": 4.097659929000942,
+      "grad_norm": 0.3154635727405548,
+      "learning_rate": 2.968390804597701e-05,
+      "loss": 0.0984,
+      "step": 3535
+    },
+    {
+      "epoch": 4.098819097297689,
+      "grad_norm": 0.7844545841217041,
+      "learning_rate": 2.9678160919540232e-05,
+      "loss": 0.1113,
+      "step": 3536
+    },
+    {
+      "epoch": 4.099978265594436,
+      "grad_norm": 0.3256261646747589,
+      "learning_rate": 2.967241379310345e-05,
+      "loss": 0.0994,
+      "step": 3537
+    },
+    {
+      "epoch": 4.101137433891183,
+      "grad_norm": 0.4824637770652771,
+      "learning_rate": 2.9666666666666672e-05,
+      "loss": 0.1117,
+      "step": 3538
+    },
+    {
+      "epoch": 4.102296602187931,
+      "grad_norm": 0.46031931042671204,
+      "learning_rate": 2.9660919540229887e-05,
+      "loss": 0.1103,
+      "step": 3539
+    },
+    {
+      "epoch": 4.103455770484677,
+      "grad_norm": 0.33370453119277954,
+      "learning_rate": 2.96551724137931e-05,
+      "loss": 0.1121,
+      "step": 3540
+    },
+    {
+      "epoch": 4.104614938781424,
+      "grad_norm": 0.4630499482154846,
+      "learning_rate": 2.9649425287356327e-05,
+      "loss": 0.1128,
+      "step": 3541
+    },
+    {
+      "epoch": 4.1057741070781715,
+      "grad_norm": 0.47765010595321655,
+      "learning_rate": 2.964367816091954e-05,
+      "loss": 0.0989,
+      "step": 3542
+    },
+    {
+      "epoch": 4.106933275374918,
+      "grad_norm": 0.2912958562374115,
+      "learning_rate": 2.9637931034482756e-05,
+      "loss": 0.0968,
+      "step": 3543
+    },
+    {
+      "epoch": 4.108092443671666,
+      "grad_norm": 0.4124128520488739,
+      "learning_rate": 2.9632183908045978e-05,
+      "loss": 0.1004,
+      "step": 3544
+    },
+    {
+      "epoch": 4.109251611968412,
+      "grad_norm": 0.38116535544395447,
+      "learning_rate": 2.9626436781609196e-05,
+      "loss": 0.1079,
+      "step": 3545
+    },
+    {
+      "epoch": 4.11041078026516,
+      "grad_norm": 0.35159099102020264,
+      "learning_rate": 2.9620689655172418e-05,
+      "loss": 0.105,
+      "step": 3546
+    },
+    {
+      "epoch": 4.111569948561907,
+      "grad_norm": 0.38807129859924316,
+      "learning_rate": 2.9614942528735633e-05,
+      "loss": 0.1012,
+      "step": 3547
+    },
+    {
+      "epoch": 4.112729116858654,
+      "grad_norm": 0.33173301815986633,
+      "learning_rate": 2.960919540229885e-05,
+      "loss": 0.1213,
+      "step": 3548
+    },
+    {
+      "epoch": 4.113888285155401,
+      "grad_norm": 0.2943272888660431,
+      "learning_rate": 2.9603448275862073e-05,
+      "loss": 0.1063,
+      "step": 3549
+    },
+    {
+      "epoch": 4.115047453452148,
+      "grad_norm": 0.4288482666015625,
+      "learning_rate": 2.9597701149425288e-05,
+      "loss": 0.1047,
+      "step": 3550
+    },
+    {
+      "epoch": 4.116206621748895,
+      "grad_norm": 0.3704341948032379,
+      "learning_rate": 2.9591954022988506e-05,
+      "loss": 0.1166,
+      "step": 3551
+    },
+    {
+      "epoch": 4.117365790045643,
+      "grad_norm": 0.34703579545021057,
+      "learning_rate": 2.9586206896551728e-05,
+      "loss": 0.1068,
+      "step": 3552
+    },
+    {
+      "epoch": 4.118524958342389,
+      "grad_norm": 0.4057508111000061,
+      "learning_rate": 2.9580459770114943e-05,
+      "loss": 0.1048,
+      "step": 3553
+    },
+    {
+      "epoch": 4.119684126639137,
+      "grad_norm": 0.30694398283958435,
+      "learning_rate": 2.957471264367816e-05,
+      "loss": 0.1057,
+      "step": 3554
+    },
+    {
+      "epoch": 4.1208432949358835,
+      "grad_norm": 0.42865851521492004,
+      "learning_rate": 2.9568965517241383e-05,
+      "loss": 0.1085,
+      "step": 3555
+    },
+    {
+      "epoch": 4.12200246323263,
+      "grad_norm": 0.3558603525161743,
+      "learning_rate": 2.9563218390804598e-05,
+      "loss": 0.1029,
+      "step": 3556
+    },
+    {
+      "epoch": 4.123161631529378,
+      "grad_norm": 0.40978485345840454,
+      "learning_rate": 2.955747126436782e-05,
+      "loss": 0.095,
+      "step": 3557
+    },
+    {
+      "epoch": 4.124320799826124,
+      "grad_norm": 0.3812257647514343,
+      "learning_rate": 2.9551724137931038e-05,
+      "loss": 0.1052,
+      "step": 3558
+    },
+    {
+      "epoch": 4.125479968122872,
+      "grad_norm": 0.3055720925331116,
+      "learning_rate": 2.9545977011494252e-05,
+      "loss": 0.1113,
+      "step": 3559
+    },
+    {
+      "epoch": 4.126639136419619,
+      "grad_norm": 0.2221103310585022,
+      "learning_rate": 2.9540229885057474e-05,
+      "loss": 0.0886,
+      "step": 3560
+    },
+    {
+      "epoch": 4.127798304716366,
+      "grad_norm": 0.3389085531234741,
+      "learning_rate": 2.953448275862069e-05,
+      "loss": 0.113,
+      "step": 3561
+    },
+    {
+      "epoch": 4.128957473013113,
+      "grad_norm": 0.3829638659954071,
+      "learning_rate": 2.9528735632183907e-05,
+      "loss": 0.1048,
+      "step": 3562
+    },
+    {
+      "epoch": 4.1301166413098604,
+      "grad_norm": 0.5124924182891846,
+      "learning_rate": 2.952298850574713e-05,
+      "loss": 0.0942,
+      "step": 3563
+    },
+    {
+      "epoch": 4.131275809606607,
+      "grad_norm": 0.2896119952201843,
+      "learning_rate": 2.9517241379310344e-05,
+      "loss": 0.1041,
+      "step": 3564
+    },
+    {
+      "epoch": 4.132434977903355,
+      "grad_norm": 0.39416196942329407,
+      "learning_rate": 2.9511494252873566e-05,
+      "loss": 0.1046,
+      "step": 3565
+    },
+    {
+      "epoch": 4.133594146200101,
+      "grad_norm": 0.4558121860027313,
+      "learning_rate": 2.9505747126436784e-05,
+      "loss": 0.1077,
+      "step": 3566
+    },
+    {
+      "epoch": 4.134753314496849,
+      "grad_norm": 0.5235021114349365,
+      "learning_rate": 2.95e-05,
+      "loss": 0.0916,
+      "step": 3567
+    },
+    {
+      "epoch": 4.135912482793596,
+      "grad_norm": 0.4148746728897095,
+      "learning_rate": 2.949425287356322e-05,
+      "loss": 0.1049,
+      "step": 3568
+    },
+    {
+      "epoch": 4.137071651090342,
+      "grad_norm": 0.3960752785205841,
+      "learning_rate": 2.948850574712644e-05,
+      "loss": 0.0955,
+      "step": 3569
+    },
+    {
+      "epoch": 4.13823081938709,
+      "grad_norm": 0.2922149896621704,
+      "learning_rate": 2.9482758620689654e-05,
+      "loss": 0.1046,
+      "step": 3570
+    },
+    {
+      "epoch": 4.1393899876838365,
+      "grad_norm": 0.461208313703537,
+      "learning_rate": 2.9477011494252875e-05,
+      "loss": 0.1062,
+      "step": 3571
+    },
+    {
+      "epoch": 4.140549155980584,
+      "grad_norm": 0.35071003437042236,
+      "learning_rate": 2.9471264367816094e-05,
+      "loss": 0.1028,
+      "step": 3572
+    },
+    {
+      "epoch": 4.141708324277331,
+      "grad_norm": 0.3092256486415863,
+      "learning_rate": 2.946551724137931e-05,
+      "loss": 0.0979,
+      "step": 3573
+    },
+    {
+      "epoch": 4.142867492574078,
+      "grad_norm": 0.36416730284690857,
+      "learning_rate": 2.945977011494253e-05,
+      "loss": 0.1096,
+      "step": 3574
+    },
+    {
+      "epoch": 4.144026660870825,
+      "grad_norm": 0.4897642135620117,
+      "learning_rate": 2.945402298850575e-05,
+      "loss": 0.0942,
+      "step": 3575
+    },
+    {
+      "epoch": 4.1451858291675725,
+      "grad_norm": 0.3243999481201172,
+      "learning_rate": 2.944827586206897e-05,
+      "loss": 0.1,
+      "step": 3576
+    },
+    {
+      "epoch": 4.146344997464319,
+      "grad_norm": 0.3881404399871826,
+      "learning_rate": 2.9442528735632185e-05,
+      "loss": 0.1136,
+      "step": 3577
+    },
+    {
+      "epoch": 4.147504165761067,
+      "grad_norm": 0.3916066884994507,
+      "learning_rate": 2.9436781609195403e-05,
+      "loss": 0.1066,
+      "step": 3578
+    },
+    {
+      "epoch": 4.148663334057813,
+      "grad_norm": 0.3146432638168335,
+      "learning_rate": 2.9431034482758625e-05,
+      "loss": 0.0874,
+      "step": 3579
+    },
+    {
+      "epoch": 4.149822502354561,
+      "grad_norm": 0.3404291272163391,
+      "learning_rate": 2.942528735632184e-05,
+      "loss": 0.095,
+      "step": 3580
+    },
+    {
+      "epoch": 4.150981670651308,
+      "grad_norm": 0.4161677658557892,
+      "learning_rate": 2.9419540229885055e-05,
+      "loss": 0.0965,
+      "step": 3581
+    },
+    {
+      "epoch": 4.152140838948055,
+      "grad_norm": 0.38748228549957275,
+      "learning_rate": 2.941379310344828e-05,
+      "loss": 0.1148,
+      "step": 3582
+    },
+    {
+      "epoch": 4.153300007244802,
+      "grad_norm": 0.3862541615962982,
+      "learning_rate": 2.9408045977011495e-05,
+      "loss": 0.0994,
+      "step": 3583
+    },
+    {
+      "epoch": 4.1544591755415485,
+      "grad_norm": 0.4990377724170685,
+      "learning_rate": 2.9402298850574716e-05,
+      "loss": 0.1023,
+      "step": 3584
+    },
+    {
+      "epoch": 4.155618343838296,
+      "grad_norm": 0.3841322362422943,
+      "learning_rate": 2.939655172413793e-05,
+      "loss": 0.1137,
+      "step": 3585
+    },
+    {
+      "epoch": 4.156777512135043,
+      "grad_norm": 0.2818942368030548,
+      "learning_rate": 2.939080459770115e-05,
+      "loss": 0.0971,
+      "step": 3586
+    },
+    {
+      "epoch": 4.15793668043179,
+      "grad_norm": 0.4144044518470764,
+      "learning_rate": 2.938505747126437e-05,
+      "loss": 0.1,
+      "step": 3587
+    },
+    {
+      "epoch": 4.159095848728537,
+      "grad_norm": 0.265598326921463,
+      "learning_rate": 2.9379310344827586e-05,
+      "loss": 0.0911,
+      "step": 3588
+    },
+    {
+      "epoch": 4.1602550170252846,
+      "grad_norm": 0.40067845582962036,
+      "learning_rate": 2.9373563218390805e-05,
+      "loss": 0.1078,
+      "step": 3589
+    },
+    {
+      "epoch": 4.161414185322031,
+      "grad_norm": 0.3143431842327118,
+      "learning_rate": 2.9367816091954026e-05,
+      "loss": 0.0986,
+      "step": 3590
+    },
+    {
+      "epoch": 4.162573353618779,
+      "grad_norm": 0.2772219181060791,
+      "learning_rate": 2.936206896551724e-05,
+      "loss": 0.0944,
+      "step": 3591
+    },
+    {
+      "epoch": 4.1637325219155255,
+      "grad_norm": 0.4990723729133606,
+      "learning_rate": 2.935632183908046e-05,
+      "loss": 0.0989,
+      "step": 3592
+    },
+    {
+      "epoch": 4.164891690212273,
+      "grad_norm": 0.3033507168292999,
+      "learning_rate": 2.935057471264368e-05,
+      "loss": 0.1002,
+      "step": 3593
+    },
+    {
+      "epoch": 4.16605085850902,
+      "grad_norm": 0.28623032569885254,
+      "learning_rate": 2.9344827586206896e-05,
+      "loss": 0.1109,
+      "step": 3594
+    },
+    {
+      "epoch": 4.167210026805767,
+      "grad_norm": 0.28263723850250244,
+      "learning_rate": 2.9339080459770118e-05,
+      "loss": 0.1128,
+      "step": 3595
+    },
+    {
+      "epoch": 4.168369195102514,
+      "grad_norm": 0.7730621099472046,
+      "learning_rate": 2.9333333333333336e-05,
+      "loss": 0.1082,
+      "step": 3596
+    },
+    {
+      "epoch": 4.169528363399261,
+      "grad_norm": 0.306794673204422,
+      "learning_rate": 2.932758620689655e-05,
+      "loss": 0.1018,
+      "step": 3597
+    },
+    {
+      "epoch": 4.170687531696008,
+      "grad_norm": 0.3806617558002472,
+      "learning_rate": 2.9321839080459773e-05,
+      "loss": 0.1206,
+      "step": 3598
+    },
+    {
+      "epoch": 4.171846699992755,
+      "grad_norm": 0.3783591687679291,
+      "learning_rate": 2.931609195402299e-05,
+      "loss": 0.105,
+      "step": 3599
+    },
+    {
+      "epoch": 4.173005868289502,
+      "grad_norm": 0.49170562624931335,
+      "learning_rate": 2.9310344827586206e-05,
+      "loss": 0.1053,
+      "step": 3600
+    },
+    {
+      "epoch": 4.174165036586249,
+      "grad_norm": 0.36278071999549866,
+      "learning_rate": 2.9304597701149427e-05,
+      "loss": 0.1006,
+      "step": 3601
+    },
+    {
+      "epoch": 4.175324204882997,
+      "grad_norm": 0.4412696957588196,
+      "learning_rate": 2.9298850574712646e-05,
+      "loss": 0.1136,
+      "step": 3602
+    },
+    {
+      "epoch": 4.176483373179743,
+      "grad_norm": 0.33460313081741333,
+      "learning_rate": 2.9293103448275867e-05,
+      "loss": 0.1022,
+      "step": 3603
+    },
+    {
+      "epoch": 4.177642541476491,
+      "grad_norm": 0.32124656438827515,
+      "learning_rate": 2.9287356321839082e-05,
+      "loss": 0.1033,
+      "step": 3604
+    },
+    {
+      "epoch": 4.1788017097732375,
+      "grad_norm": 0.3389214873313904,
+      "learning_rate": 2.9281609195402297e-05,
+      "loss": 0.1074,
+      "step": 3605
+    },
+    {
+      "epoch": 4.179960878069985,
+      "grad_norm": 0.3551189601421356,
+      "learning_rate": 2.927586206896552e-05,
+      "loss": 0.0946,
+      "step": 3606
+    },
+    {
+      "epoch": 4.181120046366732,
+      "grad_norm": 0.3766753673553467,
+      "learning_rate": 2.9270114942528737e-05,
+      "loss": 0.0939,
+      "step": 3607
+    },
+    {
+      "epoch": 4.182279214663479,
+      "grad_norm": 0.362870454788208,
+      "learning_rate": 2.9264367816091952e-05,
+      "loss": 0.1079,
+      "step": 3608
+    },
+    {
+      "epoch": 4.183438382960226,
+      "grad_norm": 0.44273027777671814,
+      "learning_rate": 2.9258620689655174e-05,
+      "loss": 0.0969,
+      "step": 3609
+    },
+    {
+      "epoch": 4.1845975512569735,
+      "grad_norm": 0.3742962181568146,
+      "learning_rate": 2.9252873563218392e-05,
+      "loss": 0.1129,
+      "step": 3610
+    },
+    {
+      "epoch": 4.18575671955372,
+      "grad_norm": 0.4322942793369293,
+      "learning_rate": 2.9247126436781614e-05,
+      "loss": 0.1058,
+      "step": 3611
+    },
+    {
+      "epoch": 4.186915887850467,
+      "grad_norm": 0.44302043318748474,
+      "learning_rate": 2.924137931034483e-05,
+      "loss": 0.1062,
+      "step": 3612
+    },
+    {
+      "epoch": 4.188075056147214,
+      "grad_norm": 0.32489603757858276,
+      "learning_rate": 2.9235632183908047e-05,
+      "loss": 0.0974,
+      "step": 3613
+    },
+    {
+      "epoch": 4.189234224443961,
+      "grad_norm": 0.282850444316864,
+      "learning_rate": 2.922988505747127e-05,
+      "loss": 0.0941,
+      "step": 3614
+    },
+    {
+      "epoch": 4.190393392740709,
+      "grad_norm": 0.27374568581581116,
+      "learning_rate": 2.9224137931034483e-05,
+      "loss": 0.0963,
+      "step": 3615
+    },
+    {
+      "epoch": 4.191552561037455,
+      "grad_norm": 0.31600022315979004,
+      "learning_rate": 2.9218390804597702e-05,
+      "loss": 0.1027,
+      "step": 3616
+    },
+    {
+      "epoch": 4.192711729334203,
+      "grad_norm": 0.24971164762973785,
+      "learning_rate": 2.9212643678160923e-05,
+      "loss": 0.1103,
+      "step": 3617
+    },
+    {
+      "epoch": 4.19387089763095,
+      "grad_norm": 0.3458114564418793,
+      "learning_rate": 2.920689655172414e-05,
+      "loss": 0.0996,
+      "step": 3618
+    },
+    {
+      "epoch": 4.195030065927697,
+      "grad_norm": 0.3708142340183258,
+      "learning_rate": 2.9201149425287357e-05,
+      "loss": 0.0953,
+      "step": 3619
+    },
+    {
+      "epoch": 4.196189234224444,
+      "grad_norm": 0.3767480254173279,
+      "learning_rate": 2.919540229885058e-05,
+      "loss": 0.1081,
+      "step": 3620
+    },
+    {
+      "epoch": 4.197348402521191,
+      "grad_norm": 0.34636861085891724,
+      "learning_rate": 2.9189655172413793e-05,
+      "loss": 0.0996,
+      "step": 3621
+    },
+    {
+      "epoch": 4.198507570817938,
+      "grad_norm": 0.4181584119796753,
+      "learning_rate": 2.9183908045977015e-05,
+      "loss": 0.0951,
+      "step": 3622
+    },
+    {
+      "epoch": 4.199666739114686,
+      "grad_norm": 0.4052351117134094,
+      "learning_rate": 2.9178160919540233e-05,
+      "loss": 0.1059,
+      "step": 3623
+    },
+    {
+      "epoch": 4.200825907411432,
+      "grad_norm": 0.28023508191108704,
+      "learning_rate": 2.9172413793103448e-05,
+      "loss": 0.0992,
+      "step": 3624
+    },
+    {
+      "epoch": 4.20198507570818,
+      "grad_norm": 0.3520253896713257,
+      "learning_rate": 2.916666666666667e-05,
+      "loss": 0.107,
+      "step": 3625
+    },
+    {
+      "epoch": 4.2031442440049265,
+      "grad_norm": 0.3649381697177887,
+      "learning_rate": 2.9160919540229885e-05,
+      "loss": 0.114,
+      "step": 3626
+    },
+    {
+      "epoch": 4.204303412301673,
+      "grad_norm": 0.36334317922592163,
+      "learning_rate": 2.9155172413793103e-05,
+      "loss": 0.1012,
+      "step": 3627
+    },
+    {
+      "epoch": 4.205462580598421,
+      "grad_norm": 0.3084128797054291,
+      "learning_rate": 2.9149425287356325e-05,
+      "loss": 0.1035,
+      "step": 3628
+    },
+    {
+      "epoch": 4.206621748895167,
+      "grad_norm": 0.31549155712127686,
+      "learning_rate": 2.914367816091954e-05,
+      "loss": 0.1022,
+      "step": 3629
+    },
+    {
+      "epoch": 4.207780917191915,
+      "grad_norm": 0.33115145564079285,
+      "learning_rate": 2.913793103448276e-05,
+      "loss": 0.1034,
+      "step": 3630
+    },
+    {
+      "epoch": 4.208940085488662,
+      "grad_norm": 0.321877121925354,
+      "learning_rate": 2.913218390804598e-05,
+      "loss": 0.0948,
+      "step": 3631
+    },
+    {
+      "epoch": 4.210099253785409,
+      "grad_norm": 0.34342601895332336,
+      "learning_rate": 2.9126436781609194e-05,
+      "loss": 0.0987,
+      "step": 3632
+    },
+    {
+      "epoch": 4.211258422082156,
+      "grad_norm": 0.3793211579322815,
+      "learning_rate": 2.9120689655172416e-05,
+      "loss": 0.1032,
+      "step": 3633
+    },
+    {
+      "epoch": 4.212417590378903,
+      "grad_norm": 0.46431970596313477,
+      "learning_rate": 2.9114942528735634e-05,
+      "loss": 0.0997,
+      "step": 3634
+    },
+    {
+      "epoch": 4.21357675867565,
+      "grad_norm": 0.27705472707748413,
+      "learning_rate": 2.910919540229885e-05,
+      "loss": 0.0957,
+      "step": 3635
+    },
+    {
+      "epoch": 4.214735926972398,
+      "grad_norm": 0.5416485667228699,
+      "learning_rate": 2.910344827586207e-05,
+      "loss": 0.1021,
+      "step": 3636
+    },
+    {
+      "epoch": 4.215895095269144,
+      "grad_norm": 0.3653283715248108,
+      "learning_rate": 2.909770114942529e-05,
+      "loss": 0.0992,
+      "step": 3637
+    },
+    {
+      "epoch": 4.217054263565892,
+      "grad_norm": 0.34338125586509705,
+      "learning_rate": 2.9091954022988504e-05,
+      "loss": 0.1045,
+      "step": 3638
+    },
+    {
+      "epoch": 4.2182134318626385,
+      "grad_norm": 0.39666372537612915,
+      "learning_rate": 2.9086206896551726e-05,
+      "loss": 0.1029,
+      "step": 3639
+    },
+    {
+      "epoch": 4.219372600159385,
+      "grad_norm": 0.46274009346961975,
+      "learning_rate": 2.9080459770114944e-05,
+      "loss": 0.1097,
+      "step": 3640
+    },
+    {
+      "epoch": 4.220531768456133,
+      "grad_norm": 0.3507300615310669,
+      "learning_rate": 2.9074712643678166e-05,
+      "loss": 0.1001,
+      "step": 3641
+    },
+    {
+      "epoch": 4.221690936752879,
+      "grad_norm": 0.5658450126647949,
+      "learning_rate": 2.906896551724138e-05,
+      "loss": 0.1062,
+      "step": 3642
+    },
+    {
+      "epoch": 4.222850105049627,
+      "grad_norm": 0.33815431594848633,
+      "learning_rate": 2.90632183908046e-05,
+      "loss": 0.1002,
+      "step": 3643
+    },
+    {
+      "epoch": 4.224009273346374,
+      "grad_norm": 0.3951635956764221,
+      "learning_rate": 2.905747126436782e-05,
+      "loss": 0.0966,
+      "step": 3644
+    },
+    {
+      "epoch": 4.225168441643121,
+      "grad_norm": 0.4805266261100769,
+      "learning_rate": 2.9051724137931036e-05,
+      "loss": 0.1122,
+      "step": 3645
+    },
+    {
+      "epoch": 4.226327609939868,
+      "grad_norm": 0.3607132136821747,
+      "learning_rate": 2.904597701149425e-05,
+      "loss": 0.0924,
+      "step": 3646
+    },
+    {
+      "epoch": 4.2274867782366155,
+      "grad_norm": 0.3858696520328522,
+      "learning_rate": 2.9040229885057476e-05,
+      "loss": 0.0954,
+      "step": 3647
+    },
+    {
+      "epoch": 4.228645946533362,
+      "grad_norm": 0.35056713223457336,
+      "learning_rate": 2.903448275862069e-05,
+      "loss": 0.1034,
+      "step": 3648
+    },
+    {
+      "epoch": 4.22980511483011,
+      "grad_norm": 0.45169034600257874,
+      "learning_rate": 2.9028735632183912e-05,
+      "loss": 0.1089,
+      "step": 3649
+    },
+    {
+      "epoch": 4.230964283126856,
+      "grad_norm": 0.33204302191734314,
+      "learning_rate": 2.9022988505747127e-05,
+      "loss": 0.1078,
+      "step": 3650
+    },
+    {
+      "epoch": 4.232123451423604,
+      "grad_norm": 0.4049871563911438,
+      "learning_rate": 2.9017241379310345e-05,
+      "loss": 0.1022,
+      "step": 3651
+    },
+    {
+      "epoch": 4.233282619720351,
+      "grad_norm": 0.37815043330192566,
+      "learning_rate": 2.9011494252873567e-05,
+      "loss": 0.1025,
+      "step": 3652
+    },
+    {
+      "epoch": 4.234441788017098,
+      "grad_norm": 0.6017569899559021,
+      "learning_rate": 2.9005747126436782e-05,
+      "loss": 0.1095,
+      "step": 3653
+    },
+    {
+      "epoch": 4.235600956313845,
+      "grad_norm": 0.44660142064094543,
+      "learning_rate": 2.9e-05,
+      "loss": 0.1027,
+      "step": 3654
+    },
+    {
+      "epoch": 4.2367601246105915,
+      "grad_norm": 0.22865894436836243,
+      "learning_rate": 2.8994252873563222e-05,
+      "loss": 0.0953,
+      "step": 3655
+    },
+    {
+      "epoch": 4.237919292907339,
+      "grad_norm": 0.28317615389823914,
+      "learning_rate": 2.8988505747126437e-05,
+      "loss": 0.1009,
+      "step": 3656
+    },
+    {
+      "epoch": 4.239078461204086,
+      "grad_norm": 0.30559006333351135,
+      "learning_rate": 2.8982758620689655e-05,
+      "loss": 0.1079,
+      "step": 3657
+    },
+    {
+      "epoch": 4.240237629500833,
+      "grad_norm": 0.3804382383823395,
+      "learning_rate": 2.8977011494252877e-05,
+      "loss": 0.115,
+      "step": 3658
+    },
+    {
+      "epoch": 4.24139679779758,
+      "grad_norm": 0.4205724000930786,
+      "learning_rate": 2.897126436781609e-05,
+      "loss": 0.1013,
+      "step": 3659
+    },
+    {
+      "epoch": 4.2425559660943275,
+      "grad_norm": 0.3758813142776489,
+      "learning_rate": 2.8965517241379313e-05,
+      "loss": 0.103,
+      "step": 3660
+    },
+    {
+      "epoch": 4.243715134391074,
+      "grad_norm": 0.6086719632148743,
+      "learning_rate": 2.895977011494253e-05,
+      "loss": 0.1231,
+      "step": 3661
+    },
+    {
+      "epoch": 4.244874302687822,
+      "grad_norm": 0.4006671607494354,
+      "learning_rate": 2.8954022988505746e-05,
+      "loss": 0.0924,
+      "step": 3662
+    },
+    {
+      "epoch": 4.246033470984568,
+      "grad_norm": 0.41011175513267517,
+      "learning_rate": 2.8948275862068968e-05,
+      "loss": 0.0992,
+      "step": 3663
+    },
+    {
+      "epoch": 4.247192639281316,
+      "grad_norm": 0.4147280156612396,
+      "learning_rate": 2.8942528735632186e-05,
+      "loss": 0.1034,
+      "step": 3664
+    },
+    {
+      "epoch": 4.248351807578063,
+      "grad_norm": 0.32018283009529114,
+      "learning_rate": 2.89367816091954e-05,
+      "loss": 0.0927,
+      "step": 3665
+    },
+    {
+      "epoch": 4.24951097587481,
+      "grad_norm": 0.3739756643772125,
+      "learning_rate": 2.8931034482758623e-05,
+      "loss": 0.11,
+      "step": 3666
+    },
+    {
+      "epoch": 4.250670144171557,
+      "grad_norm": 0.3220639228820801,
+      "learning_rate": 2.8925287356321838e-05,
+      "loss": 0.0968,
+      "step": 3667
+    },
+    {
+      "epoch": 4.251829312468304,
+      "grad_norm": 0.3207123279571533,
+      "learning_rate": 2.8919540229885063e-05,
+      "loss": 0.1002,
+      "step": 3668
+    },
+    {
+      "epoch": 4.252988480765051,
+      "grad_norm": 0.38051706552505493,
+      "learning_rate": 2.8913793103448278e-05,
+      "loss": 0.1089,
+      "step": 3669
+    },
+    {
+      "epoch": 4.254147649061798,
+      "grad_norm": 0.3652991056442261,
+      "learning_rate": 2.8908045977011493e-05,
+      "loss": 0.1044,
+      "step": 3670
+    },
+    {
+      "epoch": 4.255306817358545,
+      "grad_norm": 0.4439450204372406,
+      "learning_rate": 2.8902298850574714e-05,
+      "loss": 0.1028,
+      "step": 3671
+    },
+    {
+      "epoch": 4.256465985655292,
+      "grad_norm": 0.2599686086177826,
+      "learning_rate": 2.8896551724137933e-05,
+      "loss": 0.0919,
+      "step": 3672
+    },
+    {
+      "epoch": 4.2576251539520396,
+      "grad_norm": 0.33243459463119507,
+      "learning_rate": 2.8890804597701148e-05,
+      "loss": 0.1044,
+      "step": 3673
+    },
+    {
+      "epoch": 4.258784322248786,
+      "grad_norm": 0.305882066488266,
+      "learning_rate": 2.888505747126437e-05,
+      "loss": 0.0968,
+      "step": 3674
+    },
+    {
+      "epoch": 4.259943490545534,
+      "grad_norm": 0.403209388256073,
+      "learning_rate": 2.8879310344827588e-05,
+      "loss": 0.1124,
+      "step": 3675
+    },
+    {
+      "epoch": 4.2611026588422805,
+      "grad_norm": 0.32304686307907104,
+      "learning_rate": 2.8873563218390803e-05,
+      "loss": 0.0908,
+      "step": 3676
+    },
+    {
+      "epoch": 4.262261827139028,
+      "grad_norm": 0.3692949116230011,
+      "learning_rate": 2.8867816091954024e-05,
+      "loss": 0.1067,
+      "step": 3677
+    },
+    {
+      "epoch": 4.263420995435775,
+      "grad_norm": 0.2813093066215515,
+      "learning_rate": 2.8862068965517243e-05,
+      "loss": 0.1046,
+      "step": 3678
+    },
+    {
+      "epoch": 4.264580163732522,
+      "grad_norm": 0.28762635588645935,
+      "learning_rate": 2.8856321839080464e-05,
+      "loss": 0.0968,
+      "step": 3679
+    },
+    {
+      "epoch": 4.265739332029269,
+      "grad_norm": 0.447125643491745,
+      "learning_rate": 2.885057471264368e-05,
+      "loss": 0.115,
+      "step": 3680
+    },
+    {
+      "epoch": 4.2668985003260165,
+      "grad_norm": 0.32157212495803833,
+      "learning_rate": 2.8844827586206897e-05,
+      "loss": 0.0963,
+      "step": 3681
+    },
+    {
+      "epoch": 4.268057668622763,
+      "grad_norm": 0.5032519698143005,
+      "learning_rate": 2.883908045977012e-05,
+      "loss": 0.107,
+      "step": 3682
+    },
+    {
+      "epoch": 4.26921683691951,
+      "grad_norm": 0.2923990786075592,
+      "learning_rate": 2.8833333333333334e-05,
+      "loss": 0.0884,
+      "step": 3683
+    },
+    {
+      "epoch": 4.270376005216257,
+      "grad_norm": 0.4205268621444702,
+      "learning_rate": 2.8827586206896552e-05,
+      "loss": 0.1035,
+      "step": 3684
+    },
+    {
+      "epoch": 4.271535173513004,
+      "grad_norm": 0.343766987323761,
+      "learning_rate": 2.8821839080459774e-05,
+      "loss": 0.1016,
+      "step": 3685
+    },
+    {
+      "epoch": 4.272694341809752,
+      "grad_norm": 0.39418303966522217,
+      "learning_rate": 2.881609195402299e-05,
+      "loss": 0.1057,
+      "step": 3686
+    },
+    {
+      "epoch": 4.273853510106498,
+      "grad_norm": 0.41105031967163086,
+      "learning_rate": 2.881034482758621e-05,
+      "loss": 0.0932,
+      "step": 3687
+    },
+    {
+      "epoch": 4.275012678403246,
+      "grad_norm": 0.32232189178466797,
+      "learning_rate": 2.880459770114943e-05,
+      "loss": 0.1043,
+      "step": 3688
+    },
+    {
+      "epoch": 4.2761718466999925,
+      "grad_norm": 0.49274659156799316,
+      "learning_rate": 2.8798850574712644e-05,
+      "loss": 0.1039,
+      "step": 3689
+    },
+    {
+      "epoch": 4.27733101499674,
+      "grad_norm": 0.5565973520278931,
+      "learning_rate": 2.8793103448275865e-05,
+      "loss": 0.1054,
+      "step": 3690
+    },
+    {
+      "epoch": 4.278490183293487,
+      "grad_norm": 0.366921603679657,
+      "learning_rate": 2.878735632183908e-05,
+      "loss": 0.1087,
+      "step": 3691
+    },
+    {
+      "epoch": 4.279649351590234,
+      "grad_norm": 0.3326165974140167,
+      "learning_rate": 2.87816091954023e-05,
+      "loss": 0.0971,
+      "step": 3692
+    },
+    {
+      "epoch": 4.280808519886981,
+      "grad_norm": 0.46312418580055237,
+      "learning_rate": 2.877586206896552e-05,
+      "loss": 0.1038,
+      "step": 3693
+    },
+    {
+      "epoch": 4.2819676881837285,
+      "grad_norm": 0.2820906639099121,
+      "learning_rate": 2.8770114942528735e-05,
+      "loss": 0.0958,
+      "step": 3694
+    },
+    {
+      "epoch": 4.283126856480475,
+      "grad_norm": 0.3850758969783783,
+      "learning_rate": 2.8764367816091953e-05,
+      "loss": 0.1098,
+      "step": 3695
+    },
+    {
+      "epoch": 4.284286024777223,
+      "grad_norm": 0.4358065128326416,
+      "learning_rate": 2.8758620689655175e-05,
+      "loss": 0.1004,
+      "step": 3696
+    },
+    {
+      "epoch": 4.285445193073969,
+      "grad_norm": 0.3014400601387024,
+      "learning_rate": 2.875287356321839e-05,
+      "loss": 0.1031,
+      "step": 3697
+    },
+    {
+      "epoch": 4.286604361370716,
+      "grad_norm": 0.30815666913986206,
+      "learning_rate": 2.8747126436781612e-05,
+      "loss": 0.1043,
+      "step": 3698
+    },
+    {
+      "epoch": 4.287763529667464,
+      "grad_norm": 0.39481833577156067,
+      "learning_rate": 2.874137931034483e-05,
+      "loss": 0.1068,
+      "step": 3699
+    },
+    {
+      "epoch": 4.28892269796421,
+      "grad_norm": 0.28407761454582214,
+      "learning_rate": 2.8735632183908045e-05,
+      "loss": 0.0992,
+      "step": 3700
+    },
+    {
+      "epoch": 4.290081866260958,
+      "grad_norm": 0.3631652891635895,
+      "learning_rate": 2.8729885057471267e-05,
+      "loss": 0.1069,
+      "step": 3701
+    },
+    {
+      "epoch": 4.291241034557705,
+      "grad_norm": 0.3379894495010376,
+      "learning_rate": 2.8724137931034485e-05,
+      "loss": 0.1074,
+      "step": 3702
+    },
+    {
+      "epoch": 4.292400202854452,
+      "grad_norm": 0.29723644256591797,
+      "learning_rate": 2.87183908045977e-05,
+      "loss": 0.1051,
+      "step": 3703
+    },
+    {
+      "epoch": 4.293559371151199,
+      "grad_norm": 0.4353717863559723,
+      "learning_rate": 2.871264367816092e-05,
+      "loss": 0.1067,
+      "step": 3704
+    },
+    {
+      "epoch": 4.294718539447946,
+      "grad_norm": 0.30414873361587524,
+      "learning_rate": 2.870689655172414e-05,
+      "loss": 0.0915,
+      "step": 3705
+    },
+    {
+      "epoch": 4.295877707744693,
+      "grad_norm": 0.5142664909362793,
+      "learning_rate": 2.870114942528736e-05,
+      "loss": 0.1019,
+      "step": 3706
+    },
+    {
+      "epoch": 4.297036876041441,
+      "grad_norm": 0.47813358902931213,
+      "learning_rate": 2.8695402298850576e-05,
+      "loss": 0.0917,
+      "step": 3707
+    },
+    {
+      "epoch": 4.298196044338187,
+      "grad_norm": 0.418399840593338,
+      "learning_rate": 2.8689655172413795e-05,
+      "loss": 0.1045,
+      "step": 3708
+    },
+    {
+      "epoch": 4.299355212634935,
+      "grad_norm": 0.36528536677360535,
+      "learning_rate": 2.8683908045977016e-05,
+      "loss": 0.112,
+      "step": 3709
+    },
+    {
+      "epoch": 4.3005143809316815,
+      "grad_norm": 0.41568946838378906,
+      "learning_rate": 2.867816091954023e-05,
+      "loss": 0.0988,
+      "step": 3710
+    },
+    {
+      "epoch": 4.301673549228429,
+      "grad_norm": 0.3442577123641968,
+      "learning_rate": 2.8672413793103446e-05,
+      "loss": 0.0949,
+      "step": 3711
+    },
+    {
+      "epoch": 4.302832717525176,
+      "grad_norm": 0.4584302604198456,
+      "learning_rate": 2.8666666666666668e-05,
+      "loss": 0.1037,
+      "step": 3712
+    },
+    {
+      "epoch": 4.303991885821922,
+      "grad_norm": 0.36319372057914734,
+      "learning_rate": 2.8660919540229886e-05,
+      "loss": 0.1082,
+      "step": 3713
+    },
+    {
+      "epoch": 4.30515105411867,
+      "grad_norm": 0.2930583953857422,
+      "learning_rate": 2.86551724137931e-05,
+      "loss": 0.0955,
+      "step": 3714
+    },
+    {
+      "epoch": 4.306310222415417,
+      "grad_norm": 0.4411957859992981,
+      "learning_rate": 2.8649425287356323e-05,
+      "loss": 0.1108,
+      "step": 3715
+    },
+    {
+      "epoch": 4.307469390712164,
+      "grad_norm": 0.3296014070510864,
+      "learning_rate": 2.864367816091954e-05,
+      "loss": 0.1025,
+      "step": 3716
+    },
+    {
+      "epoch": 4.308628559008911,
+      "grad_norm": 0.3820302188396454,
+      "learning_rate": 2.8637931034482763e-05,
+      "loss": 0.1105,
+      "step": 3717
+    },
+    {
+      "epoch": 4.309787727305658,
+      "grad_norm": 0.30717596411705017,
+      "learning_rate": 2.8632183908045978e-05,
+      "loss": 0.1037,
+      "step": 3718
+    },
+    {
+      "epoch": 4.310946895602405,
+      "grad_norm": 0.4110417664051056,
+      "learning_rate": 2.8626436781609196e-05,
+      "loss": 0.1076,
+      "step": 3719
+    },
+    {
+      "epoch": 4.312106063899153,
+      "grad_norm": 0.3443779945373535,
+      "learning_rate": 2.8620689655172417e-05,
+      "loss": 0.0982,
+      "step": 3720
+    },
+    {
+      "epoch": 4.313265232195899,
+      "grad_norm": 0.2677842080593109,
+      "learning_rate": 2.8614942528735632e-05,
+      "loss": 0.1009,
+      "step": 3721
+    },
+    {
+      "epoch": 4.314424400492647,
+      "grad_norm": 0.28737112879753113,
+      "learning_rate": 2.860919540229885e-05,
+      "loss": 0.0953,
+      "step": 3722
+    },
+    {
+      "epoch": 4.3155835687893935,
+      "grad_norm": 0.3234851658344269,
+      "learning_rate": 2.8603448275862072e-05,
+      "loss": 0.0992,
+      "step": 3723
+    },
+    {
+      "epoch": 4.316742737086141,
+      "grad_norm": 0.2924320101737976,
+      "learning_rate": 2.8597701149425287e-05,
+      "loss": 0.1053,
+      "step": 3724
+    },
+    {
+      "epoch": 4.317901905382888,
+      "grad_norm": 0.42933475971221924,
+      "learning_rate": 2.859195402298851e-05,
+      "loss": 0.1063,
+      "step": 3725
+    },
+    {
+      "epoch": 4.319061073679634,
+      "grad_norm": 0.34258314967155457,
+      "learning_rate": 2.8586206896551727e-05,
+      "loss": 0.1059,
+      "step": 3726
+    },
+    {
+      "epoch": 4.320220241976382,
+      "grad_norm": 0.3477523624897003,
+      "learning_rate": 2.8580459770114942e-05,
+      "loss": 0.0947,
+      "step": 3727
+    },
+    {
+      "epoch": 4.321379410273129,
+      "grad_norm": 0.5333256125450134,
+      "learning_rate": 2.8574712643678164e-05,
+      "loss": 0.1111,
+      "step": 3728
+    },
+    {
+      "epoch": 4.322538578569876,
+      "grad_norm": 0.3772802948951721,
+      "learning_rate": 2.8568965517241382e-05,
+      "loss": 0.0991,
+      "step": 3729
+    },
+    {
+      "epoch": 4.323697746866623,
+      "grad_norm": 0.415400892496109,
+      "learning_rate": 2.8563218390804597e-05,
+      "loss": 0.108,
+      "step": 3730
+    },
+    {
+      "epoch": 4.3248569151633705,
+      "grad_norm": 0.34160271286964417,
+      "learning_rate": 2.855747126436782e-05,
+      "loss": 0.0992,
+      "step": 3731
+    },
+    {
+      "epoch": 4.326016083460117,
+      "grad_norm": 0.29234248399734497,
+      "learning_rate": 2.8551724137931034e-05,
+      "loss": 0.1012,
+      "step": 3732
+    },
+    {
+      "epoch": 4.327175251756865,
+      "grad_norm": 0.36780115962028503,
+      "learning_rate": 2.8545977011494252e-05,
+      "loss": 0.1029,
+      "step": 3733
+    },
+    {
+      "epoch": 4.328334420053611,
+      "grad_norm": 0.3559337556362152,
+      "learning_rate": 2.8540229885057474e-05,
+      "loss": 0.116,
+      "step": 3734
+    },
+    {
+      "epoch": 4.329493588350359,
+      "grad_norm": 0.35530203580856323,
+      "learning_rate": 2.853448275862069e-05,
+      "loss": 0.1117,
+      "step": 3735
+    },
+    {
+      "epoch": 4.330652756647106,
+      "grad_norm": 0.39250123500823975,
+      "learning_rate": 2.852873563218391e-05,
+      "loss": 0.1021,
+      "step": 3736
+    },
+    {
+      "epoch": 4.331811924943853,
+      "grad_norm": 0.3159255087375641,
+      "learning_rate": 2.852298850574713e-05,
+      "loss": 0.1061,
+      "step": 3737
+    },
+    {
+      "epoch": 4.3329710932406,
+      "grad_norm": 0.28667929768562317,
+      "learning_rate": 2.8517241379310343e-05,
+      "loss": 0.0963,
+      "step": 3738
+    },
+    {
+      "epoch": 4.3341302615373465,
+      "grad_norm": 0.49704504013061523,
+      "learning_rate": 2.8511494252873565e-05,
+      "loss": 0.1079,
+      "step": 3739
+    },
+    {
+      "epoch": 4.335289429834094,
+      "grad_norm": 0.383167564868927,
+      "learning_rate": 2.8505747126436783e-05,
+      "loss": 0.1031,
+      "step": 3740
+    },
+    {
+      "epoch": 4.336448598130841,
+      "grad_norm": 0.28818759322166443,
+      "learning_rate": 2.8499999999999998e-05,
+      "loss": 0.1103,
+      "step": 3741
+    },
+    {
+      "epoch": 4.337607766427588,
+      "grad_norm": 0.3614335358142853,
+      "learning_rate": 2.849425287356322e-05,
+      "loss": 0.1032,
+      "step": 3742
+    },
+    {
+      "epoch": 4.338766934724335,
+      "grad_norm": 0.3309556841850281,
+      "learning_rate": 2.8488505747126438e-05,
+      "loss": 0.1053,
+      "step": 3743
+    },
+    {
+      "epoch": 4.3399261030210825,
+      "grad_norm": 0.4003108739852905,
+      "learning_rate": 2.848275862068966e-05,
+      "loss": 0.1038,
+      "step": 3744
+    },
+    {
+      "epoch": 4.341085271317829,
+      "grad_norm": 0.30309104919433594,
+      "learning_rate": 2.8477011494252875e-05,
+      "loss": 0.1019,
+      "step": 3745
+    },
+    {
+      "epoch": 4.342244439614577,
+      "grad_norm": 0.3004578649997711,
+      "learning_rate": 2.8471264367816093e-05,
+      "loss": 0.0911,
+      "step": 3746
+    },
+    {
+      "epoch": 4.343403607911323,
+      "grad_norm": 0.3777802586555481,
+      "learning_rate": 2.8465517241379315e-05,
+      "loss": 0.0995,
+      "step": 3747
+    },
+    {
+      "epoch": 4.344562776208071,
+      "grad_norm": 0.29196664690971375,
+      "learning_rate": 2.845977011494253e-05,
+      "loss": 0.101,
+      "step": 3748
+    },
+    {
+      "epoch": 4.345721944504818,
+      "grad_norm": 0.4231297969818115,
+      "learning_rate": 2.8454022988505748e-05,
+      "loss": 0.1033,
+      "step": 3749
+    },
+    {
+      "epoch": 4.346881112801565,
+      "grad_norm": 0.38380828499794006,
+      "learning_rate": 2.844827586206897e-05,
+      "loss": 0.1193,
+      "step": 3750
+    },
+    {
+      "epoch": 4.348040281098312,
+      "grad_norm": 0.37438878417015076,
+      "learning_rate": 2.8442528735632184e-05,
+      "loss": 0.0972,
+      "step": 3751
+    },
+    {
+      "epoch": 4.349199449395059,
+      "grad_norm": 0.29754364490509033,
+      "learning_rate": 2.84367816091954e-05,
+      "loss": 0.0936,
+      "step": 3752
+    },
+    {
+      "epoch": 4.350358617691806,
+      "grad_norm": 0.2982840836048126,
+      "learning_rate": 2.8431034482758624e-05,
+      "loss": 0.0892,
+      "step": 3753
+    },
+    {
+      "epoch": 4.351517785988554,
+      "grad_norm": 0.4285561740398407,
+      "learning_rate": 2.842528735632184e-05,
+      "loss": 0.0987,
+      "step": 3754
+    },
+    {
+      "epoch": 4.3526769542853,
+      "grad_norm": 0.356584370136261,
+      "learning_rate": 2.841954022988506e-05,
+      "loss": 0.1078,
+      "step": 3755
+    },
+    {
+      "epoch": 4.353836122582047,
+      "grad_norm": 0.4146307408809662,
+      "learning_rate": 2.8413793103448276e-05,
+      "loss": 0.1098,
+      "step": 3756
+    },
+    {
+      "epoch": 4.354995290878795,
+      "grad_norm": 0.39184170961380005,
+      "learning_rate": 2.8408045977011494e-05,
+      "loss": 0.0973,
+      "step": 3757
+    },
+    {
+      "epoch": 4.356154459175541,
+      "grad_norm": 0.38053470849990845,
+      "learning_rate": 2.8402298850574716e-05,
+      "loss": 0.106,
+      "step": 3758
+    },
+    {
+      "epoch": 4.357313627472289,
+      "grad_norm": 0.4440837800502777,
+      "learning_rate": 2.839655172413793e-05,
+      "loss": 0.1195,
+      "step": 3759
+    },
+    {
+      "epoch": 4.3584727957690355,
+      "grad_norm": 0.4450792670249939,
+      "learning_rate": 2.839080459770115e-05,
+      "loss": 0.1102,
+      "step": 3760
+    },
+    {
+      "epoch": 4.359631964065783,
+      "grad_norm": 0.29842203855514526,
+      "learning_rate": 2.838505747126437e-05,
+      "loss": 0.1029,
+      "step": 3761
+    },
+    {
+      "epoch": 4.36079113236253,
+      "grad_norm": 0.35125017166137695,
+      "learning_rate": 2.8379310344827586e-05,
+      "loss": 0.1055,
+      "step": 3762
+    },
+    {
+      "epoch": 4.361950300659277,
+      "grad_norm": 0.43805792927742004,
+      "learning_rate": 2.8373563218390807e-05,
+      "loss": 0.1018,
+      "step": 3763
+    },
+    {
+      "epoch": 4.363109468956024,
+      "grad_norm": 0.4038124978542328,
+      "learning_rate": 2.8367816091954026e-05,
+      "loss": 0.1019,
+      "step": 3764
+    },
+    {
+      "epoch": 4.3642686372527715,
+      "grad_norm": 0.3258083462715149,
+      "learning_rate": 2.836206896551724e-05,
+      "loss": 0.0994,
+      "step": 3765
+    },
+    {
+      "epoch": 4.365427805549518,
+      "grad_norm": 0.3961776793003082,
+      "learning_rate": 2.8356321839080462e-05,
+      "loss": 0.1168,
+      "step": 3766
+    },
+    {
+      "epoch": 4.366586973846266,
+      "grad_norm": 0.39625585079193115,
+      "learning_rate": 2.835057471264368e-05,
+      "loss": 0.1075,
+      "step": 3767
+    },
+    {
+      "epoch": 4.367746142143012,
+      "grad_norm": 0.3018212914466858,
+      "learning_rate": 2.8344827586206895e-05,
+      "loss": 0.1109,
+      "step": 3768
+    },
+    {
+      "epoch": 4.368905310439759,
+      "grad_norm": 0.32641318440437317,
+      "learning_rate": 2.8339080459770117e-05,
+      "loss": 0.1058,
+      "step": 3769
+    },
+    {
+      "epoch": 4.370064478736507,
+      "grad_norm": 0.28830480575561523,
+      "learning_rate": 2.8333333333333335e-05,
+      "loss": 0.0888,
+      "step": 3770
+    },
+    {
+      "epoch": 4.371223647033253,
+      "grad_norm": 0.299063116312027,
+      "learning_rate": 2.8327586206896557e-05,
+      "loss": 0.0994,
+      "step": 3771
+    },
+    {
+      "epoch": 4.372382815330001,
+      "grad_norm": 0.3458064794540405,
+      "learning_rate": 2.8321839080459772e-05,
+      "loss": 0.099,
+      "step": 3772
+    },
+    {
+      "epoch": 4.3735419836267475,
+      "grad_norm": 0.44503387808799744,
+      "learning_rate": 2.8316091954022987e-05,
+      "loss": 0.1047,
+      "step": 3773
+    },
+    {
+      "epoch": 4.374701151923495,
+      "grad_norm": 0.34976091980934143,
+      "learning_rate": 2.8310344827586212e-05,
+      "loss": 0.1,
+      "step": 3774
+    },
+    {
+      "epoch": 4.375860320220242,
+      "grad_norm": 0.6220278739929199,
+      "learning_rate": 2.8304597701149427e-05,
+      "loss": 0.1055,
+      "step": 3775
+    },
+    {
+      "epoch": 4.377019488516989,
+      "grad_norm": 0.5796971917152405,
+      "learning_rate": 2.8298850574712642e-05,
+      "loss": 0.1164,
+      "step": 3776
+    },
+    {
+      "epoch": 4.378178656813736,
+      "grad_norm": 0.38796380162239075,
+      "learning_rate": 2.8293103448275863e-05,
+      "loss": 0.0984,
+      "step": 3777
+    },
+    {
+      "epoch": 4.3793378251104835,
+      "grad_norm": 0.33590441942214966,
+      "learning_rate": 2.828735632183908e-05,
+      "loss": 0.1074,
+      "step": 3778
+    },
+    {
+      "epoch": 4.38049699340723,
+      "grad_norm": 0.3619459569454193,
+      "learning_rate": 2.8281609195402297e-05,
+      "loss": 0.1002,
+      "step": 3779
+    },
+    {
+      "epoch": 4.381656161703978,
+      "grad_norm": 0.27559658885002136,
+      "learning_rate": 2.8275862068965518e-05,
+      "loss": 0.098,
+      "step": 3780
+    },
+    {
+      "epoch": 4.382815330000724,
+      "grad_norm": 0.25252318382263184,
+      "learning_rate": 2.8270114942528737e-05,
+      "loss": 0.0882,
+      "step": 3781
+    },
+    {
+      "epoch": 4.383974498297471,
+      "grad_norm": 0.35869985818862915,
+      "learning_rate": 2.8264367816091958e-05,
+      "loss": 0.106,
+      "step": 3782
+    },
+    {
+      "epoch": 4.385133666594219,
+      "grad_norm": 0.4135657250881195,
+      "learning_rate": 2.8258620689655173e-05,
+      "loss": 0.1053,
+      "step": 3783
+    },
+    {
+      "epoch": 4.386292834890965,
+      "grad_norm": 0.3181520104408264,
+      "learning_rate": 2.825287356321839e-05,
+      "loss": 0.1071,
+      "step": 3784
+    },
+    {
+      "epoch": 4.387452003187713,
+      "grad_norm": 0.5326215624809265,
+      "learning_rate": 2.8247126436781613e-05,
+      "loss": 0.1095,
+      "step": 3785
+    },
+    {
+      "epoch": 4.38861117148446,
+      "grad_norm": 0.278712660074234,
+      "learning_rate": 2.8241379310344828e-05,
+      "loss": 0.0979,
+      "step": 3786
+    },
+    {
+      "epoch": 4.389770339781207,
+      "grad_norm": 0.30752044916152954,
+      "learning_rate": 2.8235632183908046e-05,
+      "loss": 0.0938,
+      "step": 3787
+    },
+    {
+      "epoch": 4.390929508077954,
+      "grad_norm": 0.3980220854282379,
+      "learning_rate": 2.8229885057471268e-05,
+      "loss": 0.101,
+      "step": 3788
+    },
+    {
+      "epoch": 4.392088676374701,
+      "grad_norm": 0.3621883690357208,
+      "learning_rate": 2.8224137931034483e-05,
+      "loss": 0.1064,
+      "step": 3789
+    },
+    {
+      "epoch": 4.393247844671448,
+      "grad_norm": 0.3474581837654114,
+      "learning_rate": 2.8218390804597705e-05,
+      "loss": 0.1024,
+      "step": 3790
+    },
+    {
+      "epoch": 4.394407012968196,
+      "grad_norm": 0.33540961146354675,
+      "learning_rate": 2.8212643678160923e-05,
+      "loss": 0.1032,
+      "step": 3791
+    },
+    {
+      "epoch": 4.395566181264942,
+      "grad_norm": 0.39876723289489746,
+      "learning_rate": 2.8206896551724138e-05,
+      "loss": 0.0977,
+      "step": 3792
+    },
+    {
+      "epoch": 4.39672534956169,
+      "grad_norm": 0.31449171900749207,
+      "learning_rate": 2.820114942528736e-05,
+      "loss": 0.0971,
+      "step": 3793
+    },
+    {
+      "epoch": 4.3978845178584365,
+      "grad_norm": 0.30794376134872437,
+      "learning_rate": 2.8195402298850578e-05,
+      "loss": 0.11,
+      "step": 3794
+    },
+    {
+      "epoch": 4.399043686155184,
+      "grad_norm": 0.30705249309539795,
+      "learning_rate": 2.8189655172413793e-05,
+      "loss": 0.1023,
+      "step": 3795
+    },
+    {
+      "epoch": 4.400202854451931,
+      "grad_norm": 0.37450698018074036,
+      "learning_rate": 2.8183908045977014e-05,
+      "loss": 0.1003,
+      "step": 3796
+    },
+    {
+      "epoch": 4.401362022748678,
+      "grad_norm": 0.4203515350818634,
+      "learning_rate": 2.817816091954023e-05,
+      "loss": 0.1178,
+      "step": 3797
+    },
+    {
+      "epoch": 4.402521191045425,
+      "grad_norm": 0.31087568402290344,
+      "learning_rate": 2.8172413793103447e-05,
+      "loss": 0.115,
+      "step": 3798
+    },
+    {
+      "epoch": 4.403680359342172,
+      "grad_norm": 0.32186147570610046,
+      "learning_rate": 2.816666666666667e-05,
+      "loss": 0.0939,
+      "step": 3799
+    },
+    {
+      "epoch": 4.404839527638919,
+      "grad_norm": 0.38315415382385254,
+      "learning_rate": 2.8160919540229884e-05,
+      "loss": 0.0979,
+      "step": 3800
+    },
+    {
+      "epoch": 4.405998695935666,
+      "grad_norm": 0.5825996398925781,
+      "learning_rate": 2.8155172413793106e-05,
+      "loss": 0.109,
+      "step": 3801
+    },
+    {
+      "epoch": 4.407157864232413,
+      "grad_norm": 0.42345044016838074,
+      "learning_rate": 2.8149425287356324e-05,
+      "loss": 0.106,
+      "step": 3802
+    },
+    {
+      "epoch": 4.40831703252916,
+      "grad_norm": 0.2959122657775879,
+      "learning_rate": 2.814367816091954e-05,
+      "loss": 0.0996,
+      "step": 3803
+    },
+    {
+      "epoch": 4.409476200825908,
+      "grad_norm": 0.479018896818161,
+      "learning_rate": 2.813793103448276e-05,
+      "loss": 0.0952,
+      "step": 3804
+    },
+    {
+      "epoch": 4.410635369122654,
+      "grad_norm": 0.3577905297279358,
+      "learning_rate": 2.813218390804598e-05,
+      "loss": 0.1047,
+      "step": 3805
+    },
+    {
+      "epoch": 4.411794537419402,
+      "grad_norm": 0.5896329283714294,
+      "learning_rate": 2.8126436781609194e-05,
+      "loss": 0.1145,
+      "step": 3806
+    },
+    {
+      "epoch": 4.4129537057161485,
+      "grad_norm": 0.31682226061820984,
+      "learning_rate": 2.8120689655172415e-05,
+      "loss": 0.0913,
+      "step": 3807
+    },
+    {
+      "epoch": 4.414112874012896,
+      "grad_norm": 0.38672780990600586,
+      "learning_rate": 2.8114942528735634e-05,
+      "loss": 0.1036,
+      "step": 3808
+    },
+    {
+      "epoch": 4.415272042309643,
+      "grad_norm": 0.3999164402484894,
+      "learning_rate": 2.8109195402298855e-05,
+      "loss": 0.0992,
+      "step": 3809
+    },
+    {
+      "epoch": 4.41643121060639,
+      "grad_norm": 0.32970115542411804,
+      "learning_rate": 2.810344827586207e-05,
+      "loss": 0.0957,
+      "step": 3810
+    },
+    {
+      "epoch": 4.417590378903137,
+      "grad_norm": 0.4063851535320282,
+      "learning_rate": 2.809770114942529e-05,
+      "loss": 0.1041,
+      "step": 3811
+    },
+    {
+      "epoch": 4.418749547199884,
+      "grad_norm": 0.4128503203392029,
+      "learning_rate": 2.809195402298851e-05,
+      "loss": 0.1006,
+      "step": 3812
+    },
+    {
+      "epoch": 4.419908715496631,
+      "grad_norm": 0.3817613124847412,
+      "learning_rate": 2.8086206896551725e-05,
+      "loss": 0.1138,
+      "step": 3813
+    },
+    {
+      "epoch": 4.421067883793378,
+      "grad_norm": 0.4013262391090393,
+      "learning_rate": 2.8080459770114944e-05,
+      "loss": 0.1078,
+      "step": 3814
+    },
+    {
+      "epoch": 4.4222270520901255,
+      "grad_norm": 0.4090496599674225,
+      "learning_rate": 2.8074712643678165e-05,
+      "loss": 0.1052,
+      "step": 3815
+    },
+    {
+      "epoch": 4.423386220386872,
+      "grad_norm": 0.307972252368927,
+      "learning_rate": 2.806896551724138e-05,
+      "loss": 0.1008,
+      "step": 3816
+    },
+    {
+      "epoch": 4.42454538868362,
+      "grad_norm": 0.4433940052986145,
+      "learning_rate": 2.8063218390804595e-05,
+      "loss": 0.1126,
+      "step": 3817
+    },
+    {
+      "epoch": 4.425704556980366,
+      "grad_norm": 0.34657689929008484,
+      "learning_rate": 2.8057471264367817e-05,
+      "loss": 0.085,
+      "step": 3818
+    },
+    {
+      "epoch": 4.426863725277114,
+      "grad_norm": 0.4442926049232483,
+      "learning_rate": 2.8051724137931035e-05,
+      "loss": 0.1089,
+      "step": 3819
+    },
+    {
+      "epoch": 4.428022893573861,
+      "grad_norm": 0.5043272376060486,
+      "learning_rate": 2.8045977011494257e-05,
+      "loss": 0.1284,
+      "step": 3820
+    },
+    {
+      "epoch": 4.429182061870608,
+      "grad_norm": 0.34867578744888306,
+      "learning_rate": 2.804022988505747e-05,
+      "loss": 0.101,
+      "step": 3821
+    },
+    {
+      "epoch": 4.430341230167355,
+      "grad_norm": 0.3213669955730438,
+      "learning_rate": 2.803448275862069e-05,
+      "loss": 0.0922,
+      "step": 3822
+    },
+    {
+      "epoch": 4.431500398464102,
+      "grad_norm": 1.2875010967254639,
+      "learning_rate": 2.802873563218391e-05,
+      "loss": 0.1089,
+      "step": 3823
+    },
+    {
+      "epoch": 4.432659566760849,
+      "grad_norm": 0.3360012173652649,
+      "learning_rate": 2.8022988505747126e-05,
+      "loss": 0.1059,
+      "step": 3824
+    },
+    {
+      "epoch": 4.433818735057596,
+      "grad_norm": 0.31028974056243896,
+      "learning_rate": 2.8017241379310345e-05,
+      "loss": 0.1003,
+      "step": 3825
+    },
+    {
+      "epoch": 4.434977903354343,
+      "grad_norm": 0.45090678334236145,
+      "learning_rate": 2.8011494252873566e-05,
+      "loss": 0.1056,
+      "step": 3826
+    },
+    {
+      "epoch": 4.43613707165109,
+      "grad_norm": 0.29718369245529175,
+      "learning_rate": 2.800574712643678e-05,
+      "loss": 0.1012,
+      "step": 3827
+    },
+    {
+      "epoch": 4.4372962399478375,
+      "grad_norm": 0.3447516858577728,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 0.0984,
+      "step": 3828
+    },
+    {
+      "epoch": 4.438455408244584,
+      "grad_norm": 0.3421769440174103,
+      "learning_rate": 2.799425287356322e-05,
+      "loss": 0.1022,
+      "step": 3829
+    },
+    {
+      "epoch": 4.439614576541332,
+      "grad_norm": 0.4500944912433624,
+      "learning_rate": 2.7988505747126436e-05,
+      "loss": 0.1007,
+      "step": 3830
+    },
+    {
+      "epoch": 4.440773744838078,
+      "grad_norm": 0.434008926153183,
+      "learning_rate": 2.7982758620689658e-05,
+      "loss": 0.1144,
+      "step": 3831
+    },
+    {
+      "epoch": 4.441932913134826,
+      "grad_norm": 0.3270074129104614,
+      "learning_rate": 2.7977011494252876e-05,
+      "loss": 0.1051,
+      "step": 3832
+    },
+    {
+      "epoch": 4.443092081431573,
+      "grad_norm": 0.37319257855415344,
+      "learning_rate": 2.797126436781609e-05,
+      "loss": 0.1048,
+      "step": 3833
+    },
+    {
+      "epoch": 4.44425124972832,
+      "grad_norm": 0.3999885320663452,
+      "learning_rate": 2.7965517241379313e-05,
+      "loss": 0.1119,
+      "step": 3834
+    },
+    {
+      "epoch": 4.445410418025067,
+      "grad_norm": 0.3219684362411499,
+      "learning_rate": 2.795977011494253e-05,
+      "loss": 0.1173,
+      "step": 3835
+    },
+    {
+      "epoch": 4.446569586321814,
+      "grad_norm": 0.3253685534000397,
+      "learning_rate": 2.7954022988505746e-05,
+      "loss": 0.1012,
+      "step": 3836
+    },
+    {
+      "epoch": 4.447728754618561,
+      "grad_norm": 0.3479521870613098,
+      "learning_rate": 2.7948275862068968e-05,
+      "loss": 0.0977,
+      "step": 3837
+    },
+    {
+      "epoch": 4.448887922915309,
+      "grad_norm": 0.33220139145851135,
+      "learning_rate": 2.7942528735632182e-05,
+      "loss": 0.1052,
+      "step": 3838
+    },
+    {
+      "epoch": 4.450047091212055,
+      "grad_norm": 0.3189479410648346,
+      "learning_rate": 2.7936781609195408e-05,
+      "loss": 0.0937,
+      "step": 3839
+    },
+    {
+      "epoch": 4.451206259508803,
+      "grad_norm": 0.34778445959091187,
+      "learning_rate": 2.7931034482758622e-05,
+      "loss": 0.1142,
+      "step": 3840
+    },
+    {
+      "epoch": 4.45236542780555,
+      "grad_norm": 0.4988318979740143,
+      "learning_rate": 2.7925287356321837e-05,
+      "loss": 0.1025,
+      "step": 3841
+    },
+    {
+      "epoch": 4.453524596102296,
+      "grad_norm": 0.41545069217681885,
+      "learning_rate": 2.791954022988506e-05,
+      "loss": 0.1075,
+      "step": 3842
+    },
+    {
+      "epoch": 4.454683764399044,
+      "grad_norm": 0.38228297233581543,
+      "learning_rate": 2.7913793103448277e-05,
+      "loss": 0.1112,
+      "step": 3843
+    },
+    {
+      "epoch": 4.4558429326957905,
+      "grad_norm": 0.34965816140174866,
+      "learning_rate": 2.7908045977011492e-05,
+      "loss": 0.1088,
+      "step": 3844
+    },
+    {
+      "epoch": 4.457002100992538,
+      "grad_norm": 0.32655954360961914,
+      "learning_rate": 2.7902298850574714e-05,
+      "loss": 0.1071,
+      "step": 3845
+    },
+    {
+      "epoch": 4.458161269289285,
+      "grad_norm": 0.35616791248321533,
+      "learning_rate": 2.7896551724137932e-05,
+      "loss": 0.1095,
+      "step": 3846
+    },
+    {
+      "epoch": 4.459320437586032,
+      "grad_norm": 0.4080839157104492,
+      "learning_rate": 2.7890804597701154e-05,
+      "loss": 0.1029,
+      "step": 3847
+    },
+    {
+      "epoch": 4.460479605882779,
+      "grad_norm": 0.36598852276802063,
+      "learning_rate": 2.788505747126437e-05,
+      "loss": 0.0919,
+      "step": 3848
+    },
+    {
+      "epoch": 4.4616387741795265,
+      "grad_norm": 0.3768613040447235,
+      "learning_rate": 2.7879310344827587e-05,
+      "loss": 0.1065,
+      "step": 3849
+    },
+    {
+      "epoch": 4.462797942476273,
+      "grad_norm": 0.3174971044063568,
+      "learning_rate": 2.787356321839081e-05,
+      "loss": 0.1003,
+      "step": 3850
+    },
+    {
+      "epoch": 4.463957110773021,
+      "grad_norm": 0.34315162897109985,
+      "learning_rate": 2.7867816091954024e-05,
+      "loss": 0.1042,
+      "step": 3851
+    },
+    {
+      "epoch": 4.465116279069767,
+      "grad_norm": 0.29958537220954895,
+      "learning_rate": 2.7862068965517242e-05,
+      "loss": 0.0927,
+      "step": 3852
+    },
+    {
+      "epoch": 4.466275447366515,
+      "grad_norm": 0.40852096676826477,
+      "learning_rate": 2.7856321839080464e-05,
+      "loss": 0.1002,
+      "step": 3853
+    },
+    {
+      "epoch": 4.467434615663262,
+      "grad_norm": 0.32836470007896423,
+      "learning_rate": 2.785057471264368e-05,
+      "loss": 0.1069,
+      "step": 3854
+    },
+    {
+      "epoch": 4.468593783960008,
+      "grad_norm": 0.37109801173210144,
+      "learning_rate": 2.7844827586206897e-05,
+      "loss": 0.1069,
+      "step": 3855
+    },
+    {
+      "epoch": 4.469752952256756,
+      "grad_norm": 0.4152434170246124,
+      "learning_rate": 2.783908045977012e-05,
+      "loss": 0.1026,
+      "step": 3856
+    },
+    {
+      "epoch": 4.4709121205535025,
+      "grad_norm": 0.31486672163009644,
+      "learning_rate": 2.7833333333333333e-05,
+      "loss": 0.0983,
+      "step": 3857
+    },
+    {
+      "epoch": 4.47207128885025,
+      "grad_norm": 0.2974405288696289,
+      "learning_rate": 2.7827586206896555e-05,
+      "loss": 0.0975,
+      "step": 3858
+    },
+    {
+      "epoch": 4.473230457146997,
+      "grad_norm": 0.3772009015083313,
+      "learning_rate": 2.7821839080459773e-05,
+      "loss": 0.1035,
+      "step": 3859
+    },
+    {
+      "epoch": 4.474389625443744,
+      "grad_norm": 0.3580301105976105,
+      "learning_rate": 2.7816091954022988e-05,
+      "loss": 0.1001,
+      "step": 3860
+    },
+    {
+      "epoch": 4.475548793740491,
+      "grad_norm": 0.31176498532295227,
+      "learning_rate": 2.781034482758621e-05,
+      "loss": 0.0991,
+      "step": 3861
+    },
+    {
+      "epoch": 4.4767079620372385,
+      "grad_norm": 0.36399632692337036,
+      "learning_rate": 2.7804597701149425e-05,
+      "loss": 0.1106,
+      "step": 3862
+    },
+    {
+      "epoch": 4.477867130333985,
+      "grad_norm": 0.36921966075897217,
+      "learning_rate": 2.7798850574712643e-05,
+      "loss": 0.1045,
+      "step": 3863
+    },
+    {
+      "epoch": 4.479026298630733,
+      "grad_norm": 0.28541839122772217,
+      "learning_rate": 2.7793103448275865e-05,
+      "loss": 0.0879,
+      "step": 3864
+    },
+    {
+      "epoch": 4.480185466927479,
+      "grad_norm": 0.5838915109634399,
+      "learning_rate": 2.778735632183908e-05,
+      "loss": 0.1098,
+      "step": 3865
+    },
+    {
+      "epoch": 4.481344635224227,
+      "grad_norm": 0.32612380385398865,
+      "learning_rate": 2.77816091954023e-05,
+      "loss": 0.107,
+      "step": 3866
+    },
+    {
+      "epoch": 4.482503803520974,
+      "grad_norm": 0.33638033270835876,
+      "learning_rate": 2.777586206896552e-05,
+      "loss": 0.101,
+      "step": 3867
+    },
+    {
+      "epoch": 4.48366297181772,
+      "grad_norm": 0.40769678354263306,
+      "learning_rate": 2.7770114942528735e-05,
+      "loss": 0.107,
+      "step": 3868
+    },
+    {
+      "epoch": 4.484822140114468,
+      "grad_norm": 0.3535826504230499,
+      "learning_rate": 2.7764367816091956e-05,
+      "loss": 0.1087,
+      "step": 3869
+    },
+    {
+      "epoch": 4.485981308411215,
+      "grad_norm": 0.3710591495037079,
+      "learning_rate": 2.7758620689655175e-05,
+      "loss": 0.1045,
+      "step": 3870
+    },
+    {
+      "epoch": 4.487140476707962,
+      "grad_norm": 0.3905164301395416,
+      "learning_rate": 2.775287356321839e-05,
+      "loss": 0.1059,
+      "step": 3871
+    },
+    {
+      "epoch": 4.488299645004709,
+      "grad_norm": 0.27448105812072754,
+      "learning_rate": 2.774712643678161e-05,
+      "loss": 0.1021,
+      "step": 3872
+    },
+    {
+      "epoch": 4.489458813301456,
+      "grad_norm": 0.3772237002849579,
+      "learning_rate": 2.774137931034483e-05,
+      "loss": 0.1088,
+      "step": 3873
+    },
+    {
+      "epoch": 4.490617981598203,
+      "grad_norm": 0.4715537130832672,
+      "learning_rate": 2.7735632183908044e-05,
+      "loss": 0.1001,
+      "step": 3874
+    },
+    {
+      "epoch": 4.491777149894951,
+      "grad_norm": 0.4609696865081787,
+      "learning_rate": 2.7729885057471266e-05,
+      "loss": 0.1155,
+      "step": 3875
+    },
+    {
+      "epoch": 4.492936318191697,
+      "grad_norm": 0.368083119392395,
+      "learning_rate": 2.7724137931034484e-05,
+      "loss": 0.1018,
+      "step": 3876
+    },
+    {
+      "epoch": 4.494095486488445,
+      "grad_norm": 0.40971893072128296,
+      "learning_rate": 2.7718390804597706e-05,
+      "loss": 0.0994,
+      "step": 3877
+    },
+    {
+      "epoch": 4.4952546547851915,
+      "grad_norm": 0.40606266260147095,
+      "learning_rate": 2.771264367816092e-05,
+      "loss": 0.1162,
+      "step": 3878
+    },
+    {
+      "epoch": 4.496413823081939,
+      "grad_norm": 0.3284454047679901,
+      "learning_rate": 2.7706896551724136e-05,
+      "loss": 0.1001,
+      "step": 3879
+    },
+    {
+      "epoch": 4.497572991378686,
+      "grad_norm": 0.3405894935131073,
+      "learning_rate": 2.770114942528736e-05,
+      "loss": 0.089,
+      "step": 3880
+    },
+    {
+      "epoch": 4.498732159675433,
+      "grad_norm": 0.44441795349121094,
+      "learning_rate": 2.7695402298850576e-05,
+      "loss": 0.1161,
+      "step": 3881
+    },
+    {
+      "epoch": 4.49989132797218,
+      "grad_norm": 0.3282816410064697,
+      "learning_rate": 2.768965517241379e-05,
+      "loss": 0.1014,
+      "step": 3882
+    },
+    {
+      "epoch": 4.5010504962689275,
+      "grad_norm": 0.31686845421791077,
+      "learning_rate": 2.7683908045977012e-05,
+      "loss": 0.1002,
+      "step": 3883
+    },
+    {
+      "epoch": 4.502209664565674,
+      "grad_norm": 0.33207643032073975,
+      "learning_rate": 2.767816091954023e-05,
+      "loss": 0.0948,
+      "step": 3884
+    },
+    {
+      "epoch": 4.503368832862421,
+      "grad_norm": 0.3429032564163208,
+      "learning_rate": 2.7672413793103452e-05,
+      "loss": 0.1001,
+      "step": 3885
+    },
+    {
+      "epoch": 4.504528001159168,
+      "grad_norm": 0.333465576171875,
+      "learning_rate": 2.7666666666666667e-05,
+      "loss": 0.1043,
+      "step": 3886
+    },
+    {
+      "epoch": 4.505687169455915,
+      "grad_norm": 0.3548075556755066,
+      "learning_rate": 2.7660919540229885e-05,
+      "loss": 0.1056,
+      "step": 3887
+    },
+    {
+      "epoch": 4.506846337752663,
+      "grad_norm": 0.41525760293006897,
+      "learning_rate": 2.7655172413793107e-05,
+      "loss": 0.1086,
+      "step": 3888
+    },
+    {
+      "epoch": 4.508005506049409,
+      "grad_norm": 0.30972954630851746,
+      "learning_rate": 2.7649425287356322e-05,
+      "loss": 0.0972,
+      "step": 3889
+    },
+    {
+      "epoch": 4.509164674346157,
+      "grad_norm": 0.46984899044036865,
+      "learning_rate": 2.764367816091954e-05,
+      "loss": 0.1082,
+      "step": 3890
+    },
+    {
+      "epoch": 4.5103238426429035,
+      "grad_norm": 0.3709583580493927,
+      "learning_rate": 2.7637931034482762e-05,
+      "loss": 0.1007,
+      "step": 3891
+    },
+    {
+      "epoch": 4.511483010939651,
+      "grad_norm": 0.45104119181632996,
+      "learning_rate": 2.7632183908045977e-05,
+      "loss": 0.1131,
+      "step": 3892
+    },
+    {
+      "epoch": 4.512642179236398,
+      "grad_norm": 0.311952143907547,
+      "learning_rate": 2.7626436781609195e-05,
+      "loss": 0.1021,
+      "step": 3893
+    },
+    {
+      "epoch": 4.513801347533145,
+      "grad_norm": 0.3778269588947296,
+      "learning_rate": 2.7620689655172417e-05,
+      "loss": 0.1046,
+      "step": 3894
+    },
+    {
+      "epoch": 4.514960515829892,
+      "grad_norm": 0.3810199797153473,
+      "learning_rate": 2.7614942528735632e-05,
+      "loss": 0.1044,
+      "step": 3895
+    },
+    {
+      "epoch": 4.5161196841266396,
+      "grad_norm": 0.3130066394805908,
+      "learning_rate": 2.7609195402298853e-05,
+      "loss": 0.1037,
+      "step": 3896
+    },
+    {
+      "epoch": 4.517278852423386,
+      "grad_norm": 0.32273080945014954,
+      "learning_rate": 2.7603448275862072e-05,
+      "loss": 0.0984,
+      "step": 3897
+    },
+    {
+      "epoch": 4.518438020720133,
+      "grad_norm": 0.4851730763912201,
+      "learning_rate": 2.7597701149425287e-05,
+      "loss": 0.1088,
+      "step": 3898
+    },
+    {
+      "epoch": 4.5195971890168805,
+      "grad_norm": 0.30688047409057617,
+      "learning_rate": 2.759195402298851e-05,
+      "loss": 0.0997,
+      "step": 3899
+    },
+    {
+      "epoch": 4.520756357313627,
+      "grad_norm": 0.29143083095550537,
+      "learning_rate": 2.7586206896551727e-05,
+      "loss": 0.1066,
+      "step": 3900
+    },
+    {
+      "epoch": 4.521915525610375,
+      "grad_norm": 0.4881395697593689,
+      "learning_rate": 2.758045977011494e-05,
+      "loss": 0.103,
+      "step": 3901
+    },
+    {
+      "epoch": 4.523074693907121,
+      "grad_norm": 0.2735028862953186,
+      "learning_rate": 2.7574712643678163e-05,
+      "loss": 0.093,
+      "step": 3902
+    },
+    {
+      "epoch": 4.524233862203869,
+      "grad_norm": 0.3494533598423004,
+      "learning_rate": 2.7568965517241378e-05,
+      "loss": 0.1089,
+      "step": 3903
+    },
+    {
+      "epoch": 4.525393030500616,
+      "grad_norm": 0.36798563599586487,
+      "learning_rate": 2.7563218390804603e-05,
+      "loss": 0.1049,
+      "step": 3904
+    },
+    {
+      "epoch": 4.526552198797363,
+      "grad_norm": 0.3005774915218353,
+      "learning_rate": 2.7557471264367818e-05,
+      "loss": 0.1141,
+      "step": 3905
+    },
+    {
+      "epoch": 4.52771136709411,
+      "grad_norm": 0.3312096893787384,
+      "learning_rate": 2.7551724137931033e-05,
+      "loss": 0.1006,
+      "step": 3906
+    },
+    {
+      "epoch": 4.528870535390857,
+      "grad_norm": 0.3487379848957062,
+      "learning_rate": 2.7545977011494255e-05,
+      "loss": 0.1005,
+      "step": 3907
+    },
+    {
+      "epoch": 4.530029703687604,
+      "grad_norm": 0.36595210433006287,
+      "learning_rate": 2.7540229885057473e-05,
+      "loss": 0.1016,
+      "step": 3908
+    },
+    {
+      "epoch": 4.531188871984352,
+      "grad_norm": 0.2543398439884186,
+      "learning_rate": 2.7534482758620688e-05,
+      "loss": 0.0935,
+      "step": 3909
+    },
+    {
+      "epoch": 4.532348040281098,
+      "grad_norm": 0.38063111901283264,
+      "learning_rate": 2.752873563218391e-05,
+      "loss": 0.106,
+      "step": 3910
+    },
+    {
+      "epoch": 4.533507208577845,
+      "grad_norm": 0.37747347354888916,
+      "learning_rate": 2.7522988505747128e-05,
+      "loss": 0.1045,
+      "step": 3911
+    },
+    {
+      "epoch": 4.5346663768745925,
+      "grad_norm": 0.4080142676830292,
+      "learning_rate": 2.7517241379310343e-05,
+      "loss": 0.1065,
+      "step": 3912
+    },
+    {
+      "epoch": 4.535825545171339,
+      "grad_norm": 0.3274177014827728,
+      "learning_rate": 2.7511494252873564e-05,
+      "loss": 0.1012,
+      "step": 3913
+    },
+    {
+      "epoch": 4.536984713468087,
+      "grad_norm": 0.2830631136894226,
+      "learning_rate": 2.7505747126436783e-05,
+      "loss": 0.1,
+      "step": 3914
+    },
+    {
+      "epoch": 4.538143881764833,
+      "grad_norm": 0.27470821142196655,
+      "learning_rate": 2.7500000000000004e-05,
+      "loss": 0.0975,
+      "step": 3915
+    },
+    {
+      "epoch": 4.539303050061581,
+      "grad_norm": 0.37115851044654846,
+      "learning_rate": 2.749425287356322e-05,
+      "loss": 0.1079,
+      "step": 3916
+    },
+    {
+      "epoch": 4.540462218358328,
+      "grad_norm": 0.3276093900203705,
+      "learning_rate": 2.7488505747126438e-05,
+      "loss": 0.1043,
+      "step": 3917
+    },
+    {
+      "epoch": 4.541621386655075,
+      "grad_norm": 0.3176075518131256,
+      "learning_rate": 2.748275862068966e-05,
+      "loss": 0.1021,
+      "step": 3918
+    },
+    {
+      "epoch": 4.542780554951822,
+      "grad_norm": 0.32996928691864014,
+      "learning_rate": 2.7477011494252874e-05,
+      "loss": 0.095,
+      "step": 3919
+    },
+    {
+      "epoch": 4.543939723248569,
+      "grad_norm": 0.3963630497455597,
+      "learning_rate": 2.7471264367816092e-05,
+      "loss": 0.1069,
+      "step": 3920
+    },
+    {
+      "epoch": 4.545098891545316,
+      "grad_norm": 0.9190638065338135,
+      "learning_rate": 2.7465517241379314e-05,
+      "loss": 0.1153,
+      "step": 3921
+    },
+    {
+      "epoch": 4.546258059842064,
+      "grad_norm": 0.3529224991798401,
+      "learning_rate": 2.745977011494253e-05,
+      "loss": 0.092,
+      "step": 3922
+    },
+    {
+      "epoch": 4.54741722813881,
+      "grad_norm": 0.4127110540866852,
+      "learning_rate": 2.745402298850575e-05,
+      "loss": 0.1058,
+      "step": 3923
+    },
+    {
+      "epoch": 4.548576396435557,
+      "grad_norm": 0.4567547142505646,
+      "learning_rate": 2.744827586206897e-05,
+      "loss": 0.1041,
+      "step": 3924
+    },
+    {
+      "epoch": 4.549735564732305,
+      "grad_norm": 0.30119210481643677,
+      "learning_rate": 2.7442528735632184e-05,
+      "loss": 0.0998,
+      "step": 3925
+    },
+    {
+      "epoch": 4.550894733029052,
+      "grad_norm": 0.3091825842857361,
+      "learning_rate": 2.7436781609195406e-05,
+      "loss": 0.0967,
+      "step": 3926
+    },
+    {
+      "epoch": 4.552053901325799,
+      "grad_norm": 0.38511085510253906,
+      "learning_rate": 2.743103448275862e-05,
+      "loss": 0.1076,
+      "step": 3927
+    },
+    {
+      "epoch": 4.5532130696225455,
+      "grad_norm": 0.33152782917022705,
+      "learning_rate": 2.742528735632184e-05,
+      "loss": 0.0974,
+      "step": 3928
+    },
+    {
+      "epoch": 4.554372237919293,
+      "grad_norm": 0.35295701026916504,
+      "learning_rate": 2.741954022988506e-05,
+      "loss": 0.1044,
+      "step": 3929
+    },
+    {
+      "epoch": 4.55553140621604,
+      "grad_norm": 0.40102672576904297,
+      "learning_rate": 2.7413793103448275e-05,
+      "loss": 0.1007,
+      "step": 3930
+    },
+    {
+      "epoch": 4.556690574512787,
+      "grad_norm": 0.4231402575969696,
+      "learning_rate": 2.7408045977011494e-05,
+      "loss": 0.1146,
+      "step": 3931
+    },
+    {
+      "epoch": 4.557849742809534,
+      "grad_norm": 0.43630826473236084,
+      "learning_rate": 2.7402298850574715e-05,
+      "loss": 0.1009,
+      "step": 3932
+    },
+    {
+      "epoch": 4.5590089111062815,
+      "grad_norm": 0.2806771993637085,
+      "learning_rate": 2.739655172413793e-05,
+      "loss": 0.1037,
+      "step": 3933
+    },
+    {
+      "epoch": 4.560168079403028,
+      "grad_norm": 0.3072310984134674,
+      "learning_rate": 2.7390804597701152e-05,
+      "loss": 0.0999,
+      "step": 3934
+    },
+    {
+      "epoch": 4.561327247699776,
+      "grad_norm": 0.3378421366214752,
+      "learning_rate": 2.738505747126437e-05,
+      "loss": 0.1061,
+      "step": 3935
+    },
+    {
+      "epoch": 4.562486415996522,
+      "grad_norm": 0.25815141201019287,
+      "learning_rate": 2.7379310344827585e-05,
+      "loss": 0.0916,
+      "step": 3936
+    },
+    {
+      "epoch": 4.56364558429327,
+      "grad_norm": 0.3428052067756653,
+      "learning_rate": 2.7373563218390807e-05,
+      "loss": 0.1086,
+      "step": 3937
+    },
+    {
+      "epoch": 4.564804752590017,
+      "grad_norm": 0.3696090579032898,
+      "learning_rate": 2.7367816091954025e-05,
+      "loss": 0.1022,
+      "step": 3938
+    },
+    {
+      "epoch": 4.565963920886764,
+      "grad_norm": 0.31813839077949524,
+      "learning_rate": 2.736206896551724e-05,
+      "loss": 0.0936,
+      "step": 3939
+    },
+    {
+      "epoch": 4.567123089183511,
+      "grad_norm": 0.40487226843833923,
+      "learning_rate": 2.735632183908046e-05,
+      "loss": 0.103,
+      "step": 3940
+    },
+    {
+      "epoch": 4.5682822574802575,
+      "grad_norm": 0.3482483923435211,
+      "learning_rate": 2.735057471264368e-05,
+      "loss": 0.1005,
+      "step": 3941
+    },
+    {
+      "epoch": 4.569441425777005,
+      "grad_norm": 0.7645699381828308,
+      "learning_rate": 2.73448275862069e-05,
+      "loss": 0.1048,
+      "step": 3942
+    },
+    {
+      "epoch": 4.570600594073752,
+      "grad_norm": 0.2931617796421051,
+      "learning_rate": 2.7339080459770116e-05,
+      "loss": 0.0907,
+      "step": 3943
+    },
+    {
+      "epoch": 4.571759762370499,
+      "grad_norm": 0.491020530462265,
+      "learning_rate": 2.733333333333333e-05,
+      "loss": 0.1099,
+      "step": 3944
+    },
+    {
+      "epoch": 4.572918930667246,
+      "grad_norm": 0.28609660267829895,
+      "learning_rate": 2.7327586206896556e-05,
+      "loss": 0.104,
+      "step": 3945
+    },
+    {
+      "epoch": 4.5740780989639935,
+      "grad_norm": 0.4021299183368683,
+      "learning_rate": 2.732183908045977e-05,
+      "loss": 0.1071,
+      "step": 3946
+    },
+    {
+      "epoch": 4.57523726726074,
+      "grad_norm": 0.39661794900894165,
+      "learning_rate": 2.7316091954022986e-05,
+      "loss": 0.1019,
+      "step": 3947
+    },
+    {
+      "epoch": 4.576396435557488,
+      "grad_norm": 0.3372730314731598,
+      "learning_rate": 2.7310344827586208e-05,
+      "loss": 0.1109,
+      "step": 3948
+    },
+    {
+      "epoch": 4.577555603854234,
+      "grad_norm": 0.33836305141448975,
+      "learning_rate": 2.7304597701149426e-05,
+      "loss": 0.0944,
+      "step": 3949
+    },
+    {
+      "epoch": 4.578714772150982,
+      "grad_norm": 0.3433241844177246,
+      "learning_rate": 2.7298850574712648e-05,
+      "loss": 0.1087,
+      "step": 3950
+    },
+    {
+      "epoch": 4.579873940447729,
+      "grad_norm": 0.3279787600040436,
+      "learning_rate": 2.7293103448275863e-05,
+      "loss": 0.1091,
+      "step": 3951
+    },
+    {
+      "epoch": 4.581033108744476,
+      "grad_norm": 0.3588603734970093,
+      "learning_rate": 2.728735632183908e-05,
+      "loss": 0.0953,
+      "step": 3952
+    },
+    {
+      "epoch": 4.582192277041223,
+      "grad_norm": 0.2686539590358734,
+      "learning_rate": 2.7281609195402303e-05,
+      "loss": 0.099,
+      "step": 3953
+    },
+    {
+      "epoch": 4.58335144533797,
+      "grad_norm": 0.42638182640075684,
+      "learning_rate": 2.7275862068965518e-05,
+      "loss": 0.1131,
+      "step": 3954
+    },
+    {
+      "epoch": 4.584510613634717,
+      "grad_norm": 0.34486687183380127,
+      "learning_rate": 2.7270114942528736e-05,
+      "loss": 0.0993,
+      "step": 3955
+    },
+    {
+      "epoch": 4.585669781931464,
+      "grad_norm": 0.3565883934497833,
+      "learning_rate": 2.7264367816091958e-05,
+      "loss": 0.1047,
+      "step": 3956
+    },
+    {
+      "epoch": 4.586828950228211,
+      "grad_norm": 0.2582899332046509,
+      "learning_rate": 2.7258620689655173e-05,
+      "loss": 0.0902,
+      "step": 3957
+    },
+    {
+      "epoch": 4.587988118524958,
+      "grad_norm": 0.3092981278896332,
+      "learning_rate": 2.725287356321839e-05,
+      "loss": 0.1102,
+      "step": 3958
+    },
+    {
+      "epoch": 4.589147286821706,
+      "grad_norm": 0.41385966539382935,
+      "learning_rate": 2.7247126436781613e-05,
+      "loss": 0.0956,
+      "step": 3959
+    },
+    {
+      "epoch": 4.590306455118452,
+      "grad_norm": 0.4211784303188324,
+      "learning_rate": 2.7241379310344827e-05,
+      "loss": 0.1025,
+      "step": 3960
+    },
+    {
+      "epoch": 4.5914656234152,
+      "grad_norm": 0.3663957417011261,
+      "learning_rate": 2.723563218390805e-05,
+      "loss": 0.1041,
+      "step": 3961
+    },
+    {
+      "epoch": 4.5926247917119465,
+      "grad_norm": 0.30467289686203003,
+      "learning_rate": 2.7229885057471267e-05,
+      "loss": 0.1013,
+      "step": 3962
+    },
+    {
+      "epoch": 4.593783960008694,
+      "grad_norm": 0.41379502415657043,
+      "learning_rate": 2.7224137931034482e-05,
+      "loss": 0.1,
+      "step": 3963
+    },
+    {
+      "epoch": 4.594943128305441,
+      "grad_norm": 0.31340333819389343,
+      "learning_rate": 2.7218390804597704e-05,
+      "loss": 0.0935,
+      "step": 3964
+    },
+    {
+      "epoch": 4.596102296602188,
+      "grad_norm": 0.3276791274547577,
+      "learning_rate": 2.7212643678160922e-05,
+      "loss": 0.1058,
+      "step": 3965
+    },
+    {
+      "epoch": 4.597261464898935,
+      "grad_norm": 0.4072125554084778,
+      "learning_rate": 2.7206896551724137e-05,
+      "loss": 0.0965,
+      "step": 3966
+    },
+    {
+      "epoch": 4.598420633195682,
+      "grad_norm": 0.32723650336265564,
+      "learning_rate": 2.720114942528736e-05,
+      "loss": 0.1108,
+      "step": 3967
+    },
+    {
+      "epoch": 4.599579801492429,
+      "grad_norm": 0.3264493942260742,
+      "learning_rate": 2.7195402298850574e-05,
+      "loss": 0.119,
+      "step": 3968
+    },
+    {
+      "epoch": 4.600738969789177,
+      "grad_norm": 0.32702627778053284,
+      "learning_rate": 2.71896551724138e-05,
+      "loss": 0.1014,
+      "step": 3969
+    },
+    {
+      "epoch": 4.601898138085923,
+      "grad_norm": 0.5226812958717346,
+      "learning_rate": 2.7183908045977014e-05,
+      "loss": 0.1009,
+      "step": 3970
+    },
+    {
+      "epoch": 4.60305730638267,
+      "grad_norm": 0.3024195730686188,
+      "learning_rate": 2.717816091954023e-05,
+      "loss": 0.0955,
+      "step": 3971
+    },
+    {
+      "epoch": 4.604216474679418,
+      "grad_norm": 0.3582437038421631,
+      "learning_rate": 2.717241379310345e-05,
+      "loss": 0.1102,
+      "step": 3972
+    },
+    {
+      "epoch": 4.605375642976164,
+      "grad_norm": 0.38571134209632874,
+      "learning_rate": 2.716666666666667e-05,
+      "loss": 0.1012,
+      "step": 3973
+    },
+    {
+      "epoch": 4.606534811272912,
+      "grad_norm": 0.30533286929130554,
+      "learning_rate": 2.7160919540229883e-05,
+      "loss": 0.1025,
+      "step": 3974
+    },
+    {
+      "epoch": 4.6076939795696585,
+      "grad_norm": 0.38881656527519226,
+      "learning_rate": 2.7155172413793105e-05,
+      "loss": 0.0957,
+      "step": 3975
+    },
+    {
+      "epoch": 4.608853147866406,
+      "grad_norm": 0.40366286039352417,
+      "learning_rate": 2.7149425287356323e-05,
+      "loss": 0.1178,
+      "step": 3976
+    },
+    {
+      "epoch": 4.610012316163153,
+      "grad_norm": 0.3555791974067688,
+      "learning_rate": 2.714367816091954e-05,
+      "loss": 0.1039,
+      "step": 3977
+    },
+    {
+      "epoch": 4.6111714844599,
+      "grad_norm": 0.405254602432251,
+      "learning_rate": 2.713793103448276e-05,
+      "loss": 0.1036,
+      "step": 3978
+    },
+    {
+      "epoch": 4.612330652756647,
+      "grad_norm": 0.42055705189704895,
+      "learning_rate": 2.713218390804598e-05,
+      "loss": 0.1063,
+      "step": 3979
+    },
+    {
+      "epoch": 4.613489821053395,
+      "grad_norm": 0.38437414169311523,
+      "learning_rate": 2.71264367816092e-05,
+      "loss": 0.1002,
+      "step": 3980
+    },
+    {
+      "epoch": 4.614648989350141,
+      "grad_norm": 0.37615966796875,
+      "learning_rate": 2.7120689655172415e-05,
+      "loss": 0.101,
+      "step": 3981
+    },
+    {
+      "epoch": 4.615808157646889,
+      "grad_norm": 0.363473117351532,
+      "learning_rate": 2.7114942528735633e-05,
+      "loss": 0.1121,
+      "step": 3982
+    },
+    {
+      "epoch": 4.6169673259436355,
+      "grad_norm": 0.4437958896160126,
+      "learning_rate": 2.7109195402298855e-05,
+      "loss": 0.1137,
+      "step": 3983
+    },
+    {
+      "epoch": 4.618126494240382,
+      "grad_norm": 0.2698152959346771,
+      "learning_rate": 2.710344827586207e-05,
+      "loss": 0.1011,
+      "step": 3984
+    },
+    {
+      "epoch": 4.61928566253713,
+      "grad_norm": 0.360612154006958,
+      "learning_rate": 2.7097701149425288e-05,
+      "loss": 0.1123,
+      "step": 3985
+    },
+    {
+      "epoch": 4.620444830833876,
+      "grad_norm": 0.43149617314338684,
+      "learning_rate": 2.709195402298851e-05,
+      "loss": 0.1087,
+      "step": 3986
+    },
+    {
+      "epoch": 4.621603999130624,
+      "grad_norm": 0.37150686979293823,
+      "learning_rate": 2.7086206896551725e-05,
+      "loss": 0.0995,
+      "step": 3987
+    },
+    {
+      "epoch": 4.622763167427371,
+      "grad_norm": 0.3497457802295685,
+      "learning_rate": 2.7080459770114946e-05,
+      "loss": 0.1014,
+      "step": 3988
+    },
+    {
+      "epoch": 4.623922335724118,
+      "grad_norm": 0.33245381712913513,
+      "learning_rate": 2.707471264367816e-05,
+      "loss": 0.1003,
+      "step": 3989
+    },
+    {
+      "epoch": 4.625081504020865,
+      "grad_norm": 0.44494423270225525,
+      "learning_rate": 2.706896551724138e-05,
+      "loss": 0.102,
+      "step": 3990
+    },
+    {
+      "epoch": 4.626240672317612,
+      "grad_norm": 0.38262462615966797,
+      "learning_rate": 2.70632183908046e-05,
+      "loss": 0.1171,
+      "step": 3991
+    },
+    {
+      "epoch": 4.627399840614359,
+      "grad_norm": 0.4167878329753876,
+      "learning_rate": 2.7057471264367816e-05,
+      "loss": 0.1033,
+      "step": 3992
+    },
+    {
+      "epoch": 4.628559008911107,
+      "grad_norm": 0.39638566970825195,
+      "learning_rate": 2.7051724137931034e-05,
+      "loss": 0.0931,
+      "step": 3993
+    },
+    {
+      "epoch": 4.629718177207853,
+      "grad_norm": 0.45487165451049805,
+      "learning_rate": 2.7045977011494256e-05,
+      "loss": 0.1104,
+      "step": 3994
+    },
+    {
+      "epoch": 4.630877345504601,
+      "grad_norm": 0.42852312326431274,
+      "learning_rate": 2.704022988505747e-05,
+      "loss": 0.0995,
+      "step": 3995
+    },
+    {
+      "epoch": 4.6320365138013475,
+      "grad_norm": 0.388959139585495,
+      "learning_rate": 2.703448275862069e-05,
+      "loss": 0.1152,
+      "step": 3996
+    },
+    {
+      "epoch": 4.633195682098094,
+      "grad_norm": 0.33843693137168884,
+      "learning_rate": 2.702873563218391e-05,
+      "loss": 0.1018,
+      "step": 3997
+    },
+    {
+      "epoch": 4.634354850394842,
+      "grad_norm": 0.3359590768814087,
+      "learning_rate": 2.7022988505747126e-05,
+      "loss": 0.1009,
+      "step": 3998
+    },
+    {
+      "epoch": 4.635514018691588,
+      "grad_norm": 0.29807549715042114,
+      "learning_rate": 2.7017241379310348e-05,
+      "loss": 0.0949,
+      "step": 3999
+    },
+    {
+      "epoch": 4.636673186988336,
+      "grad_norm": 0.3521331250667572,
+      "learning_rate": 2.7011494252873566e-05,
+      "loss": 0.0987,
+      "step": 4000
+    },
+    {
+      "epoch": 4.637832355285083,
+      "grad_norm": 0.32744914293289185,
+      "learning_rate": 2.700574712643678e-05,
+      "loss": 0.1068,
+      "step": 4001
+    },
+    {
+      "epoch": 4.63899152358183,
+      "grad_norm": 0.33140355348587036,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 0.0941,
+      "step": 4002
+    },
+    {
+      "epoch": 4.640150691878577,
+      "grad_norm": 0.35607364773750305,
+      "learning_rate": 2.699425287356322e-05,
+      "loss": 0.1128,
+      "step": 4003
+    },
+    {
+      "epoch": 4.641309860175324,
+      "grad_norm": 0.28557470440864563,
+      "learning_rate": 2.6988505747126436e-05,
+      "loss": 0.1044,
+      "step": 4004
+    },
+    {
+      "epoch": 4.642469028472071,
+      "grad_norm": 0.4250648617744446,
+      "learning_rate": 2.6982758620689657e-05,
+      "loss": 0.1022,
+      "step": 4005
+    },
+    {
+      "epoch": 4.643628196768819,
+      "grad_norm": 0.2685200273990631,
+      "learning_rate": 2.6977011494252876e-05,
+      "loss": 0.0965,
+      "step": 4006
+    },
+    {
+      "epoch": 4.644787365065565,
+      "grad_norm": 0.3097551465034485,
+      "learning_rate": 2.6971264367816097e-05,
+      "loss": 0.11,
+      "step": 4007
+    },
+    {
+      "epoch": 4.645946533362313,
+      "grad_norm": 0.3279826045036316,
+      "learning_rate": 2.6965517241379312e-05,
+      "loss": 0.1003,
+      "step": 4008
+    },
+    {
+      "epoch": 4.64710570165906,
+      "grad_norm": 0.3129337430000305,
+      "learning_rate": 2.6959770114942527e-05,
+      "loss": 0.1062,
+      "step": 4009
+    },
+    {
+      "epoch": 4.648264869955806,
+      "grad_norm": 0.27965065836906433,
+      "learning_rate": 2.6954022988505752e-05,
+      "loss": 0.1017,
+      "step": 4010
+    },
+    {
+      "epoch": 4.649424038252554,
+      "grad_norm": 0.3252604007720947,
+      "learning_rate": 2.6948275862068967e-05,
+      "loss": 0.0935,
+      "step": 4011
+    },
+    {
+      "epoch": 4.650583206549301,
+      "grad_norm": 0.32502391934394836,
+      "learning_rate": 2.6942528735632182e-05,
+      "loss": 0.1059,
+      "step": 4012
+    },
+    {
+      "epoch": 4.651742374846048,
+      "grad_norm": 0.3370577394962311,
+      "learning_rate": 2.6936781609195404e-05,
+      "loss": 0.0959,
+      "step": 4013
+    },
+    {
+      "epoch": 4.652901543142795,
+      "grad_norm": 0.3003638684749603,
+      "learning_rate": 2.6931034482758622e-05,
+      "loss": 0.0969,
+      "step": 4014
+    },
+    {
+      "epoch": 4.654060711439542,
+      "grad_norm": 0.35102033615112305,
+      "learning_rate": 2.6925287356321837e-05,
+      "loss": 0.104,
+      "step": 4015
+    },
+    {
+      "epoch": 4.655219879736289,
+      "grad_norm": 0.355948269367218,
+      "learning_rate": 2.691954022988506e-05,
+      "loss": 0.1125,
+      "step": 4016
+    },
+    {
+      "epoch": 4.6563790480330365,
+      "grad_norm": 0.526878833770752,
+      "learning_rate": 2.6913793103448277e-05,
+      "loss": 0.1195,
+      "step": 4017
+    },
+    {
+      "epoch": 4.657538216329783,
+      "grad_norm": 0.3807942867279053,
+      "learning_rate": 2.69080459770115e-05,
+      "loss": 0.1092,
+      "step": 4018
+    },
+    {
+      "epoch": 4.658697384626531,
+      "grad_norm": 0.36606287956237793,
+      "learning_rate": 2.6902298850574713e-05,
+      "loss": 0.1061,
+      "step": 4019
+    },
+    {
+      "epoch": 4.659856552923277,
+      "grad_norm": 0.4438146650791168,
+      "learning_rate": 2.689655172413793e-05,
+      "loss": 0.1059,
+      "step": 4020
+    },
+    {
+      "epoch": 4.661015721220025,
+      "grad_norm": 0.3968319296836853,
+      "learning_rate": 2.6890804597701153e-05,
+      "loss": 0.0976,
+      "step": 4021
+    },
+    {
+      "epoch": 4.662174889516772,
+      "grad_norm": 0.34883856773376465,
+      "learning_rate": 2.6885057471264368e-05,
+      "loss": 0.1019,
+      "step": 4022
+    },
+    {
+      "epoch": 4.663334057813519,
+      "grad_norm": 0.5616782903671265,
+      "learning_rate": 2.6879310344827586e-05,
+      "loss": 0.1028,
+      "step": 4023
+    },
+    {
+      "epoch": 4.664493226110266,
+      "grad_norm": 0.3588404953479767,
+      "learning_rate": 2.6873563218390808e-05,
+      "loss": 0.0995,
+      "step": 4024
+    },
+    {
+      "epoch": 4.665652394407013,
+      "grad_norm": 0.3863297998905182,
+      "learning_rate": 2.6867816091954023e-05,
+      "loss": 0.0863,
+      "step": 4025
+    },
+    {
+      "epoch": 4.66681156270376,
+      "grad_norm": 0.2935449481010437,
+      "learning_rate": 2.6862068965517245e-05,
+      "loss": 0.0876,
+      "step": 4026
+    },
+    {
+      "epoch": 4.667970731000507,
+      "grad_norm": 0.3793972432613373,
+      "learning_rate": 2.6856321839080463e-05,
+      "loss": 0.095,
+      "step": 4027
+    },
+    {
+      "epoch": 4.669129899297254,
+      "grad_norm": 1.2369050979614258,
+      "learning_rate": 2.6850574712643678e-05,
+      "loss": 0.1036,
+      "step": 4028
+    },
+    {
+      "epoch": 4.670289067594001,
+      "grad_norm": 0.4123905599117279,
+      "learning_rate": 2.68448275862069e-05,
+      "loss": 0.0965,
+      "step": 4029
+    },
+    {
+      "epoch": 4.6714482358907485,
+      "grad_norm": 0.4004489481449127,
+      "learning_rate": 2.6839080459770118e-05,
+      "loss": 0.1136,
+      "step": 4030
+    },
+    {
+      "epoch": 4.672607404187495,
+      "grad_norm": 0.37782585620880127,
+      "learning_rate": 2.6833333333333333e-05,
+      "loss": 0.1215,
+      "step": 4031
+    },
+    {
+      "epoch": 4.673766572484243,
+      "grad_norm": 0.49208974838256836,
+      "learning_rate": 2.6827586206896554e-05,
+      "loss": 0.1135,
+      "step": 4032
+    },
+    {
+      "epoch": 4.674925740780989,
+      "grad_norm": 0.38141676783561707,
+      "learning_rate": 2.682183908045977e-05,
+      "loss": 0.0966,
+      "step": 4033
+    },
+    {
+      "epoch": 4.676084909077737,
+      "grad_norm": 0.3110532760620117,
+      "learning_rate": 2.6816091954022988e-05,
+      "loss": 0.1001,
+      "step": 4034
+    },
+    {
+      "epoch": 4.677244077374484,
+      "grad_norm": 0.38810020685195923,
+      "learning_rate": 2.681034482758621e-05,
+      "loss": 0.1021,
+      "step": 4035
+    },
+    {
+      "epoch": 4.678403245671231,
+      "grad_norm": 0.6681183576583862,
+      "learning_rate": 2.6804597701149424e-05,
+      "loss": 0.1168,
+      "step": 4036
+    },
+    {
+      "epoch": 4.679562413967978,
+      "grad_norm": 0.28798437118530273,
+      "learning_rate": 2.6798850574712646e-05,
+      "loss": 0.1115,
+      "step": 4037
+    },
+    {
+      "epoch": 4.6807215822647255,
+      "grad_norm": 0.43260279297828674,
+      "learning_rate": 2.6793103448275864e-05,
+      "loss": 0.1064,
+      "step": 4038
+    },
+    {
+      "epoch": 4.681880750561472,
+      "grad_norm": 0.3560652732849121,
+      "learning_rate": 2.678735632183908e-05,
+      "loss": 0.0963,
+      "step": 4039
+    },
+    {
+      "epoch": 4.683039918858219,
+      "grad_norm": 0.38259580731391907,
+      "learning_rate": 2.67816091954023e-05,
+      "loss": 0.1062,
+      "step": 4040
+    },
+    {
+      "epoch": 4.684199087154966,
+      "grad_norm": 0.26703354716300964,
+      "learning_rate": 2.677586206896552e-05,
+      "loss": 0.0953,
+      "step": 4041
+    },
+    {
+      "epoch": 4.685358255451713,
+      "grad_norm": 0.3569488227367401,
+      "learning_rate": 2.6770114942528734e-05,
+      "loss": 0.1066,
+      "step": 4042
+    },
+    {
+      "epoch": 4.686517423748461,
+      "grad_norm": 0.3421025574207306,
+      "learning_rate": 2.6764367816091956e-05,
+      "loss": 0.104,
+      "step": 4043
+    },
+    {
+      "epoch": 4.687676592045207,
+      "grad_norm": 0.31309759616851807,
+      "learning_rate": 2.6758620689655174e-05,
+      "loss": 0.1047,
+      "step": 4044
+    },
+    {
+      "epoch": 4.688835760341955,
+      "grad_norm": 0.6280311346054077,
+      "learning_rate": 2.6752873563218396e-05,
+      "loss": 0.1096,
+      "step": 4045
+    },
+    {
+      "epoch": 4.6899949286387015,
+      "grad_norm": 0.36928072571754456,
+      "learning_rate": 2.674712643678161e-05,
+      "loss": 0.1067,
+      "step": 4046
+    },
+    {
+      "epoch": 4.691154096935449,
+      "grad_norm": 0.2873797118663788,
+      "learning_rate": 2.674137931034483e-05,
+      "loss": 0.102,
+      "step": 4047
+    },
+    {
+      "epoch": 4.692313265232196,
+      "grad_norm": 0.3734436631202698,
+      "learning_rate": 2.673563218390805e-05,
+      "loss": 0.116,
+      "step": 4048
+    },
+    {
+      "epoch": 4.693472433528943,
+      "grad_norm": 0.33200186491012573,
+      "learning_rate": 2.6729885057471265e-05,
+      "loss": 0.0983,
+      "step": 4049
+    },
+    {
+      "epoch": 4.69463160182569,
+      "grad_norm": 0.3504514992237091,
+      "learning_rate": 2.672413793103448e-05,
+      "loss": 0.1072,
+      "step": 4050
+    },
+    {
+      "epoch": 4.6957907701224375,
+      "grad_norm": 0.4115615487098694,
+      "learning_rate": 2.6718390804597705e-05,
+      "loss": 0.1073,
+      "step": 4051
+    },
+    {
+      "epoch": 4.696949938419184,
+      "grad_norm": 0.35126233100891113,
+      "learning_rate": 2.671264367816092e-05,
+      "loss": 0.1048,
+      "step": 4052
+    },
+    {
+      "epoch": 4.698109106715931,
+      "grad_norm": 0.2556185722351074,
+      "learning_rate": 2.6706896551724135e-05,
+      "loss": 0.1022,
+      "step": 4053
+    },
+    {
+      "epoch": 4.699268275012678,
+      "grad_norm": 0.38733574748039246,
+      "learning_rate": 2.6701149425287357e-05,
+      "loss": 0.1198,
+      "step": 4054
+    },
+    {
+      "epoch": 4.700427443309426,
+      "grad_norm": 0.45527157187461853,
+      "learning_rate": 2.6695402298850575e-05,
+      "loss": 0.1059,
+      "step": 4055
+    },
+    {
+      "epoch": 4.701586611606173,
+      "grad_norm": 0.5728080868721008,
+      "learning_rate": 2.6689655172413797e-05,
+      "loss": 0.0996,
+      "step": 4056
+    },
+    {
+      "epoch": 4.702745779902919,
+      "grad_norm": 0.3340456187725067,
+      "learning_rate": 2.6683908045977012e-05,
+      "loss": 0.108,
+      "step": 4057
+    },
+    {
+      "epoch": 4.703904948199667,
+      "grad_norm": 0.35559847950935364,
+      "learning_rate": 2.667816091954023e-05,
+      "loss": 0.1056,
+      "step": 4058
+    },
+    {
+      "epoch": 4.7050641164964135,
+      "grad_norm": 0.2684958279132843,
+      "learning_rate": 2.667241379310345e-05,
+      "loss": 0.0996,
+      "step": 4059
+    },
+    {
+      "epoch": 4.706223284793161,
+      "grad_norm": 0.31825652718544006,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 0.1091,
+      "step": 4060
+    },
+    {
+      "epoch": 4.707382453089908,
+      "grad_norm": 0.3788098990917206,
+      "learning_rate": 2.6660919540229885e-05,
+      "loss": 0.0985,
+      "step": 4061
+    },
+    {
+      "epoch": 4.708541621386655,
+      "grad_norm": 0.31583863496780396,
+      "learning_rate": 2.6655172413793107e-05,
+      "loss": 0.1064,
+      "step": 4062
+    },
+    {
+      "epoch": 4.709700789683402,
+      "grad_norm": 0.5462097525596619,
+      "learning_rate": 2.664942528735632e-05,
+      "loss": 0.0976,
+      "step": 4063
+    },
+    {
+      "epoch": 4.71085995798015,
+      "grad_norm": 0.4478006660938263,
+      "learning_rate": 2.6643678160919543e-05,
+      "loss": 0.1024,
+      "step": 4064
+    },
+    {
+      "epoch": 4.712019126276896,
+      "grad_norm": 0.41845348477363586,
+      "learning_rate": 2.663793103448276e-05,
+      "loss": 0.1003,
+      "step": 4065
+    },
+    {
+      "epoch": 4.713178294573644,
+      "grad_norm": 0.3364732265472412,
+      "learning_rate": 2.6632183908045976e-05,
+      "loss": 0.0996,
+      "step": 4066
+    },
+    {
+      "epoch": 4.7143374628703905,
+      "grad_norm": 0.3141407370567322,
+      "learning_rate": 2.6626436781609198e-05,
+      "loss": 0.0995,
+      "step": 4067
+    },
+    {
+      "epoch": 4.715496631167138,
+      "grad_norm": 0.2963089942932129,
+      "learning_rate": 2.6620689655172416e-05,
+      "loss": 0.1043,
+      "step": 4068
+    },
+    {
+      "epoch": 4.716655799463885,
+      "grad_norm": 0.3780442774295807,
+      "learning_rate": 2.661494252873563e-05,
+      "loss": 0.1161,
+      "step": 4069
+    },
+    {
+      "epoch": 4.717814967760631,
+      "grad_norm": 0.4881404936313629,
+      "learning_rate": 2.6609195402298853e-05,
+      "loss": 0.1049,
+      "step": 4070
+    },
+    {
+      "epoch": 4.718974136057379,
+      "grad_norm": 0.39198312163352966,
+      "learning_rate": 2.660344827586207e-05,
+      "loss": 0.1134,
+      "step": 4071
+    },
+    {
+      "epoch": 4.720133304354126,
+      "grad_norm": 0.35181596875190735,
+      "learning_rate": 2.6597701149425286e-05,
+      "loss": 0.1013,
+      "step": 4072
+    },
+    {
+      "epoch": 4.721292472650873,
+      "grad_norm": 0.3244793713092804,
+      "learning_rate": 2.6591954022988508e-05,
+      "loss": 0.098,
+      "step": 4073
+    },
+    {
+      "epoch": 4.72245164094762,
+      "grad_norm": 0.3857901692390442,
+      "learning_rate": 2.6586206896551723e-05,
+      "loss": 0.1198,
+      "step": 4074
+    },
+    {
+      "epoch": 4.723610809244367,
+      "grad_norm": 0.3005591034889221,
+      "learning_rate": 2.6580459770114948e-05,
+      "loss": 0.0984,
+      "step": 4075
+    },
+    {
+      "epoch": 4.724769977541114,
+      "grad_norm": 0.34256380796432495,
+      "learning_rate": 2.6574712643678163e-05,
+      "loss": 0.106,
+      "step": 4076
+    },
+    {
+      "epoch": 4.725929145837862,
+      "grad_norm": 0.31626880168914795,
+      "learning_rate": 2.6568965517241378e-05,
+      "loss": 0.1079,
+      "step": 4077
+    },
+    {
+      "epoch": 4.727088314134608,
+      "grad_norm": 0.3006041944026947,
+      "learning_rate": 2.65632183908046e-05,
+      "loss": 0.1084,
+      "step": 4078
+    },
+    {
+      "epoch": 4.728247482431356,
+      "grad_norm": 0.31316351890563965,
+      "learning_rate": 2.6557471264367817e-05,
+      "loss": 0.1048,
+      "step": 4079
+    },
+    {
+      "epoch": 4.7294066507281025,
+      "grad_norm": 0.30015018582344055,
+      "learning_rate": 2.6551724137931032e-05,
+      "loss": 0.1159,
+      "step": 4080
+    },
+    {
+      "epoch": 4.73056581902485,
+      "grad_norm": 0.3358341455459595,
+      "learning_rate": 2.6545977011494254e-05,
+      "loss": 0.1077,
+      "step": 4081
+    },
+    {
+      "epoch": 4.731724987321597,
+      "grad_norm": 0.3661000728607178,
+      "learning_rate": 2.6540229885057472e-05,
+      "loss": 0.1164,
+      "step": 4082
+    },
+    {
+      "epoch": 4.732884155618343,
+      "grad_norm": 0.30234283208847046,
+      "learning_rate": 2.6534482758620694e-05,
+      "loss": 0.1018,
+      "step": 4083
+    },
+    {
+      "epoch": 4.734043323915091,
+      "grad_norm": 0.29934728145599365,
+      "learning_rate": 2.652873563218391e-05,
+      "loss": 0.1044,
+      "step": 4084
+    },
+    {
+      "epoch": 4.735202492211838,
+      "grad_norm": 0.3741641640663147,
+      "learning_rate": 2.6522988505747127e-05,
+      "loss": 0.1084,
+      "step": 4085
+    },
+    {
+      "epoch": 4.736361660508585,
+      "grad_norm": 0.35596325993537903,
+      "learning_rate": 2.651724137931035e-05,
+      "loss": 0.1069,
+      "step": 4086
+    },
+    {
+      "epoch": 4.737520828805332,
+      "grad_norm": 0.3338968753814697,
+      "learning_rate": 2.6511494252873564e-05,
+      "loss": 0.1055,
+      "step": 4087
+    },
+    {
+      "epoch": 4.738679997102079,
+      "grad_norm": 0.290626585483551,
+      "learning_rate": 2.6505747126436782e-05,
+      "loss": 0.0997,
+      "step": 4088
+    },
+    {
+      "epoch": 4.739839165398826,
+      "grad_norm": 0.6464786529541016,
+      "learning_rate": 2.6500000000000004e-05,
+      "loss": 0.1179,
+      "step": 4089
+    },
+    {
+      "epoch": 4.740998333695574,
+      "grad_norm": 0.5353714227676392,
+      "learning_rate": 2.649425287356322e-05,
+      "loss": 0.1145,
+      "step": 4090
+    },
+    {
+      "epoch": 4.74215750199232,
+      "grad_norm": 0.39538970589637756,
+      "learning_rate": 2.6488505747126437e-05,
+      "loss": 0.0998,
+      "step": 4091
+    },
+    {
+      "epoch": 4.743316670289068,
+      "grad_norm": 0.36018985509872437,
+      "learning_rate": 2.648275862068966e-05,
+      "loss": 0.1087,
+      "step": 4092
+    },
+    {
+      "epoch": 4.744475838585815,
+      "grad_norm": 0.4141652286052704,
+      "learning_rate": 2.6477011494252874e-05,
+      "loss": 0.1129,
+      "step": 4093
+    },
+    {
+      "epoch": 4.745635006882562,
+      "grad_norm": 0.4767494201660156,
+      "learning_rate": 2.6471264367816095e-05,
+      "loss": 0.116,
+      "step": 4094
+    },
+    {
+      "epoch": 4.746794175179309,
+      "grad_norm": 0.37823235988616943,
+      "learning_rate": 2.646551724137931e-05,
+      "loss": 0.1107,
+      "step": 4095
+    },
+    {
+      "epoch": 4.7479533434760555,
+      "grad_norm": 0.3563480079174042,
+      "learning_rate": 2.645977011494253e-05,
+      "loss": 0.1083,
+      "step": 4096
+    },
+    {
+      "epoch": 4.749112511772803,
+      "grad_norm": 0.27948468923568726,
+      "learning_rate": 2.645402298850575e-05,
+      "loss": 0.1067,
+      "step": 4097
+    },
+    {
+      "epoch": 4.750271680069551,
+      "grad_norm": 0.311713844537735,
+      "learning_rate": 2.6448275862068965e-05,
+      "loss": 0.1077,
+      "step": 4098
+    },
+    {
+      "epoch": 4.751430848366297,
+      "grad_norm": 0.2752847969532013,
+      "learning_rate": 2.6442528735632183e-05,
+      "loss": 0.0956,
+      "step": 4099
+    },
+    {
+      "epoch": 4.752590016663044,
+      "grad_norm": 0.3526310622692108,
+      "learning_rate": 2.6436781609195405e-05,
+      "loss": 0.1152,
+      "step": 4100
+    },
+    {
+      "epoch": 4.7537491849597915,
+      "grad_norm": 0.38243040442466736,
+      "learning_rate": 2.643103448275862e-05,
+      "loss": 0.1043,
+      "step": 4101
+    },
+    {
+      "epoch": 4.754908353256538,
+      "grad_norm": 0.32026922702789307,
+      "learning_rate": 2.642528735632184e-05,
+      "loss": 0.1129,
+      "step": 4102
+    },
+    {
+      "epoch": 4.756067521553286,
+      "grad_norm": 0.2896019220352173,
+      "learning_rate": 2.641954022988506e-05,
+      "loss": 0.0993,
+      "step": 4103
+    },
+    {
+      "epoch": 4.757226689850032,
+      "grad_norm": 0.33304235339164734,
+      "learning_rate": 2.6413793103448275e-05,
+      "loss": 0.1097,
+      "step": 4104
+    },
+    {
+      "epoch": 4.75838585814678,
+      "grad_norm": 0.2931871712207794,
+      "learning_rate": 2.6408045977011496e-05,
+      "loss": 0.0938,
+      "step": 4105
+    },
+    {
+      "epoch": 4.759545026443527,
+      "grad_norm": 0.39289239048957825,
+      "learning_rate": 2.6402298850574715e-05,
+      "loss": 0.116,
+      "step": 4106
+    },
+    {
+      "epoch": 4.760704194740274,
+      "grad_norm": 0.45948904752731323,
+      "learning_rate": 2.639655172413793e-05,
+      "loss": 0.1125,
+      "step": 4107
+    },
+    {
+      "epoch": 4.761863363037021,
+      "grad_norm": 0.3527555465698242,
+      "learning_rate": 2.639080459770115e-05,
+      "loss": 0.0992,
+      "step": 4108
+    },
+    {
+      "epoch": 4.763022531333768,
+      "grad_norm": 0.3574606478214264,
+      "learning_rate": 2.638505747126437e-05,
+      "loss": 0.1097,
+      "step": 4109
+    },
+    {
+      "epoch": 4.764181699630515,
+      "grad_norm": 0.41160982847213745,
+      "learning_rate": 2.637931034482759e-05,
+      "loss": 0.0972,
+      "step": 4110
+    },
+    {
+      "epoch": 4.765340867927263,
+      "grad_norm": 0.5347371697425842,
+      "learning_rate": 2.6373563218390806e-05,
+      "loss": 0.101,
+      "step": 4111
+    },
+    {
+      "epoch": 4.766500036224009,
+      "grad_norm": 0.3305839002132416,
+      "learning_rate": 2.6367816091954024e-05,
+      "loss": 0.0983,
+      "step": 4112
+    },
+    {
+      "epoch": 4.767659204520756,
+      "grad_norm": 0.4040708541870117,
+      "learning_rate": 2.6362068965517246e-05,
+      "loss": 0.1072,
+      "step": 4113
+    },
+    {
+      "epoch": 4.7688183728175035,
+      "grad_norm": 0.3733755648136139,
+      "learning_rate": 2.635632183908046e-05,
+      "loss": 0.0923,
+      "step": 4114
+    },
+    {
+      "epoch": 4.76997754111425,
+      "grad_norm": 0.4366565942764282,
+      "learning_rate": 2.6350574712643676e-05,
+      "loss": 0.0999,
+      "step": 4115
+    },
+    {
+      "epoch": 4.771136709410998,
+      "grad_norm": 0.3389378786087036,
+      "learning_rate": 2.63448275862069e-05,
+      "loss": 0.1016,
+      "step": 4116
+    },
+    {
+      "epoch": 4.772295877707744,
+      "grad_norm": 0.2860844135284424,
+      "learning_rate": 2.6339080459770116e-05,
+      "loss": 0.0976,
+      "step": 4117
+    },
+    {
+      "epoch": 4.773455046004492,
+      "grad_norm": 0.4258897304534912,
+      "learning_rate": 2.633333333333333e-05,
+      "loss": 0.105,
+      "step": 4118
+    },
+    {
+      "epoch": 4.774614214301239,
+      "grad_norm": 0.36527156829833984,
+      "learning_rate": 2.6327586206896552e-05,
+      "loss": 0.1089,
+      "step": 4119
+    },
+    {
+      "epoch": 4.775773382597986,
+      "grad_norm": 0.4984228312969208,
+      "learning_rate": 2.632183908045977e-05,
+      "loss": 0.1178,
+      "step": 4120
+    },
+    {
+      "epoch": 4.776932550894733,
+      "grad_norm": 0.3134072721004486,
+      "learning_rate": 2.6316091954022992e-05,
+      "loss": 0.0994,
+      "step": 4121
+    },
+    {
+      "epoch": 4.7780917191914805,
+      "grad_norm": 0.28975674510002136,
+      "learning_rate": 2.6310344827586207e-05,
+      "loss": 0.112,
+      "step": 4122
+    },
+    {
+      "epoch": 4.779250887488227,
+      "grad_norm": 0.4431288242340088,
+      "learning_rate": 2.6304597701149426e-05,
+      "loss": 0.1108,
+      "step": 4123
+    },
+    {
+      "epoch": 4.780410055784975,
+      "grad_norm": 0.3906228542327881,
+      "learning_rate": 2.6298850574712647e-05,
+      "loss": 0.1187,
+      "step": 4124
+    },
+    {
+      "epoch": 4.781569224081721,
+      "grad_norm": 0.32078051567077637,
+      "learning_rate": 2.6293103448275862e-05,
+      "loss": 0.1081,
+      "step": 4125
+    },
+    {
+      "epoch": 4.782728392378468,
+      "grad_norm": 0.34437334537506104,
+      "learning_rate": 2.628735632183908e-05,
+      "loss": 0.1045,
+      "step": 4126
+    },
+    {
+      "epoch": 4.783887560675216,
+      "grad_norm": 0.37716683745384216,
+      "learning_rate": 2.6281609195402302e-05,
+      "loss": 0.1112,
+      "step": 4127
+    },
+    {
+      "epoch": 4.785046728971962,
+      "grad_norm": 0.3816121220588684,
+      "learning_rate": 2.6275862068965517e-05,
+      "loss": 0.1015,
+      "step": 4128
+    },
+    {
+      "epoch": 4.78620589726871,
+      "grad_norm": 0.30392956733703613,
+      "learning_rate": 2.627011494252874e-05,
+      "loss": 0.096,
+      "step": 4129
+    },
+    {
+      "epoch": 4.7873650655654565,
+      "grad_norm": 0.37450024485588074,
+      "learning_rate": 2.6264367816091957e-05,
+      "loss": 0.1037,
+      "step": 4130
+    },
+    {
+      "epoch": 4.788524233862204,
+      "grad_norm": 0.3201805055141449,
+      "learning_rate": 2.6258620689655172e-05,
+      "loss": 0.0916,
+      "step": 4131
+    },
+    {
+      "epoch": 4.789683402158951,
+      "grad_norm": 0.35114505887031555,
+      "learning_rate": 2.6252873563218394e-05,
+      "loss": 0.0957,
+      "step": 4132
+    },
+    {
+      "epoch": 4.790842570455698,
+      "grad_norm": 0.3193678855895996,
+      "learning_rate": 2.6247126436781612e-05,
+      "loss": 0.0976,
+      "step": 4133
+    },
+    {
+      "epoch": 4.792001738752445,
+      "grad_norm": 0.3565538227558136,
+      "learning_rate": 2.6241379310344827e-05,
+      "loss": 0.0974,
+      "step": 4134
+    },
+    {
+      "epoch": 4.7931609070491925,
+      "grad_norm": 0.425341933965683,
+      "learning_rate": 2.623563218390805e-05,
+      "loss": 0.1083,
+      "step": 4135
+    },
+    {
+      "epoch": 4.794320075345939,
+      "grad_norm": 0.3331052362918854,
+      "learning_rate": 2.6229885057471267e-05,
+      "loss": 0.1,
+      "step": 4136
+    },
+    {
+      "epoch": 4.795479243642687,
+      "grad_norm": 0.3157724142074585,
+      "learning_rate": 2.6224137931034482e-05,
+      "loss": 0.103,
+      "step": 4137
+    },
+    {
+      "epoch": 4.796638411939433,
+      "grad_norm": 0.3534112572669983,
+      "learning_rate": 2.6218390804597703e-05,
+      "loss": 0.1032,
+      "step": 4138
+    },
+    {
+      "epoch": 4.79779758023618,
+      "grad_norm": 0.2811379134654999,
+      "learning_rate": 2.6212643678160918e-05,
+      "loss": 0.0926,
+      "step": 4139
+    },
+    {
+      "epoch": 4.798956748532928,
+      "grad_norm": 0.3642650246620178,
+      "learning_rate": 2.620689655172414e-05,
+      "loss": 0.1109,
+      "step": 4140
+    },
+    {
+      "epoch": 4.800115916829675,
+      "grad_norm": 0.4630174934864044,
+      "learning_rate": 2.6201149425287358e-05,
+      "loss": 0.1029,
+      "step": 4141
+    },
+    {
+      "epoch": 4.801275085126422,
+      "grad_norm": 0.48048609495162964,
+      "learning_rate": 2.6195402298850573e-05,
+      "loss": 0.1111,
+      "step": 4142
+    },
+    {
+      "epoch": 4.8024342534231685,
+      "grad_norm": 0.47064340114593506,
+      "learning_rate": 2.6189655172413795e-05,
+      "loss": 0.1048,
+      "step": 4143
+    },
+    {
+      "epoch": 4.803593421719916,
+      "grad_norm": 0.30796119570732117,
+      "learning_rate": 2.6183908045977013e-05,
+      "loss": 0.1083,
+      "step": 4144
+    },
+    {
+      "epoch": 4.804752590016663,
+      "grad_norm": 0.325536847114563,
+      "learning_rate": 2.6178160919540228e-05,
+      "loss": 0.1034,
+      "step": 4145
+    },
+    {
+      "epoch": 4.80591175831341,
+      "grad_norm": 0.40425774455070496,
+      "learning_rate": 2.617241379310345e-05,
+      "loss": 0.106,
+      "step": 4146
+    },
+    {
+      "epoch": 4.807070926610157,
+      "grad_norm": 0.33187299966812134,
+      "learning_rate": 2.6166666666666668e-05,
+      "loss": 0.1018,
+      "step": 4147
+    },
+    {
+      "epoch": 4.808230094906905,
+      "grad_norm": 0.35282886028289795,
+      "learning_rate": 2.616091954022989e-05,
+      "loss": 0.1052,
+      "step": 4148
+    },
+    {
+      "epoch": 4.809389263203651,
+      "grad_norm": 0.3740447163581848,
+      "learning_rate": 2.6155172413793105e-05,
+      "loss": 0.106,
+      "step": 4149
+    },
+    {
+      "epoch": 4.810548431500399,
+      "grad_norm": 0.30205562710762024,
+      "learning_rate": 2.6149425287356323e-05,
+      "loss": 0.0938,
+      "step": 4150
+    },
+    {
+      "epoch": 4.8117075997971455,
+      "grad_norm": 0.2860054671764374,
+      "learning_rate": 2.6143678160919545e-05,
+      "loss": 0.0934,
+      "step": 4151
+    },
+    {
+      "epoch": 4.812866768093892,
+      "grad_norm": 0.465196430683136,
+      "learning_rate": 2.613793103448276e-05,
+      "loss": 0.098,
+      "step": 4152
+    },
+    {
+      "epoch": 4.81402593639064,
+      "grad_norm": 0.3192795217037201,
+      "learning_rate": 2.6132183908045978e-05,
+      "loss": 0.0972,
+      "step": 4153
+    },
+    {
+      "epoch": 4.815185104687387,
+      "grad_norm": 0.42049890756607056,
+      "learning_rate": 2.61264367816092e-05,
+      "loss": 0.1133,
+      "step": 4154
+    },
+    {
+      "epoch": 4.816344272984134,
+      "grad_norm": 0.38218989968299866,
+      "learning_rate": 2.6120689655172414e-05,
+      "loss": 0.1014,
+      "step": 4155
+    },
+    {
+      "epoch": 4.817503441280881,
+      "grad_norm": 0.46369317173957825,
+      "learning_rate": 2.611494252873563e-05,
+      "loss": 0.1097,
+      "step": 4156
+    },
+    {
+      "epoch": 4.818662609577628,
+      "grad_norm": 0.3875451982021332,
+      "learning_rate": 2.6109195402298854e-05,
+      "loss": 0.1063,
+      "step": 4157
+    },
+    {
+      "epoch": 4.819821777874375,
+      "grad_norm": 0.582574725151062,
+      "learning_rate": 2.610344827586207e-05,
+      "loss": 0.0954,
+      "step": 4158
+    },
+    {
+      "epoch": 4.820980946171122,
+      "grad_norm": 0.2862607538700104,
+      "learning_rate": 2.609770114942529e-05,
+      "loss": 0.0948,
+      "step": 4159
+    },
+    {
+      "epoch": 4.822140114467869,
+      "grad_norm": 0.47772908210754395,
+      "learning_rate": 2.6091954022988506e-05,
+      "loss": 0.0984,
+      "step": 4160
+    },
+    {
+      "epoch": 4.823299282764617,
+      "grad_norm": 0.4275151789188385,
+      "learning_rate": 2.6086206896551724e-05,
+      "loss": 0.103,
+      "step": 4161
+    },
+    {
+      "epoch": 4.824458451061363,
+      "grad_norm": 0.3913339674472809,
+      "learning_rate": 2.6080459770114946e-05,
+      "loss": 0.1175,
+      "step": 4162
+    },
+    {
+      "epoch": 4.825617619358111,
+      "grad_norm": 0.37874650955200195,
+      "learning_rate": 2.607471264367816e-05,
+      "loss": 0.1093,
+      "step": 4163
+    },
+    {
+      "epoch": 4.8267767876548575,
+      "grad_norm": 0.33609622716903687,
+      "learning_rate": 2.606896551724138e-05,
+      "loss": 0.1113,
+      "step": 4164
+    },
+    {
+      "epoch": 4.827935955951605,
+      "grad_norm": 0.32703521847724915,
+      "learning_rate": 2.60632183908046e-05,
+      "loss": 0.1167,
+      "step": 4165
+    },
+    {
+      "epoch": 4.829095124248352,
+      "grad_norm": 0.45330101251602173,
+      "learning_rate": 2.6057471264367815e-05,
+      "loss": 0.0996,
+      "step": 4166
+    },
+    {
+      "epoch": 4.830254292545099,
+      "grad_norm": 0.3833322823047638,
+      "learning_rate": 2.6051724137931037e-05,
+      "loss": 0.1106,
+      "step": 4167
+    },
+    {
+      "epoch": 4.831413460841846,
+      "grad_norm": 0.416790634393692,
+      "learning_rate": 2.6045977011494255e-05,
+      "loss": 0.1077,
+      "step": 4168
+    },
+    {
+      "epoch": 4.832572629138593,
+      "grad_norm": 0.40256428718566895,
+      "learning_rate": 2.604022988505747e-05,
+      "loss": 0.0983,
+      "step": 4169
+    },
+    {
+      "epoch": 4.83373179743534,
+      "grad_norm": 0.3929414451122284,
+      "learning_rate": 2.6034482758620692e-05,
+      "loss": 0.1143,
+      "step": 4170
+    },
+    {
+      "epoch": 4.834890965732087,
+      "grad_norm": 0.3878033757209778,
+      "learning_rate": 2.602873563218391e-05,
+      "loss": 0.0966,
+      "step": 4171
+    },
+    {
+      "epoch": 4.836050134028834,
+      "grad_norm": 0.40418297052383423,
+      "learning_rate": 2.6022988505747125e-05,
+      "loss": 0.1041,
+      "step": 4172
+    },
+    {
+      "epoch": 4.837209302325581,
+      "grad_norm": 0.3119466006755829,
+      "learning_rate": 2.6017241379310347e-05,
+      "loss": 0.104,
+      "step": 4173
+    },
+    {
+      "epoch": 4.838368470622329,
+      "grad_norm": 0.31403571367263794,
+      "learning_rate": 2.6011494252873565e-05,
+      "loss": 0.1008,
+      "step": 4174
+    },
+    {
+      "epoch": 4.839527638919075,
+      "grad_norm": 0.32932233810424805,
+      "learning_rate": 2.600574712643678e-05,
+      "loss": 0.1069,
+      "step": 4175
+    },
+    {
+      "epoch": 4.840686807215823,
+      "grad_norm": 0.3228423595428467,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 0.1007,
+      "step": 4176
+    },
+    {
+      "epoch": 4.84184597551257,
+      "grad_norm": 0.3488909900188446,
+      "learning_rate": 2.599425287356322e-05,
+      "loss": 0.1019,
+      "step": 4177
+    },
+    {
+      "epoch": 4.843005143809317,
+      "grad_norm": 0.4107312858104706,
+      "learning_rate": 2.5988505747126442e-05,
+      "loss": 0.1235,
+      "step": 4178
+    },
+    {
+      "epoch": 4.844164312106064,
+      "grad_norm": 0.2800889015197754,
+      "learning_rate": 2.5982758620689657e-05,
+      "loss": 0.1017,
+      "step": 4179
+    },
+    {
+      "epoch": 4.845323480402811,
+      "grad_norm": 0.2643517851829529,
+      "learning_rate": 2.597701149425287e-05,
+      "loss": 0.1084,
+      "step": 4180
+    },
+    {
+      "epoch": 4.846482648699558,
+      "grad_norm": 0.40645238757133484,
+      "learning_rate": 2.5971264367816097e-05,
+      "loss": 0.1026,
+      "step": 4181
+    },
+    {
+      "epoch": 4.847641816996305,
+      "grad_norm": 0.5874878168106079,
+      "learning_rate": 2.596551724137931e-05,
+      "loss": 0.0999,
+      "step": 4182
+    },
+    {
+      "epoch": 4.848800985293052,
+      "grad_norm": 0.3134619891643524,
+      "learning_rate": 2.5959770114942526e-05,
+      "loss": 0.1144,
+      "step": 4183
+    },
+    {
+      "epoch": 4.8499601535898,
+      "grad_norm": 0.33643344044685364,
+      "learning_rate": 2.5954022988505748e-05,
+      "loss": 0.1035,
+      "step": 4184
+    },
+    {
+      "epoch": 4.8511193218865465,
+      "grad_norm": 0.39385950565338135,
+      "learning_rate": 2.5948275862068966e-05,
+      "loss": 0.1082,
+      "step": 4185
+    },
+    {
+      "epoch": 4.852278490183293,
+      "grad_norm": 0.32700175046920776,
+      "learning_rate": 2.5942528735632188e-05,
+      "loss": 0.1048,
+      "step": 4186
+    },
+    {
+      "epoch": 4.853437658480041,
+      "grad_norm": 0.4382705092430115,
+      "learning_rate": 2.5936781609195403e-05,
+      "loss": 0.0977,
+      "step": 4187
+    },
+    {
+      "epoch": 4.854596826776787,
+      "grad_norm": 0.4250156879425049,
+      "learning_rate": 2.593103448275862e-05,
+      "loss": 0.1147,
+      "step": 4188
+    },
+    {
+      "epoch": 4.855755995073535,
+      "grad_norm": 0.29872676730155945,
+      "learning_rate": 2.5925287356321843e-05,
+      "loss": 0.0983,
+      "step": 4189
+    },
+    {
+      "epoch": 4.856915163370282,
+      "grad_norm": 0.33435744047164917,
+      "learning_rate": 2.5919540229885058e-05,
+      "loss": 0.099,
+      "step": 4190
+    },
+    {
+      "epoch": 4.858074331667029,
+      "grad_norm": 0.49045172333717346,
+      "learning_rate": 2.5913793103448276e-05,
+      "loss": 0.0921,
+      "step": 4191
+    },
+    {
+      "epoch": 4.859233499963776,
+      "grad_norm": 0.3408951759338379,
+      "learning_rate": 2.5908045977011498e-05,
+      "loss": 0.0991,
+      "step": 4192
+    },
+    {
+      "epoch": 4.860392668260523,
+      "grad_norm": 0.2995651960372925,
+      "learning_rate": 2.5902298850574713e-05,
+      "loss": 0.0979,
+      "step": 4193
+    },
+    {
+      "epoch": 4.86155183655727,
+      "grad_norm": 0.462162584066391,
+      "learning_rate": 2.589655172413793e-05,
+      "loss": 0.1083,
+      "step": 4194
+    },
+    {
+      "epoch": 4.862711004854017,
+      "grad_norm": 0.4822317659854889,
+      "learning_rate": 2.5890804597701153e-05,
+      "loss": 0.0976,
+      "step": 4195
+    },
+    {
+      "epoch": 4.863870173150764,
+      "grad_norm": 0.34353649616241455,
+      "learning_rate": 2.5885057471264368e-05,
+      "loss": 0.1036,
+      "step": 4196
+    },
+    {
+      "epoch": 4.865029341447512,
+      "grad_norm": 0.32048463821411133,
+      "learning_rate": 2.587931034482759e-05,
+      "loss": 0.1164,
+      "step": 4197
+    },
+    {
+      "epoch": 4.8661885097442585,
+      "grad_norm": 0.37083297967910767,
+      "learning_rate": 2.5873563218390808e-05,
+      "loss": 0.107,
+      "step": 4198
+    },
+    {
+      "epoch": 4.867347678041005,
+      "grad_norm": 0.2568010091781616,
+      "learning_rate": 2.5867816091954022e-05,
+      "loss": 0.0924,
+      "step": 4199
+    },
+    {
+      "epoch": 4.868506846337753,
+      "grad_norm": 0.3095264136791229,
+      "learning_rate": 2.5862068965517244e-05,
+      "loss": 0.0984,
+      "step": 4200
+    },
+    {
+      "epoch": 4.869666014634499,
+      "grad_norm": 0.34810638427734375,
+      "learning_rate": 2.585632183908046e-05,
+      "loss": 0.1055,
+      "step": 4201
+    },
+    {
+      "epoch": 4.870825182931247,
+      "grad_norm": 0.3801257014274597,
+      "learning_rate": 2.5850574712643677e-05,
+      "loss": 0.1107,
+      "step": 4202
+    },
+    {
+      "epoch": 4.871984351227994,
+      "grad_norm": 0.31729915738105774,
+      "learning_rate": 2.58448275862069e-05,
+      "loss": 0.1158,
+      "step": 4203
+    },
+    {
+      "epoch": 4.873143519524741,
+      "grad_norm": 0.4292013347148895,
+      "learning_rate": 2.5839080459770114e-05,
+      "loss": 0.1062,
+      "step": 4204
+    },
+    {
+      "epoch": 4.874302687821488,
+      "grad_norm": 0.35838496685028076,
+      "learning_rate": 2.5833333333333336e-05,
+      "loss": 0.1062,
+      "step": 4205
+    },
+    {
+      "epoch": 4.8754618561182355,
+      "grad_norm": 0.479670912027359,
+      "learning_rate": 2.5827586206896554e-05,
+      "loss": 0.0975,
+      "step": 4206
+    },
+    {
+      "epoch": 4.876621024414982,
+      "grad_norm": 0.3806205689907074,
+      "learning_rate": 2.582183908045977e-05,
+      "loss": 0.1128,
+      "step": 4207
+    },
+    {
+      "epoch": 4.87778019271173,
+      "grad_norm": 0.2937345802783966,
+      "learning_rate": 2.581609195402299e-05,
+      "loss": 0.1024,
+      "step": 4208
+    },
+    {
+      "epoch": 4.878939361008476,
+      "grad_norm": 0.28283917903900146,
+      "learning_rate": 2.581034482758621e-05,
+      "loss": 0.0977,
+      "step": 4209
+    },
+    {
+      "epoch": 4.880098529305224,
+      "grad_norm": 0.4149821400642395,
+      "learning_rate": 2.5804597701149424e-05,
+      "loss": 0.1125,
+      "step": 4210
+    },
+    {
+      "epoch": 4.881257697601971,
+      "grad_norm": 0.41813716292381287,
+      "learning_rate": 2.5798850574712645e-05,
+      "loss": 0.1096,
+      "step": 4211
+    },
+    {
+      "epoch": 4.882416865898717,
+      "grad_norm": 0.40485164523124695,
+      "learning_rate": 2.5793103448275864e-05,
+      "loss": 0.1119,
+      "step": 4212
+    },
+    {
+      "epoch": 4.883576034195465,
+      "grad_norm": 0.36861032247543335,
+      "learning_rate": 2.578735632183908e-05,
+      "loss": 0.0978,
+      "step": 4213
+    },
+    {
+      "epoch": 4.8847352024922115,
+      "grad_norm": 0.4352358281612396,
+      "learning_rate": 2.57816091954023e-05,
+      "loss": 0.1026,
+      "step": 4214
+    },
+    {
+      "epoch": 4.885894370788959,
+      "grad_norm": 0.3257162272930145,
+      "learning_rate": 2.577586206896552e-05,
+      "loss": 0.1098,
+      "step": 4215
+    },
+    {
+      "epoch": 4.887053539085706,
+      "grad_norm": 0.3483160734176636,
+      "learning_rate": 2.577011494252874e-05,
+      "loss": 0.1115,
+      "step": 4216
+    },
+    {
+      "epoch": 4.888212707382453,
+      "grad_norm": 0.30593305826187134,
+      "learning_rate": 2.5764367816091955e-05,
+      "loss": 0.1081,
+      "step": 4217
+    },
+    {
+      "epoch": 4.8893718756792,
+      "grad_norm": 0.30475834012031555,
+      "learning_rate": 2.5758620689655173e-05,
+      "loss": 0.0919,
+      "step": 4218
+    },
+    {
+      "epoch": 4.8905310439759475,
+      "grad_norm": 0.296648770570755,
+      "learning_rate": 2.5752873563218395e-05,
+      "loss": 0.0985,
+      "step": 4219
+    },
+    {
+      "epoch": 4.891690212272694,
+      "grad_norm": 0.3274528980255127,
+      "learning_rate": 2.574712643678161e-05,
+      "loss": 0.1018,
+      "step": 4220
+    },
+    {
+      "epoch": 4.892849380569442,
+      "grad_norm": 0.3239176571369171,
+      "learning_rate": 2.5741379310344825e-05,
+      "loss": 0.1053,
+      "step": 4221
+    },
+    {
+      "epoch": 4.894008548866188,
+      "grad_norm": 0.3093695342540741,
+      "learning_rate": 2.573563218390805e-05,
+      "loss": 0.1006,
+      "step": 4222
+    },
+    {
+      "epoch": 4.895167717162936,
+      "grad_norm": 0.29413002729415894,
+      "learning_rate": 2.5729885057471265e-05,
+      "loss": 0.1035,
+      "step": 4223
+    },
+    {
+      "epoch": 4.896326885459683,
+      "grad_norm": 0.3807078003883362,
+      "learning_rate": 2.5724137931034486e-05,
+      "loss": 0.1111,
+      "step": 4224
+    },
+    {
+      "epoch": 4.897486053756429,
+      "grad_norm": 0.357345312833786,
+      "learning_rate": 2.57183908045977e-05,
+      "loss": 0.1102,
+      "step": 4225
+    },
+    {
+      "epoch": 4.898645222053177,
+      "grad_norm": 0.45739462971687317,
+      "learning_rate": 2.571264367816092e-05,
+      "loss": 0.0942,
+      "step": 4226
+    },
+    {
+      "epoch": 4.8998043903499235,
+      "grad_norm": 0.325125515460968,
+      "learning_rate": 2.570689655172414e-05,
+      "loss": 0.0932,
+      "step": 4227
+    },
+    {
+      "epoch": 4.900963558646671,
+      "grad_norm": 0.39498814940452576,
+      "learning_rate": 2.5701149425287356e-05,
+      "loss": 0.1001,
+      "step": 4228
+    },
+    {
+      "epoch": 4.902122726943418,
+      "grad_norm": 0.3661857843399048,
+      "learning_rate": 2.5695402298850575e-05,
+      "loss": 0.1013,
+      "step": 4229
+    },
+    {
+      "epoch": 4.903281895240165,
+      "grad_norm": 0.4122169315814972,
+      "learning_rate": 2.5689655172413796e-05,
+      "loss": 0.1106,
+      "step": 4230
+    },
+    {
+      "epoch": 4.904441063536912,
+      "grad_norm": 0.6558185815811157,
+      "learning_rate": 2.568390804597701e-05,
+      "loss": 0.1162,
+      "step": 4231
+    },
+    {
+      "epoch": 4.90560023183366,
+      "grad_norm": 0.2946203052997589,
+      "learning_rate": 2.567816091954023e-05,
+      "loss": 0.1005,
+      "step": 4232
+    },
+    {
+      "epoch": 4.906759400130406,
+      "grad_norm": 0.3813852071762085,
+      "learning_rate": 2.567241379310345e-05,
+      "loss": 0.1039,
+      "step": 4233
+    },
+    {
+      "epoch": 4.907918568427154,
+      "grad_norm": 0.33533862233161926,
+      "learning_rate": 2.5666666666666666e-05,
+      "loss": 0.0955,
+      "step": 4234
+    },
+    {
+      "epoch": 4.9090777367239005,
+      "grad_norm": 0.3679860830307007,
+      "learning_rate": 2.5660919540229888e-05,
+      "loss": 0.1084,
+      "step": 4235
+    },
+    {
+      "epoch": 4.910236905020648,
+      "grad_norm": 0.32838982343673706,
+      "learning_rate": 2.5655172413793106e-05,
+      "loss": 0.1006,
+      "step": 4236
+    },
+    {
+      "epoch": 4.911396073317395,
+      "grad_norm": 0.40485119819641113,
+      "learning_rate": 2.564942528735632e-05,
+      "loss": 0.1078,
+      "step": 4237
+    },
+    {
+      "epoch": 4.912555241614141,
+      "grad_norm": 0.43411320447921753,
+      "learning_rate": 2.5643678160919543e-05,
+      "loss": 0.1015,
+      "step": 4238
+    },
+    {
+      "epoch": 4.913714409910889,
+      "grad_norm": 0.622022807598114,
+      "learning_rate": 2.563793103448276e-05,
+      "loss": 0.1006,
+      "step": 4239
+    },
+    {
+      "epoch": 4.9148735782076365,
+      "grad_norm": 0.37312477827072144,
+      "learning_rate": 2.5632183908045976e-05,
+      "loss": 0.0984,
+      "step": 4240
+    },
+    {
+      "epoch": 4.916032746504383,
+      "grad_norm": 0.4039802849292755,
+      "learning_rate": 2.5626436781609197e-05,
+      "loss": 0.1012,
+      "step": 4241
+    },
+    {
+      "epoch": 4.91719191480113,
+      "grad_norm": 0.28577470779418945,
+      "learning_rate": 2.5620689655172416e-05,
+      "loss": 0.1062,
+      "step": 4242
+    },
+    {
+      "epoch": 4.918351083097877,
+      "grad_norm": 0.4363808333873749,
+      "learning_rate": 2.5614942528735637e-05,
+      "loss": 0.1079,
+      "step": 4243
+    },
+    {
+      "epoch": 4.919510251394624,
+      "grad_norm": 0.31524115800857544,
+      "learning_rate": 2.5609195402298852e-05,
+      "loss": 0.1027,
+      "step": 4244
+    },
+    {
+      "epoch": 4.920669419691372,
+      "grad_norm": 0.324258416891098,
+      "learning_rate": 2.5603448275862067e-05,
+      "loss": 0.1011,
+      "step": 4245
+    },
+    {
+      "epoch": 4.921828587988118,
+      "grad_norm": 0.3237408697605133,
+      "learning_rate": 2.559770114942529e-05,
+      "loss": 0.0999,
+      "step": 4246
+    },
+    {
+      "epoch": 4.922987756284866,
+      "grad_norm": 0.3803741931915283,
+      "learning_rate": 2.5591954022988507e-05,
+      "loss": 0.1069,
+      "step": 4247
+    },
+    {
+      "epoch": 4.9241469245816125,
+      "grad_norm": 0.36005181074142456,
+      "learning_rate": 2.5586206896551722e-05,
+      "loss": 0.1102,
+      "step": 4248
+    },
+    {
+      "epoch": 4.92530609287836,
+      "grad_norm": 0.3981182873249054,
+      "learning_rate": 2.5580459770114944e-05,
+      "loss": 0.1076,
+      "step": 4249
+    },
+    {
+      "epoch": 4.926465261175107,
+      "grad_norm": 0.37391379475593567,
+      "learning_rate": 2.5574712643678162e-05,
+      "loss": 0.0985,
+      "step": 4250
+    },
+    {
+      "epoch": 4.927624429471854,
+      "grad_norm": 0.38138946890830994,
+      "learning_rate": 2.5568965517241377e-05,
+      "loss": 0.0987,
+      "step": 4251
+    },
+    {
+      "epoch": 4.928783597768601,
+      "grad_norm": 0.5195475816726685,
+      "learning_rate": 2.55632183908046e-05,
+      "loss": 0.0944,
+      "step": 4252
+    },
+    {
+      "epoch": 4.9299427660653485,
+      "grad_norm": 0.4090215265750885,
+      "learning_rate": 2.5557471264367817e-05,
+      "loss": 0.1082,
+      "step": 4253
+    },
+    {
+      "epoch": 4.931101934362095,
+      "grad_norm": 0.2851259410381317,
+      "learning_rate": 2.555172413793104e-05,
+      "loss": 0.0993,
+      "step": 4254
+    },
+    {
+      "epoch": 4.932261102658842,
+      "grad_norm": 0.37689948081970215,
+      "learning_rate": 2.5545977011494253e-05,
+      "loss": 0.108,
+      "step": 4255
+    },
+    {
+      "epoch": 4.933420270955589,
+      "grad_norm": 0.3453246057033539,
+      "learning_rate": 2.5540229885057472e-05,
+      "loss": 0.1001,
+      "step": 4256
+    },
+    {
+      "epoch": 4.934579439252336,
+      "grad_norm": 0.4019702970981598,
+      "learning_rate": 2.5534482758620693e-05,
+      "loss": 0.1084,
+      "step": 4257
+    },
+    {
+      "epoch": 4.935738607549084,
+      "grad_norm": 0.30191949009895325,
+      "learning_rate": 2.552873563218391e-05,
+      "loss": 0.0991,
+      "step": 4258
+    },
+    {
+      "epoch": 4.93689777584583,
+      "grad_norm": 0.34570759534835815,
+      "learning_rate": 2.5522988505747127e-05,
+      "loss": 0.0976,
+      "step": 4259
+    },
+    {
+      "epoch": 4.938056944142578,
+      "grad_norm": 0.32790854573249817,
+      "learning_rate": 2.551724137931035e-05,
+      "loss": 0.088,
+      "step": 4260
+    },
+    {
+      "epoch": 4.939216112439325,
+      "grad_norm": 0.32591426372528076,
+      "learning_rate": 2.5511494252873563e-05,
+      "loss": 0.0989,
+      "step": 4261
+    },
+    {
+      "epoch": 4.940375280736072,
+      "grad_norm": 0.3843526542186737,
+      "learning_rate": 2.5505747126436785e-05,
+      "loss": 0.103,
+      "step": 4262
+    },
+    {
+      "epoch": 4.941534449032819,
+      "grad_norm": 0.38071221113204956,
+      "learning_rate": 2.5500000000000003e-05,
+      "loss": 0.1045,
+      "step": 4263
+    },
+    {
+      "epoch": 4.942693617329566,
+      "grad_norm": 0.3825156092643738,
+      "learning_rate": 2.5494252873563218e-05,
+      "loss": 0.0969,
+      "step": 4264
+    },
+    {
+      "epoch": 4.943852785626313,
+      "grad_norm": 0.3230441212654114,
+      "learning_rate": 2.548850574712644e-05,
+      "loss": 0.1042,
+      "step": 4265
+    },
+    {
+      "epoch": 4.945011953923061,
+      "grad_norm": 0.35447943210601807,
+      "learning_rate": 2.5482758620689655e-05,
+      "loss": 0.1133,
+      "step": 4266
+    },
+    {
+      "epoch": 4.946171122219807,
+      "grad_norm": 0.2828787863254547,
+      "learning_rate": 2.5477011494252873e-05,
+      "loss": 0.0917,
+      "step": 4267
+    },
+    {
+      "epoch": 4.947330290516554,
+      "grad_norm": 0.2748185694217682,
+      "learning_rate": 2.5471264367816095e-05,
+      "loss": 0.0973,
+      "step": 4268
+    },
+    {
+      "epoch": 4.9484894588133015,
+      "grad_norm": 0.321551114320755,
+      "learning_rate": 2.546551724137931e-05,
+      "loss": 0.0973,
+      "step": 4269
+    },
+    {
+      "epoch": 4.949648627110048,
+      "grad_norm": 0.3038749396800995,
+      "learning_rate": 2.545977011494253e-05,
+      "loss": 0.0888,
+      "step": 4270
+    },
+    {
+      "epoch": 4.950807795406796,
+      "grad_norm": 0.4207051396369934,
+      "learning_rate": 2.545402298850575e-05,
+      "loss": 0.0991,
+      "step": 4271
+    },
+    {
+      "epoch": 4.951966963703542,
+      "grad_norm": 0.44856876134872437,
+      "learning_rate": 2.5448275862068964e-05,
+      "loss": 0.1067,
+      "step": 4272
+    },
+    {
+      "epoch": 4.95312613200029,
+      "grad_norm": 0.3357430398464203,
+      "learning_rate": 2.5442528735632186e-05,
+      "loss": 0.0877,
+      "step": 4273
+    },
+    {
+      "epoch": 4.954285300297037,
+      "grad_norm": 0.33568447828292847,
+      "learning_rate": 2.5436781609195404e-05,
+      "loss": 0.0988,
+      "step": 4274
+    },
+    {
+      "epoch": 4.955444468593784,
+      "grad_norm": 0.3613547384738922,
+      "learning_rate": 2.543103448275862e-05,
+      "loss": 0.1052,
+      "step": 4275
+    },
+    {
+      "epoch": 4.956603636890531,
+      "grad_norm": 0.3529345393180847,
+      "learning_rate": 2.542528735632184e-05,
+      "loss": 0.1049,
+      "step": 4276
+    },
+    {
+      "epoch": 4.957762805187278,
+      "grad_norm": 0.35080334544181824,
+      "learning_rate": 2.541954022988506e-05,
+      "loss": 0.1034,
+      "step": 4277
+    },
+    {
+      "epoch": 4.958921973484025,
+      "grad_norm": 0.68172287940979,
+      "learning_rate": 2.5413793103448274e-05,
+      "loss": 0.1033,
+      "step": 4278
+    },
+    {
+      "epoch": 4.960081141780773,
+      "grad_norm": 0.3911987543106079,
+      "learning_rate": 2.5408045977011496e-05,
+      "loss": 0.1144,
+      "step": 4279
+    },
+    {
+      "epoch": 4.961240310077519,
+      "grad_norm": 0.3486397862434387,
+      "learning_rate": 2.5402298850574714e-05,
+      "loss": 0.1069,
+      "step": 4280
+    },
+    {
+      "epoch": 4.962399478374266,
+      "grad_norm": 0.4205693006515503,
+      "learning_rate": 2.5396551724137936e-05,
+      "loss": 0.1115,
+      "step": 4281
+    },
+    {
+      "epoch": 4.9635586466710135,
+      "grad_norm": 0.4175969660282135,
+      "learning_rate": 2.539080459770115e-05,
+      "loss": 0.106,
+      "step": 4282
+    },
+    {
+      "epoch": 4.964717814967761,
+      "grad_norm": 0.3387162685394287,
+      "learning_rate": 2.538505747126437e-05,
+      "loss": 0.1113,
+      "step": 4283
+    },
+    {
+      "epoch": 4.965876983264508,
+      "grad_norm": 0.3226720988750458,
+      "learning_rate": 2.537931034482759e-05,
+      "loss": 0.0975,
+      "step": 4284
+    },
+    {
+      "epoch": 4.967036151561254,
+      "grad_norm": 0.3947923183441162,
+      "learning_rate": 2.5373563218390806e-05,
+      "loss": 0.0999,
+      "step": 4285
+    },
+    {
+      "epoch": 4.968195319858002,
+      "grad_norm": 0.40375250577926636,
+      "learning_rate": 2.536781609195402e-05,
+      "loss": 0.1104,
+      "step": 4286
+    },
+    {
+      "epoch": 4.969354488154749,
+      "grad_norm": 0.44277888536453247,
+      "learning_rate": 2.5362068965517246e-05,
+      "loss": 0.1052,
+      "step": 4287
+    },
+    {
+      "epoch": 4.970513656451496,
+      "grad_norm": 0.38605281710624695,
+      "learning_rate": 2.535632183908046e-05,
+      "loss": 0.1062,
+      "step": 4288
+    },
+    {
+      "epoch": 4.971672824748243,
+      "grad_norm": 0.3479140102863312,
+      "learning_rate": 2.5350574712643682e-05,
+      "loss": 0.1085,
+      "step": 4289
+    },
+    {
+      "epoch": 4.9728319930449905,
+      "grad_norm": 0.38351091742515564,
+      "learning_rate": 2.5344827586206897e-05,
+      "loss": 0.1046,
+      "step": 4290
+    },
+    {
+      "epoch": 4.973991161341737,
+      "grad_norm": 0.3506777882575989,
+      "learning_rate": 2.5339080459770115e-05,
+      "loss": 0.0964,
+      "step": 4291
+    },
+    {
+      "epoch": 4.975150329638485,
+      "grad_norm": 0.4924359619617462,
+      "learning_rate": 2.5333333333333337e-05,
+      "loss": 0.116,
+      "step": 4292
+    },
+    {
+      "epoch": 4.976309497935231,
+      "grad_norm": 0.31600138545036316,
+      "learning_rate": 2.5327586206896552e-05,
+      "loss": 0.1058,
+      "step": 4293
+    },
+    {
+      "epoch": 4.977468666231979,
+      "grad_norm": 0.3318912386894226,
+      "learning_rate": 2.532183908045977e-05,
+      "loss": 0.0918,
+      "step": 4294
+    },
+    {
+      "epoch": 4.978627834528726,
+      "grad_norm": 0.361688494682312,
+      "learning_rate": 2.5316091954022992e-05,
+      "loss": 0.1061,
+      "step": 4295
+    },
+    {
+      "epoch": 4.979787002825473,
+      "grad_norm": 0.37395092844963074,
+      "learning_rate": 2.5310344827586207e-05,
+      "loss": 0.1067,
+      "step": 4296
+    },
+    {
+      "epoch": 4.98094617112222,
+      "grad_norm": 0.48089030385017395,
+      "learning_rate": 2.5304597701149425e-05,
+      "loss": 0.1015,
+      "step": 4297
+    },
+    {
+      "epoch": 4.9821053394189665,
+      "grad_norm": 0.305130273103714,
+      "learning_rate": 2.5298850574712647e-05,
+      "loss": 0.0966,
+      "step": 4298
+    },
+    {
+      "epoch": 4.983264507715714,
+      "grad_norm": 0.30088353157043457,
+      "learning_rate": 2.529310344827586e-05,
+      "loss": 0.1099,
+      "step": 4299
+    },
+    {
+      "epoch": 4.984423676012461,
+      "grad_norm": 0.4635865092277527,
+      "learning_rate": 2.5287356321839083e-05,
+      "loss": 0.1118,
+      "step": 4300
+    },
+    {
+      "epoch": 4.985582844309208,
+      "grad_norm": 0.34878647327423096,
+      "learning_rate": 2.52816091954023e-05,
+      "loss": 0.1004,
+      "step": 4301
+    },
+    {
+      "epoch": 4.986742012605955,
+      "grad_norm": 0.31814780831336975,
+      "learning_rate": 2.5275862068965516e-05,
+      "loss": 0.1032,
+      "step": 4302
+    },
+    {
+      "epoch": 4.9879011809027025,
+      "grad_norm": 0.311321884393692,
+      "learning_rate": 2.5270114942528738e-05,
+      "loss": 0.0982,
+      "step": 4303
+    },
+    {
+      "epoch": 4.989060349199449,
+      "grad_norm": 0.3587198853492737,
+      "learning_rate": 2.5264367816091956e-05,
+      "loss": 0.0981,
+      "step": 4304
+    },
+    {
+      "epoch": 4.990219517496197,
+      "grad_norm": 0.33750030398368835,
+      "learning_rate": 2.525862068965517e-05,
+      "loss": 0.0996,
+      "step": 4305
+    },
+    {
+      "epoch": 4.991378685792943,
+      "grad_norm": 0.3702969253063202,
+      "learning_rate": 2.5252873563218393e-05,
+      "loss": 0.1126,
+      "step": 4306
+    },
+    {
+      "epoch": 4.992537854089691,
+      "grad_norm": 0.33090633153915405,
+      "learning_rate": 2.5247126436781608e-05,
+      "loss": 0.1072,
+      "step": 4307
+    },
+    {
+      "epoch": 4.993697022386438,
+      "grad_norm": 0.4805144965648651,
+      "learning_rate": 2.5241379310344833e-05,
+      "loss": 0.0977,
+      "step": 4308
+    },
+    {
+      "epoch": 4.994856190683185,
+      "grad_norm": 0.3834288418292999,
+      "learning_rate": 2.5235632183908048e-05,
+      "loss": 0.0978,
+      "step": 4309
+    },
+    {
+      "epoch": 4.996015358979932,
+      "grad_norm": 0.39906617999076843,
+      "learning_rate": 2.5229885057471263e-05,
+      "loss": 0.1049,
+      "step": 4310
+    },
+    {
+      "epoch": 4.9971745272766785,
+      "grad_norm": 0.2854554355144501,
+      "learning_rate": 2.5224137931034484e-05,
+      "loss": 0.1047,
+      "step": 4311
+    },
+    {
+      "epoch": 4.998333695573426,
+      "grad_norm": 0.37985438108444214,
+      "learning_rate": 2.5218390804597703e-05,
+      "loss": 0.101,
+      "step": 4312
+    },
+    {
+      "epoch": 4.999492863870173,
+      "grad_norm": 0.4384147822856903,
+      "learning_rate": 2.5212643678160918e-05,
+      "loss": 0.1011,
+      "step": 4313
+    },
+    {
+      "epoch": 4.999492863870173,
+      "eval_loss": 0.1395179033279419,
+      "eval_runtime": 265.5301,
+      "eval_samples_per_second": 5.777,
+      "eval_steps_per_second": 5.777,
+      "step": 4313
+    },
+    {
+      "epoch": 5.00065203216692,
+      "grad_norm": 0.49006325006484985,
+      "learning_rate": 2.520689655172414e-05,
+      "loss": 0.1007,
+      "step": 4314
+    },
+    {
+      "epoch": 5.001811200463667,
+      "grad_norm": 0.2520461082458496,
+      "learning_rate": 2.5201149425287358e-05,
+      "loss": 0.0905,
+      "step": 4315
+    },
+    {
+      "epoch": 5.002970368760415,
+      "grad_norm": 0.2542027235031128,
+      "learning_rate": 2.5195402298850573e-05,
+      "loss": 0.0858,
+      "step": 4316
+    },
+    {
+      "epoch": 5.004129537057161,
+      "grad_norm": 0.3190763592720032,
+      "learning_rate": 2.5189655172413794e-05,
+      "loss": 0.0818,
+      "step": 4317
+    },
+    {
+      "epoch": 5.005288705353909,
+      "grad_norm": 0.2673119008541107,
+      "learning_rate": 2.5183908045977013e-05,
+      "loss": 0.0943,
+      "step": 4318
+    },
+    {
+      "epoch": 5.0064478736506555,
+      "grad_norm": 0.31075116991996765,
+      "learning_rate": 2.5178160919540234e-05,
+      "loss": 0.0966,
+      "step": 4319
+    },
+    {
+      "epoch": 5.007607041947403,
+      "grad_norm": 0.3015635311603546,
+      "learning_rate": 2.517241379310345e-05,
+      "loss": 0.0971,
+      "step": 4320
+    },
+    {
+      "epoch": 5.00876621024415,
+      "grad_norm": 0.43081480264663696,
+      "learning_rate": 2.5166666666666667e-05,
+      "loss": 0.0955,
+      "step": 4321
+    },
+    {
+      "epoch": 5.009925378540897,
+      "grad_norm": 0.3676944971084595,
+      "learning_rate": 2.516091954022989e-05,
+      "loss": 0.0966,
+      "step": 4322
+    },
+    {
+      "epoch": 5.011084546837644,
+      "grad_norm": 0.34852761030197144,
+      "learning_rate": 2.5155172413793104e-05,
+      "loss": 0.0881,
+      "step": 4323
+    },
+    {
+      "epoch": 5.0122437151343915,
+      "grad_norm": 0.3189998269081116,
+      "learning_rate": 2.5149425287356322e-05,
+      "loss": 0.0981,
+      "step": 4324
+    },
+    {
+      "epoch": 5.013402883431138,
+      "grad_norm": 0.5241860151290894,
+      "learning_rate": 2.5143678160919544e-05,
+      "loss": 0.0923,
+      "step": 4325
+    },
+    {
+      "epoch": 5.014562051727885,
+      "grad_norm": 0.4675620198249817,
+      "learning_rate": 2.513793103448276e-05,
+      "loss": 0.0899,
+      "step": 4326
+    },
+    {
+      "epoch": 5.015721220024632,
+      "grad_norm": 0.37763670086860657,
+      "learning_rate": 2.513218390804598e-05,
+      "loss": 0.084,
+      "step": 4327
+    },
+    {
+      "epoch": 5.016880388321379,
+      "grad_norm": 0.4259863495826721,
+      "learning_rate": 2.51264367816092e-05,
+      "loss": 0.1091,
+      "step": 4328
+    },
+    {
+      "epoch": 5.018039556618127,
+      "grad_norm": 0.6199847459793091,
+      "learning_rate": 2.5120689655172414e-05,
+      "loss": 0.0982,
+      "step": 4329
+    },
+    {
+      "epoch": 5.019198724914873,
+      "grad_norm": 0.4678066670894623,
+      "learning_rate": 2.5114942528735635e-05,
+      "loss": 0.0968,
+      "step": 4330
+    },
+    {
+      "epoch": 5.020357893211621,
+      "grad_norm": 0.46299999952316284,
+      "learning_rate": 2.510919540229885e-05,
+      "loss": 0.0903,
+      "step": 4331
+    },
+    {
+      "epoch": 5.0215170615083675,
+      "grad_norm": 0.3906437158584595,
+      "learning_rate": 2.510344827586207e-05,
+      "loss": 0.0929,
+      "step": 4332
+    },
+    {
+      "epoch": 5.022676229805115,
+      "grad_norm": 0.4142506420612335,
+      "learning_rate": 2.509770114942529e-05,
+      "loss": 0.1009,
+      "step": 4333
+    },
+    {
+      "epoch": 5.023835398101862,
+      "grad_norm": 0.3510461151599884,
+      "learning_rate": 2.5091954022988505e-05,
+      "loss": 0.0879,
+      "step": 4334
+    },
+    {
+      "epoch": 5.024994566398609,
+      "grad_norm": 0.3772026598453522,
+      "learning_rate": 2.5086206896551723e-05,
+      "loss": 0.0947,
+      "step": 4335
+    },
+    {
+      "epoch": 5.026153734695356,
+      "grad_norm": 0.3540421426296234,
+      "learning_rate": 2.5080459770114945e-05,
+      "loss": 0.0897,
+      "step": 4336
+    },
+    {
+      "epoch": 5.0273129029921035,
+      "grad_norm": 0.39619699120521545,
+      "learning_rate": 2.507471264367816e-05,
+      "loss": 0.0932,
+      "step": 4337
+    },
+    {
+      "epoch": 5.02847207128885,
+      "grad_norm": 0.4173181354999542,
+      "learning_rate": 2.5068965517241382e-05,
+      "loss": 0.0857,
+      "step": 4338
+    },
+    {
+      "epoch": 5.029631239585598,
+      "grad_norm": 0.4821605682373047,
+      "learning_rate": 2.50632183908046e-05,
+      "loss": 0.0936,
+      "step": 4339
+    },
+    {
+      "epoch": 5.030790407882344,
+      "grad_norm": 0.3579258918762207,
+      "learning_rate": 2.5057471264367815e-05,
+      "loss": 0.0945,
+      "step": 4340
+    },
+    {
+      "epoch": 5.031949576179091,
+      "grad_norm": 0.3280121982097626,
+      "learning_rate": 2.5051724137931037e-05,
+      "loss": 0.0958,
+      "step": 4341
+    },
+    {
+      "epoch": 5.033108744475839,
+      "grad_norm": 0.4641885757446289,
+      "learning_rate": 2.5045977011494255e-05,
+      "loss": 0.0982,
+      "step": 4342
+    },
+    {
+      "epoch": 5.034267912772585,
+      "grad_norm": 0.4874429404735565,
+      "learning_rate": 2.504022988505747e-05,
+      "loss": 0.0996,
+      "step": 4343
+    },
+    {
+      "epoch": 5.035427081069333,
+      "grad_norm": 0.45620620250701904,
+      "learning_rate": 2.503448275862069e-05,
+      "loss": 0.0894,
+      "step": 4344
+    },
+    {
+      "epoch": 5.03658624936608,
+      "grad_norm": 0.42721596360206604,
+      "learning_rate": 2.502873563218391e-05,
+      "loss": 0.1,
+      "step": 4345
+    },
+    {
+      "epoch": 5.037745417662827,
+      "grad_norm": 0.5702162981033325,
+      "learning_rate": 2.502298850574713e-05,
+      "loss": 0.0954,
+      "step": 4346
+    },
+    {
+      "epoch": 5.038904585959574,
+      "grad_norm": 0.2864901125431061,
+      "learning_rate": 2.5017241379310346e-05,
+      "loss": 0.0902,
+      "step": 4347
+    },
+    {
+      "epoch": 5.040063754256321,
+      "grad_norm": 0.31084322929382324,
+      "learning_rate": 2.5011494252873565e-05,
+      "loss": 0.0924,
+      "step": 4348
+    },
+    {
+      "epoch": 5.041222922553068,
+      "grad_norm": 0.5150142312049866,
+      "learning_rate": 2.5005747126436786e-05,
+      "loss": 0.0936,
+      "step": 4349
+    },
+    {
+      "epoch": 5.042382090849816,
+      "grad_norm": 0.572697103023529,
+      "learning_rate": 2.5e-05,
+      "loss": 0.0926,
+      "step": 4350
+    },
+    {
+      "epoch": 5.043541259146562,
+      "grad_norm": 0.4093453288078308,
+      "learning_rate": 2.499425287356322e-05,
+      "loss": 0.0913,
+      "step": 4351
+    },
+    {
+      "epoch": 5.04470042744331,
+      "grad_norm": 0.3991166949272156,
+      "learning_rate": 2.4988505747126438e-05,
+      "loss": 0.0965,
+      "step": 4352
+    },
+    {
+      "epoch": 5.0458595957400565,
+      "grad_norm": 0.6385246515274048,
+      "learning_rate": 2.4982758620689656e-05,
+      "loss": 0.0879,
+      "step": 4353
+    },
+    {
+      "epoch": 5.047018764036803,
+      "grad_norm": 0.4218897223472595,
+      "learning_rate": 2.4977011494252874e-05,
+      "loss": 0.0874,
+      "step": 4354
+    },
+    {
+      "epoch": 5.048177932333551,
+      "grad_norm": 0.3741064667701721,
+      "learning_rate": 2.4971264367816093e-05,
+      "loss": 0.0987,
+      "step": 4355
+    },
+    {
+      "epoch": 5.049337100630297,
+      "grad_norm": 0.4053811728954315,
+      "learning_rate": 2.496551724137931e-05,
+      "loss": 0.0972,
+      "step": 4356
+    },
+    {
+      "epoch": 5.050496268927045,
+      "grad_norm": 0.521132230758667,
+      "learning_rate": 2.495977011494253e-05,
+      "loss": 0.0903,
+      "step": 4357
+    },
+    {
+      "epoch": 5.051655437223792,
+      "grad_norm": 0.3791821002960205,
+      "learning_rate": 2.4954022988505748e-05,
+      "loss": 0.0827,
+      "step": 4358
+    },
+    {
+      "epoch": 5.052814605520539,
+      "grad_norm": 0.5845307111740112,
+      "learning_rate": 2.494827586206897e-05,
+      "loss": 0.1025,
+      "step": 4359
+    },
+    {
+      "epoch": 5.053973773817286,
+      "grad_norm": 0.5024420022964478,
+      "learning_rate": 2.4942528735632184e-05,
+      "loss": 0.0903,
+      "step": 4360
+    },
+    {
+      "epoch": 5.055132942114033,
+      "grad_norm": 0.3732399046421051,
+      "learning_rate": 2.4936781609195402e-05,
+      "loss": 0.0925,
+      "step": 4361
+    },
+    {
+      "epoch": 5.05629211041078,
+      "grad_norm": 0.3285994827747345,
+      "learning_rate": 2.493103448275862e-05,
+      "loss": 0.0922,
+      "step": 4362
+    },
+    {
+      "epoch": 5.057451278707528,
+      "grad_norm": 0.4758022129535675,
+      "learning_rate": 2.4925287356321842e-05,
+      "loss": 0.0955,
+      "step": 4363
+    },
+    {
+      "epoch": 5.058610447004274,
+      "grad_norm": 0.558384358882904,
+      "learning_rate": 2.4919540229885057e-05,
+      "loss": 0.0947,
+      "step": 4364
+    },
+    {
+      "epoch": 5.059769615301022,
+      "grad_norm": 0.7286369800567627,
+      "learning_rate": 2.4913793103448276e-05,
+      "loss": 0.0965,
+      "step": 4365
+    },
+    {
+      "epoch": 5.0609287835977685,
+      "grad_norm": 0.35651078820228577,
+      "learning_rate": 2.4908045977011497e-05,
+      "loss": 0.0874,
+      "step": 4366
+    },
+    {
+      "epoch": 5.062087951894516,
+      "grad_norm": 0.5121673345565796,
+      "learning_rate": 2.4902298850574716e-05,
+      "loss": 0.0997,
+      "step": 4367
+    },
+    {
+      "epoch": 5.063247120191263,
+      "grad_norm": 0.506806492805481,
+      "learning_rate": 2.489655172413793e-05,
+      "loss": 0.0869,
+      "step": 4368
+    },
+    {
+      "epoch": 5.0644062884880094,
+      "grad_norm": 1.1122589111328125,
+      "learning_rate": 2.4890804597701152e-05,
+      "loss": 0.1026,
+      "step": 4369
+    },
+    {
+      "epoch": 5.065565456784757,
+      "grad_norm": 0.42941662669181824,
+      "learning_rate": 2.488505747126437e-05,
+      "loss": 0.0868,
+      "step": 4370
+    },
+    {
+      "epoch": 5.066724625081504,
+      "grad_norm": 0.4219612181186676,
+      "learning_rate": 2.4879310344827585e-05,
+      "loss": 0.0924,
+      "step": 4371
+    },
+    {
+      "epoch": 5.067883793378251,
+      "grad_norm": 0.3930709660053253,
+      "learning_rate": 2.4873563218390804e-05,
+      "loss": 0.0868,
+      "step": 4372
+    },
+    {
+      "epoch": 5.069042961674998,
+      "grad_norm": 0.3940015137195587,
+      "learning_rate": 2.4867816091954025e-05,
+      "loss": 0.0815,
+      "step": 4373
+    },
+    {
+      "epoch": 5.0702021299717455,
+      "grad_norm": 0.496271014213562,
+      "learning_rate": 2.4862068965517244e-05,
+      "loss": 0.0987,
+      "step": 4374
+    },
+    {
+      "epoch": 5.071361298268492,
+      "grad_norm": 0.3629682958126068,
+      "learning_rate": 2.485632183908046e-05,
+      "loss": 0.0875,
+      "step": 4375
+    },
+    {
+      "epoch": 5.07252046656524,
+      "grad_norm": 0.5022166967391968,
+      "learning_rate": 2.485057471264368e-05,
+      "loss": 0.0989,
+      "step": 4376
+    },
+    {
+      "epoch": 5.073679634861986,
+      "grad_norm": 0.7410409450531006,
+      "learning_rate": 2.48448275862069e-05,
+      "loss": 0.0991,
+      "step": 4377
+    },
+    {
+      "epoch": 5.074838803158734,
+      "grad_norm": 0.30620214343070984,
+      "learning_rate": 2.4839080459770117e-05,
+      "loss": 0.0834,
+      "step": 4378
+    },
+    {
+      "epoch": 5.075997971455481,
+      "grad_norm": 0.5494264960289001,
+      "learning_rate": 2.4833333333333335e-05,
+      "loss": 0.0888,
+      "step": 4379
+    },
+    {
+      "epoch": 5.077157139752228,
+      "grad_norm": 0.45978859066963196,
+      "learning_rate": 2.4827586206896553e-05,
+      "loss": 0.1035,
+      "step": 4380
+    },
+    {
+      "epoch": 5.078316308048975,
+      "grad_norm": 0.4293060004711151,
+      "learning_rate": 2.482183908045977e-05,
+      "loss": 0.0898,
+      "step": 4381
+    },
+    {
+      "epoch": 5.079475476345722,
+      "grad_norm": 0.3662157654762268,
+      "learning_rate": 2.481609195402299e-05,
+      "loss": 0.1001,
+      "step": 4382
+    },
+    {
+      "epoch": 5.080634644642469,
+      "grad_norm": 0.3850663900375366,
+      "learning_rate": 2.4810344827586208e-05,
+      "loss": 0.0901,
+      "step": 4383
+    },
+    {
+      "epoch": 5.081793812939216,
+      "grad_norm": 0.32764530181884766,
+      "learning_rate": 2.4804597701149426e-05,
+      "loss": 0.0922,
+      "step": 4384
+    },
+    {
+      "epoch": 5.082952981235963,
+      "grad_norm": 0.8065647482872009,
+      "learning_rate": 2.4798850574712645e-05,
+      "loss": 0.0949,
+      "step": 4385
+    },
+    {
+      "epoch": 5.08411214953271,
+      "grad_norm": 0.4645923376083374,
+      "learning_rate": 2.4793103448275863e-05,
+      "loss": 0.0899,
+      "step": 4386
+    },
+    {
+      "epoch": 5.0852713178294575,
+      "grad_norm": 0.3961159288883209,
+      "learning_rate": 2.478735632183908e-05,
+      "loss": 0.0918,
+      "step": 4387
+    },
+    {
+      "epoch": 5.086430486126204,
+      "grad_norm": 0.3519403636455536,
+      "learning_rate": 2.47816091954023e-05,
+      "loss": 0.0942,
+      "step": 4388
+    },
+    {
+      "epoch": 5.087589654422952,
+      "grad_norm": 0.45820310711860657,
+      "learning_rate": 2.4775862068965518e-05,
+      "loss": 0.0977,
+      "step": 4389
+    },
+    {
+      "epoch": 5.088748822719698,
+      "grad_norm": 0.810590386390686,
+      "learning_rate": 2.4770114942528736e-05,
+      "loss": 0.0983,
+      "step": 4390
+    },
+    {
+      "epoch": 5.089907991016446,
+      "grad_norm": 0.4400874972343445,
+      "learning_rate": 2.4764367816091954e-05,
+      "loss": 0.0962,
+      "step": 4391
+    },
+    {
+      "epoch": 5.091067159313193,
+      "grad_norm": 0.3362007737159729,
+      "learning_rate": 2.4758620689655173e-05,
+      "loss": 0.0957,
+      "step": 4392
+    },
+    {
+      "epoch": 5.09222632760994,
+      "grad_norm": 0.5532455444335938,
+      "learning_rate": 2.4752873563218394e-05,
+      "loss": 0.1131,
+      "step": 4393
+    },
+    {
+      "epoch": 5.093385495906687,
+      "grad_norm": 0.3802591860294342,
+      "learning_rate": 2.474712643678161e-05,
+      "loss": 0.091,
+      "step": 4394
+    },
+    {
+      "epoch": 5.094544664203434,
+      "grad_norm": 0.45833706855773926,
+      "learning_rate": 2.4741379310344828e-05,
+      "loss": 0.0942,
+      "step": 4395
+    },
+    {
+      "epoch": 5.095703832500181,
+      "grad_norm": 0.3089923858642578,
+      "learning_rate": 2.4735632183908046e-05,
+      "loss": 0.0918,
+      "step": 4396
+    },
+    {
+      "epoch": 5.096863000796928,
+      "grad_norm": 0.43163448572158813,
+      "learning_rate": 2.4729885057471268e-05,
+      "loss": 0.0963,
+      "step": 4397
+    },
+    {
+      "epoch": 5.098022169093675,
+      "grad_norm": 0.388334721326828,
+      "learning_rate": 2.4724137931034483e-05,
+      "loss": 0.0901,
+      "step": 4398
+    },
+    {
+      "epoch": 5.099181337390422,
+      "grad_norm": 0.5277796387672424,
+      "learning_rate": 2.47183908045977e-05,
+      "loss": 0.0934,
+      "step": 4399
+    },
+    {
+      "epoch": 5.10034050568717,
+      "grad_norm": 0.42713913321495056,
+      "learning_rate": 2.4712643678160922e-05,
+      "loss": 0.1081,
+      "step": 4400
+    },
+    {
+      "epoch": 5.101499673983916,
+      "grad_norm": 0.4533218741416931,
+      "learning_rate": 2.470689655172414e-05,
+      "loss": 0.088,
+      "step": 4401
+    },
+    {
+      "epoch": 5.102658842280664,
+      "grad_norm": 0.4103233516216278,
+      "learning_rate": 2.4701149425287356e-05,
+      "loss": 0.1027,
+      "step": 4402
+    },
+    {
+      "epoch": 5.1038180105774105,
+      "grad_norm": 0.5257367491722107,
+      "learning_rate": 2.4695402298850577e-05,
+      "loss": 0.0954,
+      "step": 4403
+    },
+    {
+      "epoch": 5.104977178874158,
+      "grad_norm": 0.5127593874931335,
+      "learning_rate": 2.4689655172413796e-05,
+      "loss": 0.0903,
+      "step": 4404
+    },
+    {
+      "epoch": 5.106136347170905,
+      "grad_norm": 0.4334225058555603,
+      "learning_rate": 2.4683908045977014e-05,
+      "loss": 0.0976,
+      "step": 4405
+    },
+    {
+      "epoch": 5.107295515467652,
+      "grad_norm": 0.5956823229789734,
+      "learning_rate": 2.467816091954023e-05,
+      "loss": 0.1057,
+      "step": 4406
+    },
+    {
+      "epoch": 5.108454683764399,
+      "grad_norm": 0.448223352432251,
+      "learning_rate": 2.467241379310345e-05,
+      "loss": 0.0955,
+      "step": 4407
+    },
+    {
+      "epoch": 5.1096138520611465,
+      "grad_norm": 0.5333186984062195,
+      "learning_rate": 2.466666666666667e-05,
+      "loss": 0.0998,
+      "step": 4408
+    },
+    {
+      "epoch": 5.110773020357893,
+      "grad_norm": 0.5879203081130981,
+      "learning_rate": 2.4660919540229887e-05,
+      "loss": 0.0954,
+      "step": 4409
+    },
+    {
+      "epoch": 5.111932188654641,
+      "grad_norm": 0.4079529047012329,
+      "learning_rate": 2.4655172413793105e-05,
+      "loss": 0.0925,
+      "step": 4410
+    },
+    {
+      "epoch": 5.113091356951387,
+      "grad_norm": 0.3445991277694702,
+      "learning_rate": 2.4649425287356324e-05,
+      "loss": 0.0894,
+      "step": 4411
+    },
+    {
+      "epoch": 5.114250525248134,
+      "grad_norm": 0.371430903673172,
+      "learning_rate": 2.4643678160919542e-05,
+      "loss": 0.0904,
+      "step": 4412
+    },
+    {
+      "epoch": 5.115409693544882,
+      "grad_norm": 0.4925328195095062,
+      "learning_rate": 2.4637931034482757e-05,
+      "loss": 0.0954,
+      "step": 4413
+    },
+    {
+      "epoch": 5.116568861841628,
+      "grad_norm": 0.4289555847644806,
+      "learning_rate": 2.463218390804598e-05,
+      "loss": 0.0909,
+      "step": 4414
+    },
+    {
+      "epoch": 5.117728030138376,
+      "grad_norm": 0.6553305983543396,
+      "learning_rate": 2.4626436781609197e-05,
+      "loss": 0.1009,
+      "step": 4415
+    },
+    {
+      "epoch": 5.1188871984351225,
+      "grad_norm": 0.4029340445995331,
+      "learning_rate": 2.4620689655172415e-05,
+      "loss": 0.0964,
+      "step": 4416
+    },
+    {
+      "epoch": 5.12004636673187,
+      "grad_norm": 0.48213991522789,
+      "learning_rate": 2.4614942528735633e-05,
+      "loss": 0.099,
+      "step": 4417
+    },
+    {
+      "epoch": 5.121205535028617,
+      "grad_norm": 0.5332958698272705,
+      "learning_rate": 2.4609195402298852e-05,
+      "loss": 0.0959,
+      "step": 4418
+    },
+    {
+      "epoch": 5.122364703325364,
+      "grad_norm": 0.48427054286003113,
+      "learning_rate": 2.460344827586207e-05,
+      "loss": 0.0936,
+      "step": 4419
+    },
+    {
+      "epoch": 5.123523871622111,
+      "grad_norm": 0.32984548807144165,
+      "learning_rate": 2.4597701149425288e-05,
+      "loss": 0.0927,
+      "step": 4420
+    },
+    {
+      "epoch": 5.1246830399188585,
+      "grad_norm": 0.6007506847381592,
+      "learning_rate": 2.4591954022988507e-05,
+      "loss": 0.0944,
+      "step": 4421
+    },
+    {
+      "epoch": 5.125842208215605,
+      "grad_norm": 0.4519351124763489,
+      "learning_rate": 2.4586206896551725e-05,
+      "loss": 0.0904,
+      "step": 4422
+    },
+    {
+      "epoch": 5.127001376512353,
+      "grad_norm": 0.44407084584236145,
+      "learning_rate": 2.4580459770114943e-05,
+      "loss": 0.0938,
+      "step": 4423
+    },
+    {
+      "epoch": 5.128160544809099,
+      "grad_norm": 0.4065387547016144,
+      "learning_rate": 2.4574712643678165e-05,
+      "loss": 0.0949,
+      "step": 4424
+    },
+    {
+      "epoch": 5.129319713105847,
+      "grad_norm": 0.42463329434394836,
+      "learning_rate": 2.456896551724138e-05,
+      "loss": 0.0859,
+      "step": 4425
+    },
+    {
+      "epoch": 5.130478881402594,
+      "grad_norm": 0.5019070506095886,
+      "learning_rate": 2.4563218390804598e-05,
+      "loss": 0.1039,
+      "step": 4426
+    },
+    {
+      "epoch": 5.13163804969934,
+      "grad_norm": 0.463072806596756,
+      "learning_rate": 2.4557471264367816e-05,
+      "loss": 0.1004,
+      "step": 4427
+    },
+    {
+      "epoch": 5.132797217996088,
+      "grad_norm": 0.32865774631500244,
+      "learning_rate": 2.4551724137931038e-05,
+      "loss": 0.0915,
+      "step": 4428
+    },
+    {
+      "epoch": 5.133956386292835,
+      "grad_norm": 0.26548025012016296,
+      "learning_rate": 2.4545977011494253e-05,
+      "loss": 0.0837,
+      "step": 4429
+    },
+    {
+      "epoch": 5.135115554589582,
+      "grad_norm": 0.3461511731147766,
+      "learning_rate": 2.454022988505747e-05,
+      "loss": 0.0882,
+      "step": 4430
+    },
+    {
+      "epoch": 5.136274722886329,
+      "grad_norm": 0.816858172416687,
+      "learning_rate": 2.4534482758620693e-05,
+      "loss": 0.0924,
+      "step": 4431
+    },
+    {
+      "epoch": 5.137433891183076,
+      "grad_norm": 0.3309417963027954,
+      "learning_rate": 2.4528735632183908e-05,
+      "loss": 0.0926,
+      "step": 4432
+    },
+    {
+      "epoch": 5.138593059479823,
+      "grad_norm": 0.6483179926872253,
+      "learning_rate": 2.4522988505747126e-05,
+      "loss": 0.0908,
+      "step": 4433
+    },
+    {
+      "epoch": 5.139752227776571,
+      "grad_norm": 0.3910187780857086,
+      "learning_rate": 2.4517241379310348e-05,
+      "loss": 0.0956,
+      "step": 4434
+    },
+    {
+      "epoch": 5.140911396073317,
+      "grad_norm": 0.4869095981121063,
+      "learning_rate": 2.4511494252873566e-05,
+      "loss": 0.0881,
+      "step": 4435
+    },
+    {
+      "epoch": 5.142070564370065,
+      "grad_norm": 0.6011160612106323,
+      "learning_rate": 2.450574712643678e-05,
+      "loss": 0.1055,
+      "step": 4436
+    },
+    {
+      "epoch": 5.1432297326668115,
+      "grad_norm": 0.38875725865364075,
+      "learning_rate": 2.45e-05,
+      "loss": 0.0919,
+      "step": 4437
+    },
+    {
+      "epoch": 5.144388900963559,
+      "grad_norm": 0.28228649497032166,
+      "learning_rate": 2.449425287356322e-05,
+      "loss": 0.0942,
+      "step": 4438
+    },
+    {
+      "epoch": 5.145548069260306,
+      "grad_norm": 0.3258858919143677,
+      "learning_rate": 2.448850574712644e-05,
+      "loss": 0.0845,
+      "step": 4439
+    },
+    {
+      "epoch": 5.146707237557052,
+      "grad_norm": 0.3680334985256195,
+      "learning_rate": 2.4482758620689654e-05,
+      "loss": 0.0922,
+      "step": 4440
+    },
+    {
+      "epoch": 5.1478664058538,
+      "grad_norm": 0.38462620973587036,
+      "learning_rate": 2.4477011494252876e-05,
+      "loss": 0.1046,
+      "step": 4441
+    },
+    {
+      "epoch": 5.149025574150547,
+      "grad_norm": 0.48732998967170715,
+      "learning_rate": 2.4471264367816094e-05,
+      "loss": 0.1,
+      "step": 4442
+    },
+    {
+      "epoch": 5.150184742447294,
+      "grad_norm": 0.45189419388771057,
+      "learning_rate": 2.4465517241379312e-05,
+      "loss": 0.0962,
+      "step": 4443
+    },
+    {
+      "epoch": 5.151343910744041,
+      "grad_norm": 0.4464831054210663,
+      "learning_rate": 2.445977011494253e-05,
+      "loss": 0.0986,
+      "step": 4444
+    },
+    {
+      "epoch": 5.152503079040788,
+      "grad_norm": 0.3762010931968689,
+      "learning_rate": 2.445402298850575e-05,
+      "loss": 0.1022,
+      "step": 4445
+    },
+    {
+      "epoch": 5.153662247337535,
+      "grad_norm": 0.3228253424167633,
+      "learning_rate": 2.4448275862068967e-05,
+      "loss": 0.0903,
+      "step": 4446
+    },
+    {
+      "epoch": 5.154821415634283,
+      "grad_norm": 0.40066200494766235,
+      "learning_rate": 2.4442528735632185e-05,
+      "loss": 0.103,
+      "step": 4447
+    },
+    {
+      "epoch": 5.155980583931029,
+      "grad_norm": 0.42761850357055664,
+      "learning_rate": 2.4436781609195404e-05,
+      "loss": 0.0986,
+      "step": 4448
+    },
+    {
+      "epoch": 5.157139752227777,
+      "grad_norm": 0.3788778483867645,
+      "learning_rate": 2.4431034482758622e-05,
+      "loss": 0.0954,
+      "step": 4449
+    },
+    {
+      "epoch": 5.1582989205245235,
+      "grad_norm": 0.35833922028541565,
+      "learning_rate": 2.442528735632184e-05,
+      "loss": 0.0897,
+      "step": 4450
+    },
+    {
+      "epoch": 5.159458088821271,
+      "grad_norm": 0.522954523563385,
+      "learning_rate": 2.441954022988506e-05,
+      "loss": 0.0864,
+      "step": 4451
+    },
+    {
+      "epoch": 5.160617257118018,
+      "grad_norm": 0.43595683574676514,
+      "learning_rate": 2.4413793103448277e-05,
+      "loss": 0.0897,
+      "step": 4452
+    },
+    {
+      "epoch": 5.161776425414765,
+      "grad_norm": 0.4088694155216217,
+      "learning_rate": 2.4408045977011495e-05,
+      "loss": 0.1002,
+      "step": 4453
+    },
+    {
+      "epoch": 5.162935593711512,
+      "grad_norm": 0.4870903491973877,
+      "learning_rate": 2.4402298850574714e-05,
+      "loss": 0.0898,
+      "step": 4454
+    },
+    {
+      "epoch": 5.164094762008259,
+      "grad_norm": 0.3365190029144287,
+      "learning_rate": 2.4396551724137932e-05,
+      "loss": 0.0888,
+      "step": 4455
+    },
+    {
+      "epoch": 5.165253930305006,
+      "grad_norm": 0.5863637328147888,
+      "learning_rate": 2.439080459770115e-05,
+      "loss": 0.1003,
+      "step": 4456
+    },
+    {
+      "epoch": 5.166413098601753,
+      "grad_norm": 0.3489425778388977,
+      "learning_rate": 2.438505747126437e-05,
+      "loss": 0.0891,
+      "step": 4457
+    },
+    {
+      "epoch": 5.1675722668985005,
+      "grad_norm": 0.42141303420066833,
+      "learning_rate": 2.4379310344827587e-05,
+      "loss": 0.0868,
+      "step": 4458
+    },
+    {
+      "epoch": 5.168731435195247,
+      "grad_norm": 0.5264760255813599,
+      "learning_rate": 2.4373563218390805e-05,
+      "loss": 0.1003,
+      "step": 4459
+    },
+    {
+      "epoch": 5.169890603491995,
+      "grad_norm": 0.4587344825267792,
+      "learning_rate": 2.4367816091954023e-05,
+      "loss": 0.0844,
+      "step": 4460
+    },
+    {
+      "epoch": 5.171049771788741,
+      "grad_norm": 0.5122568607330322,
+      "learning_rate": 2.436206896551724e-05,
+      "loss": 0.0842,
+      "step": 4461
+    },
+    {
+      "epoch": 5.172208940085489,
+      "grad_norm": 0.38229185342788696,
+      "learning_rate": 2.4356321839080463e-05,
+      "loss": 0.0903,
+      "step": 4462
+    },
+    {
+      "epoch": 5.173368108382236,
+      "grad_norm": 0.5392225980758667,
+      "learning_rate": 2.4350574712643678e-05,
+      "loss": 0.0971,
+      "step": 4463
+    },
+    {
+      "epoch": 5.174527276678983,
+      "grad_norm": 0.4981425106525421,
+      "learning_rate": 2.4344827586206896e-05,
+      "loss": 0.0872,
+      "step": 4464
+    },
+    {
+      "epoch": 5.17568644497573,
+      "grad_norm": 0.4396236538887024,
+      "learning_rate": 2.4339080459770118e-05,
+      "loss": 0.0926,
+      "step": 4465
+    },
+    {
+      "epoch": 5.176845613272477,
+      "grad_norm": 0.42154166102409363,
+      "learning_rate": 2.4333333333333336e-05,
+      "loss": 0.0882,
+      "step": 4466
+    },
+    {
+      "epoch": 5.178004781569224,
+      "grad_norm": 0.4124399721622467,
+      "learning_rate": 2.432758620689655e-05,
+      "loss": 0.092,
+      "step": 4467
+    },
+    {
+      "epoch": 5.179163949865972,
+      "grad_norm": 0.378948450088501,
+      "learning_rate": 2.432183908045977e-05,
+      "loss": 0.0951,
+      "step": 4468
+    },
+    {
+      "epoch": 5.180323118162718,
+      "grad_norm": 0.3714349567890167,
+      "learning_rate": 2.431609195402299e-05,
+      "loss": 0.0981,
+      "step": 4469
+    },
+    {
+      "epoch": 5.181482286459465,
+      "grad_norm": 0.3504725396633148,
+      "learning_rate": 2.4310344827586206e-05,
+      "loss": 0.0973,
+      "step": 4470
+    },
+    {
+      "epoch": 5.1826414547562125,
+      "grad_norm": 0.4103260636329651,
+      "learning_rate": 2.4304597701149424e-05,
+      "loss": 0.0957,
+      "step": 4471
+    },
+    {
+      "epoch": 5.183800623052959,
+      "grad_norm": 0.5066686868667603,
+      "learning_rate": 2.4298850574712646e-05,
+      "loss": 0.0986,
+      "step": 4472
+    },
+    {
+      "epoch": 5.184959791349707,
+      "grad_norm": 0.6200160980224609,
+      "learning_rate": 2.4293103448275864e-05,
+      "loss": 0.1011,
+      "step": 4473
+    },
+    {
+      "epoch": 5.186118959646453,
+      "grad_norm": 0.4090159833431244,
+      "learning_rate": 2.428735632183908e-05,
+      "loss": 0.0977,
+      "step": 4474
+    },
+    {
+      "epoch": 5.187278127943201,
+      "grad_norm": 0.45171108841896057,
+      "learning_rate": 2.42816091954023e-05,
+      "loss": 0.0855,
+      "step": 4475
+    },
+    {
+      "epoch": 5.188437296239948,
+      "grad_norm": 0.3407220244407654,
+      "learning_rate": 2.427586206896552e-05,
+      "loss": 0.0917,
+      "step": 4476
+    },
+    {
+      "epoch": 5.189596464536695,
+      "grad_norm": 0.3243277370929718,
+      "learning_rate": 2.4270114942528738e-05,
+      "loss": 0.0855,
+      "step": 4477
+    },
+    {
+      "epoch": 5.190755632833442,
+      "grad_norm": 0.5616956949234009,
+      "learning_rate": 2.4264367816091952e-05,
+      "loss": 0.0855,
+      "step": 4478
+    },
+    {
+      "epoch": 5.191914801130189,
+      "grad_norm": 0.369253009557724,
+      "learning_rate": 2.4258620689655174e-05,
+      "loss": 0.0832,
+      "step": 4479
+    },
+    {
+      "epoch": 5.193073969426936,
+      "grad_norm": 0.3325676918029785,
+      "learning_rate": 2.4252873563218392e-05,
+      "loss": 0.0883,
+      "step": 4480
+    },
+    {
+      "epoch": 5.194233137723684,
+      "grad_norm": 0.4389870762825012,
+      "learning_rate": 2.424712643678161e-05,
+      "loss": 0.1078,
+      "step": 4481
+    },
+    {
+      "epoch": 5.19539230602043,
+      "grad_norm": 0.3416706919670105,
+      "learning_rate": 2.424137931034483e-05,
+      "loss": 0.0885,
+      "step": 4482
+    },
+    {
+      "epoch": 5.196551474317177,
+      "grad_norm": 0.3958418369293213,
+      "learning_rate": 2.4235632183908047e-05,
+      "loss": 0.0959,
+      "step": 4483
+    },
+    {
+      "epoch": 5.197710642613925,
+      "grad_norm": 0.41249221563339233,
+      "learning_rate": 2.4229885057471266e-05,
+      "loss": 0.0972,
+      "step": 4484
+    },
+    {
+      "epoch": 5.198869810910671,
+      "grad_norm": 0.3895527720451355,
+      "learning_rate": 2.4224137931034484e-05,
+      "loss": 0.0847,
+      "step": 4485
+    },
+    {
+      "epoch": 5.200028979207419,
+      "grad_norm": 0.3692401349544525,
+      "learning_rate": 2.4218390804597702e-05,
+      "loss": 0.095,
+      "step": 4486
+    },
+    {
+      "epoch": 5.2011881475041655,
+      "grad_norm": 0.5421669483184814,
+      "learning_rate": 2.421264367816092e-05,
+      "loss": 0.0923,
+      "step": 4487
+    },
+    {
+      "epoch": 5.202347315800913,
+      "grad_norm": 0.40526431798934937,
+      "learning_rate": 2.420689655172414e-05,
+      "loss": 0.0903,
+      "step": 4488
+    },
+    {
+      "epoch": 5.20350648409766,
+      "grad_norm": 0.35065868496894836,
+      "learning_rate": 2.420114942528736e-05,
+      "loss": 0.0945,
+      "step": 4489
+    },
+    {
+      "epoch": 5.204665652394407,
+      "grad_norm": 0.34505459666252136,
+      "learning_rate": 2.4195402298850575e-05,
+      "loss": 0.0952,
+      "step": 4490
+    },
+    {
+      "epoch": 5.205824820691154,
+      "grad_norm": 0.43336015939712524,
+      "learning_rate": 2.4189655172413794e-05,
+      "loss": 0.0989,
+      "step": 4491
+    },
+    {
+      "epoch": 5.2069839889879015,
+      "grad_norm": 0.3147088885307312,
+      "learning_rate": 2.4183908045977012e-05,
+      "loss": 0.089,
+      "step": 4492
+    },
+    {
+      "epoch": 5.208143157284648,
+      "grad_norm": 0.567955493927002,
+      "learning_rate": 2.417816091954023e-05,
+      "loss": 0.088,
+      "step": 4493
+    },
+    {
+      "epoch": 5.209302325581396,
+      "grad_norm": 0.4204098582267761,
+      "learning_rate": 2.417241379310345e-05,
+      "loss": 0.0927,
+      "step": 4494
+    },
+    {
+      "epoch": 5.210461493878142,
+      "grad_norm": 0.4277987778186798,
+      "learning_rate": 2.4166666666666667e-05,
+      "loss": 0.0955,
+      "step": 4495
+    },
+    {
+      "epoch": 5.211620662174889,
+      "grad_norm": 0.44593873620033264,
+      "learning_rate": 2.416091954022989e-05,
+      "loss": 0.1008,
+      "step": 4496
+    },
+    {
+      "epoch": 5.212779830471637,
+      "grad_norm": 0.39049631357192993,
+      "learning_rate": 2.4155172413793103e-05,
+      "loss": 0.0957,
+      "step": 4497
+    },
+    {
+      "epoch": 5.213938998768383,
+      "grad_norm": 0.3604249954223633,
+      "learning_rate": 2.414942528735632e-05,
+      "loss": 0.1004,
+      "step": 4498
+    },
+    {
+      "epoch": 5.215098167065131,
+      "grad_norm": 0.3670724630355835,
+      "learning_rate": 2.4143678160919543e-05,
+      "loss": 0.0904,
+      "step": 4499
+    },
+    {
+      "epoch": 5.2162573353618775,
+      "grad_norm": 0.34518417716026306,
+      "learning_rate": 2.413793103448276e-05,
+      "loss": 0.0893,
+      "step": 4500
+    },
+    {
+      "epoch": 5.217416503658625,
+      "grad_norm": 0.39651983976364136,
+      "learning_rate": 2.4132183908045977e-05,
+      "loss": 0.0948,
+      "step": 4501
+    },
+    {
+      "epoch": 5.218575671955372,
+      "grad_norm": 0.4444870948791504,
+      "learning_rate": 2.4126436781609195e-05,
+      "loss": 0.0863,
+      "step": 4502
+    },
+    {
+      "epoch": 5.219734840252119,
+      "grad_norm": 0.5257106423377991,
+      "learning_rate": 2.4120689655172417e-05,
+      "loss": 0.0919,
+      "step": 4503
+    },
+    {
+      "epoch": 5.220894008548866,
+      "grad_norm": 0.45813608169555664,
+      "learning_rate": 2.4114942528735635e-05,
+      "loss": 0.1027,
+      "step": 4504
+    },
+    {
+      "epoch": 5.2220531768456135,
+      "grad_norm": 0.4071040749549866,
+      "learning_rate": 2.410919540229885e-05,
+      "loss": 0.0958,
+      "step": 4505
+    },
+    {
+      "epoch": 5.22321234514236,
+      "grad_norm": 0.4606774151325226,
+      "learning_rate": 2.410344827586207e-05,
+      "loss": 0.0926,
+      "step": 4506
+    },
+    {
+      "epoch": 5.224371513439108,
+      "grad_norm": 0.47253385186195374,
+      "learning_rate": 2.409770114942529e-05,
+      "loss": 0.1084,
+      "step": 4507
+    },
+    {
+      "epoch": 5.225530681735854,
+      "grad_norm": 0.4298096299171448,
+      "learning_rate": 2.4091954022988508e-05,
+      "loss": 0.0866,
+      "step": 4508
+    },
+    {
+      "epoch": 5.226689850032602,
+      "grad_norm": 0.843309760093689,
+      "learning_rate": 2.4086206896551726e-05,
+      "loss": 0.104,
+      "step": 4509
+    },
+    {
+      "epoch": 5.227849018329349,
+      "grad_norm": 0.5684033632278442,
+      "learning_rate": 2.4080459770114945e-05,
+      "loss": 0.0958,
+      "step": 4510
+    },
+    {
+      "epoch": 5.229008186626096,
+      "grad_norm": 0.6145815849304199,
+      "learning_rate": 2.4074712643678163e-05,
+      "loss": 0.0957,
+      "step": 4511
+    },
+    {
+      "epoch": 5.230167354922843,
+      "grad_norm": 0.5087504386901855,
+      "learning_rate": 2.4068965517241378e-05,
+      "loss": 0.0846,
+      "step": 4512
+    },
+    {
+      "epoch": 5.23132652321959,
+      "grad_norm": 0.49333977699279785,
+      "learning_rate": 2.40632183908046e-05,
+      "loss": 0.088,
+      "step": 4513
+    },
+    {
+      "epoch": 5.232485691516337,
+      "grad_norm": 0.29817384481430054,
+      "learning_rate": 2.4057471264367818e-05,
+      "loss": 0.086,
+      "step": 4514
+    },
+    {
+      "epoch": 5.233644859813084,
+      "grad_norm": 0.3775709867477417,
+      "learning_rate": 2.4051724137931036e-05,
+      "loss": 0.0947,
+      "step": 4515
+    },
+    {
+      "epoch": 5.234804028109831,
+      "grad_norm": 0.31703516840934753,
+      "learning_rate": 2.4045977011494254e-05,
+      "loss": 0.0883,
+      "step": 4516
+    },
+    {
+      "epoch": 5.235963196406578,
+      "grad_norm": 0.42791876196861267,
+      "learning_rate": 2.4040229885057473e-05,
+      "loss": 0.0933,
+      "step": 4517
+    },
+    {
+      "epoch": 5.237122364703326,
+      "grad_norm": 0.5185449123382568,
+      "learning_rate": 2.403448275862069e-05,
+      "loss": 0.1011,
+      "step": 4518
+    },
+    {
+      "epoch": 5.238281533000072,
+      "grad_norm": 0.4122971296310425,
+      "learning_rate": 2.402873563218391e-05,
+      "loss": 0.1013,
+      "step": 4519
+    },
+    {
+      "epoch": 5.23944070129682,
+      "grad_norm": 0.3461478054523468,
+      "learning_rate": 2.4022988505747127e-05,
+      "loss": 0.0909,
+      "step": 4520
+    },
+    {
+      "epoch": 5.2405998695935665,
+      "grad_norm": 0.38688674569129944,
+      "learning_rate": 2.4017241379310346e-05,
+      "loss": 0.0908,
+      "step": 4521
+    },
+    {
+      "epoch": 5.241759037890314,
+      "grad_norm": 0.3087342381477356,
+      "learning_rate": 2.4011494252873564e-05,
+      "loss": 0.0818,
+      "step": 4522
+    },
+    {
+      "epoch": 5.242918206187061,
+      "grad_norm": 0.46722832322120667,
+      "learning_rate": 2.4005747126436782e-05,
+      "loss": 0.09,
+      "step": 4523
+    },
+    {
+      "epoch": 5.244077374483808,
+      "grad_norm": 0.4601225256919861,
+      "learning_rate": 2.4e-05,
+      "loss": 0.1053,
+      "step": 4524
+    },
+    {
+      "epoch": 5.245236542780555,
+      "grad_norm": 0.3325468599796295,
+      "learning_rate": 2.399425287356322e-05,
+      "loss": 0.0875,
+      "step": 4525
+    },
+    {
+      "epoch": 5.246395711077302,
+      "grad_norm": 0.4317132830619812,
+      "learning_rate": 2.3988505747126437e-05,
+      "loss": 0.0961,
+      "step": 4526
+    },
+    {
+      "epoch": 5.247554879374049,
+      "grad_norm": 0.3786173462867737,
+      "learning_rate": 2.398275862068966e-05,
+      "loss": 0.088,
+      "step": 4527
+    },
+    {
+      "epoch": 5.248714047670796,
+      "grad_norm": 0.34360384941101074,
+      "learning_rate": 2.3977011494252874e-05,
+      "loss": 0.0882,
+      "step": 4528
+    },
+    {
+      "epoch": 5.249873215967543,
+      "grad_norm": 0.3245190382003784,
+      "learning_rate": 2.3971264367816092e-05,
+      "loss": 0.0888,
+      "step": 4529
+    },
+    {
+      "epoch": 5.25103238426429,
+      "grad_norm": 0.3508089780807495,
+      "learning_rate": 2.3965517241379314e-05,
+      "loss": 0.0838,
+      "step": 4530
+    },
+    {
+      "epoch": 5.252191552561038,
+      "grad_norm": 0.7918804287910461,
+      "learning_rate": 2.395977011494253e-05,
+      "loss": 0.1023,
+      "step": 4531
+    },
+    {
+      "epoch": 5.253350720857784,
+      "grad_norm": 0.5230761766433716,
+      "learning_rate": 2.3954022988505747e-05,
+      "loss": 0.0951,
+      "step": 4532
+    },
+    {
+      "epoch": 5.254509889154532,
+      "grad_norm": 0.3835771381855011,
+      "learning_rate": 2.3948275862068965e-05,
+      "loss": 0.0938,
+      "step": 4533
+    },
+    {
+      "epoch": 5.2556690574512785,
+      "grad_norm": 0.28895124793052673,
+      "learning_rate": 2.3942528735632187e-05,
+      "loss": 0.081,
+      "step": 4534
+    },
+    {
+      "epoch": 5.256828225748026,
+      "grad_norm": 0.36572563648223877,
+      "learning_rate": 2.3936781609195402e-05,
+      "loss": 0.0896,
+      "step": 4535
+    },
+    {
+      "epoch": 5.257987394044773,
+      "grad_norm": 0.4853745996952057,
+      "learning_rate": 2.393103448275862e-05,
+      "loss": 0.0952,
+      "step": 4536
+    },
+    {
+      "epoch": 5.25914656234152,
+      "grad_norm": 0.37866243720054626,
+      "learning_rate": 2.3925287356321842e-05,
+      "loss": 0.0936,
+      "step": 4537
+    },
+    {
+      "epoch": 5.260305730638267,
+      "grad_norm": 0.42627018690109253,
+      "learning_rate": 2.391954022988506e-05,
+      "loss": 0.0868,
+      "step": 4538
+    },
+    {
+      "epoch": 5.261464898935014,
+      "grad_norm": 0.3555832803249359,
+      "learning_rate": 2.3913793103448275e-05,
+      "loss": 0.0904,
+      "step": 4539
+    },
+    {
+      "epoch": 5.262624067231761,
+      "grad_norm": 0.4143363833427429,
+      "learning_rate": 2.3908045977011497e-05,
+      "loss": 0.0939,
+      "step": 4540
+    },
+    {
+      "epoch": 5.263783235528508,
+      "grad_norm": 0.3686874806880951,
+      "learning_rate": 2.3902298850574715e-05,
+      "loss": 0.0899,
+      "step": 4541
+    },
+    {
+      "epoch": 5.2649424038252555,
+      "grad_norm": 0.34536075592041016,
+      "learning_rate": 2.3896551724137933e-05,
+      "loss": 0.0869,
+      "step": 4542
+    },
+    {
+      "epoch": 5.266101572122002,
+      "grad_norm": 0.6285473704338074,
+      "learning_rate": 2.3890804597701148e-05,
+      "loss": 0.1043,
+      "step": 4543
+    },
+    {
+      "epoch": 5.26726074041875,
+      "grad_norm": 0.4580352306365967,
+      "learning_rate": 2.388505747126437e-05,
+      "loss": 0.0931,
+      "step": 4544
+    },
+    {
+      "epoch": 5.268419908715496,
+      "grad_norm": 0.3974003493785858,
+      "learning_rate": 2.3879310344827588e-05,
+      "loss": 0.0933,
+      "step": 4545
+    },
+    {
+      "epoch": 5.269579077012244,
+      "grad_norm": 0.4980361759662628,
+      "learning_rate": 2.3873563218390806e-05,
+      "loss": 0.095,
+      "step": 4546
+    },
+    {
+      "epoch": 5.270738245308991,
+      "grad_norm": 0.3423739969730377,
+      "learning_rate": 2.3867816091954025e-05,
+      "loss": 0.0922,
+      "step": 4547
+    },
+    {
+      "epoch": 5.271897413605738,
+      "grad_norm": 0.47561824321746826,
+      "learning_rate": 2.3862068965517243e-05,
+      "loss": 0.0896,
+      "step": 4548
+    },
+    {
+      "epoch": 5.273056581902485,
+      "grad_norm": 0.36675018072128296,
+      "learning_rate": 2.385632183908046e-05,
+      "loss": 0.0984,
+      "step": 4549
+    },
+    {
+      "epoch": 5.274215750199232,
+      "grad_norm": 0.507520854473114,
+      "learning_rate": 2.385057471264368e-05,
+      "loss": 0.092,
+      "step": 4550
+    },
+    {
+      "epoch": 5.275374918495979,
+      "grad_norm": 0.43633562326431274,
+      "learning_rate": 2.3844827586206898e-05,
+      "loss": 0.0979,
+      "step": 4551
+    },
+    {
+      "epoch": 5.276534086792727,
+      "grad_norm": 0.41792401671409607,
+      "learning_rate": 2.3839080459770116e-05,
+      "loss": 0.0897,
+      "step": 4552
+    },
+    {
+      "epoch": 5.277693255089473,
+      "grad_norm": 0.44738197326660156,
+      "learning_rate": 2.3833333333333334e-05,
+      "loss": 0.0915,
+      "step": 4553
+    },
+    {
+      "epoch": 5.278852423386221,
+      "grad_norm": 0.49772951006889343,
+      "learning_rate": 2.3827586206896553e-05,
+      "loss": 0.0982,
+      "step": 4554
+    },
+    {
+      "epoch": 5.2800115916829675,
+      "grad_norm": 0.5207464098930359,
+      "learning_rate": 2.382183908045977e-05,
+      "loss": 0.0929,
+      "step": 4555
+    },
+    {
+      "epoch": 5.281170759979714,
+      "grad_norm": 0.8150321841239929,
+      "learning_rate": 2.381609195402299e-05,
+      "loss": 0.1011,
+      "step": 4556
+    },
+    {
+      "epoch": 5.282329928276462,
+      "grad_norm": 0.4668427109718323,
+      "learning_rate": 2.3810344827586208e-05,
+      "loss": 0.0894,
+      "step": 4557
+    },
+    {
+      "epoch": 5.283489096573208,
+      "grad_norm": 0.42127272486686707,
+      "learning_rate": 2.3804597701149426e-05,
+      "loss": 0.0844,
+      "step": 4558
+    },
+    {
+      "epoch": 5.284648264869956,
+      "grad_norm": 0.6911650896072388,
+      "learning_rate": 2.3798850574712644e-05,
+      "loss": 0.1004,
+      "step": 4559
+    },
+    {
+      "epoch": 5.285807433166703,
+      "grad_norm": 0.37736475467681885,
+      "learning_rate": 2.3793103448275862e-05,
+      "loss": 0.0964,
+      "step": 4560
+    },
+    {
+      "epoch": 5.28696660146345,
+      "grad_norm": 0.4046756327152252,
+      "learning_rate": 2.3787356321839084e-05,
+      "loss": 0.0899,
+      "step": 4561
+    },
+    {
+      "epoch": 5.288125769760197,
+      "grad_norm": 0.3828308582305908,
+      "learning_rate": 2.37816091954023e-05,
+      "loss": 0.1046,
+      "step": 4562
+    },
+    {
+      "epoch": 5.289284938056944,
+      "grad_norm": 0.3368671238422394,
+      "learning_rate": 2.3775862068965517e-05,
+      "loss": 0.0895,
+      "step": 4563
+    },
+    {
+      "epoch": 5.290444106353691,
+      "grad_norm": 0.3442758321762085,
+      "learning_rate": 2.3770114942528736e-05,
+      "loss": 0.0895,
+      "step": 4564
+    },
+    {
+      "epoch": 5.291603274650439,
+      "grad_norm": 0.4719181954860687,
+      "learning_rate": 2.3764367816091957e-05,
+      "loss": 0.1,
+      "step": 4565
+    },
+    {
+      "epoch": 5.292762442947185,
+      "grad_norm": 0.37224331498146057,
+      "learning_rate": 2.3758620689655172e-05,
+      "loss": 0.1004,
+      "step": 4566
+    },
+    {
+      "epoch": 5.293921611243933,
+      "grad_norm": 0.4549705982208252,
+      "learning_rate": 2.375287356321839e-05,
+      "loss": 0.0958,
+      "step": 4567
+    },
+    {
+      "epoch": 5.29508077954068,
+      "grad_norm": 0.3962538242340088,
+      "learning_rate": 2.3747126436781612e-05,
+      "loss": 0.0984,
+      "step": 4568
+    },
+    {
+      "epoch": 5.296239947837426,
+      "grad_norm": 0.42686334252357483,
+      "learning_rate": 2.374137931034483e-05,
+      "loss": 0.0917,
+      "step": 4569
+    },
+    {
+      "epoch": 5.297399116134174,
+      "grad_norm": 0.4089353382587433,
+      "learning_rate": 2.3735632183908045e-05,
+      "loss": 0.0938,
+      "step": 4570
+    },
+    {
+      "epoch": 5.2985582844309205,
+      "grad_norm": 0.43393051624298096,
+      "learning_rate": 2.3729885057471267e-05,
+      "loss": 0.0894,
+      "step": 4571
+    },
+    {
+      "epoch": 5.299717452727668,
+      "grad_norm": 0.4832133650779724,
+      "learning_rate": 2.3724137931034485e-05,
+      "loss": 0.1016,
+      "step": 4572
+    },
+    {
+      "epoch": 5.300876621024415,
+      "grad_norm": 0.4066667854785919,
+      "learning_rate": 2.37183908045977e-05,
+      "loss": 0.0954,
+      "step": 4573
+    },
+    {
+      "epoch": 5.302035789321162,
+      "grad_norm": 0.40719154477119446,
+      "learning_rate": 2.371264367816092e-05,
+      "loss": 0.0924,
+      "step": 4574
+    },
+    {
+      "epoch": 5.303194957617909,
+      "grad_norm": 0.38603299856185913,
+      "learning_rate": 2.370689655172414e-05,
+      "loss": 0.0946,
+      "step": 4575
+    },
+    {
+      "epoch": 5.3043541259146565,
+      "grad_norm": 0.45713695883750916,
+      "learning_rate": 2.370114942528736e-05,
+      "loss": 0.0946,
+      "step": 4576
+    },
+    {
+      "epoch": 5.305513294211403,
+      "grad_norm": 0.43153274059295654,
+      "learning_rate": 2.3695402298850573e-05,
+      "loss": 0.0941,
+      "step": 4577
+    },
+    {
+      "epoch": 5.306672462508151,
+      "grad_norm": 0.5558722019195557,
+      "learning_rate": 2.3689655172413795e-05,
+      "loss": 0.0991,
+      "step": 4578
+    },
+    {
+      "epoch": 5.307831630804897,
+      "grad_norm": 0.3364887833595276,
+      "learning_rate": 2.3683908045977013e-05,
+      "loss": 0.0903,
+      "step": 4579
+    },
+    {
+      "epoch": 5.308990799101645,
+      "grad_norm": 0.39872419834136963,
+      "learning_rate": 2.367816091954023e-05,
+      "loss": 0.1015,
+      "step": 4580
+    },
+    {
+      "epoch": 5.310149967398392,
+      "grad_norm": 0.4545663297176361,
+      "learning_rate": 2.367241379310345e-05,
+      "loss": 0.1017,
+      "step": 4581
+    },
+    {
+      "epoch": 5.311309135695138,
+      "grad_norm": 0.4072364866733551,
+      "learning_rate": 2.3666666666666668e-05,
+      "loss": 0.0888,
+      "step": 4582
+    },
+    {
+      "epoch": 5.312468303991886,
+      "grad_norm": 0.373927503824234,
+      "learning_rate": 2.3660919540229886e-05,
+      "loss": 0.0973,
+      "step": 4583
+    },
+    {
+      "epoch": 5.3136274722886325,
+      "grad_norm": 0.35121285915374756,
+      "learning_rate": 2.3655172413793105e-05,
+      "loss": 0.0938,
+      "step": 4584
+    },
+    {
+      "epoch": 5.31478664058538,
+      "grad_norm": 0.4112243950366974,
+      "learning_rate": 2.3649425287356323e-05,
+      "loss": 0.1031,
+      "step": 4585
+    },
+    {
+      "epoch": 5.315945808882127,
+      "grad_norm": 0.4238256812095642,
+      "learning_rate": 2.364367816091954e-05,
+      "loss": 0.0929,
+      "step": 4586
+    },
+    {
+      "epoch": 5.317104977178874,
+      "grad_norm": 0.531015157699585,
+      "learning_rate": 2.363793103448276e-05,
+      "loss": 0.0988,
+      "step": 4587
+    },
+    {
+      "epoch": 5.318264145475621,
+      "grad_norm": 0.4540468752384186,
+      "learning_rate": 2.3632183908045978e-05,
+      "loss": 0.1005,
+      "step": 4588
+    },
+    {
+      "epoch": 5.3194233137723685,
+      "grad_norm": 0.30282753705978394,
+      "learning_rate": 2.3626436781609196e-05,
+      "loss": 0.0845,
+      "step": 4589
+    },
+    {
+      "epoch": 5.320582482069115,
+      "grad_norm": 0.3676490783691406,
+      "learning_rate": 2.3620689655172415e-05,
+      "loss": 0.0846,
+      "step": 4590
+    },
+    {
+      "epoch": 5.321741650365863,
+      "grad_norm": 0.4564397931098938,
+      "learning_rate": 2.3614942528735633e-05,
+      "loss": 0.086,
+      "step": 4591
+    },
+    {
+      "epoch": 5.3229008186626094,
+      "grad_norm": 0.3975692689418793,
+      "learning_rate": 2.360919540229885e-05,
+      "loss": 0.0934,
+      "step": 4592
+    },
+    {
+      "epoch": 5.324059986959357,
+      "grad_norm": 0.35405778884887695,
+      "learning_rate": 2.360344827586207e-05,
+      "loss": 0.0997,
+      "step": 4593
+    },
+    {
+      "epoch": 5.325219155256104,
+      "grad_norm": 0.3428707718849182,
+      "learning_rate": 2.3597701149425288e-05,
+      "loss": 0.0859,
+      "step": 4594
+    },
+    {
+      "epoch": 5.326378323552851,
+      "grad_norm": 0.41354939341545105,
+      "learning_rate": 2.359195402298851e-05,
+      "loss": 0.0982,
+      "step": 4595
+    },
+    {
+      "epoch": 5.327537491849598,
+      "grad_norm": 0.5002196431159973,
+      "learning_rate": 2.3586206896551724e-05,
+      "loss": 0.0968,
+      "step": 4596
+    },
+    {
+      "epoch": 5.3286966601463455,
+      "grad_norm": 0.40941765904426575,
+      "learning_rate": 2.3580459770114943e-05,
+      "loss": 0.0923,
+      "step": 4597
+    },
+    {
+      "epoch": 5.329855828443092,
+      "grad_norm": 0.37814703583717346,
+      "learning_rate": 2.357471264367816e-05,
+      "loss": 0.0916,
+      "step": 4598
+    },
+    {
+      "epoch": 5.331014996739839,
+      "grad_norm": 0.3758867681026459,
+      "learning_rate": 2.3568965517241383e-05,
+      "loss": 0.0851,
+      "step": 4599
+    },
+    {
+      "epoch": 5.332174165036586,
+      "grad_norm": 0.48879507184028625,
+      "learning_rate": 2.3563218390804597e-05,
+      "loss": 0.1009,
+      "step": 4600
+    },
+    {
+      "epoch": 5.333333333333333,
+      "grad_norm": 0.6215201616287231,
+      "learning_rate": 2.3557471264367816e-05,
+      "loss": 0.098,
+      "step": 4601
+    },
+    {
+      "epoch": 5.334492501630081,
+      "grad_norm": 0.48262956738471985,
+      "learning_rate": 2.3551724137931037e-05,
+      "loss": 0.0904,
+      "step": 4602
+    },
+    {
+      "epoch": 5.335651669926827,
+      "grad_norm": 0.40028294920921326,
+      "learning_rate": 2.3545977011494256e-05,
+      "loss": 0.0851,
+      "step": 4603
+    },
+    {
+      "epoch": 5.336810838223575,
+      "grad_norm": 0.47819679975509644,
+      "learning_rate": 2.354022988505747e-05,
+      "loss": 0.096,
+      "step": 4604
+    },
+    {
+      "epoch": 5.3379700065203215,
+      "grad_norm": 0.36950936913490295,
+      "learning_rate": 2.3534482758620692e-05,
+      "loss": 0.0909,
+      "step": 4605
+    },
+    {
+      "epoch": 5.339129174817069,
+      "grad_norm": 0.42461180686950684,
+      "learning_rate": 2.352873563218391e-05,
+      "loss": 0.0992,
+      "step": 4606
+    },
+    {
+      "epoch": 5.340288343113816,
+      "grad_norm": 0.3869633674621582,
+      "learning_rate": 2.352298850574713e-05,
+      "loss": 0.0915,
+      "step": 4607
+    },
+    {
+      "epoch": 5.341447511410563,
+      "grad_norm": 0.36021578311920166,
+      "learning_rate": 2.3517241379310344e-05,
+      "loss": 0.0952,
+      "step": 4608
+    },
+    {
+      "epoch": 5.34260667970731,
+      "grad_norm": 0.3799790143966675,
+      "learning_rate": 2.3511494252873565e-05,
+      "loss": 0.0881,
+      "step": 4609
+    },
+    {
+      "epoch": 5.3437658480040575,
+      "grad_norm": 0.4257470667362213,
+      "learning_rate": 2.3505747126436784e-05,
+      "loss": 0.1011,
+      "step": 4610
+    },
+    {
+      "epoch": 5.344925016300804,
+      "grad_norm": 0.5010349154472351,
+      "learning_rate": 2.35e-05,
+      "loss": 0.0897,
+      "step": 4611
+    },
+    {
+      "epoch": 5.346084184597551,
+      "grad_norm": 0.3810718357563019,
+      "learning_rate": 2.349425287356322e-05,
+      "loss": 0.0969,
+      "step": 4612
+    },
+    {
+      "epoch": 5.347243352894298,
+      "grad_norm": 0.4444500505924225,
+      "learning_rate": 2.348850574712644e-05,
+      "loss": 0.0941,
+      "step": 4613
+    },
+    {
+      "epoch": 5.348402521191045,
+      "grad_norm": 0.5119409561157227,
+      "learning_rate": 2.3482758620689657e-05,
+      "loss": 0.1,
+      "step": 4614
+    },
+    {
+      "epoch": 5.349561689487793,
+      "grad_norm": 0.31901487708091736,
+      "learning_rate": 2.3477011494252875e-05,
+      "loss": 0.0996,
+      "step": 4615
+    },
+    {
+      "epoch": 5.350720857784539,
+      "grad_norm": 0.33912405371665955,
+      "learning_rate": 2.3471264367816093e-05,
+      "loss": 0.0938,
+      "step": 4616
+    },
+    {
+      "epoch": 5.351880026081287,
+      "grad_norm": 0.4147634208202362,
+      "learning_rate": 2.3465517241379312e-05,
+      "loss": 0.0975,
+      "step": 4617
+    },
+    {
+      "epoch": 5.3530391943780335,
+      "grad_norm": 0.33564382791519165,
+      "learning_rate": 2.345977011494253e-05,
+      "loss": 0.0827,
+      "step": 4618
+    },
+    {
+      "epoch": 5.354198362674781,
+      "grad_norm": 0.47823843359947205,
+      "learning_rate": 2.345402298850575e-05,
+      "loss": 0.1004,
+      "step": 4619
+    },
+    {
+      "epoch": 5.355357530971528,
+      "grad_norm": 0.43068960309028625,
+      "learning_rate": 2.3448275862068967e-05,
+      "loss": 0.0898,
+      "step": 4620
+    },
+    {
+      "epoch": 5.356516699268275,
+      "grad_norm": 0.4093489646911621,
+      "learning_rate": 2.3442528735632185e-05,
+      "loss": 0.099,
+      "step": 4621
+    },
+    {
+      "epoch": 5.357675867565022,
+      "grad_norm": 0.4601415991783142,
+      "learning_rate": 2.3436781609195403e-05,
+      "loss": 0.0959,
+      "step": 4622
+    },
+    {
+      "epoch": 5.35883503586177,
+      "grad_norm": 0.42657336592674255,
+      "learning_rate": 2.343103448275862e-05,
+      "loss": 0.0862,
+      "step": 4623
+    },
+    {
+      "epoch": 5.359994204158516,
+      "grad_norm": 0.41843754053115845,
+      "learning_rate": 2.342528735632184e-05,
+      "loss": 0.0896,
+      "step": 4624
+    },
+    {
+      "epoch": 5.361153372455263,
+      "grad_norm": 0.4212128818035126,
+      "learning_rate": 2.3419540229885058e-05,
+      "loss": 0.0981,
+      "step": 4625
+    },
+    {
+      "epoch": 5.3623125407520105,
+      "grad_norm": 0.3441298305988312,
+      "learning_rate": 2.341379310344828e-05,
+      "loss": 0.0958,
+      "step": 4626
+    },
+    {
+      "epoch": 5.363471709048757,
+      "grad_norm": 0.398131787776947,
+      "learning_rate": 2.3408045977011495e-05,
+      "loss": 0.0944,
+      "step": 4627
+    },
+    {
+      "epoch": 5.364630877345505,
+      "grad_norm": 0.5075594186782837,
+      "learning_rate": 2.3402298850574713e-05,
+      "loss": 0.0958,
+      "step": 4628
+    },
+    {
+      "epoch": 5.365790045642251,
+      "grad_norm": 0.3779784142971039,
+      "learning_rate": 2.339655172413793e-05,
+      "loss": 0.1002,
+      "step": 4629
+    },
+    {
+      "epoch": 5.366949213938999,
+      "grad_norm": 0.39865589141845703,
+      "learning_rate": 2.339080459770115e-05,
+      "loss": 0.0853,
+      "step": 4630
+    },
+    {
+      "epoch": 5.368108382235746,
+      "grad_norm": 0.7066544890403748,
+      "learning_rate": 2.3385057471264368e-05,
+      "loss": 0.0903,
+      "step": 4631
+    },
+    {
+      "epoch": 5.369267550532493,
+      "grad_norm": 0.3441803455352783,
+      "learning_rate": 2.3379310344827586e-05,
+      "loss": 0.084,
+      "step": 4632
+    },
+    {
+      "epoch": 5.37042671882924,
+      "grad_norm": 0.4163109362125397,
+      "learning_rate": 2.3373563218390808e-05,
+      "loss": 0.095,
+      "step": 4633
+    },
+    {
+      "epoch": 5.371585887125987,
+      "grad_norm": 0.43912622332572937,
+      "learning_rate": 2.3367816091954023e-05,
+      "loss": 0.0922,
+      "step": 4634
+    },
+    {
+      "epoch": 5.372745055422734,
+      "grad_norm": 0.304113507270813,
+      "learning_rate": 2.336206896551724e-05,
+      "loss": 0.0875,
+      "step": 4635
+    },
+    {
+      "epoch": 5.373904223719482,
+      "grad_norm": 0.3352859318256378,
+      "learning_rate": 2.3356321839080463e-05,
+      "loss": 0.0966,
+      "step": 4636
+    },
+    {
+      "epoch": 5.375063392016228,
+      "grad_norm": 0.4194989800453186,
+      "learning_rate": 2.335057471264368e-05,
+      "loss": 0.1005,
+      "step": 4637
+    },
+    {
+      "epoch": 5.376222560312976,
+      "grad_norm": 0.3426196873188019,
+      "learning_rate": 2.3344827586206896e-05,
+      "loss": 0.0988,
+      "step": 4638
+    },
+    {
+      "epoch": 5.3773817286097225,
+      "grad_norm": 0.35888779163360596,
+      "learning_rate": 2.3339080459770114e-05,
+      "loss": 0.0897,
+      "step": 4639
+    },
+    {
+      "epoch": 5.378540896906469,
+      "grad_norm": 0.35968539118766785,
+      "learning_rate": 2.3333333333333336e-05,
+      "loss": 0.0873,
+      "step": 4640
+    },
+    {
+      "epoch": 5.379700065203217,
+      "grad_norm": 0.4251978099346161,
+      "learning_rate": 2.3327586206896554e-05,
+      "loss": 0.0904,
+      "step": 4641
+    },
+    {
+      "epoch": 5.380859233499963,
+      "grad_norm": 0.5857329964637756,
+      "learning_rate": 2.332183908045977e-05,
+      "loss": 0.0879,
+      "step": 4642
+    },
+    {
+      "epoch": 5.382018401796711,
+      "grad_norm": 0.5313498377799988,
+      "learning_rate": 2.331609195402299e-05,
+      "loss": 0.0975,
+      "step": 4643
+    },
+    {
+      "epoch": 5.383177570093458,
+      "grad_norm": 0.4486619234085083,
+      "learning_rate": 2.331034482758621e-05,
+      "loss": 0.09,
+      "step": 4644
+    },
+    {
+      "epoch": 5.384336738390205,
+      "grad_norm": 0.40548866987228394,
+      "learning_rate": 2.3304597701149427e-05,
+      "loss": 0.0929,
+      "step": 4645
+    },
+    {
+      "epoch": 5.385495906686952,
+      "grad_norm": 0.5880759954452515,
+      "learning_rate": 2.3298850574712646e-05,
+      "loss": 0.0945,
+      "step": 4646
+    },
+    {
+      "epoch": 5.386655074983699,
+      "grad_norm": 0.5098415017127991,
+      "learning_rate": 2.3293103448275864e-05,
+      "loss": 0.0874,
+      "step": 4647
+    },
+    {
+      "epoch": 5.387814243280446,
+      "grad_norm": 0.42802801728248596,
+      "learning_rate": 2.3287356321839082e-05,
+      "loss": 0.0853,
+      "step": 4648
+    },
+    {
+      "epoch": 5.388973411577194,
+      "grad_norm": 0.5465481281280518,
+      "learning_rate": 2.32816091954023e-05,
+      "loss": 0.0944,
+      "step": 4649
+    },
+    {
+      "epoch": 5.39013257987394,
+      "grad_norm": 0.4456057548522949,
+      "learning_rate": 2.327586206896552e-05,
+      "loss": 0.1031,
+      "step": 4650
+    },
+    {
+      "epoch": 5.391291748170688,
+      "grad_norm": 0.36883285641670227,
+      "learning_rate": 2.3270114942528737e-05,
+      "loss": 0.0992,
+      "step": 4651
+    },
+    {
+      "epoch": 5.392450916467435,
+      "grad_norm": 0.39192837476730347,
+      "learning_rate": 2.3264367816091955e-05,
+      "loss": 0.0943,
+      "step": 4652
+    },
+    {
+      "epoch": 5.393610084764182,
+      "grad_norm": 0.36094704270362854,
+      "learning_rate": 2.3258620689655174e-05,
+      "loss": 0.0896,
+      "step": 4653
+    },
+    {
+      "epoch": 5.394769253060929,
+      "grad_norm": 0.8563114404678345,
+      "learning_rate": 2.3252873563218392e-05,
+      "loss": 0.0834,
+      "step": 4654
+    },
+    {
+      "epoch": 5.3959284213576755,
+      "grad_norm": 0.4293108880519867,
+      "learning_rate": 2.324712643678161e-05,
+      "loss": 0.1027,
+      "step": 4655
+    },
+    {
+      "epoch": 5.397087589654423,
+      "grad_norm": 0.5573574900627136,
+      "learning_rate": 2.324137931034483e-05,
+      "loss": 0.0978,
+      "step": 4656
+    },
+    {
+      "epoch": 5.39824675795117,
+      "grad_norm": 0.3921107351779938,
+      "learning_rate": 2.3235632183908047e-05,
+      "loss": 0.09,
+      "step": 4657
+    },
+    {
+      "epoch": 5.399405926247917,
+      "grad_norm": 0.41322219371795654,
+      "learning_rate": 2.3229885057471265e-05,
+      "loss": 0.0815,
+      "step": 4658
+    },
+    {
+      "epoch": 5.400565094544664,
+      "grad_norm": 0.41785478591918945,
+      "learning_rate": 2.3224137931034483e-05,
+      "loss": 0.0907,
+      "step": 4659
+    },
+    {
+      "epoch": 5.4017242628414115,
+      "grad_norm": 0.39730241894721985,
+      "learning_rate": 2.3218390804597705e-05,
+      "loss": 0.098,
+      "step": 4660
+    },
+    {
+      "epoch": 5.402883431138158,
+      "grad_norm": 0.46797794103622437,
+      "learning_rate": 2.321264367816092e-05,
+      "loss": 0.0967,
+      "step": 4661
+    },
+    {
+      "epoch": 5.404042599434906,
+      "grad_norm": 0.42924076318740845,
+      "learning_rate": 2.3206896551724138e-05,
+      "loss": 0.0872,
+      "step": 4662
+    },
+    {
+      "epoch": 5.405201767731652,
+      "grad_norm": 0.4479629695415497,
+      "learning_rate": 2.3201149425287356e-05,
+      "loss": 0.0882,
+      "step": 4663
+    },
+    {
+      "epoch": 5.4063609360284,
+      "grad_norm": 0.32181426882743835,
+      "learning_rate": 2.3195402298850578e-05,
+      "loss": 0.0851,
+      "step": 4664
+    },
+    {
+      "epoch": 5.407520104325147,
+      "grad_norm": 0.44183018803596497,
+      "learning_rate": 2.3189655172413793e-05,
+      "loss": 0.0861,
+      "step": 4665
+    },
+    {
+      "epoch": 5.408679272621894,
+      "grad_norm": 0.4371998608112335,
+      "learning_rate": 2.318390804597701e-05,
+      "loss": 0.0945,
+      "step": 4666
+    },
+    {
+      "epoch": 5.409838440918641,
+      "grad_norm": 0.3353933095932007,
+      "learning_rate": 2.3178160919540233e-05,
+      "loss": 0.0867,
+      "step": 4667
+    },
+    {
+      "epoch": 5.4109976092153875,
+      "grad_norm": 0.35332971811294556,
+      "learning_rate": 2.317241379310345e-05,
+      "loss": 0.0795,
+      "step": 4668
+    },
+    {
+      "epoch": 5.412156777512135,
+      "grad_norm": 0.3787877857685089,
+      "learning_rate": 2.3166666666666666e-05,
+      "loss": 0.0977,
+      "step": 4669
+    },
+    {
+      "epoch": 5.413315945808882,
+      "grad_norm": 0.3937995135784149,
+      "learning_rate": 2.3160919540229885e-05,
+      "loss": 0.101,
+      "step": 4670
+    },
+    {
+      "epoch": 5.414475114105629,
+      "grad_norm": 0.3958158493041992,
+      "learning_rate": 2.3155172413793106e-05,
+      "loss": 0.0962,
+      "step": 4671
+    },
+    {
+      "epoch": 5.415634282402376,
+      "grad_norm": 0.5695372819900513,
+      "learning_rate": 2.314942528735632e-05,
+      "loss": 0.1024,
+      "step": 4672
+    },
+    {
+      "epoch": 5.4167934506991235,
+      "grad_norm": 0.4417628049850464,
+      "learning_rate": 2.314367816091954e-05,
+      "loss": 0.0931,
+      "step": 4673
+    },
+    {
+      "epoch": 5.41795261899587,
+      "grad_norm": 0.44972625374794006,
+      "learning_rate": 2.313793103448276e-05,
+      "loss": 0.1041,
+      "step": 4674
+    },
+    {
+      "epoch": 5.419111787292618,
+      "grad_norm": 0.375545471906662,
+      "learning_rate": 2.313218390804598e-05,
+      "loss": 0.0948,
+      "step": 4675
+    },
+    {
+      "epoch": 5.4202709555893644,
+      "grad_norm": 0.4629718065261841,
+      "learning_rate": 2.3126436781609194e-05,
+      "loss": 0.09,
+      "step": 4676
+    },
+    {
+      "epoch": 5.421430123886112,
+      "grad_norm": 0.462091326713562,
+      "learning_rate": 2.3120689655172416e-05,
+      "loss": 0.0955,
+      "step": 4677
+    },
+    {
+      "epoch": 5.422589292182859,
+      "grad_norm": 0.43810778856277466,
+      "learning_rate": 2.3114942528735634e-05,
+      "loss": 0.0946,
+      "step": 4678
+    },
+    {
+      "epoch": 5.423748460479606,
+      "grad_norm": 0.48332980275154114,
+      "learning_rate": 2.3109195402298853e-05,
+      "loss": 0.097,
+      "step": 4679
+    },
+    {
+      "epoch": 5.424907628776353,
+      "grad_norm": 0.4389806091785431,
+      "learning_rate": 2.3103448275862067e-05,
+      "loss": 0.0881,
+      "step": 4680
+    },
+    {
+      "epoch": 5.4260667970731,
+      "grad_norm": 0.4335387349128723,
+      "learning_rate": 2.309770114942529e-05,
+      "loss": 0.0946,
+      "step": 4681
+    },
+    {
+      "epoch": 5.427225965369847,
+      "grad_norm": 0.4705444872379303,
+      "learning_rate": 2.3091954022988507e-05,
+      "loss": 0.0989,
+      "step": 4682
+    },
+    {
+      "epoch": 5.428385133666594,
+      "grad_norm": 0.4291810989379883,
+      "learning_rate": 2.3086206896551726e-05,
+      "loss": 0.0942,
+      "step": 4683
+    },
+    {
+      "epoch": 5.429544301963341,
+      "grad_norm": 0.6216928958892822,
+      "learning_rate": 2.3080459770114944e-05,
+      "loss": 0.096,
+      "step": 4684
+    },
+    {
+      "epoch": 5.430703470260088,
+      "grad_norm": 0.3836531937122345,
+      "learning_rate": 2.3074712643678162e-05,
+      "loss": 0.088,
+      "step": 4685
+    },
+    {
+      "epoch": 5.431862638556836,
+      "grad_norm": 0.46026742458343506,
+      "learning_rate": 2.306896551724138e-05,
+      "loss": 0.0907,
+      "step": 4686
+    },
+    {
+      "epoch": 5.433021806853582,
+      "grad_norm": 0.45022183656692505,
+      "learning_rate": 2.30632183908046e-05,
+      "loss": 0.0981,
+      "step": 4687
+    },
+    {
+      "epoch": 5.43418097515033,
+      "grad_norm": 0.3708306550979614,
+      "learning_rate": 2.3057471264367817e-05,
+      "loss": 0.0984,
+      "step": 4688
+    },
+    {
+      "epoch": 5.4353401434470765,
+      "grad_norm": 0.3784964084625244,
+      "learning_rate": 2.3051724137931035e-05,
+      "loss": 0.0918,
+      "step": 4689
+    },
+    {
+      "epoch": 5.436499311743824,
+      "grad_norm": 0.6409464478492737,
+      "learning_rate": 2.3045977011494254e-05,
+      "loss": 0.0958,
+      "step": 4690
+    },
+    {
+      "epoch": 5.437658480040571,
+      "grad_norm": 0.5641512274742126,
+      "learning_rate": 2.3040229885057472e-05,
+      "loss": 0.0927,
+      "step": 4691
+    },
+    {
+      "epoch": 5.438817648337318,
+      "grad_norm": 0.42215609550476074,
+      "learning_rate": 2.303448275862069e-05,
+      "loss": 0.0903,
+      "step": 4692
+    },
+    {
+      "epoch": 5.439976816634065,
+      "grad_norm": 0.5272184610366821,
+      "learning_rate": 2.302873563218391e-05,
+      "loss": 0.0919,
+      "step": 4693
+    },
+    {
+      "epoch": 5.4411359849308125,
+      "grad_norm": 0.37865811586380005,
+      "learning_rate": 2.3022988505747127e-05,
+      "loss": 0.0928,
+      "step": 4694
+    },
+    {
+      "epoch": 5.442295153227559,
+      "grad_norm": 0.48235297203063965,
+      "learning_rate": 2.3017241379310345e-05,
+      "loss": 0.0885,
+      "step": 4695
+    },
+    {
+      "epoch": 5.443454321524307,
+      "grad_norm": 0.3501775562763214,
+      "learning_rate": 2.3011494252873563e-05,
+      "loss": 0.0847,
+      "step": 4696
+    },
+    {
+      "epoch": 5.444613489821053,
+      "grad_norm": 0.5579177141189575,
+      "learning_rate": 2.3005747126436782e-05,
+      "loss": 0.1053,
+      "step": 4697
+    },
+    {
+      "epoch": 5.4457726581178,
+      "grad_norm": 0.423701673746109,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 0.0932,
+      "step": 4698
+    },
+    {
+      "epoch": 5.446931826414548,
+      "grad_norm": 0.4396893084049225,
+      "learning_rate": 2.299425287356322e-05,
+      "loss": 0.0968,
+      "step": 4699
+    },
+    {
+      "epoch": 5.448090994711294,
+      "grad_norm": 0.3700418472290039,
+      "learning_rate": 2.2988505747126437e-05,
+      "loss": 0.0908,
+      "step": 4700
+    },
+    {
+      "epoch": 5.449250163008042,
+      "grad_norm": 0.3896975815296173,
+      "learning_rate": 2.2982758620689658e-05,
+      "loss": 0.0947,
+      "step": 4701
+    },
+    {
+      "epoch": 5.4504093313047886,
+      "grad_norm": 0.3217882812023163,
+      "learning_rate": 2.2977011494252877e-05,
+      "loss": 0.0881,
+      "step": 4702
+    },
+    {
+      "epoch": 5.451568499601536,
+      "grad_norm": 0.47387391328811646,
+      "learning_rate": 2.297126436781609e-05,
+      "loss": 0.1008,
+      "step": 4703
+    },
+    {
+      "epoch": 5.452727667898283,
+      "grad_norm": 0.2585291862487793,
+      "learning_rate": 2.296551724137931e-05,
+      "loss": 0.0952,
+      "step": 4704
+    },
+    {
+      "epoch": 5.45388683619503,
+      "grad_norm": 0.4016747772693634,
+      "learning_rate": 2.295977011494253e-05,
+      "loss": 0.0852,
+      "step": 4705
+    },
+    {
+      "epoch": 5.455046004491777,
+      "grad_norm": 0.29344117641448975,
+      "learning_rate": 2.295402298850575e-05,
+      "loss": 0.0946,
+      "step": 4706
+    },
+    {
+      "epoch": 5.456205172788525,
+      "grad_norm": 0.334484338760376,
+      "learning_rate": 2.2948275862068965e-05,
+      "loss": 0.088,
+      "step": 4707
+    },
+    {
+      "epoch": 5.457364341085271,
+      "grad_norm": 0.5377522706985474,
+      "learning_rate": 2.2942528735632186e-05,
+      "loss": 0.0911,
+      "step": 4708
+    },
+    {
+      "epoch": 5.458523509382019,
+      "grad_norm": 0.3403340280056,
+      "learning_rate": 2.2936781609195405e-05,
+      "loss": 0.0961,
+      "step": 4709
+    },
+    {
+      "epoch": 5.4596826776787655,
+      "grad_norm": 0.370523601770401,
+      "learning_rate": 2.293103448275862e-05,
+      "loss": 0.0951,
+      "step": 4710
+    },
+    {
+      "epoch": 5.460841845975512,
+      "grad_norm": 0.3723559081554413,
+      "learning_rate": 2.292528735632184e-05,
+      "loss": 0.0967,
+      "step": 4711
+    },
+    {
+      "epoch": 5.46200101427226,
+      "grad_norm": 0.4718587398529053,
+      "learning_rate": 2.291954022988506e-05,
+      "loss": 0.0917,
+      "step": 4712
+    },
+    {
+      "epoch": 5.463160182569006,
+      "grad_norm": 0.3738349974155426,
+      "learning_rate": 2.2913793103448278e-05,
+      "loss": 0.0973,
+      "step": 4713
+    },
+    {
+      "epoch": 5.464319350865754,
+      "grad_norm": 0.4696074426174164,
+      "learning_rate": 2.2908045977011493e-05,
+      "loss": 0.095,
+      "step": 4714
+    },
+    {
+      "epoch": 5.465478519162501,
+      "grad_norm": 0.39924389123916626,
+      "learning_rate": 2.2902298850574714e-05,
+      "loss": 0.0899,
+      "step": 4715
+    },
+    {
+      "epoch": 5.466637687459248,
+      "grad_norm": 0.4714980125427246,
+      "learning_rate": 2.2896551724137933e-05,
+      "loss": 0.0892,
+      "step": 4716
+    },
+    {
+      "epoch": 5.467796855755995,
+      "grad_norm": 0.3208000361919403,
+      "learning_rate": 2.289080459770115e-05,
+      "loss": 0.0917,
+      "step": 4717
+    },
+    {
+      "epoch": 5.468956024052742,
+      "grad_norm": 0.5692875385284424,
+      "learning_rate": 2.288505747126437e-05,
+      "loss": 0.0917,
+      "step": 4718
+    },
+    {
+      "epoch": 5.470115192349489,
+      "grad_norm": 0.5145659446716309,
+      "learning_rate": 2.2879310344827587e-05,
+      "loss": 0.0994,
+      "step": 4719
+    },
+    {
+      "epoch": 5.471274360646237,
+      "grad_norm": 0.4095035791397095,
+      "learning_rate": 2.2873563218390806e-05,
+      "loss": 0.0857,
+      "step": 4720
+    },
+    {
+      "epoch": 5.472433528942983,
+      "grad_norm": 0.38376545906066895,
+      "learning_rate": 2.2867816091954024e-05,
+      "loss": 0.0959,
+      "step": 4721
+    },
+    {
+      "epoch": 5.473592697239731,
+      "grad_norm": 0.5485108494758606,
+      "learning_rate": 2.2862068965517242e-05,
+      "loss": 0.0857,
+      "step": 4722
+    },
+    {
+      "epoch": 5.4747518655364775,
+      "grad_norm": 0.6178170442581177,
+      "learning_rate": 2.285632183908046e-05,
+      "loss": 0.1075,
+      "step": 4723
+    },
+    {
+      "epoch": 5.475911033833224,
+      "grad_norm": 0.44804081320762634,
+      "learning_rate": 2.285057471264368e-05,
+      "loss": 0.0933,
+      "step": 4724
+    },
+    {
+      "epoch": 5.477070202129972,
+      "grad_norm": 0.7503616213798523,
+      "learning_rate": 2.2844827586206897e-05,
+      "loss": 0.0994,
+      "step": 4725
+    },
+    {
+      "epoch": 5.478229370426718,
+      "grad_norm": 0.4127942621707916,
+      "learning_rate": 2.2839080459770116e-05,
+      "loss": 0.0947,
+      "step": 4726
+    },
+    {
+      "epoch": 5.479388538723466,
+      "grad_norm": 0.4280507266521454,
+      "learning_rate": 2.2833333333333334e-05,
+      "loss": 0.0837,
+      "step": 4727
+    },
+    {
+      "epoch": 5.480547707020213,
+      "grad_norm": 0.41026708483695984,
+      "learning_rate": 2.2827586206896552e-05,
+      "loss": 0.0915,
+      "step": 4728
+    },
+    {
+      "epoch": 5.48170687531696,
+      "grad_norm": 0.3558439016342163,
+      "learning_rate": 2.2821839080459774e-05,
+      "loss": 0.0843,
+      "step": 4729
+    },
+    {
+      "epoch": 5.482866043613707,
+      "grad_norm": 0.3199481964111328,
+      "learning_rate": 2.281609195402299e-05,
+      "loss": 0.088,
+      "step": 4730
+    },
+    {
+      "epoch": 5.4840252119104544,
+      "grad_norm": 0.3802203834056854,
+      "learning_rate": 2.2810344827586207e-05,
+      "loss": 0.0955,
+      "step": 4731
+    },
+    {
+      "epoch": 5.485184380207201,
+      "grad_norm": 0.6358181238174438,
+      "learning_rate": 2.280459770114943e-05,
+      "loss": 0.1024,
+      "step": 4732
+    },
+    {
+      "epoch": 5.486343548503949,
+      "grad_norm": 0.3913860619068146,
+      "learning_rate": 2.2798850574712644e-05,
+      "loss": 0.0922,
+      "step": 4733
+    },
+    {
+      "epoch": 5.487502716800695,
+      "grad_norm": 0.4372271001338959,
+      "learning_rate": 2.2793103448275862e-05,
+      "loss": 0.0905,
+      "step": 4734
+    },
+    {
+      "epoch": 5.488661885097443,
+      "grad_norm": 0.4437418282032013,
+      "learning_rate": 2.278735632183908e-05,
+      "loss": 0.0815,
+      "step": 4735
+    },
+    {
+      "epoch": 5.48982105339419,
+      "grad_norm": 0.38980501890182495,
+      "learning_rate": 2.2781609195402302e-05,
+      "loss": 0.0888,
+      "step": 4736
+    },
+    {
+      "epoch": 5.490980221690937,
+      "grad_norm": 0.6082010865211487,
+      "learning_rate": 2.2775862068965517e-05,
+      "loss": 0.1109,
+      "step": 4737
+    },
+    {
+      "epoch": 5.492139389987684,
+      "grad_norm": 0.4929609000682831,
+      "learning_rate": 2.2770114942528735e-05,
+      "loss": 0.1048,
+      "step": 4738
+    },
+    {
+      "epoch": 5.493298558284431,
+      "grad_norm": 0.3931887745857239,
+      "learning_rate": 2.2764367816091957e-05,
+      "loss": 0.087,
+      "step": 4739
+    },
+    {
+      "epoch": 5.494457726581178,
+      "grad_norm": 0.4002040922641754,
+      "learning_rate": 2.2758620689655175e-05,
+      "loss": 0.0938,
+      "step": 4740
+    },
+    {
+      "epoch": 5.495616894877925,
+      "grad_norm": 0.6055549383163452,
+      "learning_rate": 2.275287356321839e-05,
+      "loss": 0.1004,
+      "step": 4741
+    },
+    {
+      "epoch": 5.496776063174672,
+      "grad_norm": 0.4150696098804474,
+      "learning_rate": 2.274712643678161e-05,
+      "loss": 0.0971,
+      "step": 4742
+    },
+    {
+      "epoch": 5.497935231471419,
+      "grad_norm": 0.43481725454330444,
+      "learning_rate": 2.274137931034483e-05,
+      "loss": 0.0936,
+      "step": 4743
+    },
+    {
+      "epoch": 5.4990943997681665,
+      "grad_norm": 0.4212779998779297,
+      "learning_rate": 2.2735632183908048e-05,
+      "loss": 0.0863,
+      "step": 4744
+    },
+    {
+      "epoch": 5.500253568064913,
+      "grad_norm": 0.3982880711555481,
+      "learning_rate": 2.2729885057471263e-05,
+      "loss": 0.092,
+      "step": 4745
+    },
+    {
+      "epoch": 5.501412736361661,
+      "grad_norm": 0.7657057642936707,
+      "learning_rate": 2.2724137931034485e-05,
+      "loss": 0.0915,
+      "step": 4746
+    },
+    {
+      "epoch": 5.502571904658407,
+      "grad_norm": 0.47100207209587097,
+      "learning_rate": 2.2718390804597703e-05,
+      "loss": 0.0896,
+      "step": 4747
+    },
+    {
+      "epoch": 5.503731072955155,
+      "grad_norm": 0.41695326566696167,
+      "learning_rate": 2.271264367816092e-05,
+      "loss": 0.1028,
+      "step": 4748
+    },
+    {
+      "epoch": 5.504890241251902,
+      "grad_norm": 0.38724851608276367,
+      "learning_rate": 2.270689655172414e-05,
+      "loss": 0.0922,
+      "step": 4749
+    },
+    {
+      "epoch": 5.506049409548649,
+      "grad_norm": 0.32382097840309143,
+      "learning_rate": 2.2701149425287358e-05,
+      "loss": 0.0894,
+      "step": 4750
+    },
+    {
+      "epoch": 5.507208577845396,
+      "grad_norm": 0.7724230289459229,
+      "learning_rate": 2.2695402298850576e-05,
+      "loss": 0.0947,
+      "step": 4751
+    },
+    {
+      "epoch": 5.508367746142143,
+      "grad_norm": 0.35658082365989685,
+      "learning_rate": 2.2689655172413794e-05,
+      "loss": 0.086,
+      "step": 4752
+    },
+    {
+      "epoch": 5.50952691443889,
+      "grad_norm": 0.38609573245048523,
+      "learning_rate": 2.2683908045977013e-05,
+      "loss": 0.0961,
+      "step": 4753
+    },
+    {
+      "epoch": 5.510686082735637,
+      "grad_norm": 0.37133869528770447,
+      "learning_rate": 2.267816091954023e-05,
+      "loss": 0.0916,
+      "step": 4754
+    },
+    {
+      "epoch": 5.511845251032384,
+      "grad_norm": 0.34857529401779175,
+      "learning_rate": 2.267241379310345e-05,
+      "loss": 0.0952,
+      "step": 4755
+    },
+    {
+      "epoch": 5.513004419329131,
+      "grad_norm": 0.3999149799346924,
+      "learning_rate": 2.2666666666666668e-05,
+      "loss": 0.1025,
+      "step": 4756
+    },
+    {
+      "epoch": 5.5141635876258785,
+      "grad_norm": 0.38613250851631165,
+      "learning_rate": 2.2660919540229886e-05,
+      "loss": 0.0912,
+      "step": 4757
+    },
+    {
+      "epoch": 5.515322755922625,
+      "grad_norm": 0.4043678939342499,
+      "learning_rate": 2.2655172413793104e-05,
+      "loss": 0.0991,
+      "step": 4758
+    },
+    {
+      "epoch": 5.516481924219373,
+      "grad_norm": 0.460321843624115,
+      "learning_rate": 2.2649425287356322e-05,
+      "loss": 0.0861,
+      "step": 4759
+    },
+    {
+      "epoch": 5.5176410925161194,
+      "grad_norm": 0.5290979146957397,
+      "learning_rate": 2.264367816091954e-05,
+      "loss": 0.0946,
+      "step": 4760
+    },
+    {
+      "epoch": 5.518800260812867,
+      "grad_norm": 0.4141453802585602,
+      "learning_rate": 2.263793103448276e-05,
+      "loss": 0.0918,
+      "step": 4761
+    },
+    {
+      "epoch": 5.519959429109614,
+      "grad_norm": 0.5237475633621216,
+      "learning_rate": 2.2632183908045977e-05,
+      "loss": 0.0966,
+      "step": 4762
+    },
+    {
+      "epoch": 5.521118597406361,
+      "grad_norm": 0.392376571893692,
+      "learning_rate": 2.26264367816092e-05,
+      "loss": 0.1033,
+      "step": 4763
+    },
+    {
+      "epoch": 5.522277765703108,
+      "grad_norm": 0.43914276361465454,
+      "learning_rate": 2.2620689655172414e-05,
+      "loss": 0.09,
+      "step": 4764
+    },
+    {
+      "epoch": 5.5234369339998555,
+      "grad_norm": 0.5471196174621582,
+      "learning_rate": 2.2614942528735632e-05,
+      "loss": 0.0959,
+      "step": 4765
+    },
+    {
+      "epoch": 5.524596102296602,
+      "grad_norm": 0.47647860646247864,
+      "learning_rate": 2.2609195402298854e-05,
+      "loss": 0.0975,
+      "step": 4766
+    },
+    {
+      "epoch": 5.525755270593349,
+      "grad_norm": 0.5217294096946716,
+      "learning_rate": 2.2603448275862072e-05,
+      "loss": 0.1029,
+      "step": 4767
+    },
+    {
+      "epoch": 5.526914438890096,
+      "grad_norm": 0.3615458309650421,
+      "learning_rate": 2.2597701149425287e-05,
+      "loss": 0.0837,
+      "step": 4768
+    },
+    {
+      "epoch": 5.528073607186844,
+      "grad_norm": 0.3933470547199249,
+      "learning_rate": 2.2591954022988505e-05,
+      "loss": 0.0959,
+      "step": 4769
+    },
+    {
+      "epoch": 5.529232775483591,
+      "grad_norm": 0.32372987270355225,
+      "learning_rate": 2.2586206896551727e-05,
+      "loss": 0.0963,
+      "step": 4770
+    },
+    {
+      "epoch": 5.530391943780337,
+      "grad_norm": 0.3662286102771759,
+      "learning_rate": 2.2580459770114942e-05,
+      "loss": 0.103,
+      "step": 4771
+    },
+    {
+      "epoch": 5.531551112077085,
+      "grad_norm": 0.5479937791824341,
+      "learning_rate": 2.257471264367816e-05,
+      "loss": 0.1051,
+      "step": 4772
+    },
+    {
+      "epoch": 5.5327102803738315,
+      "grad_norm": 0.43857720494270325,
+      "learning_rate": 2.2568965517241382e-05,
+      "loss": 0.0945,
+      "step": 4773
+    },
+    {
+      "epoch": 5.533869448670579,
+      "grad_norm": 0.45895788073539734,
+      "learning_rate": 2.25632183908046e-05,
+      "loss": 0.0961,
+      "step": 4774
+    },
+    {
+      "epoch": 5.535028616967326,
+      "grad_norm": 0.3341083526611328,
+      "learning_rate": 2.2557471264367815e-05,
+      "loss": 0.0899,
+      "step": 4775
+    },
+    {
+      "epoch": 5.536187785264073,
+      "grad_norm": 0.45365577936172485,
+      "learning_rate": 2.2551724137931033e-05,
+      "loss": 0.0899,
+      "step": 4776
+    },
+    {
+      "epoch": 5.53734695356082,
+      "grad_norm": 0.3847201466560364,
+      "learning_rate": 2.2545977011494255e-05,
+      "loss": 0.1055,
+      "step": 4777
+    },
+    {
+      "epoch": 5.5385061218575675,
+      "grad_norm": 0.570408821105957,
+      "learning_rate": 2.2540229885057473e-05,
+      "loss": 0.0844,
+      "step": 4778
+    },
+    {
+      "epoch": 5.539665290154314,
+      "grad_norm": 0.4045121669769287,
+      "learning_rate": 2.2534482758620688e-05,
+      "loss": 0.0953,
+      "step": 4779
+    },
+    {
+      "epoch": 5.540824458451062,
+      "grad_norm": 0.321380615234375,
+      "learning_rate": 2.252873563218391e-05,
+      "loss": 0.0908,
+      "step": 4780
+    },
+    {
+      "epoch": 5.541983626747808,
+      "grad_norm": 0.40816688537597656,
+      "learning_rate": 2.2522988505747128e-05,
+      "loss": 0.0952,
+      "step": 4781
+    },
+    {
+      "epoch": 5.543142795044556,
+      "grad_norm": 0.7444669604301453,
+      "learning_rate": 2.2517241379310347e-05,
+      "loss": 0.1046,
+      "step": 4782
+    },
+    {
+      "epoch": 5.544301963341303,
+      "grad_norm": 0.41582420468330383,
+      "learning_rate": 2.2511494252873565e-05,
+      "loss": 0.094,
+      "step": 4783
+    },
+    {
+      "epoch": 5.545461131638049,
+      "grad_norm": 0.2793687880039215,
+      "learning_rate": 2.2505747126436783e-05,
+      "loss": 0.0919,
+      "step": 4784
+    },
+    {
+      "epoch": 5.546620299934797,
+      "grad_norm": 0.530275821685791,
+      "learning_rate": 2.25e-05,
+      "loss": 0.0953,
+      "step": 4785
+    },
+    {
+      "epoch": 5.5477794682315436,
+      "grad_norm": 0.484336793422699,
+      "learning_rate": 2.249425287356322e-05,
+      "loss": 0.0929,
+      "step": 4786
+    },
+    {
+      "epoch": 5.548938636528291,
+      "grad_norm": 0.4025559723377228,
+      "learning_rate": 2.2488505747126438e-05,
+      "loss": 0.0982,
+      "step": 4787
+    },
+    {
+      "epoch": 5.550097804825038,
+      "grad_norm": 0.639001727104187,
+      "learning_rate": 2.2482758620689656e-05,
+      "loss": 0.0926,
+      "step": 4788
+    },
+    {
+      "epoch": 5.551256973121785,
+      "grad_norm": 0.4068092107772827,
+      "learning_rate": 2.2477011494252875e-05,
+      "loss": 0.0983,
+      "step": 4789
+    },
+    {
+      "epoch": 5.552416141418532,
+      "grad_norm": 0.4974704384803772,
+      "learning_rate": 2.2471264367816093e-05,
+      "loss": 0.0971,
+      "step": 4790
+    },
+    {
+      "epoch": 5.55357530971528,
+      "grad_norm": 0.5406764149665833,
+      "learning_rate": 2.246551724137931e-05,
+      "loss": 0.0927,
+      "step": 4791
+    },
+    {
+      "epoch": 5.554734478012026,
+      "grad_norm": 0.3397584855556488,
+      "learning_rate": 2.245977011494253e-05,
+      "loss": 0.0869,
+      "step": 4792
+    },
+    {
+      "epoch": 5.555893646308774,
+      "grad_norm": 0.3097866475582123,
+      "learning_rate": 2.2454022988505748e-05,
+      "loss": 0.0877,
+      "step": 4793
+    },
+    {
+      "epoch": 5.5570528146055205,
+      "grad_norm": 0.4206595718860626,
+      "learning_rate": 2.2448275862068966e-05,
+      "loss": 0.0911,
+      "step": 4794
+    },
+    {
+      "epoch": 5.558211982902268,
+      "grad_norm": 0.2879236042499542,
+      "learning_rate": 2.2442528735632184e-05,
+      "loss": 0.1,
+      "step": 4795
+    },
+    {
+      "epoch": 5.559371151199015,
+      "grad_norm": 0.3888731300830841,
+      "learning_rate": 2.2436781609195403e-05,
+      "loss": 0.0943,
+      "step": 4796
+    },
+    {
+      "epoch": 5.560530319495761,
+      "grad_norm": 0.3849053382873535,
+      "learning_rate": 2.2431034482758624e-05,
+      "loss": 0.0925,
+      "step": 4797
+    },
+    {
+      "epoch": 5.561689487792509,
+      "grad_norm": 0.464656263589859,
+      "learning_rate": 2.242528735632184e-05,
+      "loss": 0.1005,
+      "step": 4798
+    },
+    {
+      "epoch": 5.562848656089256,
+      "grad_norm": 0.4027196764945984,
+      "learning_rate": 2.2419540229885057e-05,
+      "loss": 0.1009,
+      "step": 4799
+    },
+    {
+      "epoch": 5.564007824386003,
+      "grad_norm": 0.47038623690605164,
+      "learning_rate": 2.2413793103448276e-05,
+      "loss": 0.0972,
+      "step": 4800
+    },
+    {
+      "epoch": 5.56516699268275,
+      "grad_norm": 0.35763296484947205,
+      "learning_rate": 2.2408045977011497e-05,
+      "loss": 0.0836,
+      "step": 4801
+    },
+    {
+      "epoch": 5.566326160979497,
+      "grad_norm": 0.4369265139102936,
+      "learning_rate": 2.2402298850574712e-05,
+      "loss": 0.0853,
+      "step": 4802
+    },
+    {
+      "epoch": 5.567485329276244,
+      "grad_norm": 0.43006211519241333,
+      "learning_rate": 2.239655172413793e-05,
+      "loss": 0.0973,
+      "step": 4803
+    },
+    {
+      "epoch": 5.568644497572992,
+      "grad_norm": 0.4107172191143036,
+      "learning_rate": 2.2390804597701152e-05,
+      "loss": 0.0965,
+      "step": 4804
+    },
+    {
+      "epoch": 5.569803665869738,
+      "grad_norm": 0.3978167176246643,
+      "learning_rate": 2.238505747126437e-05,
+      "loss": 0.0923,
+      "step": 4805
+    },
+    {
+      "epoch": 5.570962834166486,
+      "grad_norm": 0.49095186591148376,
+      "learning_rate": 2.2379310344827586e-05,
+      "loss": 0.0946,
+      "step": 4806
+    },
+    {
+      "epoch": 5.5721220024632325,
+      "grad_norm": 0.32563725113868713,
+      "learning_rate": 2.2373563218390807e-05,
+      "loss": 0.092,
+      "step": 4807
+    },
+    {
+      "epoch": 5.57328117075998,
+      "grad_norm": 0.42100727558135986,
+      "learning_rate": 2.2367816091954025e-05,
+      "loss": 0.0942,
+      "step": 4808
+    },
+    {
+      "epoch": 5.574440339056727,
+      "grad_norm": 0.39792951941490173,
+      "learning_rate": 2.236206896551724e-05,
+      "loss": 0.09,
+      "step": 4809
+    },
+    {
+      "epoch": 5.575599507353473,
+      "grad_norm": 0.5242913365364075,
+      "learning_rate": 2.235632183908046e-05,
+      "loss": 0.0943,
+      "step": 4810
+    },
+    {
+      "epoch": 5.576758675650221,
+      "grad_norm": 0.5006229281425476,
+      "learning_rate": 2.235057471264368e-05,
+      "loss": 0.0987,
+      "step": 4811
+    },
+    {
+      "epoch": 5.5779178439469685,
+      "grad_norm": 0.37062880396842957,
+      "learning_rate": 2.23448275862069e-05,
+      "loss": 0.0929,
+      "step": 4812
+    },
+    {
+      "epoch": 5.579077012243715,
+      "grad_norm": 0.5827232003211975,
+      "learning_rate": 2.2339080459770114e-05,
+      "loss": 0.1032,
+      "step": 4813
+    },
+    {
+      "epoch": 5.580236180540462,
+      "grad_norm": 0.30347374081611633,
+      "learning_rate": 2.2333333333333335e-05,
+      "loss": 0.0965,
+      "step": 4814
+    },
+    {
+      "epoch": 5.5813953488372094,
+      "grad_norm": 0.5079871416091919,
+      "learning_rate": 2.2327586206896554e-05,
+      "loss": 0.0942,
+      "step": 4815
+    },
+    {
+      "epoch": 5.582554517133956,
+      "grad_norm": 0.3671127259731293,
+      "learning_rate": 2.2321839080459772e-05,
+      "loss": 0.0885,
+      "step": 4816
+    },
+    {
+      "epoch": 5.583713685430704,
+      "grad_norm": 0.4037294089794159,
+      "learning_rate": 2.231609195402299e-05,
+      "loss": 0.0965,
+      "step": 4817
+    },
+    {
+      "epoch": 5.58487285372745,
+      "grad_norm": 0.44176700711250305,
+      "learning_rate": 2.231034482758621e-05,
+      "loss": 0.0977,
+      "step": 4818
+    },
+    {
+      "epoch": 5.586032022024198,
+      "grad_norm": 0.5387149453163147,
+      "learning_rate": 2.2304597701149427e-05,
+      "loss": 0.0961,
+      "step": 4819
+    },
+    {
+      "epoch": 5.587191190320945,
+      "grad_norm": 0.4820229411125183,
+      "learning_rate": 2.2298850574712645e-05,
+      "loss": 0.1048,
+      "step": 4820
+    },
+    {
+      "epoch": 5.588350358617692,
+      "grad_norm": 0.46356868743896484,
+      "learning_rate": 2.2293103448275863e-05,
+      "loss": 0.0917,
+      "step": 4821
+    },
+    {
+      "epoch": 5.589509526914439,
+      "grad_norm": 0.3790029287338257,
+      "learning_rate": 2.228735632183908e-05,
+      "loss": 0.0897,
+      "step": 4822
+    },
+    {
+      "epoch": 5.5906686952111855,
+      "grad_norm": 0.4080239534378052,
+      "learning_rate": 2.22816091954023e-05,
+      "loss": 0.0932,
+      "step": 4823
+    },
+    {
+      "epoch": 5.591827863507933,
+      "grad_norm": 0.5119190216064453,
+      "learning_rate": 2.2275862068965518e-05,
+      "loss": 0.099,
+      "step": 4824
+    },
+    {
+      "epoch": 5.592987031804681,
+      "grad_norm": 0.40316665172576904,
+      "learning_rate": 2.2270114942528736e-05,
+      "loss": 0.1001,
+      "step": 4825
+    },
+    {
+      "epoch": 5.594146200101427,
+      "grad_norm": 0.3797335922718048,
+      "learning_rate": 2.2264367816091955e-05,
+      "loss": 0.0963,
+      "step": 4826
+    },
+    {
+      "epoch": 5.595305368398174,
+      "grad_norm": 0.4197828471660614,
+      "learning_rate": 2.2258620689655173e-05,
+      "loss": 0.0921,
+      "step": 4827
+    },
+    {
+      "epoch": 5.5964645366949215,
+      "grad_norm": 0.41356906294822693,
+      "learning_rate": 2.2252873563218395e-05,
+      "loss": 0.0948,
+      "step": 4828
+    },
+    {
+      "epoch": 5.597623704991668,
+      "grad_norm": 0.5795599818229675,
+      "learning_rate": 2.224712643678161e-05,
+      "loss": 0.0966,
+      "step": 4829
+    },
+    {
+      "epoch": 5.598782873288416,
+      "grad_norm": 0.3532811999320984,
+      "learning_rate": 2.2241379310344828e-05,
+      "loss": 0.0926,
+      "step": 4830
+    },
+    {
+      "epoch": 5.599942041585162,
+      "grad_norm": 0.32101044058799744,
+      "learning_rate": 2.2235632183908046e-05,
+      "loss": 0.0955,
+      "step": 4831
+    },
+    {
+      "epoch": 5.60110120988191,
+      "grad_norm": 0.3435097634792328,
+      "learning_rate": 2.2229885057471264e-05,
+      "loss": 0.1026,
+      "step": 4832
+    },
+    {
+      "epoch": 5.602260378178657,
+      "grad_norm": 0.42557913064956665,
+      "learning_rate": 2.2224137931034483e-05,
+      "loss": 0.0857,
+      "step": 4833
+    },
+    {
+      "epoch": 5.603419546475404,
+      "grad_norm": 0.36306434869766235,
+      "learning_rate": 2.22183908045977e-05,
+      "loss": 0.0925,
+      "step": 4834
+    },
+    {
+      "epoch": 5.604578714772151,
+      "grad_norm": 0.45969343185424805,
+      "learning_rate": 2.2212643678160923e-05,
+      "loss": 0.0987,
+      "step": 4835
+    },
+    {
+      "epoch": 5.605737883068898,
+      "grad_norm": 0.39553216099739075,
+      "learning_rate": 2.2206896551724138e-05,
+      "loss": 0.0932,
+      "step": 4836
+    },
+    {
+      "epoch": 5.606897051365645,
+      "grad_norm": 0.6770229339599609,
+      "learning_rate": 2.2201149425287356e-05,
+      "loss": 0.0989,
+      "step": 4837
+    },
+    {
+      "epoch": 5.608056219662393,
+      "grad_norm": 0.5006446838378906,
+      "learning_rate": 2.2195402298850578e-05,
+      "loss": 0.0959,
+      "step": 4838
+    },
+    {
+      "epoch": 5.609215387959139,
+      "grad_norm": 0.33049294352531433,
+      "learning_rate": 2.2189655172413796e-05,
+      "loss": 0.0861,
+      "step": 4839
+    },
+    {
+      "epoch": 5.610374556255886,
+      "grad_norm": 0.5273866653442383,
+      "learning_rate": 2.218390804597701e-05,
+      "loss": 0.1108,
+      "step": 4840
+    },
+    {
+      "epoch": 5.6115337245526336,
+      "grad_norm": 0.45553502440452576,
+      "learning_rate": 2.217816091954023e-05,
+      "loss": 0.0959,
+      "step": 4841
+    },
+    {
+      "epoch": 5.61269289284938,
+      "grad_norm": 0.3138861358165741,
+      "learning_rate": 2.217241379310345e-05,
+      "loss": 0.086,
+      "step": 4842
+    },
+    {
+      "epoch": 5.613852061146128,
+      "grad_norm": 0.5423036217689514,
+      "learning_rate": 2.216666666666667e-05,
+      "loss": 0.102,
+      "step": 4843
+    },
+    {
+      "epoch": 5.6150112294428745,
+      "grad_norm": 0.5598729848861694,
+      "learning_rate": 2.2160919540229884e-05,
+      "loss": 0.0991,
+      "step": 4844
+    },
+    {
+      "epoch": 5.616170397739622,
+      "grad_norm": 0.5099043846130371,
+      "learning_rate": 2.2155172413793106e-05,
+      "loss": 0.0908,
+      "step": 4845
+    },
+    {
+      "epoch": 5.617329566036369,
+      "grad_norm": 0.40814903378486633,
+      "learning_rate": 2.2149425287356324e-05,
+      "loss": 0.0911,
+      "step": 4846
+    },
+    {
+      "epoch": 5.618488734333116,
+      "grad_norm": 0.3231344223022461,
+      "learning_rate": 2.2143678160919542e-05,
+      "loss": 0.0867,
+      "step": 4847
+    },
+    {
+      "epoch": 5.619647902629863,
+      "grad_norm": 0.3704584538936615,
+      "learning_rate": 2.213793103448276e-05,
+      "loss": 0.0857,
+      "step": 4848
+    },
+    {
+      "epoch": 5.6208070709266105,
+      "grad_norm": 0.49821195006370544,
+      "learning_rate": 2.213218390804598e-05,
+      "loss": 0.1002,
+      "step": 4849
+    },
+    {
+      "epoch": 5.621966239223357,
+      "grad_norm": 0.35286957025527954,
+      "learning_rate": 2.2126436781609197e-05,
+      "loss": 0.0895,
+      "step": 4850
+    },
+    {
+      "epoch": 5.623125407520105,
+      "grad_norm": 0.46417170763015747,
+      "learning_rate": 2.2120689655172412e-05,
+      "loss": 0.0919,
+      "step": 4851
+    },
+    {
+      "epoch": 5.624284575816851,
+      "grad_norm": 0.4302671253681183,
+      "learning_rate": 2.2114942528735634e-05,
+      "loss": 0.0954,
+      "step": 4852
+    },
+    {
+      "epoch": 5.625443744113598,
+      "grad_norm": 0.3605821132659912,
+      "learning_rate": 2.2109195402298852e-05,
+      "loss": 0.0913,
+      "step": 4853
+    },
+    {
+      "epoch": 5.626602912410346,
+      "grad_norm": 0.4404606223106384,
+      "learning_rate": 2.210344827586207e-05,
+      "loss": 0.0899,
+      "step": 4854
+    },
+    {
+      "epoch": 5.627762080707092,
+      "grad_norm": 0.4587152898311615,
+      "learning_rate": 2.209770114942529e-05,
+      "loss": 0.0897,
+      "step": 4855
+    },
+    {
+      "epoch": 5.62892124900384,
+      "grad_norm": 0.4051382839679718,
+      "learning_rate": 2.2091954022988507e-05,
+      "loss": 0.1017,
+      "step": 4856
+    },
+    {
+      "epoch": 5.6300804173005865,
+      "grad_norm": 0.4511359930038452,
+      "learning_rate": 2.2086206896551725e-05,
+      "loss": 0.0966,
+      "step": 4857
+    },
+    {
+      "epoch": 5.631239585597334,
+      "grad_norm": 0.43057432770729065,
+      "learning_rate": 2.2080459770114943e-05,
+      "loss": 0.1004,
+      "step": 4858
+    },
+    {
+      "epoch": 5.632398753894081,
+      "grad_norm": 0.41975724697113037,
+      "learning_rate": 2.207471264367816e-05,
+      "loss": 0.1025,
+      "step": 4859
+    },
+    {
+      "epoch": 5.633557922190828,
+      "grad_norm": 0.4029335677623749,
+      "learning_rate": 2.206896551724138e-05,
+      "loss": 0.0968,
+      "step": 4860
+    },
+    {
+      "epoch": 5.634717090487575,
+      "grad_norm": 0.296252965927124,
+      "learning_rate": 2.2063218390804598e-05,
+      "loss": 0.0929,
+      "step": 4861
+    },
+    {
+      "epoch": 5.6358762587843225,
+      "grad_norm": 0.3576434552669525,
+      "learning_rate": 2.205747126436782e-05,
+      "loss": 0.0906,
+      "step": 4862
+    },
+    {
+      "epoch": 5.637035427081069,
+      "grad_norm": 0.37971681356430054,
+      "learning_rate": 2.2051724137931035e-05,
+      "loss": 0.0954,
+      "step": 4863
+    },
+    {
+      "epoch": 5.638194595377817,
+      "grad_norm": 0.4097626805305481,
+      "learning_rate": 2.2045977011494253e-05,
+      "loss": 0.0987,
+      "step": 4864
+    },
+    {
+      "epoch": 5.639353763674563,
+      "grad_norm": 0.34203460812568665,
+      "learning_rate": 2.204022988505747e-05,
+      "loss": 0.0954,
+      "step": 4865
+    },
+    {
+      "epoch": 5.64051293197131,
+      "grad_norm": 0.4253942370414734,
+      "learning_rate": 2.2034482758620693e-05,
+      "loss": 0.0997,
+      "step": 4866
+    },
+    {
+      "epoch": 5.641672100268058,
+      "grad_norm": 0.5293704271316528,
+      "learning_rate": 2.2028735632183908e-05,
+      "loss": 0.0933,
+      "step": 4867
+    },
+    {
+      "epoch": 5.642831268564805,
+      "grad_norm": 0.36057615280151367,
+      "learning_rate": 2.2022988505747126e-05,
+      "loss": 0.0883,
+      "step": 4868
+    },
+    {
+      "epoch": 5.643990436861552,
+      "grad_norm": 0.44225749373435974,
+      "learning_rate": 2.2017241379310348e-05,
+      "loss": 0.0927,
+      "step": 4869
+    },
+    {
+      "epoch": 5.6451496051582986,
+      "grad_norm": 0.34227925539016724,
+      "learning_rate": 2.2011494252873563e-05,
+      "loss": 0.0883,
+      "step": 4870
+    },
+    {
+      "epoch": 5.646308773455046,
+      "grad_norm": 0.44135892391204834,
+      "learning_rate": 2.200574712643678e-05,
+      "loss": 0.094,
+      "step": 4871
+    },
+    {
+      "epoch": 5.647467941751793,
+      "grad_norm": 0.5417119860649109,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 0.0943,
+      "step": 4872
+    },
+    {
+      "epoch": 5.64862711004854,
+      "grad_norm": 0.47763824462890625,
+      "learning_rate": 2.199425287356322e-05,
+      "loss": 0.1042,
+      "step": 4873
+    },
+    {
+      "epoch": 5.649786278345287,
+      "grad_norm": 0.4434187710285187,
+      "learning_rate": 2.1988505747126436e-05,
+      "loss": 0.098,
+      "step": 4874
+    },
+    {
+      "epoch": 5.650945446642035,
+      "grad_norm": 0.466353178024292,
+      "learning_rate": 2.1982758620689654e-05,
+      "loss": 0.109,
+      "step": 4875
+    },
+    {
+      "epoch": 5.652104614938781,
+      "grad_norm": 0.5113919377326965,
+      "learning_rate": 2.1977011494252876e-05,
+      "loss": 0.0935,
+      "step": 4876
+    },
+    {
+      "epoch": 5.653263783235529,
+      "grad_norm": 0.43153828382492065,
+      "learning_rate": 2.1971264367816094e-05,
+      "loss": 0.1003,
+      "step": 4877
+    },
+    {
+      "epoch": 5.6544229515322755,
+      "grad_norm": 0.40823087096214294,
+      "learning_rate": 2.196551724137931e-05,
+      "loss": 0.0954,
+      "step": 4878
+    },
+    {
+      "epoch": 5.655582119829023,
+      "grad_norm": 0.4650181531906128,
+      "learning_rate": 2.195977011494253e-05,
+      "loss": 0.0951,
+      "step": 4879
+    },
+    {
+      "epoch": 5.65674128812577,
+      "grad_norm": 0.41202208399772644,
+      "learning_rate": 2.195402298850575e-05,
+      "loss": 0.1001,
+      "step": 4880
+    },
+    {
+      "epoch": 5.657900456422517,
+      "grad_norm": 0.5160025954246521,
+      "learning_rate": 2.1948275862068967e-05,
+      "loss": 0.0914,
+      "step": 4881
+    },
+    {
+      "epoch": 5.659059624719264,
+      "grad_norm": 0.5344350337982178,
+      "learning_rate": 2.1942528735632186e-05,
+      "loss": 0.0995,
+      "step": 4882
+    },
+    {
+      "epoch": 5.660218793016011,
+      "grad_norm": 1.5580607652664185,
+      "learning_rate": 2.1936781609195404e-05,
+      "loss": 0.1032,
+      "step": 4883
+    },
+    {
+      "epoch": 5.661377961312758,
+      "grad_norm": 0.43727943301200867,
+      "learning_rate": 2.1931034482758622e-05,
+      "loss": 0.1002,
+      "step": 4884
+    },
+    {
+      "epoch": 5.662537129609505,
+      "grad_norm": 0.35894694924354553,
+      "learning_rate": 2.192528735632184e-05,
+      "loss": 0.1022,
+      "step": 4885
+    },
+    {
+      "epoch": 5.663696297906252,
+      "grad_norm": 0.46932268142700195,
+      "learning_rate": 2.191954022988506e-05,
+      "loss": 0.0913,
+      "step": 4886
+    },
+    {
+      "epoch": 5.664855466202999,
+      "grad_norm": 0.5814934372901917,
+      "learning_rate": 2.1913793103448277e-05,
+      "loss": 0.1021,
+      "step": 4887
+    },
+    {
+      "epoch": 5.666014634499747,
+      "grad_norm": 0.41701653599739075,
+      "learning_rate": 2.1908045977011495e-05,
+      "loss": 0.0984,
+      "step": 4888
+    },
+    {
+      "epoch": 5.667173802796493,
+      "grad_norm": 0.564414381980896,
+      "learning_rate": 2.1902298850574714e-05,
+      "loss": 0.1031,
+      "step": 4889
+    },
+    {
+      "epoch": 5.668332971093241,
+      "grad_norm": 0.40304380655288696,
+      "learning_rate": 2.1896551724137932e-05,
+      "loss": 0.0933,
+      "step": 4890
+    },
+    {
+      "epoch": 5.6694921393899875,
+      "grad_norm": 0.40032997727394104,
+      "learning_rate": 2.189080459770115e-05,
+      "loss": 0.0899,
+      "step": 4891
+    },
+    {
+      "epoch": 5.670651307686735,
+      "grad_norm": 0.48019522428512573,
+      "learning_rate": 2.188505747126437e-05,
+      "loss": 0.0989,
+      "step": 4892
+    },
+    {
+      "epoch": 5.671810475983482,
+      "grad_norm": 0.5299116969108582,
+      "learning_rate": 2.1879310344827587e-05,
+      "loss": 0.1001,
+      "step": 4893
+    },
+    {
+      "epoch": 5.672969644280229,
+      "grad_norm": 0.3729042112827301,
+      "learning_rate": 2.1873563218390805e-05,
+      "loss": 0.0886,
+      "step": 4894
+    },
+    {
+      "epoch": 5.674128812576976,
+      "grad_norm": 0.4367959201335907,
+      "learning_rate": 2.1867816091954023e-05,
+      "loss": 0.0877,
+      "step": 4895
+    },
+    {
+      "epoch": 5.675287980873723,
+      "grad_norm": 0.36830776929855347,
+      "learning_rate": 2.1862068965517242e-05,
+      "loss": 0.0926,
+      "step": 4896
+    },
+    {
+      "epoch": 5.67644714917047,
+      "grad_norm": 0.37894847989082336,
+      "learning_rate": 2.185632183908046e-05,
+      "loss": 0.0922,
+      "step": 4897
+    },
+    {
+      "epoch": 5.677606317467217,
+      "grad_norm": 0.4589332640171051,
+      "learning_rate": 2.185057471264368e-05,
+      "loss": 0.0984,
+      "step": 4898
+    },
+    {
+      "epoch": 5.6787654857639644,
+      "grad_norm": 0.3483442962169647,
+      "learning_rate": 2.1844827586206897e-05,
+      "loss": 0.0949,
+      "step": 4899
+    },
+    {
+      "epoch": 5.679924654060711,
+      "grad_norm": 0.4337550103664398,
+      "learning_rate": 2.183908045977012e-05,
+      "loss": 0.1002,
+      "step": 4900
+    },
+    {
+      "epoch": 5.681083822357459,
+      "grad_norm": 0.42401912808418274,
+      "learning_rate": 2.1833333333333333e-05,
+      "loss": 0.087,
+      "step": 4901
+    },
+    {
+      "epoch": 5.682242990654205,
+      "grad_norm": 0.40039029717445374,
+      "learning_rate": 2.182758620689655e-05,
+      "loss": 0.1024,
+      "step": 4902
+    },
+    {
+      "epoch": 5.683402158950953,
+      "grad_norm": 0.3798970878124237,
+      "learning_rate": 2.1821839080459773e-05,
+      "loss": 0.0913,
+      "step": 4903
+    },
+    {
+      "epoch": 5.6845613272477,
+      "grad_norm": 0.41023609042167664,
+      "learning_rate": 2.181609195402299e-05,
+      "loss": 0.0864,
+      "step": 4904
+    },
+    {
+      "epoch": 5.685720495544447,
+      "grad_norm": 0.5003272891044617,
+      "learning_rate": 2.1810344827586206e-05,
+      "loss": 0.096,
+      "step": 4905
+    },
+    {
+      "epoch": 5.686879663841194,
+      "grad_norm": 0.673970103263855,
+      "learning_rate": 2.1804597701149425e-05,
+      "loss": 0.0967,
+      "step": 4906
+    },
+    {
+      "epoch": 5.688038832137941,
+      "grad_norm": 0.4125644564628601,
+      "learning_rate": 2.1798850574712646e-05,
+      "loss": 0.097,
+      "step": 4907
+    },
+    {
+      "epoch": 5.689198000434688,
+      "grad_norm": 0.4496941566467285,
+      "learning_rate": 2.1793103448275865e-05,
+      "loss": 0.0917,
+      "step": 4908
+    },
+    {
+      "epoch": 5.690357168731435,
+      "grad_norm": 0.37504181265830994,
+      "learning_rate": 2.178735632183908e-05,
+      "loss": 0.0965,
+      "step": 4909
+    },
+    {
+      "epoch": 5.691516337028182,
+      "grad_norm": 0.4096089005470276,
+      "learning_rate": 2.17816091954023e-05,
+      "loss": 0.0958,
+      "step": 4910
+    },
+    {
+      "epoch": 5.69267550532493,
+      "grad_norm": 0.43474438786506653,
+      "learning_rate": 2.177586206896552e-05,
+      "loss": 0.0888,
+      "step": 4911
+    },
+    {
+      "epoch": 5.6938346736216765,
+      "grad_norm": 0.39973193407058716,
+      "learning_rate": 2.1770114942528734e-05,
+      "loss": 0.0946,
+      "step": 4912
+    },
+    {
+      "epoch": 5.694993841918423,
+      "grad_norm": 0.5031002163887024,
+      "learning_rate": 2.1764367816091956e-05,
+      "loss": 0.0978,
+      "step": 4913
+    },
+    {
+      "epoch": 5.696153010215171,
+      "grad_norm": 0.5082476139068604,
+      "learning_rate": 2.1758620689655174e-05,
+      "loss": 0.0917,
+      "step": 4914
+    },
+    {
+      "epoch": 5.697312178511917,
+      "grad_norm": 0.48969969153404236,
+      "learning_rate": 2.1752873563218393e-05,
+      "loss": 0.1036,
+      "step": 4915
+    },
+    {
+      "epoch": 5.698471346808665,
+      "grad_norm": 0.41776567697525024,
+      "learning_rate": 2.1747126436781608e-05,
+      "loss": 0.0936,
+      "step": 4916
+    },
+    {
+      "epoch": 5.699630515105412,
+      "grad_norm": 0.43276071548461914,
+      "learning_rate": 2.174137931034483e-05,
+      "loss": 0.0969,
+      "step": 4917
+    },
+    {
+      "epoch": 5.700789683402159,
+      "grad_norm": 0.44773709774017334,
+      "learning_rate": 2.1735632183908048e-05,
+      "loss": 0.0923,
+      "step": 4918
+    },
+    {
+      "epoch": 5.701948851698906,
+      "grad_norm": 0.3338507115840912,
+      "learning_rate": 2.1729885057471266e-05,
+      "loss": 0.0912,
+      "step": 4919
+    },
+    {
+      "epoch": 5.703108019995653,
+      "grad_norm": 0.6209862232208252,
+      "learning_rate": 2.1724137931034484e-05,
+      "loss": 0.096,
+      "step": 4920
+    },
+    {
+      "epoch": 5.7042671882924,
+      "grad_norm": 0.482950896024704,
+      "learning_rate": 2.1718390804597702e-05,
+      "loss": 0.0981,
+      "step": 4921
+    },
+    {
+      "epoch": 5.705426356589148,
+      "grad_norm": 0.4542859196662903,
+      "learning_rate": 2.171264367816092e-05,
+      "loss": 0.0949,
+      "step": 4922
+    },
+    {
+      "epoch": 5.706585524885894,
+      "grad_norm": 0.4882400333881378,
+      "learning_rate": 2.170689655172414e-05,
+      "loss": 0.0961,
+      "step": 4923
+    },
+    {
+      "epoch": 5.707744693182642,
+      "grad_norm": 0.4771440923213959,
+      "learning_rate": 2.1701149425287357e-05,
+      "loss": 0.1034,
+      "step": 4924
+    },
+    {
+      "epoch": 5.7089038614793886,
+      "grad_norm": 0.5598927140235901,
+      "learning_rate": 2.1695402298850576e-05,
+      "loss": 0.0996,
+      "step": 4925
+    },
+    {
+      "epoch": 5.710063029776135,
+      "grad_norm": 0.45021867752075195,
+      "learning_rate": 2.1689655172413794e-05,
+      "loss": 0.1006,
+      "step": 4926
+    },
+    {
+      "epoch": 5.711222198072883,
+      "grad_norm": 0.38733410835266113,
+      "learning_rate": 2.1683908045977016e-05,
+      "loss": 0.0896,
+      "step": 4927
+    },
+    {
+      "epoch": 5.7123813663696295,
+      "grad_norm": 0.3161242604255676,
+      "learning_rate": 2.167816091954023e-05,
+      "loss": 0.0883,
+      "step": 4928
+    },
+    {
+      "epoch": 5.713540534666377,
+      "grad_norm": 0.43492892384529114,
+      "learning_rate": 2.167241379310345e-05,
+      "loss": 0.0994,
+      "step": 4929
+    },
+    {
+      "epoch": 5.714699702963124,
+      "grad_norm": 0.34343111515045166,
+      "learning_rate": 2.1666666666666667e-05,
+      "loss": 0.0947,
+      "step": 4930
+    },
+    {
+      "epoch": 5.715858871259871,
+      "grad_norm": 1.0595885515213013,
+      "learning_rate": 2.1660919540229885e-05,
+      "loss": 0.1021,
+      "step": 4931
+    },
+    {
+      "epoch": 5.717018039556618,
+      "grad_norm": 0.4083106815814972,
+      "learning_rate": 2.1655172413793104e-05,
+      "loss": 0.0991,
+      "step": 4932
+    },
+    {
+      "epoch": 5.7181772078533655,
+      "grad_norm": 0.3929539918899536,
+      "learning_rate": 2.1649425287356322e-05,
+      "loss": 0.0956,
+      "step": 4933
+    },
+    {
+      "epoch": 5.719336376150112,
+      "grad_norm": 0.3941984176635742,
+      "learning_rate": 2.1643678160919544e-05,
+      "loss": 0.1015,
+      "step": 4934
+    },
+    {
+      "epoch": 5.72049554444686,
+      "grad_norm": 0.5016578435897827,
+      "learning_rate": 2.163793103448276e-05,
+      "loss": 0.1034,
+      "step": 4935
+    },
+    {
+      "epoch": 5.721654712743606,
+      "grad_norm": 0.41408535838127136,
+      "learning_rate": 2.1632183908045977e-05,
+      "loss": 0.0969,
+      "step": 4936
+    },
+    {
+      "epoch": 5.722813881040354,
+      "grad_norm": 0.3475428521633148,
+      "learning_rate": 2.1626436781609195e-05,
+      "loss": 0.0932,
+      "step": 4937
+    },
+    {
+      "epoch": 5.723973049337101,
+      "grad_norm": 0.315900057554245,
+      "learning_rate": 2.1620689655172417e-05,
+      "loss": 0.0907,
+      "step": 4938
+    },
+    {
+      "epoch": 5.725132217633847,
+      "grad_norm": 0.5354477763175964,
+      "learning_rate": 2.161494252873563e-05,
+      "loss": 0.0927,
+      "step": 4939
+    },
+    {
+      "epoch": 5.726291385930595,
+      "grad_norm": 0.40579307079315186,
+      "learning_rate": 2.160919540229885e-05,
+      "loss": 0.1103,
+      "step": 4940
+    },
+    {
+      "epoch": 5.7274505542273415,
+      "grad_norm": 0.3445356786251068,
+      "learning_rate": 2.160344827586207e-05,
+      "loss": 0.1003,
+      "step": 4941
+    },
+    {
+      "epoch": 5.728609722524089,
+      "grad_norm": 0.5686594247817993,
+      "learning_rate": 2.159770114942529e-05,
+      "loss": 0.1108,
+      "step": 4942
+    },
+    {
+      "epoch": 5.729768890820836,
+      "grad_norm": 0.47493574023246765,
+      "learning_rate": 2.1591954022988505e-05,
+      "loss": 0.1049,
+      "step": 4943
+    },
+    {
+      "epoch": 5.730928059117583,
+      "grad_norm": 0.3111097514629364,
+      "learning_rate": 2.1586206896551726e-05,
+      "loss": 0.0867,
+      "step": 4944
+    },
+    {
+      "epoch": 5.73208722741433,
+      "grad_norm": 0.4092866778373718,
+      "learning_rate": 2.1580459770114945e-05,
+      "loss": 0.0945,
+      "step": 4945
+    },
+    {
+      "epoch": 5.7332463957110775,
+      "grad_norm": 0.4263165593147278,
+      "learning_rate": 2.1574712643678163e-05,
+      "loss": 0.0818,
+      "step": 4946
+    },
+    {
+      "epoch": 5.734405564007824,
+      "grad_norm": 0.3667573928833008,
+      "learning_rate": 2.1568965517241378e-05,
+      "loss": 0.0882,
+      "step": 4947
+    },
+    {
+      "epoch": 5.735564732304572,
+      "grad_norm": 0.505990207195282,
+      "learning_rate": 2.15632183908046e-05,
+      "loss": 0.0843,
+      "step": 4948
+    },
+    {
+      "epoch": 5.736723900601318,
+      "grad_norm": 0.4376978278160095,
+      "learning_rate": 2.1557471264367818e-05,
+      "loss": 0.0911,
+      "step": 4949
+    },
+    {
+      "epoch": 5.737883068898066,
+      "grad_norm": 0.4979074001312256,
+      "learning_rate": 2.1551724137931033e-05,
+      "loss": 0.1033,
+      "step": 4950
+    },
+    {
+      "epoch": 5.739042237194813,
+      "grad_norm": 0.48788338899612427,
+      "learning_rate": 2.1545977011494255e-05,
+      "loss": 0.1081,
+      "step": 4951
+    },
+    {
+      "epoch": 5.740201405491559,
+      "grad_norm": 0.6209906935691833,
+      "learning_rate": 2.1540229885057473e-05,
+      "loss": 0.1015,
+      "step": 4952
+    },
+    {
+      "epoch": 5.741360573788307,
+      "grad_norm": 0.40411826968193054,
+      "learning_rate": 2.153448275862069e-05,
+      "loss": 0.0942,
+      "step": 4953
+    },
+    {
+      "epoch": 5.7425197420850544,
+      "grad_norm": 0.4680006206035614,
+      "learning_rate": 2.152873563218391e-05,
+      "loss": 0.0898,
+      "step": 4954
+    },
+    {
+      "epoch": 5.743678910381801,
+      "grad_norm": 0.4377489984035492,
+      "learning_rate": 2.1522988505747128e-05,
+      "loss": 0.0927,
+      "step": 4955
+    },
+    {
+      "epoch": 5.744838078678548,
+      "grad_norm": 0.6448618173599243,
+      "learning_rate": 2.1517241379310346e-05,
+      "loss": 0.0984,
+      "step": 4956
+    },
+    {
+      "epoch": 5.745997246975295,
+      "grad_norm": 0.44574832916259766,
+      "learning_rate": 2.1511494252873564e-05,
+      "loss": 0.1009,
+      "step": 4957
+    },
+    {
+      "epoch": 5.747156415272042,
+      "grad_norm": 0.5613508224487305,
+      "learning_rate": 2.1505747126436783e-05,
+      "loss": 0.0987,
+      "step": 4958
+    },
+    {
+      "epoch": 5.74831558356879,
+      "grad_norm": 0.5421482920646667,
+      "learning_rate": 2.15e-05,
+      "loss": 0.0975,
+      "step": 4959
+    },
+    {
+      "epoch": 5.749474751865536,
+      "grad_norm": 0.4239422082901001,
+      "learning_rate": 2.149425287356322e-05,
+      "loss": 0.0974,
+      "step": 4960
+    },
+    {
+      "epoch": 5.750633920162284,
+      "grad_norm": 0.7376547455787659,
+      "learning_rate": 2.1488505747126437e-05,
+      "loss": 0.0881,
+      "step": 4961
+    },
+    {
+      "epoch": 5.7517930884590305,
+      "grad_norm": 0.449830561876297,
+      "learning_rate": 2.1482758620689656e-05,
+      "loss": 0.098,
+      "step": 4962
+    },
+    {
+      "epoch": 5.752952256755778,
+      "grad_norm": 0.6881018280982971,
+      "learning_rate": 2.1477011494252874e-05,
+      "loss": 0.1013,
+      "step": 4963
+    },
+    {
+      "epoch": 5.754111425052525,
+      "grad_norm": 0.4860606789588928,
+      "learning_rate": 2.1471264367816092e-05,
+      "loss": 0.0979,
+      "step": 4964
+    },
+    {
+      "epoch": 5.755270593349272,
+      "grad_norm": 0.3773813247680664,
+      "learning_rate": 2.1465517241379314e-05,
+      "loss": 0.0878,
+      "step": 4965
+    },
+    {
+      "epoch": 5.756429761646019,
+      "grad_norm": 0.3880038857460022,
+      "learning_rate": 2.145977011494253e-05,
+      "loss": 0.0956,
+      "step": 4966
+    },
+    {
+      "epoch": 5.7575889299427665,
+      "grad_norm": 0.6223691701889038,
+      "learning_rate": 2.1454022988505747e-05,
+      "loss": 0.1005,
+      "step": 4967
+    },
+    {
+      "epoch": 5.758748098239513,
+      "grad_norm": 0.3865233361721039,
+      "learning_rate": 2.144827586206897e-05,
+      "loss": 0.1,
+      "step": 4968
+    },
+    {
+      "epoch": 5.75990726653626,
+      "grad_norm": 0.426666259765625,
+      "learning_rate": 2.1442528735632184e-05,
+      "loss": 0.0984,
+      "step": 4969
+    },
+    {
+      "epoch": 5.761066434833007,
+      "grad_norm": 0.3938763439655304,
+      "learning_rate": 2.1436781609195402e-05,
+      "loss": 0.096,
+      "step": 4970
+    },
+    {
+      "epoch": 5.762225603129754,
+      "grad_norm": 0.4771715998649597,
+      "learning_rate": 2.143103448275862e-05,
+      "loss": 0.0964,
+      "step": 4971
+    },
+    {
+      "epoch": 5.763384771426502,
+      "grad_norm": 0.3489798307418823,
+      "learning_rate": 2.1425287356321842e-05,
+      "loss": 0.0877,
+      "step": 4972
+    },
+    {
+      "epoch": 5.764543939723248,
+      "grad_norm": 0.36042097210884094,
+      "learning_rate": 2.1419540229885057e-05,
+      "loss": 0.0889,
+      "step": 4973
+    },
+    {
+      "epoch": 5.765703108019996,
+      "grad_norm": 0.41603612899780273,
+      "learning_rate": 2.1413793103448275e-05,
+      "loss": 0.095,
+      "step": 4974
+    },
+    {
+      "epoch": 5.7668622763167425,
+      "grad_norm": 0.4094390273094177,
+      "learning_rate": 2.1408045977011497e-05,
+      "loss": 0.0955,
+      "step": 4975
+    },
+    {
+      "epoch": 5.76802144461349,
+      "grad_norm": 0.4778774678707123,
+      "learning_rate": 2.1402298850574715e-05,
+      "loss": 0.1005,
+      "step": 4976
+    },
+    {
+      "epoch": 5.769180612910237,
+      "grad_norm": 0.4297574758529663,
+      "learning_rate": 2.139655172413793e-05,
+      "loss": 0.0954,
+      "step": 4977
+    },
+    {
+      "epoch": 5.770339781206984,
+      "grad_norm": 0.3127455711364746,
+      "learning_rate": 2.1390804597701152e-05,
+      "loss": 0.0896,
+      "step": 4978
+    },
+    {
+      "epoch": 5.771498949503731,
+      "grad_norm": 0.37766167521476746,
+      "learning_rate": 2.138505747126437e-05,
+      "loss": 0.0906,
+      "step": 4979
+    },
+    {
+      "epoch": 5.7726581178004785,
+      "grad_norm": 0.49125292897224426,
+      "learning_rate": 2.137931034482759e-05,
+      "loss": 0.0963,
+      "step": 4980
+    },
+    {
+      "epoch": 5.773817286097225,
+      "grad_norm": 0.4246441423892975,
+      "learning_rate": 2.1373563218390803e-05,
+      "loss": 0.0919,
+      "step": 4981
+    },
+    {
+      "epoch": 5.774976454393972,
+      "grad_norm": 0.366273432970047,
+      "learning_rate": 2.1367816091954025e-05,
+      "loss": 0.0896,
+      "step": 4982
+    },
+    {
+      "epoch": 5.7761356226907195,
+      "grad_norm": 0.4927791655063629,
+      "learning_rate": 2.1362068965517243e-05,
+      "loss": 0.1007,
+      "step": 4983
+    },
+    {
+      "epoch": 5.777294790987466,
+      "grad_norm": 0.29492560029029846,
+      "learning_rate": 2.135632183908046e-05,
+      "loss": 0.0959,
+      "step": 4984
+    },
+    {
+      "epoch": 5.778453959284214,
+      "grad_norm": 0.5132672786712646,
+      "learning_rate": 2.135057471264368e-05,
+      "loss": 0.0917,
+      "step": 4985
+    },
+    {
+      "epoch": 5.77961312758096,
+      "grad_norm": 0.3356097340583801,
+      "learning_rate": 2.1344827586206898e-05,
+      "loss": 0.083,
+      "step": 4986
+    },
+    {
+      "epoch": 5.780772295877708,
+      "grad_norm": 0.43617239594459534,
+      "learning_rate": 2.1339080459770116e-05,
+      "loss": 0.1005,
+      "step": 4987
+    },
+    {
+      "epoch": 5.781931464174455,
+      "grad_norm": 0.4654097259044647,
+      "learning_rate": 2.1333333333333335e-05,
+      "loss": 0.0922,
+      "step": 4988
+    },
+    {
+      "epoch": 5.783090632471202,
+      "grad_norm": 0.42662137746810913,
+      "learning_rate": 2.1327586206896553e-05,
+      "loss": 0.099,
+      "step": 4989
+    },
+    {
+      "epoch": 5.784249800767949,
+      "grad_norm": 0.4461085796356201,
+      "learning_rate": 2.132183908045977e-05,
+      "loss": 0.0907,
+      "step": 4990
+    },
+    {
+      "epoch": 5.785408969064696,
+      "grad_norm": 0.6144752502441406,
+      "learning_rate": 2.131609195402299e-05,
+      "loss": 0.0959,
+      "step": 4991
+    },
+    {
+      "epoch": 5.786568137361443,
+      "grad_norm": 0.4535113275051117,
+      "learning_rate": 2.1310344827586208e-05,
+      "loss": 0.0941,
+      "step": 4992
+    },
+    {
+      "epoch": 5.787727305658191,
+      "grad_norm": 0.556451141834259,
+      "learning_rate": 2.1304597701149426e-05,
+      "loss": 0.1009,
+      "step": 4993
+    },
+    {
+      "epoch": 5.788886473954937,
+      "grad_norm": 0.40173500776290894,
+      "learning_rate": 2.1298850574712644e-05,
+      "loss": 0.0952,
+      "step": 4994
+    },
+    {
+      "epoch": 5.790045642251684,
+      "grad_norm": 0.485877126455307,
+      "learning_rate": 2.1293103448275863e-05,
+      "loss": 0.0937,
+      "step": 4995
+    },
+    {
+      "epoch": 5.7912048105484315,
+      "grad_norm": 0.425545334815979,
+      "learning_rate": 2.128735632183908e-05,
+      "loss": 0.0954,
+      "step": 4996
+    },
+    {
+      "epoch": 5.792363978845179,
+      "grad_norm": 0.463670939207077,
+      "learning_rate": 2.12816091954023e-05,
+      "loss": 0.09,
+      "step": 4997
+    },
+    {
+      "epoch": 5.793523147141926,
+      "grad_norm": 0.3836846947669983,
+      "learning_rate": 2.1275862068965518e-05,
+      "loss": 0.0934,
+      "step": 4998
+    },
+    {
+      "epoch": 5.794682315438672,
+      "grad_norm": 0.7452934980392456,
+      "learning_rate": 2.127011494252874e-05,
+      "loss": 0.094,
+      "step": 4999
+    },
+    {
+      "epoch": 5.79584148373542,
+      "grad_norm": 0.364729106426239,
+      "learning_rate": 2.1264367816091954e-05,
+      "loss": 0.092,
+      "step": 5000
+    },
+    {
+      "epoch": 5.797000652032167,
+      "grad_norm": 0.560750424861908,
+      "learning_rate": 2.1258620689655172e-05,
+      "loss": 0.0987,
+      "step": 5001
+    },
+    {
+      "epoch": 5.798159820328914,
+      "grad_norm": 0.3720407485961914,
+      "learning_rate": 2.125287356321839e-05,
+      "loss": 0.0894,
+      "step": 5002
+    },
+    {
+      "epoch": 5.799318988625661,
+      "grad_norm": 0.3824548125267029,
+      "learning_rate": 2.1247126436781612e-05,
+      "loss": 0.0929,
+      "step": 5003
+    },
+    {
+      "epoch": 5.800478156922408,
+      "grad_norm": 0.3994615972042084,
+      "learning_rate": 2.1241379310344827e-05,
+      "loss": 0.1029,
+      "step": 5004
+    },
+    {
+      "epoch": 5.801637325219155,
+      "grad_norm": 0.4771389067173004,
+      "learning_rate": 2.1235632183908046e-05,
+      "loss": 0.0842,
+      "step": 5005
+    },
+    {
+      "epoch": 5.802796493515903,
+      "grad_norm": 0.4739455580711365,
+      "learning_rate": 2.1229885057471267e-05,
+      "loss": 0.1055,
+      "step": 5006
+    },
+    {
+      "epoch": 5.803955661812649,
+      "grad_norm": 0.5355760455131531,
+      "learning_rate": 2.1224137931034486e-05,
+      "loss": 0.0892,
+      "step": 5007
+    },
+    {
+      "epoch": 5.805114830109397,
+      "grad_norm": 0.5140407681465149,
+      "learning_rate": 2.12183908045977e-05,
+      "loss": 0.0986,
+      "step": 5008
+    },
+    {
+      "epoch": 5.8062739984061436,
+      "grad_norm": 0.3573043644428253,
+      "learning_rate": 2.1212643678160922e-05,
+      "loss": 0.0904,
+      "step": 5009
+    },
+    {
+      "epoch": 5.807433166702891,
+      "grad_norm": 0.3605383634567261,
+      "learning_rate": 2.120689655172414e-05,
+      "loss": 0.0984,
+      "step": 5010
+    },
+    {
+      "epoch": 5.808592334999638,
+      "grad_norm": 0.3857939541339874,
+      "learning_rate": 2.1201149425287355e-05,
+      "loss": 0.0911,
+      "step": 5011
+    },
+    {
+      "epoch": 5.8097515032963845,
+      "grad_norm": 0.3786557614803314,
+      "learning_rate": 2.1195402298850574e-05,
+      "loss": 0.0969,
+      "step": 5012
+    },
+    {
+      "epoch": 5.810910671593132,
+      "grad_norm": 0.4758943021297455,
+      "learning_rate": 2.1189655172413795e-05,
+      "loss": 0.1068,
+      "step": 5013
+    },
+    {
+      "epoch": 5.812069839889879,
+      "grad_norm": 0.5058685541152954,
+      "learning_rate": 2.1183908045977014e-05,
+      "loss": 0.0927,
+      "step": 5014
+    },
+    {
+      "epoch": 5.813229008186626,
+      "grad_norm": 0.31769004464149475,
+      "learning_rate": 2.117816091954023e-05,
+      "loss": 0.0904,
+      "step": 5015
+    },
+    {
+      "epoch": 5.814388176483373,
+      "grad_norm": 0.354096919298172,
+      "learning_rate": 2.117241379310345e-05,
+      "loss": 0.1018,
+      "step": 5016
+    },
+    {
+      "epoch": 5.8155473447801205,
+      "grad_norm": 0.41636335849761963,
+      "learning_rate": 2.116666666666667e-05,
+      "loss": 0.0969,
+      "step": 5017
+    },
+    {
+      "epoch": 5.816706513076867,
+      "grad_norm": 0.46119603514671326,
+      "learning_rate": 2.1160919540229887e-05,
+      "loss": 0.1032,
+      "step": 5018
+    },
+    {
+      "epoch": 5.817865681373615,
+      "grad_norm": 0.38647714257240295,
+      "learning_rate": 2.1155172413793105e-05,
+      "loss": 0.0972,
+      "step": 5019
+    },
+    {
+      "epoch": 5.819024849670361,
+      "grad_norm": 0.3750416338443756,
+      "learning_rate": 2.1149425287356323e-05,
+      "loss": 0.0897,
+      "step": 5020
+    },
+    {
+      "epoch": 5.820184017967109,
+      "grad_norm": 0.4357173442840576,
+      "learning_rate": 2.114367816091954e-05,
+      "loss": 0.0988,
+      "step": 5021
+    },
+    {
+      "epoch": 5.821343186263856,
+      "grad_norm": 0.2895772457122803,
+      "learning_rate": 2.113793103448276e-05,
+      "loss": 0.0955,
+      "step": 5022
+    },
+    {
+      "epoch": 5.822502354560603,
+      "grad_norm": 0.3872501850128174,
+      "learning_rate": 2.1132183908045978e-05,
+      "loss": 0.103,
+      "step": 5023
+    },
+    {
+      "epoch": 5.82366152285735,
+      "grad_norm": 0.40317487716674805,
+      "learning_rate": 2.1126436781609196e-05,
+      "loss": 0.0928,
+      "step": 5024
+    },
+    {
+      "epoch": 5.8248206911540965,
+      "grad_norm": 0.49441027641296387,
+      "learning_rate": 2.1120689655172415e-05,
+      "loss": 0.0944,
+      "step": 5025
+    },
+    {
+      "epoch": 5.825979859450844,
+      "grad_norm": 0.39836275577545166,
+      "learning_rate": 2.1114942528735633e-05,
+      "loss": 0.0926,
+      "step": 5026
+    },
+    {
+      "epoch": 5.827139027747591,
+      "grad_norm": 0.5158929824829102,
+      "learning_rate": 2.110919540229885e-05,
+      "loss": 0.0888,
+      "step": 5027
+    },
+    {
+      "epoch": 5.828298196044338,
+      "grad_norm": 0.4455549716949463,
+      "learning_rate": 2.110344827586207e-05,
+      "loss": 0.0907,
+      "step": 5028
+    },
+    {
+      "epoch": 5.829457364341085,
+      "grad_norm": 0.4579765498638153,
+      "learning_rate": 2.1097701149425288e-05,
+      "loss": 0.0903,
+      "step": 5029
+    },
+    {
+      "epoch": 5.8306165326378325,
+      "grad_norm": 0.6447916030883789,
+      "learning_rate": 2.1091954022988506e-05,
+      "loss": 0.0945,
+      "step": 5030
+    },
+    {
+      "epoch": 5.831775700934579,
+      "grad_norm": 0.44061580300331116,
+      "learning_rate": 2.1086206896551724e-05,
+      "loss": 0.1007,
+      "step": 5031
+    },
+    {
+      "epoch": 5.832934869231327,
+      "grad_norm": 0.46343931555747986,
+      "learning_rate": 2.1080459770114943e-05,
+      "loss": 0.1078,
+      "step": 5032
+    },
+    {
+      "epoch": 5.834094037528073,
+      "grad_norm": 0.4852432310581207,
+      "learning_rate": 2.1074712643678164e-05,
+      "loss": 0.1009,
+      "step": 5033
+    },
+    {
+      "epoch": 5.835253205824821,
+      "grad_norm": 0.2604130208492279,
+      "learning_rate": 2.106896551724138e-05,
+      "loss": 0.09,
+      "step": 5034
+    },
+    {
+      "epoch": 5.836412374121568,
+      "grad_norm": 0.6115668416023254,
+      "learning_rate": 2.1063218390804598e-05,
+      "loss": 0.1032,
+      "step": 5035
+    },
+    {
+      "epoch": 5.837571542418315,
+      "grad_norm": 0.4292698800563812,
+      "learning_rate": 2.1057471264367816e-05,
+      "loss": 0.0984,
+      "step": 5036
+    },
+    {
+      "epoch": 5.838730710715062,
+      "grad_norm": 0.42464151978492737,
+      "learning_rate": 2.1051724137931038e-05,
+      "loss": 0.0854,
+      "step": 5037
+    },
+    {
+      "epoch": 5.839889879011809,
+      "grad_norm": 0.36907604336738586,
+      "learning_rate": 2.1045977011494253e-05,
+      "loss": 0.0948,
+      "step": 5038
+    },
+    {
+      "epoch": 5.841049047308556,
+      "grad_norm": 0.43384525179862976,
+      "learning_rate": 2.104022988505747e-05,
+      "loss": 0.0955,
+      "step": 5039
+    },
+    {
+      "epoch": 5.842208215605304,
+      "grad_norm": 0.40040022134780884,
+      "learning_rate": 2.1034482758620692e-05,
+      "loss": 0.0971,
+      "step": 5040
+    },
+    {
+      "epoch": 5.84336738390205,
+      "grad_norm": 0.4272783100605011,
+      "learning_rate": 2.102873563218391e-05,
+      "loss": 0.0898,
+      "step": 5041
+    },
+    {
+      "epoch": 5.844526552198797,
+      "grad_norm": 0.3518434762954712,
+      "learning_rate": 2.1022988505747126e-05,
+      "loss": 0.0905,
+      "step": 5042
+    },
+    {
+      "epoch": 5.845685720495545,
+      "grad_norm": 0.39061328768730164,
+      "learning_rate": 2.1017241379310344e-05,
+      "loss": 0.0924,
+      "step": 5043
+    },
+    {
+      "epoch": 5.846844888792291,
+      "grad_norm": 0.4481010437011719,
+      "learning_rate": 2.1011494252873566e-05,
+      "loss": 0.0922,
+      "step": 5044
+    },
+    {
+      "epoch": 5.848004057089039,
+      "grad_norm": 0.5339946746826172,
+      "learning_rate": 2.1005747126436784e-05,
+      "loss": 0.1021,
+      "step": 5045
+    },
+    {
+      "epoch": 5.8491632253857855,
+      "grad_norm": 0.6023551821708679,
+      "learning_rate": 2.1e-05,
+      "loss": 0.0971,
+      "step": 5046
+    },
+    {
+      "epoch": 5.850322393682533,
+      "grad_norm": 0.41510429978370667,
+      "learning_rate": 2.099425287356322e-05,
+      "loss": 0.1031,
+      "step": 5047
+    },
+    {
+      "epoch": 5.85148156197928,
+      "grad_norm": 0.4894249439239502,
+      "learning_rate": 2.098850574712644e-05,
+      "loss": 0.0986,
+      "step": 5048
+    },
+    {
+      "epoch": 5.852640730276027,
+      "grad_norm": 0.3392792046070099,
+      "learning_rate": 2.0982758620689654e-05,
+      "loss": 0.0904,
+      "step": 5049
+    },
+    {
+      "epoch": 5.853799898572774,
+      "grad_norm": 0.43131858110427856,
+      "learning_rate": 2.0977011494252875e-05,
+      "loss": 0.0933,
+      "step": 5050
+    },
+    {
+      "epoch": 5.8549590668695215,
+      "grad_norm": 0.3799227476119995,
+      "learning_rate": 2.0971264367816094e-05,
+      "loss": 0.0892,
+      "step": 5051
+    },
+    {
+      "epoch": 5.856118235166268,
+      "grad_norm": 0.44021299481391907,
+      "learning_rate": 2.0965517241379312e-05,
+      "loss": 0.0978,
+      "step": 5052
+    },
+    {
+      "epoch": 5.857277403463016,
+      "grad_norm": 0.39485037326812744,
+      "learning_rate": 2.0959770114942527e-05,
+      "loss": 0.0918,
+      "step": 5053
+    },
+    {
+      "epoch": 5.858436571759762,
+      "grad_norm": 0.367180734872818,
+      "learning_rate": 2.095402298850575e-05,
+      "loss": 0.1015,
+      "step": 5054
+    },
+    {
+      "epoch": 5.859595740056509,
+      "grad_norm": 0.6080685257911682,
+      "learning_rate": 2.0948275862068967e-05,
+      "loss": 0.0916,
+      "step": 5055
+    },
+    {
+      "epoch": 5.860754908353257,
+      "grad_norm": 0.3705676198005676,
+      "learning_rate": 2.0942528735632185e-05,
+      "loss": 0.0935,
+      "step": 5056
+    },
+    {
+      "epoch": 5.861914076650003,
+      "grad_norm": 0.41616174578666687,
+      "learning_rate": 2.0936781609195403e-05,
+      "loss": 0.096,
+      "step": 5057
+    },
+    {
+      "epoch": 5.863073244946751,
+      "grad_norm": 0.3993401825428009,
+      "learning_rate": 2.0931034482758622e-05,
+      "loss": 0.0943,
+      "step": 5058
+    },
+    {
+      "epoch": 5.8642324132434975,
+      "grad_norm": 0.4943077564239502,
+      "learning_rate": 2.092528735632184e-05,
+      "loss": 0.096,
+      "step": 5059
+    },
+    {
+      "epoch": 5.865391581540245,
+      "grad_norm": 0.3290475010871887,
+      "learning_rate": 2.0919540229885058e-05,
+      "loss": 0.0948,
+      "step": 5060
+    },
+    {
+      "epoch": 5.866550749836992,
+      "grad_norm": 0.39300331473350525,
+      "learning_rate": 2.0913793103448277e-05,
+      "loss": 0.1004,
+      "step": 5061
+    },
+    {
+      "epoch": 5.867709918133739,
+      "grad_norm": 0.4835030138492584,
+      "learning_rate": 2.0908045977011495e-05,
+      "loss": 0.0884,
+      "step": 5062
+    },
+    {
+      "epoch": 5.868869086430486,
+      "grad_norm": 0.5800598859786987,
+      "learning_rate": 2.0902298850574713e-05,
+      "loss": 0.0988,
+      "step": 5063
+    },
+    {
+      "epoch": 5.8700282547272336,
+      "grad_norm": 0.4258168041706085,
+      "learning_rate": 2.0896551724137935e-05,
+      "loss": 0.1035,
+      "step": 5064
+    },
+    {
+      "epoch": 5.87118742302398,
+      "grad_norm": 0.4930982291698456,
+      "learning_rate": 2.089080459770115e-05,
+      "loss": 0.0923,
+      "step": 5065
+    },
+    {
+      "epoch": 5.872346591320728,
+      "grad_norm": 0.3994540870189667,
+      "learning_rate": 2.0885057471264368e-05,
+      "loss": 0.0916,
+      "step": 5066
+    },
+    {
+      "epoch": 5.8735057596174745,
+      "grad_norm": 0.35855406522750854,
+      "learning_rate": 2.0879310344827586e-05,
+      "loss": 0.0976,
+      "step": 5067
+    },
+    {
+      "epoch": 5.874664927914221,
+      "grad_norm": 0.37041381001472473,
+      "learning_rate": 2.0873563218390808e-05,
+      "loss": 0.0863,
+      "step": 5068
+    },
+    {
+      "epoch": 5.875824096210969,
+      "grad_norm": 0.44822248816490173,
+      "learning_rate": 2.0867816091954023e-05,
+      "loss": 0.0963,
+      "step": 5069
+    },
+    {
+      "epoch": 5.876983264507715,
+      "grad_norm": 0.3357725739479065,
+      "learning_rate": 2.086206896551724e-05,
+      "loss": 0.0877,
+      "step": 5070
+    },
+    {
+      "epoch": 5.878142432804463,
+      "grad_norm": 0.37162381410598755,
+      "learning_rate": 2.0856321839080463e-05,
+      "loss": 0.0874,
+      "step": 5071
+    },
+    {
+      "epoch": 5.87930160110121,
+      "grad_norm": 0.4773227274417877,
+      "learning_rate": 2.0850574712643678e-05,
+      "loss": 0.0965,
+      "step": 5072
+    },
+    {
+      "epoch": 5.880460769397957,
+      "grad_norm": 0.47884953022003174,
+      "learning_rate": 2.0844827586206896e-05,
+      "loss": 0.0946,
+      "step": 5073
+    },
+    {
+      "epoch": 5.881619937694704,
+      "grad_norm": 0.3373884856700897,
+      "learning_rate": 2.0839080459770118e-05,
+      "loss": 0.0949,
+      "step": 5074
+    },
+    {
+      "epoch": 5.882779105991451,
+      "grad_norm": 0.399009644985199,
+      "learning_rate": 2.0833333333333336e-05,
+      "loss": 0.0953,
+      "step": 5075
+    },
+    {
+      "epoch": 5.883938274288198,
+      "grad_norm": 0.44031986594200134,
+      "learning_rate": 2.082758620689655e-05,
+      "loss": 0.1024,
+      "step": 5076
+    },
+    {
+      "epoch": 5.885097442584946,
+      "grad_norm": 0.3464363217353821,
+      "learning_rate": 2.082183908045977e-05,
+      "loss": 0.1056,
+      "step": 5077
+    },
+    {
+      "epoch": 5.886256610881692,
+      "grad_norm": 0.4489310383796692,
+      "learning_rate": 2.081609195402299e-05,
+      "loss": 0.0995,
+      "step": 5078
+    },
+    {
+      "epoch": 5.88741577917844,
+      "grad_norm": 0.43739187717437744,
+      "learning_rate": 2.081034482758621e-05,
+      "loss": 0.0965,
+      "step": 5079
+    },
+    {
+      "epoch": 5.8885749474751865,
+      "grad_norm": 0.35651272535324097,
+      "learning_rate": 2.0804597701149424e-05,
+      "loss": 0.0944,
+      "step": 5080
+    },
+    {
+      "epoch": 5.889734115771933,
+      "grad_norm": 0.4365938603878021,
+      "learning_rate": 2.0798850574712646e-05,
+      "loss": 0.101,
+      "step": 5081
+    },
+    {
+      "epoch": 5.890893284068681,
+      "grad_norm": 0.4614748954772949,
+      "learning_rate": 2.0793103448275864e-05,
+      "loss": 0.0986,
+      "step": 5082
+    },
+    {
+      "epoch": 5.892052452365428,
+      "grad_norm": 0.3819468915462494,
+      "learning_rate": 2.0787356321839082e-05,
+      "loss": 0.0882,
+      "step": 5083
+    },
+    {
+      "epoch": 5.893211620662175,
+      "grad_norm": 0.5456834435462952,
+      "learning_rate": 2.07816091954023e-05,
+      "loss": 0.0931,
+      "step": 5084
+    },
+    {
+      "epoch": 5.894370788958922,
+      "grad_norm": 0.48817238211631775,
+      "learning_rate": 2.077586206896552e-05,
+      "loss": 0.0946,
+      "step": 5085
+    },
+    {
+      "epoch": 5.895529957255669,
+      "grad_norm": 0.5912837386131287,
+      "learning_rate": 2.0770114942528737e-05,
+      "loss": 0.0907,
+      "step": 5086
+    },
+    {
+      "epoch": 5.896689125552416,
+      "grad_norm": 0.38435614109039307,
+      "learning_rate": 2.0764367816091956e-05,
+      "loss": 0.0919,
+      "step": 5087
+    },
+    {
+      "epoch": 5.897848293849163,
+      "grad_norm": 0.3552255928516388,
+      "learning_rate": 2.0758620689655174e-05,
+      "loss": 0.0925,
+      "step": 5088
+    },
+    {
+      "epoch": 5.89900746214591,
+      "grad_norm": 0.3824387192726135,
+      "learning_rate": 2.0752873563218392e-05,
+      "loss": 0.1059,
+      "step": 5089
+    },
+    {
+      "epoch": 5.900166630442658,
+      "grad_norm": 0.43972963094711304,
+      "learning_rate": 2.074712643678161e-05,
+      "loss": 0.0966,
+      "step": 5090
+    },
+    {
+      "epoch": 5.901325798739404,
+      "grad_norm": 0.3910900354385376,
+      "learning_rate": 2.074137931034483e-05,
+      "loss": 0.0889,
+      "step": 5091
+    },
+    {
+      "epoch": 5.902484967036152,
+      "grad_norm": 0.45151054859161377,
+      "learning_rate": 2.0735632183908047e-05,
+      "loss": 0.1027,
+      "step": 5092
+    },
+    {
+      "epoch": 5.903644135332899,
+      "grad_norm": 0.3880411982536316,
+      "learning_rate": 2.0729885057471265e-05,
+      "loss": 0.0947,
+      "step": 5093
+    },
+    {
+      "epoch": 5.904803303629646,
+      "grad_norm": 0.40984082221984863,
+      "learning_rate": 2.0724137931034484e-05,
+      "loss": 0.0929,
+      "step": 5094
+    },
+    {
+      "epoch": 5.905962471926393,
+      "grad_norm": 0.32482361793518066,
+      "learning_rate": 2.0718390804597702e-05,
+      "loss": 0.0848,
+      "step": 5095
+    },
+    {
+      "epoch": 5.90712164022314,
+      "grad_norm": 0.438496470451355,
+      "learning_rate": 2.071264367816092e-05,
+      "loss": 0.088,
+      "step": 5096
+    },
+    {
+      "epoch": 5.908280808519887,
+      "grad_norm": 0.47613009810447693,
+      "learning_rate": 2.070689655172414e-05,
+      "loss": 0.0942,
+      "step": 5097
+    },
+    {
+      "epoch": 5.909439976816634,
+      "grad_norm": 0.3807111978530884,
+      "learning_rate": 2.0701149425287357e-05,
+      "loss": 0.096,
+      "step": 5098
+    },
+    {
+      "epoch": 5.910599145113381,
+      "grad_norm": 0.4180908799171448,
+      "learning_rate": 2.0695402298850575e-05,
+      "loss": 0.0892,
+      "step": 5099
+    },
+    {
+      "epoch": 5.911758313410128,
+      "grad_norm": 0.5334672927856445,
+      "learning_rate": 2.0689655172413793e-05,
+      "loss": 0.0952,
+      "step": 5100
+    },
+    {
+      "epoch": 5.9129174817068755,
+      "grad_norm": 0.4528830945491791,
+      "learning_rate": 2.068390804597701e-05,
+      "loss": 0.0967,
+      "step": 5101
+    },
+    {
+      "epoch": 5.914076650003622,
+      "grad_norm": 0.4010438323020935,
+      "learning_rate": 2.0678160919540233e-05,
+      "loss": 0.0969,
+      "step": 5102
+    },
+    {
+      "epoch": 5.91523581830037,
+      "grad_norm": 0.3919377624988556,
+      "learning_rate": 2.0672413793103448e-05,
+      "loss": 0.0836,
+      "step": 5103
+    },
+    {
+      "epoch": 5.916394986597116,
+      "grad_norm": 0.3746892511844635,
+      "learning_rate": 2.0666666666666666e-05,
+      "loss": 0.0957,
+      "step": 5104
+    },
+    {
+      "epoch": 5.917554154893864,
+      "grad_norm": 0.5136653780937195,
+      "learning_rate": 2.0660919540229888e-05,
+      "loss": 0.0989,
+      "step": 5105
+    },
+    {
+      "epoch": 5.918713323190611,
+      "grad_norm": 0.4311780631542206,
+      "learning_rate": 2.0655172413793106e-05,
+      "loss": 0.0998,
+      "step": 5106
+    },
+    {
+      "epoch": 5.919872491487358,
+      "grad_norm": 0.4143439829349518,
+      "learning_rate": 2.064942528735632e-05,
+      "loss": 0.0921,
+      "step": 5107
+    },
+    {
+      "epoch": 5.921031659784105,
+      "grad_norm": 0.857010543346405,
+      "learning_rate": 2.064367816091954e-05,
+      "loss": 0.0987,
+      "step": 5108
+    },
+    {
+      "epoch": 5.922190828080852,
+      "grad_norm": 0.5835907459259033,
+      "learning_rate": 2.063793103448276e-05,
+      "loss": 0.1125,
+      "step": 5109
+    },
+    {
+      "epoch": 5.923349996377599,
+      "grad_norm": 0.3759761154651642,
+      "learning_rate": 2.0632183908045976e-05,
+      "loss": 0.0884,
+      "step": 5110
+    },
+    {
+      "epoch": 5.924509164674346,
+      "grad_norm": 0.4615977704524994,
+      "learning_rate": 2.0626436781609194e-05,
+      "loss": 0.0969,
+      "step": 5111
+    },
+    {
+      "epoch": 5.925668332971093,
+      "grad_norm": 0.3750910758972168,
+      "learning_rate": 2.0620689655172416e-05,
+      "loss": 0.0927,
+      "step": 5112
+    },
+    {
+      "epoch": 5.92682750126784,
+      "grad_norm": 0.42511269450187683,
+      "learning_rate": 2.0614942528735634e-05,
+      "loss": 0.0899,
+      "step": 5113
+    },
+    {
+      "epoch": 5.9279866695645875,
+      "grad_norm": 0.5754655003547668,
+      "learning_rate": 2.060919540229885e-05,
+      "loss": 0.0914,
+      "step": 5114
+    },
+    {
+      "epoch": 5.929145837861334,
+      "grad_norm": 0.43884530663490295,
+      "learning_rate": 2.060344827586207e-05,
+      "loss": 0.0984,
+      "step": 5115
+    },
+    {
+      "epoch": 5.930305006158082,
+      "grad_norm": 0.3772994577884674,
+      "learning_rate": 2.059770114942529e-05,
+      "loss": 0.0964,
+      "step": 5116
+    },
+    {
+      "epoch": 5.931464174454828,
+      "grad_norm": 0.46012642979621887,
+      "learning_rate": 2.0591954022988508e-05,
+      "loss": 0.0927,
+      "step": 5117
+    },
+    {
+      "epoch": 5.932623342751576,
+      "grad_norm": 0.39248010516166687,
+      "learning_rate": 2.0586206896551722e-05,
+      "loss": 0.0925,
+      "step": 5118
+    },
+    {
+      "epoch": 5.933782511048323,
+      "grad_norm": 0.42726606130599976,
+      "learning_rate": 2.0580459770114944e-05,
+      "loss": 0.1009,
+      "step": 5119
+    },
+    {
+      "epoch": 5.93494167934507,
+      "grad_norm": 0.4554595947265625,
+      "learning_rate": 2.0574712643678162e-05,
+      "loss": 0.1023,
+      "step": 5120
+    },
+    {
+      "epoch": 5.936100847641817,
+      "grad_norm": 0.34201934933662415,
+      "learning_rate": 2.056896551724138e-05,
+      "loss": 0.095,
+      "step": 5121
+    },
+    {
+      "epoch": 5.9372600159385645,
+      "grad_norm": 0.4734445810317993,
+      "learning_rate": 2.05632183908046e-05,
+      "loss": 0.0948,
+      "step": 5122
+    },
+    {
+      "epoch": 5.938419184235311,
+      "grad_norm": 0.4069608449935913,
+      "learning_rate": 2.0557471264367817e-05,
+      "loss": 0.0912,
+      "step": 5123
+    },
+    {
+      "epoch": 5.939578352532058,
+      "grad_norm": 0.39140281081199646,
+      "learning_rate": 2.0551724137931036e-05,
+      "loss": 0.09,
+      "step": 5124
+    },
+    {
+      "epoch": 5.940737520828805,
+      "grad_norm": 0.547514796257019,
+      "learning_rate": 2.0545977011494254e-05,
+      "loss": 0.0972,
+      "step": 5125
+    },
+    {
+      "epoch": 5.941896689125553,
+      "grad_norm": 0.34587743878364563,
+      "learning_rate": 2.0540229885057472e-05,
+      "loss": 0.0912,
+      "step": 5126
+    },
+    {
+      "epoch": 5.9430558574223,
+      "grad_norm": 0.5674780607223511,
+      "learning_rate": 2.053448275862069e-05,
+      "loss": 0.1001,
+      "step": 5127
+    },
+    {
+      "epoch": 5.944215025719046,
+      "grad_norm": 0.3166492283344269,
+      "learning_rate": 2.052873563218391e-05,
+      "loss": 0.0838,
+      "step": 5128
+    },
+    {
+      "epoch": 5.945374194015794,
+      "grad_norm": 0.3704341948032379,
+      "learning_rate": 2.0522988505747127e-05,
+      "loss": 0.0883,
+      "step": 5129
+    },
+    {
+      "epoch": 5.9465333623125405,
+      "grad_norm": 0.3320915699005127,
+      "learning_rate": 2.0517241379310345e-05,
+      "loss": 0.0836,
+      "step": 5130
+    },
+    {
+      "epoch": 5.947692530609288,
+      "grad_norm": 0.40248364210128784,
+      "learning_rate": 2.0511494252873564e-05,
+      "loss": 0.0946,
+      "step": 5131
+    },
+    {
+      "epoch": 5.948851698906035,
+      "grad_norm": 0.523185670375824,
+      "learning_rate": 2.0505747126436782e-05,
+      "loss": 0.0928,
+      "step": 5132
+    },
+    {
+      "epoch": 5.950010867202782,
+      "grad_norm": 0.3942788243293762,
+      "learning_rate": 2.05e-05,
+      "loss": 0.0931,
+      "step": 5133
+    },
+    {
+      "epoch": 5.951170035499529,
+      "grad_norm": 0.39801275730133057,
+      "learning_rate": 2.049425287356322e-05,
+      "loss": 0.096,
+      "step": 5134
+    },
+    {
+      "epoch": 5.9523292037962765,
+      "grad_norm": 0.5294638276100159,
+      "learning_rate": 2.0488505747126437e-05,
+      "loss": 0.0982,
+      "step": 5135
+    },
+    {
+      "epoch": 5.953488372093023,
+      "grad_norm": 0.732921838760376,
+      "learning_rate": 2.048275862068966e-05,
+      "loss": 0.1096,
+      "step": 5136
+    },
+    {
+      "epoch": 5.95464754038977,
+      "grad_norm": 0.39925023913383484,
+      "learning_rate": 2.0477011494252873e-05,
+      "loss": 0.0964,
+      "step": 5137
+    },
+    {
+      "epoch": 5.955806708686517,
+      "grad_norm": 0.47731050848960876,
+      "learning_rate": 2.047126436781609e-05,
+      "loss": 0.0994,
+      "step": 5138
+    },
+    {
+      "epoch": 5.956965876983265,
+      "grad_norm": 0.48555245995521545,
+      "learning_rate": 2.0465517241379313e-05,
+      "loss": 0.0893,
+      "step": 5139
+    },
+    {
+      "epoch": 5.958125045280012,
+      "grad_norm": 0.33314958214759827,
+      "learning_rate": 2.045977011494253e-05,
+      "loss": 0.0874,
+      "step": 5140
+    },
+    {
+      "epoch": 5.959284213576758,
+      "grad_norm": 0.5869859457015991,
+      "learning_rate": 2.0454022988505747e-05,
+      "loss": 0.0995,
+      "step": 5141
+    },
+    {
+      "epoch": 5.960443381873506,
+      "grad_norm": 0.3285931348800659,
+      "learning_rate": 2.0448275862068965e-05,
+      "loss": 0.0937,
+      "step": 5142
+    },
+    {
+      "epoch": 5.9616025501702525,
+      "grad_norm": 0.3716854751110077,
+      "learning_rate": 2.0442528735632187e-05,
+      "loss": 0.0976,
+      "step": 5143
+    },
+    {
+      "epoch": 5.962761718467,
+      "grad_norm": 0.3806571960449219,
+      "learning_rate": 2.0436781609195405e-05,
+      "loss": 0.096,
+      "step": 5144
+    },
+    {
+      "epoch": 5.963920886763747,
+      "grad_norm": 0.4217926859855652,
+      "learning_rate": 2.043103448275862e-05,
+      "loss": 0.0966,
+      "step": 5145
+    },
+    {
+      "epoch": 5.965080055060494,
+      "grad_norm": 0.3270195424556732,
+      "learning_rate": 2.042528735632184e-05,
+      "loss": 0.0873,
+      "step": 5146
+    },
+    {
+      "epoch": 5.966239223357241,
+      "grad_norm": 0.4076954126358032,
+      "learning_rate": 2.041954022988506e-05,
+      "loss": 0.0909,
+      "step": 5147
+    },
+    {
+      "epoch": 5.9673983916539886,
+      "grad_norm": 0.491375595331192,
+      "learning_rate": 2.0413793103448278e-05,
+      "loss": 0.092,
+      "step": 5148
+    },
+    {
+      "epoch": 5.968557559950735,
+      "grad_norm": 0.42915698885917664,
+      "learning_rate": 2.0408045977011493e-05,
+      "loss": 0.0966,
+      "step": 5149
+    },
+    {
+      "epoch": 5.969716728247483,
+      "grad_norm": 0.38473838567733765,
+      "learning_rate": 2.0402298850574715e-05,
+      "loss": 0.0911,
+      "step": 5150
+    },
+    {
+      "epoch": 5.9708758965442295,
+      "grad_norm": 0.4680832028388977,
+      "learning_rate": 2.0396551724137933e-05,
+      "loss": 0.0904,
+      "step": 5151
+    },
+    {
+      "epoch": 5.972035064840977,
+      "grad_norm": 0.4412045180797577,
+      "learning_rate": 2.0390804597701148e-05,
+      "loss": 0.1048,
+      "step": 5152
+    },
+    {
+      "epoch": 5.973194233137724,
+      "grad_norm": 0.4312341809272766,
+      "learning_rate": 2.038505747126437e-05,
+      "loss": 0.096,
+      "step": 5153
+    },
+    {
+      "epoch": 5.97435340143447,
+      "grad_norm": 0.5846208333969116,
+      "learning_rate": 2.0379310344827588e-05,
+      "loss": 0.0986,
+      "step": 5154
+    },
+    {
+      "epoch": 5.975512569731218,
+      "grad_norm": 0.3806995749473572,
+      "learning_rate": 2.0373563218390806e-05,
+      "loss": 0.091,
+      "step": 5155
+    },
+    {
+      "epoch": 5.976671738027965,
+      "grad_norm": 0.3785600960254669,
+      "learning_rate": 2.0367816091954024e-05,
+      "loss": 0.0969,
+      "step": 5156
+    },
+    {
+      "epoch": 5.977830906324712,
+      "grad_norm": 0.3626193106174469,
+      "learning_rate": 2.0362068965517243e-05,
+      "loss": 0.09,
+      "step": 5157
+    },
+    {
+      "epoch": 5.978990074621459,
+      "grad_norm": 0.33364805579185486,
+      "learning_rate": 2.035632183908046e-05,
+      "loss": 0.0933,
+      "step": 5158
+    },
+    {
+      "epoch": 5.980149242918206,
+      "grad_norm": 0.4983973503112793,
+      "learning_rate": 2.035057471264368e-05,
+      "loss": 0.1003,
+      "step": 5159
+    },
+    {
+      "epoch": 5.981308411214953,
+      "grad_norm": 0.5828344225883484,
+      "learning_rate": 2.0344827586206897e-05,
+      "loss": 0.0986,
+      "step": 5160
+    },
+    {
+      "epoch": 5.982467579511701,
+      "grad_norm": 0.3898671269416809,
+      "learning_rate": 2.0339080459770116e-05,
+      "loss": 0.0839,
+      "step": 5161
+    },
+    {
+      "epoch": 5.983626747808447,
+      "grad_norm": 0.3521953523159027,
+      "learning_rate": 2.0333333333333334e-05,
+      "loss": 0.0837,
+      "step": 5162
+    },
+    {
+      "epoch": 5.984785916105195,
+      "grad_norm": 0.40425774455070496,
+      "learning_rate": 2.0327586206896552e-05,
+      "loss": 0.095,
+      "step": 5163
+    },
+    {
+      "epoch": 5.9859450844019415,
+      "grad_norm": 0.45948293805122375,
+      "learning_rate": 2.032183908045977e-05,
+      "loss": 0.0974,
+      "step": 5164
+    },
+    {
+      "epoch": 5.987104252698689,
+      "grad_norm": 0.40446847677230835,
+      "learning_rate": 2.031609195402299e-05,
+      "loss": 0.0932,
+      "step": 5165
+    },
+    {
+      "epoch": 5.988263420995436,
+      "grad_norm": 0.49766218662261963,
+      "learning_rate": 2.0310344827586207e-05,
+      "loss": 0.0998,
+      "step": 5166
+    },
+    {
+      "epoch": 5.989422589292182,
+      "grad_norm": 0.5645685195922852,
+      "learning_rate": 2.030459770114943e-05,
+      "loss": 0.1052,
+      "step": 5167
+    },
+    {
+      "epoch": 5.99058175758893,
+      "grad_norm": 0.28001391887664795,
+      "learning_rate": 2.0298850574712644e-05,
+      "loss": 0.0839,
+      "step": 5168
+    },
+    {
+      "epoch": 5.991740925885677,
+      "grad_norm": 0.3594200611114502,
+      "learning_rate": 2.0293103448275862e-05,
+      "loss": 0.0934,
+      "step": 5169
+    },
+    {
+      "epoch": 5.992900094182424,
+      "grad_norm": 0.34900590777397156,
+      "learning_rate": 2.0287356321839084e-05,
+      "loss": 0.0856,
+      "step": 5170
+    },
+    {
+      "epoch": 5.994059262479171,
+      "grad_norm": 0.4731251001358032,
+      "learning_rate": 2.02816091954023e-05,
+      "loss": 0.1025,
+      "step": 5171
+    },
+    {
+      "epoch": 5.995218430775918,
+      "grad_norm": 0.3666514754295349,
+      "learning_rate": 2.0275862068965517e-05,
+      "loss": 0.0929,
+      "step": 5172
+    },
+    {
+      "epoch": 5.996377599072665,
+      "grad_norm": 0.348015695810318,
+      "learning_rate": 2.0270114942528735e-05,
+      "loss": 0.0902,
+      "step": 5173
+    },
+    {
+      "epoch": 5.997536767369413,
+      "grad_norm": 0.3184606432914734,
+      "learning_rate": 2.0264367816091957e-05,
+      "loss": 0.0891,
+      "step": 5174
+    },
+    {
+      "epoch": 5.998695935666159,
+      "grad_norm": 0.7606669664382935,
+      "learning_rate": 2.0258620689655172e-05,
+      "loss": 0.1031,
+      "step": 5175
+    },
+    {
+      "epoch": 5.999855103962907,
+      "grad_norm": 0.6448565721511841,
+      "learning_rate": 2.025287356321839e-05,
+      "loss": 0.1023,
+      "step": 5176
+    },
+    {
+      "epoch": 5.999855103962907,
+      "eval_loss": 0.148734450340271,
+      "eval_runtime": 265.936,
+      "eval_samples_per_second": 5.768,
+      "eval_steps_per_second": 5.768,
+      "step": 5176
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 8620,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.3198416576181862e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}